Organisation mondiale de la santé (OMS) · Publications

WHO guidelines on hepatitis B and C testing

16 février 2017 Organisation mondiale de la santé

Le texte intégral est hébergé par l’organisation qui le publie. lawenc.com indexe les métadonnées et renvoie vers la source officielle.

Retour à la vue par article

Texte intégral

GUIDELINES ON HEPATITIS B AND C TESTING FEBRUARY 2017

GUIDELINES

GUIDELINES ON HEPATITIS B AND C TESTING FEBRUARY 2017

GUIDELINES

WHO guidelines on hepatitis B and C testing ISBN 978-92-4-154998-1 © World Health Organization 2017 Some rights reserved. This work is available under the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 IGO licence (CC BY-NC-SA 3.0 IGO; https://creativecommons.org/licenses/by-nc-sa/3.0/igo). Under the terms of this licence, you may copy, redistribute and adapt the work for noncommercial purposes, provided the work is appropriately cited, as indicated below. In any use of this work, there should be no suggestion that WHO endorses any specific organization, products or services. The use of the WHO logo is not permitted. If you adapt the work, then you must license your work under the same or equivalent Creative Commons licence. If you create a translation of this work, you should add the following disclaimer along with the suggested citation: “This translation was not created by the World Health Organization (WHO). WHO is not responsible for the content or accuracy of this translation. The original English edition shall be the binding and authentic edition”. Any mediation relating to disputes arising under the licence shall be conducted in accordance with the mediation rules of the World Intellectual Property Organization. Suggested citation. WHO guidelines on hepatitis B and C testing. Geneva: World Health Organization; 2017. Licence: CC BY-NC-SA 3.0 IGO. Cataloguing-in-Publication (CIP) data. CIP data are available at http://apps.who.int/iris. Sales, rights and licensing. To purchase WHO publications, see http://apps.who.int/bookorders. To submit requests for commercial use and queries on rights and licensing, see http://www. who.int/about/licensing. Third-party materials. If you wish to reuse material from this work that is attributed to a third party, such as tables, figures or images, it is your responsibility to determine whether permission is needed for that reuse and to obtain permission from the copyright holder. The risk of claims resulting from infringement of any third-party-owned component in the work rests solely with the user. General disclaimers. The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of WHO concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted and dashed lines on maps represent approximate border lines for which there may not yet be full agreement. The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or recommended by WHO in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters. All reasonable precautions have been taken by WHO to verify the information contained in this publication. However, the published material is being distributed without warranty of any kind, either expressed or implied. The responsibility for the interpretation and use of the material lies with the reader. In no event shall WHO be liable for damages arising from its use. Printed in China Design and layout: www.blossoming.it

CONTENTS ACKNOWLEDGEMENTS ABBREVIATIONS AND ACRONYMS GLOSSARY OF TERMS EXECUTIVE SUMMARY SUMMARY ALGORITHMS SUMMARY OF RECOMMENDATIONS x xiv xvi xxi xxvi xxvii

PART 1: BACKGROUND 1. INTRODUCTION 1.1. Current challenges in viral hepatitis testing 1.2. Goals of viral hepatitis testing 1.3. Why are testing guidelines needed? 1.4. Goals and objectives of the guidelines 1.5. Scope of the guidelines 1.6. Target audience 1.7. Related WHO materials and guidelines 2. GUIDING PRINCIPLES 2.1. Promoting human rights and equity in access to hepatitis testing 2.2. The public health approach along the continuum of care 2.3. The WHO “5 Cs” 2.4. Accurate testing 3. METHODOLOGY AND PROCESS OF DEVELOPING THE GUIDELINES 3.1. WHO guideline development process 3.2. Systematic reviews and additional background work 3.3. Grading of quality of evidence and strength of recommendations 3.4. Formulation of recommendations 3.5. Declaration and management of conflicts of interest 3.6. Updating, disseminating and monitoring implementation of the guidelines

1 2 2 3 4 5 5 6 6 8 8 8 9 9 10 10 10 11 14 15 15

4. BACKGROUND – EPIDEMIOLOGY AND NATURAL HISTORY 16 4.1. Hepatitis B infection19 4.1.1. 4.1.2. 4.1.3. 4.1.4. 4.1.5. 4.1.6. Epidemiology of hepatitis B infection Transmission of hepatitis B infection Natural history of HBV infection Time course and interpretation of serological markers of HBV infection Preventing hepatitis B infection through vaccination Treatment of hepatitis B infection 19 20 20 21 24 24

4.2. 4.2.1. 4.2.2. 4.2.3. 4.2.4. 4.2.5. 4.2.6.

Hepatitis C infection24 Epidemiology of hepatitis C infection Transmission of hepatitis C infection Natural history of hepatitis C infection Time course of serological markers for HCV infection Prevention of hepatitis C infection Treatment of hepatitis C infection 24 25 27 27 28 28

5. BACKGROUND – DIAGNOSTICS FOR TESTING FOR HEPATITIS B AND C INFECTION 5.1. Types of viral hepatitis assays 5.2 Serological assays 5.3. Nucleic acid testing (NAT) technologies 5.4. Choice of serological assays 5.5. Selection of a one or two assay serological testing strategy

30 30 30 31 32 32

PART 2: RECOMMENDATIONS 6. WHO TO TEST FOR CHRONIC HEPATITIS B OR C INFECTION – testing approaches and service delivery 6.1. Recommendations 6.2. Background 6A TESTING APPROACHES TO DETECT CHRONIC HEPATITIS B 6.3. Summary of the evidence 6.4. Rationale for the recommendations 6B TESTING APPROACHES TO DETECT CHRONIC HEPATITIS C 6.5. Summary of the evidence 6.6. Rationale for the recommendations 6C SERVICE DELIVERY OF HEPATITIS B AND C TESTING 6.7. Rationale for the recommendations on community-based testing 6.8. Rationale for the recommendations on facility-based testing 6.9. Implementation considerations for HBV and HCV testing approaches 7. HOW TO TEST FOR CHRONIC HEPATITIS B INFECTION – choice of serological assay and testing strategy 7.1. Recommendations 7.2. Background 7.3. Summary of the evidence 7.4. Rationale for the recommendations on which assay to use 7.5. Rationale for the recommendations on testing strategy

35 36 36 38 40 40 41 45 45 46 49 49 50 51 52 52 53 54 57 59

vii

8. HOW TO TEST FOR CURRENT OR PAST HCV INFECTION (HCV EXPOSURE) – choice of serological assay and testing strategy 8.1. Recommendations 8.2. Background 8.3. Summary of the evidence 8.4. Rationale for the recommendations on which assay to use 8.5. Rationale for the recommendation for a one-serological assay testing strategy 8.6. Implementation considerations for HBsAg and HCV antibody serological testing 9. DETECTION OF VIRAEMIC HBV INFECTION – to guide who to treat or not treat 9.1. Recommendation 9.2. Background 9.3. Rationale for the recommendations (WHO 2015 HBV guidelines) 9.4. Implementation considerations 10. MONITORING FOR HBV TREATMENT RESPONSE AND DISEASE PROGRESSION 10.1. Recommendations 10.2. Background – goals of monitoring 10.3. Rationale for the recommendations (WHO 2015 HBV guidelines ) 10.4. Implementation considerations 11. DETECTION OF VIRAEMIC HCV INFECTION – to guide who to treat 11.1. Recommendations 11.2. Background 11.3. Summary of the evidence 11.4. Rationale for the recommendations 11.5. Implementation considerations 12. ASSESSMENT OF HCV TREATMENT RESPONSE 12.1. Recommendation 12.2. Background 12.3. Summary of the evidence 12.4. Rationale for the recommendations 12.5. Implementation considerations 13. USE OF DRIED BLOOD SPOT SPECIMENS FOR SEROLOGICAL AND VIROLOGICAL TESTING 13.1. Recommendations 13.2. Background 13.3. Summary of the evidence 13.4. Rationale for the recommendations 13.5. Implementation considerations

61 61 62 62 65 67 68 70 70 72 72 73 74 74 74 75 76 77 77 77 78 80 81 83 83 83 83 84 85 86 86 86 87 89 92

viii

14. IMPROVING THE UPTAKE OF TESTING AND LINKAGE TO CARE AND PREVENTION 14.1. Recommendations 14.2. Background 14.3. Summary of the evidence 14.4. Rationale for the recommendations 14.5. Implementation considerations

95 95 95 96 98 100

PART 3: IMPLEMENTATION

101

15. IMPLEMENTING LABORATORY TESTING SERVICES FOR VIRAL HEPATITIS 102 15.1. Key elements for national testing services 102 15.2. National framework for viral hepatitis testing 103 15.3. Building capacity for testing services 104 15.4. Product selection 106 15.5. Assuring the quality of testing services 109 15.6. Assuring the safety of testing services 111 15.7. Other practical considerations for testing 111 16. PRE-TEST AND POST-TEST COUNSELLING 16.1. Promoting testing awareness 16.2. Creating an enabling environment 16.3. The WHO 5 “Cs” 16.4. Providing pre-test information 16.5. Post-test counseling and services 17. SERVICE DELIVERY APPROACHES TO VIRAL HEPATITIS TESTING – examples from the field 17.1. Health-care facility-based testing 17.2. Community-based testing 17.3. Good practices for delivery of effective viral hepatitis testing services 17.4. Diagnostic innovations to promote access to testing 18. TESTING ISSUES IN SPECIFIC POPULATIONS 18.1 Principles for testing in all populations 18.2 Principles for testing in key populations 18.3. Persons living with HIV 18.4. Tuberculosis-infected populations 18.5. Migrant and mobile populations 18.6. Health-care workers 18.7. Couples, partners, family members and household contacts 18.8. Pregnant women 18.9. Children 18.10. Adolescents 114 114 115 115 116 117 119 120 123 127 130 132 132 132 135 135 136 136 137 138 139 141

19. STRATEGIC PLANNING FOR IMPLEMENTING TESTING SERVICES AND APPROACHES 143 Step 1: Review national and subnational epidemiology 146 Step 2: Set testing (and treatment) coverage targets 147 Step 3: Review the effectiveness of and identify gaps in hepatitis testing service delivery 147 Step 4: Assess costs and review cost–effectiveness of different testing approaches 149 Step 5: Adjust programmes and monitor 150 REFERENCES Web annexes All annexes will be available on the WHO hepatitis website. Annex 1: The Global Hepatitis Health Sector Strategy – global targets Annex 2: Guidelines for the prevention, care and treatment of persons with chronic hepatitis B infection – summary of recommendations Annex 3: Guidelines for the screening, care and treatment of persons with chronic hepatitis C infection – summary of recommendations Annex 4: PICO questions and decision-making tables Annex 5: Systematic reviews and evidence summaries Annex 6: Predictive modelling analysis Annex 7: Summary of declared interests Annex 8: Systematic review teams, Guideline Steering Group, Guideline Development Group, External Review Group 152

ACKNOWLEDGEMENTS Many professionals from a range of backgrounds and specialties have contributed to the development of this guidance. WHO is sincerely grateful for their time and support.

Guidelines Development Group The chair of the Guidelines Development Group was Margaret Hellard (Burnet Institute, Melbourne Australia). Roger Chou (Oregon Health and Science University, Portland, USA) was the guidelines methodologist. The following experts served on the Guidelines Development Group: Jacinto Amandua (Ministry of Health, Uganda); Isabelle Andrieux-Meyer (Médecins Sans Frontières, Geneva, Switzerland); Manal Hamdy El-Sayed (Egypt National Hepatitis Committee, Cairo, Egypt); Charles Gore (World Hepatitis Alliance, London, UK); Niklas Luhmann (Médicins du Monde, Paris, France); Michael Ninburg (Hepatitis Education Project, Seattle, USA); Richard Njouom (Centre Pasteur, Cameroon); John Parry (Public Health England, London, UK); Trevor Peter (Clinton Health Access Initiative, New York, USA); Teri Roberts (Foundation for Innovative New Diagnostics, Geneva, Switzerland); Giten Khwairakpam Singh (TREAT Asia/amFAR, Bangkok, Thailand); Lara Tavoschi (European Center for Disease Prevention and Control, Stockholm, Sweden); Richard Tedder – unable to attend (Public Health England, London, UK). WHO regions: Fabian Ndenzako (WHO Regional Office for Africa), Nicole Simone Seguy (WHO Country Office, India), Nick Walsh (WHO Regional Office for the Western Pacific).

Contributors to the systematic reviews We would like to credit the following researchers for conducting the systematic reviews, evidence profiles and GRADE tables: Ali Amini (London School of Hygiene and Tropical Medicine, London, UK); Debra Boeras (London School of Hygiene and Tropical Medicine, London, UK); Wen Chen (London School of Hygiene and Tropical Medicine, London, UK); Jennifer Cohn (Médecins Sans Frontières, Geneva, Switzerland – Team leader); Claudia Denkinger (Foundation for Innovative New Diagnostics, Geneva, Switzerland – Team leader); Jane Falconer (London School of Hygiene and Tropical Medicine, London, UK); Thomas Fitzpatrick (University of Washington, School of Medicine, Seattle, USA – Team leader); Timothy Hallet

(Imperial College, London, UK – Team leader); Helen Kelly (London School of Hygiene and Tropical Medicine, London, UK); Mellanye Lackey (University of Utah, Salt Lake City, USA); Berit Lange (University Hospital, Freiburg, Germany); Ying-Ru Lo (WHO Regional Office for the Western Pacific); Shevanthi Nayagam (Imperial College, London, UK); Rosanna Peeling (London School of Hygiene and Tropical Medicine, London, UK – Team leader); Teri Roberts (Foundation for Innovative New Diagnostics, Geneva, Switzerland); Julia Scott (WHO Regional Office for the Western Pacific); Weiming Tang (London School of Hygiene and Tropical Medicine, London, UK); Edouard Tuaillon (Montpellier Teaching Hospital, France); Joseph Tucker (UNC Project-China, University of North Carolina, USA – Team leader); Philippe Van de Perre (Université Montpellier, Montpellier, France); Olivia Varsaneux (London School of Hygiene and Tropical Medicine, London, UK); Nick Walsh (WHO Regional Office for the Western Pacific); Ji Young Kim (WHO Regional Office for the Western Pacific); Kali Zhou (University of California, Department of Medicine, San Francisco, USA – Team leader).

Contributors to supporting evidence Predictive modelling: Benjamin Linas (Boston University School of Medicine, Boston, USA – Team leader); Jake Morgan (Boston University School of Medicine, Boston, USA); John Parry (Public Health England, London, UK). Values and preferences survey: Elena Ivanova, Teri Roberts, and Alessandra Trianni (Foundation for Innovative New Diagnostics, Geneva, Switzerland). Feasibility survey: Julie Bouscaillou and Niklas Luhmann (Médicins du Monde, Paris, France); Philippa Easterbrook and Azumi Ishizaki (WHO headquarters). Case studies: John Best (University of California, San Francisco, USA); Kathrine Myers (Aaron Diamond AIDS Research Center, USA); and Joseph Tucker (University of North Carolina, and the International Diagnostics Centre, China).

External peer review group The following experts served as external peer reviewers of the draft guidelines: Tanya Applegate (Kirby Institute, Australia); Susan Best (National Serology Reference Laboratory, Australia); Yap Boum (Epicentre, Medecins sans Frontière, France); Alaa Gad Hashish (Al Shams University, Egypt); Joumana Hermez (WHO, Egypt); Cami Graham (Beth Israel Deaconess Medical Center, USA); Stephen Locarnini (Doherty Institute, Australia); Jean-Bosco Ndinokubwayo (WHO Regional Office for Africa); Ponsiano Ocama (Makerere University, Uganda); Jilian Sacks (Clinton Health Access Initiative, USA); Jules Mugabo Semahore (WHO, Rwanda); Mark Sonderup (University of Cape Town, South Africa); Gilles Wandeler (University of Bern, Swizerland).

xii

Hepatitis testing innovation contest: Steering Group: Joseph Tucker (University of North Carolina, SESH, and the International Diagnostics Centre, China), Kathrine Myers (Aaron Diamond AIDS Research Center, USA), John Best (University of California, San Francisco, USA), and Philippa Easterbrook (Global Hepatitis Programme, WHO HQ, Switzerland). Judging panel: Isabelle Andrieux-Meyer (Médecins Sans Frontières, Switzerland); Betty Apica (Makerere University College of Health Sciences, Uganda); Tasnim Azim (International Centre for Diarrhoeal Disease Research, Bangladesh); Carmen Figueroa (WHO, Switzerland); Charles Gore (Hepatitis C Trust and World Hepatitis Alliance, United Kingdom); Azumi Ishizaki (WHO, Switzerland); Kenneth Kabagambe (The National Organisation for People Living with Hepatitis B Uganda); Karyn Kaplan (Treatment Action Group, USA); Medhi Karkouri (Association de Lutte Contre le Sida, Morocco); Giten Khwairakpam Singh (TREAT Asia/amfAR, Thailand); Othman Mellouk (ITPC Global); Veronica Miller (Forum for Collaborative HIV Research, USA); Antons Mozalevskis (WHO Regional Office for Europe, Copenhagen, Denmark); Michael Ninburg (Hepatitis Education Project, USA); Ponsiano Ocama (Makerere University, College of Health Sciences, Uganda); Rosanna Peeling (London School of Hygiene & Tropical Medicine and International Diagnostics Centre, United Kingdom); Razia Pendse (WHO Regional Office for South-East Asia, India); Gabriele Riedner (WHO Regional Office for the Eastern Mediterranean, Egypt); Patricia Velez (Guatemalan Liver Association); Nick Walsh (WHO Regional Office for the Western Pacific).

Steering Committee The following WHO staff formed the Guidelines Steering Committee: Rachel Baggaley (HIV Key Populations and Innovative Prevention); Cheryl Johnson (HIV Key Populations and Innovative Prevention); Anita Sands (Essential Medicines and Health Products), Willy Urassa (Essential Medicines and Health Products); Shaffiq Essajee (Treatment and Care, HIV/AIDS); Marco Vitoria (Treatment and Care, HIV/AIDS); Nicolas Campion Clark (Mental Health and Substance Abuse); Junping Yu (Blood and Transfusion Safety, Service Delivery and Safety). Guidelines writing was led by Philippa Easterbrook (WHO headquarters), Surjo De and Sam Lattimore (Public Health England, University College London Hospital, London, UK); Elizabeth Peach (Melbourne, Australia); Anita Sands (WHO headquarters), and Jilian Sacks (Clinton Health Access Initiative [CHAI]). Drafts were reviewed and input provided by members of the Systematic Review teams, Guidelines Development Group, peer reviewers and WHO Secretariat staff.

xiii

We extend our gratitude to the following individuals of the WHO Secretariat and regional WHO offices for excellent support to the Steering Committee and the Guidelines Development Group: Azumi Ishizaki, Judith van Holten, Sarah Hess, Marc Bulterys, Stefan Wiktor, Andrew Ball, Gottfried Hirnschall, Yvan Hutin, Hande Harmanci, Taner Jonathan Bertuna, Bandana Malhotra, Oyuntungalag Namjilsuren, Han Qin, Nick Walsh, Fabian Ndenzako, Nicole Simone Seguy.

Overall coordination Philippa Easterbrook (Global Hepatitis Programme).

Funding Funding for the development of these guidelines was provided by United States Centers for Disease Control and Prevention (CDC).

xiv

ABBREVIATIONS AND ACRONYMS ALP alkaline phosphatase ALT alanine aminotransferase ANC antenatal clinic APRI aminotransferase/platelet ratio index ART antiretroviral therapy ARV antiretroviral (drug) AST aspartate aminotransferase CDC U.S. Centers for Disease Control and Prevention CG Cockcroft–Gault CHB chronic hepatitis B CI confidence interval CLIA chemiluminescence immunoassay CrCl creatinine clearance DAA direct-acting antiviral (drug) DALY disability-adjusted life year DBS dried blood spot (specimen) ECL electrochemiluminescence immunoassay eGFR estimated glomerular filtration rate EIA enzyme immunoassay ELISA enzyme-linked immunosorbent assay EQAS external quality assessment scheme FBC full blood count FDA U.S. Food and Drug Administration FIB-4 fibrosis-4 score GDP gross domestic product gGT gamma glutamyl transpeptidase GHTF Global Harmonization Task Force GHSS Global Health Sector Strategy GRADE Grading of Recommendations Assessment, Development and Evaluation HBcAg hepatitis B core antigen HBeAg hepatitis B e antigen HBIG hepatitis B immunoglobulin HBsAg hepatitis B surface antigen HBV hepatitis B virus HCC hepatocellular carcinoma HCV hepatitis C virus HCVcAg hepatitis C virus core antigen HDV hepatitis D virus HIC high-income country HIV human immunodeficiency virus ICER incremental cost–effectiveness ratio IFN interferon

IQR interquartile range IVD in vitro diagnostic (medical device) LMICs low- and middle-income countries LoD limit of detection LY life year M&E monitoring and evaluation MSM men who have sex with men MTCT mother-to-child transmission NAT nucleic acid testing NGO nongovernmental organization NIT non-invasive test NPV negative predictive value NRTI nucleos(t)ide reverse transcriptase inhibitor OR odds ratio OST opioid substitution therapy PCR polymerase chain reaction PEG-IFN pegylated interferon PICO population, intervention, comparison, outcomes PITC provider-initiated testing and counselling PMTCT prevention of mother-to-child transmission PPV positive predictive value PQ (WHO) prequalification PWID people who inject drugs QA quality assurance QALY quality-adjusted life year QC quality control QI quality improvement RBV ribavirin RCT randomized controlled trial RDT rapid diagnostic test RNA ribonucleic acid RR relative risk SOP standard operating procedure SSA sub-Saharan Africa STI sexually transmitted infection SVR sustained virological response TDF tenofovir ULN upper limit of normal UNAIDS Joint United Nations Programme on HIV/AIDS UNICEF United Nations Children’s Fund UNODC United Nations Office on Drugs and Crime VCT voluntary counselling and testing WHO World Health Organization WHO GHP WHO Global Hepatitis Programme

xvi

GLOSSARY OF TERMS Markers for HBV infection HB surface antigen (HBsAg) HB core antigen (HBcAg) HB e antigen (HBeAg) HBV envelope protein often produced in excess and detectable in the blood in acute and chronic HBV infection HBV core protein. The core protein is coated with HBsAg and therefore not found free in serum Viral protein found in the high replicative phase of HBV. HBeAg is usually a marker of high levels of replication with wild-type virus but is not essential for viral replication

HB surface Antibody to HBsAg. Develops in response to hepatitis B vaccination and antibody (anti-HBs) during recovery from hepatitis B, denoting past infection and immunity HB core antibody (anti-HBc) anti-HBc IgM HBV e antibody (anti-HBe) HBV DNA Antibody to HBV core (capsid) protein. Anti-HBc antibodies are non neutralizing antibodies and are detected in both acute and chronic infection Subclass of anti-HBc. Detected in recent HBV infection but can be detected by sensitive assays in chronic HBV infection Antibody to HBeAg. Detected in persons with lower levels of HBV replication but also in HBeAg-negative disease (i.e. HBV that does not express HBeAg) HBV viral genomes that can be detected and quantified in serum by nucleic acid testing (NAT)

Markers for HCV infection Anti-HCV antibody Antibody to HCV, which can be detected in the blood usually within two or three months of HCV infection or exposure. The terms HCV antibody and anti-HCV antibody are equivalent, but in these guidelines, HCV antibody is used throughout. HCV RNA HCV core antigen (HCVcAg) HCV viral genomes that can be detected and quantified in serum by nucleic acid testing (NAT). Nucleocapsid peptide 22 [p22] of HCV, which is released into plasma during viral assembly and can be detected from early on and throughout the course of infection

Natural history of viral hepatitis Chronic HBV infection Persistence of HBsAg for at least six months. The persistence of HBsAg in two specimens six months apart is frequently used in clinical practice to confirm chronic hepatitis B infection.

xvii

Chronic HCV infection Viraemic infection

The presence of viraemic HCV RNA or HCVcAg in association with positive serology for HCV antibody. Hepatitis B or C infection associated with presence of virus in the blood (as measured by HBV DNA or HCV RNA), and often referred to as active, ongoing or current infection. HBsAg negative but HBV DNA positive, although at very low levels (invariably <200 IU/mL). Most are also anti-HBc positive. An advanced stage of liver disease characterized by extensive hepatic fibrosis, nodularity of the liver, alteration of liver architecture and disrupted hepatic circulation. Clinical features are portal hypertension (ascites, variceal haemorrhage and hepatic encephalopathy), coagulopathy, or liver insufficiency (jaundice). Other clinical features of advanced liver disease/cirrhosis may include: hepatomegaly, splenomegaly, pruritus, fatigue, arthralgia, palmar erythema, and oedema. Primary cancer of the liver arising from the hepatocytes and may be a complication of chronic hepatitis B or C infection

Occult HBV infection Cirrhosis

Decompensated cirrhosis

Hepatocellular carcinoma (HCC)

Measures of treatment response HCV sustained virological response (SVR) Undetectable HCV RNA in the blood at defined time point after the end of treatment, usually at 12 or 24 weeks (SVR12 or 24)

HCV non-response Detectable HCV RNA in the blood throughout treatment HCV relapse HCV viral breakthrough HBV treatment failure Undetectable HCV RNA during treatment and/or at end of treatment, but subsequent detectable HCV RNA following treatment cessation Undetectable HCV RNA during treatment followed by detectable HCV RNA despite continued treatment May be primary or secondary. Primary antiviral treatment failure may be defined as failure of an antiviral drug to reduce HBV DNA levels by ≥1 x log10 IU/mL within 3 months of initiating therapy. Secondary antiviral treatment failure may be defined as a rebound of HBV DNA levels of ≥1 x log10 IU/mL from the nadir in persons with an initial antiviral treatment effect (≥1 x log10 IU/mL decrease in serum HBV DNA).

Diagnostic testing for hepatitis B and hepatitis C Serological assays Assays that detect the presence of either antigens or antibodies, typically in serum or plasma but also in capillary/venous whole blood and oral fluid. These include rapid diagnostic tests (RDTs), and laboratory-based immunoassays, e.g. enzyme immunoassays (EIAs), chemiluminescence immunoassays (CLIAs), and electrochemiluminescence immunoassays (ECLs).

xviii

Rapid diagnostic test (RDT)

Immunoassays that detect antibodies or antigens and can give a result in less than 30 minutes. Most RDTs can be performed with capillary whole blood collected by finger-stick sampling.

Enzyme Laboratory-based serological immunoassays that detect antibodies, immunoassay (EIA) antigens, or a combination of both Nucleic acid testing (NAT) A molecular technology, for example, polymerase chain reaction (PCR) or nucleic acid sequence-based amplification (NASBA) that can detect very small quantities of viral nucleic acid (RNA or DNA), either qualitatively or quantitatively. Refers to testing using one specimen in the same test device (or reagent cartridge) that can detect other infections (e.g. HIV, syphilis, hepatitis C, hepatitis B)

Multiplex or multidisease testing

Measures of test performance Clinical/diagnostic sensitivity of a test The ability of a test to correctly identify those with the infection or disease (i.e. true positives/true positives + false negatives)

Clinical/diagnostic The ability of a test to correctly identify those without the infection or specificity of a test disease (i.e. true negatives/true negatives + false positives) Sensitivity and specificity are usually expressed as point estimates accompanied by confidence intervals. Positive predictive value (PPV) Negative predictive value (NPV) Analytical sensitivity/Limit of detection (LoD) Testing terminology Testing algorithm Testing approach The combination and sequence of specific assays used within hepatitis B and C testing strategies In the context of these guidelines, the testing approach describes both “who to test” i.e. different populations and “where to test” i.e different settings. Testing approaches include general population testing, focused testing of high-risk groups, “birth-cohort” testing or of antenatal clinics. These can be delivered through either health-facility or community-based testing. The probability that when a person’s test result is positive, they truly have the infection/disease The probability that when a person’s test result is negative, they truly do not have the infection/disease Predictive values are influenced by the prevalence of the disease in the population. The lowest concentration of measurement that can be consistently detected in 95% of specimens tested under routine laboratory conditions. It defines the analytical sensitivity in contrast to the clinical or diagnostic sensitivity.

xix

Testing strategy

A general sequence of assays for a specific testing objective or approach, taking into consideration the presumed disease prevalence in the population being tested. A one-assay serological testing strategy involves a single serological assay. A two-assay serological testing strategy involves two different serological assays used sequentially.

Testing approaches terminology Key populations Groups of people who due to specific high-risk behaviours, are at increased risk for HIV infection irrespective of the epidemic type or local context. This may also apply to HBV and/or HCV infection. Key populations often have legal and social issues related to their behaviours that increase their vulnerability to HIV, HBV and HCV infection. These guidelines refer to the following groups as key populations: men who have sex with men (MSM); people who inject drugs (PWID); people in prisons and other closed settings; sex workers; and transgender people. Groups of people who are particularly vulnerable to HBV/HCV infection in certain situations or contexts. These guidelines refer to the following groups as vulnerable populations: migrant and mobile workers, and indigenous populations. This approach refers to routine testing throughout the entire population without attempting to identify high-risk behaviours or characteristics. It means that all members of the population should have potential access to the testing programme. This approach means routine testing among easily identified age or demographic groups (i.e. specific “birth cohorts”) known to have a high HCV prevalence due to past generalized exposures that have since been identified and removed. This approach means routine testing of pregnant women especially in settings where there is an intermediate or high seroprevalence, to identify women in need of antiviral treatment for their own health and additional interventions to reduce mother-to-child transmission (MTCT) Includes using outreach (mobile) approaches in general and key populations; home-based testing (or door-to-door outreach); testing in workplaces, places of worship, parks, bars and other venues; in schools and other educational establishments; as well as through campaigns Includes testing in primary care clinics, inpatient wards and outpatient clinics, including specialist dedicated clinics such as HIV, STI and TB clinics, in district, provincial or regional hospitals and their laboratories, and in private clinical services.

Vulnerable populations

General population testing

“Birth cohort” testing

Antenatal clinic testing

Community-based testing

Facility-based testing

Service delivery terminology Integration The co-location and sharing of services and resources across different disease areas. In the context of hepatitis B or C infection, this may include the provision of testing, prevention, care and treatment services alongside other health services, such as HIV, tuberculosis (TB), sexually transmitted infections (STI), antenatal clinic (ANC), contraceptive and other family planning services. The process of delegating significant authority and resources to lower levels of the health system (provincial, regional, district, sub-district, primary health care and community The rational redistribution of tasks from “higher-level” cadres of healthcare providers to other cadres, such as trained lay providers Any person who performs functions related to health-care delivery and has been trained to deliver services but has received no formal professional or paraprofessional certificate or tertiary education degree A process of actions and activities that support people testing for HBV/HCV to engage with prevention, treatment and care services as appropriate for their hepatitis B and C status.

Decentralization

Task-shifting/ sharing Lay provider

Linkage to care

xxi

EXECUTIVE SUMMARY Background Hepatitis B virus (HBV) and hepatitis C virus (HCV) infection are major causes of acute and chronic liver disease (e.g. cirrhosis and hepatocellular carcinoma) globally, and cause an estimated 1.4 million deaths annually. It is estimated that, at present, 248 million people are living with chronic HBV infection, and that 110 million persons are HCV-antibody positive, of which 80 million have active viraemic infection. The burden of chronic HBV and HCV remains disproportionately high in low- and middle-income countries (LMICs), particularly in Asia and Africa. Additionally, even in low-prevalence areas, certain populations have high levels of HCV and HBV infection, such as persons who inject drugs (PWID), men who have sex with men (MSM), people with HIV, as well as those belonging to certain indigenous communities. The development of highly effective, well-tolerated oral direct acting antiviral (DAA) treatment regimens with high rates of cure after 8–12 weeks of treatment has revolutionized the treatment of chronic HCV infection, although the high prices of these new medicines remain a major barrier to access in many countries. Effective long-term antiviral treatment with tenofovir or entecavir is also available for people with chronic HBV infection. However, despite the high global burden of disease due to chronic HBV and HCV infection, and the advances and opportunities for treatment, most people infected with HBV and/ or HCV remain unaware of their infection and therefore frequently present with advanced disease and may transmit infection to others. There are several key reasons for this low rate of hepatitis testing. These include the limited facilities or services for hepatitis testing, lack of effective testing policies or national guidelines, complex diagnostic algorithms, and poor laboratory capacity and quality assurance systems. Testing and diagnosis of hepatitis B and C infection is the gateway for access to both prevention and treatment services, and is a crucial component of an effective response to the hepatitis epidemic. Early identification of persons with chronic HBV or HCV infection enables them to receive the necessary care and treatment to prevent or delay progression of liver disease. Testing also provides an opportunity to link people to interventions to reduce transmission, through counselling on risk behaviours and provision of prevention commodities (such as sterile needles and syringes) and hepatitis B vaccination.

xxii

About the guidelines These are the first WHO guidelines on testing for chronic HBV and HCV infection and complement published guidance by WHO on the prevention, care and treatment of chronic hepatitis C and hepatitis B infection1,2. These guidelines outline the public health approach to strengthening and expanding current testing practices for HBV and HCV, and are intended for use across age groups and populations. The primary audience for these WHO guidelines are country programme managers and health-care providers, particularly in LMICs, responsible for planning and implementing hepatitis testing, prevention, care and treatment services. The document is organized into three distinct sections: Introduction – Part 1: Introductory chapters on epidemiology, natural history and in vitro diagnostic assays for hepatitis B and C virus infection. Recommendations – Part 2: Nine chapters with summary of recommendations, evidence and rationale for recommendations covering: • • • • • • who to test for chronic hepatitis B and C infection (testing approaches) how to test serologically for chronic hepatitis B and C infection (testing strategies) how to confirm viraemic HBV and HCV infection to guide treatment decisions how to assess response to antiviral treatment for chronic hepatitis B and C infection use of dried blood spot (DBS) specimens for serology testing and virological testing for chronic hepatitis B and C infection interventions to promote uptake of testing and linkage to care.

Implementation – Part 3: Guidance to support implementation of these recommendations at country level which include a framework for country decision-making and planning in two key areas: how to organize hepatitis testing laboratory services (systems for selection and evaluation of assays and quality assurance systems) and how to plan the best strategic mix of testing approaches. There is also guidance on different service delivery models for testing; pre and post-test counselling; and tailored testing approaches in specific populations (e.g. PWID, prisoners, pregnant women, children and adolescents).

Guidelines for the screening, care and treatment of persons with chronic hepatitis C infection. Updated version, April 2016. Geneva: World Health Organization; 2016. Guidelines for the prevention, care and treatment of persons with chronic hepatitis B infection. Geneva: World Health Organization; 2015.

xxiii

FIG. 1. Organization of the guidelines along the continuum of care Part 1: Background chapters • • • Introduction: Guiding principles, guideline methodology Background: Epidemiology and natural history Background: In vitro diagnostics for HBV and HCV infection Promoting uptake of testing and linkage to care Dried blood spots for HBV serological and virological testing Dried blood spots for HCV serological and virological testing Other interventions to promote uptake of testing and linkage to care

Part 2: Recommendation chapters along with continuum of care for testing Who to test? How to test? How to confirm current/viraemic infection? Confirmation of viraemic HBV infection (HBV DNA) Confirmation of viraemic HCV infection (HCV RNA or cAg) Monitoring treatment response Monitoring of HBV treatment response and disease progression Assessment of HCV treatment response (test of cure)

HBV

Who to test

How to test (HBsAg)

HCV

Who to test

How to test (anti-HCV antibody)

Pre- and post-test counselling

Part 3: Implementation chapters • • Laboratory (How to test): How to organize laboratory testing services for viral hepatitis Service delivery (Who and where to test): Pre- and post-test counselling Sevice delivery approaches for viral hepatitis Testing issues in specific populations Strategic planning for testing services and approaches

Summary of recommendations Table 1 summarizes the recommendations on who to test (i.e. testing approaches); how to test (i.e. testing strategies), and interventions to promote uptake of testing and linkage to care. Figures 2 and 3 show summary algorithms for diagnosis, monitoring and management of chronic hepatitis B and C infection. Who to test for HBV and HCV infection – testing approaches The guidelines recommend offering focused testing to individuals from populations most affected by HBV or HCV infection (i.e. who are either part of a population with higher seroprevalence or who have a history of exposure to or high-risk behaviours for HBV or HCV infection). In settings with a ≥2% or ≥5% seroprevalence of hepatitis B surface antigen (HBsAg) or HCV antibody (anti-HCV) (based on existing published thresholds for intermediate or high seroprevalence, respectively), it is recommended that all adults have routine access to and be offered testing (i.e. a general population testing approach), or use “birth cohort” testing for specific age groups with higher anti-HCV seroprevalence. However, the threshold used by a country will depend on other country considerations and epidemiological context. Overall, these different testing approaches should make use of existing facility-based (such as antenatal clinics, HIV or TB services) or

xxiv

community- based testing opportunities and programmes. How to test for HBV and HCV infection – serological assays and testing strategies Overall, the guidelines recommend the use of a single quality-assured serological in vitro diagnostic test (i.e. either a laboratory-based immunoassay [enzyme immunoassay or chemiluminiscence immunoassay] or rapid diagnostic test [RDT]) to detect HBsAg and HCV antibody. RDTs used should meet minimum performance standards, and be delivered at the point of care to improve access and linkage to care and treatment. Confirming viraemic infection and monitoring for treatment response Following a reactive HCV antibody serological test result, a quantitative or qualitative RNA NAT is recommended as the preferred testing strategy to diagnose viraemic infection. Detection of core HCV antigen, where the assay has comparable clinical sensitivity to NAT technologies, may be considered as an alternative. The use of HBV DNA NAT following a reactive HBsAg serological test result, is recommended to help further guide who to treat or not treat if there is no evidence of cirrhosis, and to monitor for treatment response, based on existing recommendations from the 2015 WHO HBV management guidelines. Use of dried blood spot sampling and other strategies to promote testing uptake and linkage to care The use of capillary whole blood DBS specimens for both serological and NAT technologies for HBV and HCV infection may be considered to facilitate access to testing in certain settings where there are either no facilities or expertise to take venous blood specimens, in persons with poor venous access, or where qualityassured RDTs are not available or their use is not feasible. Programmes should consider only the use of assays that have been validated by their manufacturer for use with DBS specimens. Other recommended interventions to promote uptake of hepatitis testing and linkage to care include peer and lay health worker support in community- based settings, clinician reminders in facilities, and testing as part of integrated services within drug treatment and communitybased harm reduction services. The development of these guidelines was conducted in accordance with procedures established by the WHO Guidelines Review Committee. Clinical recommendations were formulated by a regionally representative and multidisciplinary Guidelines Development Group at a meeting held in September 2015. The GRADE (Grading of Recommendations Assessment, Development and Evaluation) approach was used to formulate and categorize strength of recommendations (strong or conditional), and was adapted for diagnostic tests. This includes an assessment of the quality of evidence (high, moderate, low or

xxv

very low), consideration of overall balance of benefits and harms (at individual and population levels), patient/health worker values and preferences, resource use, cost–effectiveness and consideration of feasibility and effectiveness across a variety of resource-limited settings, including where access to laboratory infrastructure and specialized tests is limited. There was a very limited evidence base to guide recommendations on testing approaches (i.e. who to test and service delivery approaches) and an absence of evidence on patient-important outcomes in evaluation of performance of diagnostic tests and testing strategies. The process also identified key gaps in knowledge that will guide the future research agenda. Most of the evidence was based on published studies in adults from Asia, North America and Western Europe; there is a lack of data from subSaharan Africa, and in children. Implementation of these recommendations pose practical challenges to policymakers and implementers in LMICs, particularly in sub-Saharan Africa, where there is currently very limited access to diagnostic tests, antiviral therapies and appropriate laboratory infrastructure. These guidelines also provide the framework for country decision-making and planning for hepatitis laboratory testing programmes to ensure the quality and accuracy of hepatitis testing, as well as approaches to delivery of testing services, including opportunities to integrate hepatitis testing with existing services, where appropriate. These guidelines and recommendations provide a major opportunity to improve identification and treatment of persons with chronic hepatitis B and C, and achieve the Global Hepatitis Health Sector Strategy (GHSS) on Viral Hepatitis 3 targets, including those on testing (i.e. identify 30% of persons living with HBV and HCV by 2020 and 90% by 2030). This in turn will improve clinical outcomes and save lives, as well as facilitate prevention, reducing hepatitis transmission and new infections.

WHO Global health sector strategy on viral hepatitis 2016–2021. Geneva: World Health Organization; 2016.

xxvi

SUMMARY ALGORITHMS FIG.2. Summary algorithm for diagnosis, treatment and monitoring1 of chronic HBV infection HEPATITIS B SURFACE ANTIGEN (HBsAg) Single RDT2 or laboratory-based immunoassay3

SEROLOGICAL TESTING

HBsAg + (reactive) Report positive

HBsAg – (non-reactive) Report negative

Compatible with HBV infection

No serological evidence of HBV infection

(using clinical criteria4 and/or non-invasive tests (NITs) for presence of cirrhosis, i.e. APRI score5 >2 or based on TE6)

ASSESSMENT OF STAGE OF LIVER DISEASE

HBV DNA NUCLEIC ACID TEST (NAT) (quantitative) ASSESSMENT FOR TREATMENT

(to further guide who to treat and not treat, if no evidence of cirrhosis) Yes PRESENCE OF CIRRHOSIS

ALL AGES >30 years (in particular) ALT7,8 Persistently abnormal HBV DNA >20 000 IU/mL HBV DNA 2000– 20 000 IU/mL ALT7 Intermittently abnormal HBV DNA <2000 IU/mL ALT7 Persistently normal

AGE ≤30 years ALT7 Persistently normal HBV DNA <2000 IU/mL

INITIATE ANTIVIRAL THERAPY9 AND MONITOR • Tenofovir or entecavir • Entecavir in children aged 2–11 years

DEFER TREATMENT AND MONITOR

DETECTION OF HCC in persons with cirrhosis or HCC family history (every 6 months) MONITORING • Ultrasound and serum AFP

TREATMENT RESPONSE AND/OR DISEASE PROGRESSION • Adherence at each visit, if on treatment • ALT, HBV DNA and HBeAg • Staging of liver disease (clinical criteria and NITs (e.g. APRI in adults or TE)

(every 12 months)

TOXICITY MONITORING in persons on treatment • Renal function and risk factors for renal dysfunction

(baseline and every 12 months)

Abbreviations: RDT: rapid diagnostic test; ALT: alanine aminotransferase; APRI: aspartase aminotransferase-to-platelet ratio index; TE: transient elastography; HCC: hepatocellular carcinoma; AFP: apha fetoprotein 1 Guidelines for the prevention, care and treatment of persons with chronic hepatitis B infection. Geneva: World Health Organization; 2015. 2. In settings or populations with a low HBsAg seroprevalence <0.4%, confirmation of HBsAg positivity on the same immunoassay with a neutralization step or a second different RDT assay for detection of HBsAg may be considered. 3 Laboratory-based Immunoassays include enzyme immunoassay (EIA), chemoluminescence immunoassay (CLIA), and electrochemoluminescence assay (ECL) 4 Decompensated cirrhosis is defined by the development of portal hypertension (ascites, variceal haemorrhage and hepatic encephalopathy), coagulopathy, or liver insufficiency (jaundice). Other clinical features of advanced liver disease/cirrhosis may include: hepatomegaly, splenomegaly, pruritus, fatigue, arthralgia, palmar erythema, and oedema. 5 Aspartate aminotransferase (AST)-to-platelet ratio index (APRI) is a simple index for estimating hepatic fibrosis based on a formula derived from AST and platelet concentrations. The formula for calculating the APRI score is: APRI = (AST/AST ULN) x 100) /platelet count (109/L). Most recommend using 40 IU/L as the value for AST upper limit of normal (ULN). An online calculator can be found at: http://www.hepatitisc.uw.edu/page/clinical-calculators/apri 6 Transient elastography (Fibroscan): a technique to measure liver stiffness (as a surrogate for fibrosis) 7 ALT levels fluctuate in persons with chronic hepatitis B and require longitudinal monitoring to determine the trend. Upper limits for normal ALT have been defined as below 30 U/L for men and 19 U/L for women, though local laboratory normal ranges should be applied. Persistently normal/abnormal may be defined as three ALT determinations below or above the upper limit of normal, made at unspecified intervals during a 6–12-month period or predefined intervals during a 12-month period. 8 Where HBV DNA testing is not available, treatment may be considered based on persistently abnormal ALT levels, but other common causes of persistently raised ALT levels such as impaired glucose tolerance, dyslipidaemia and fatty liver should be excluded. 9 Initiate antiviral therapy with tenofovir alone only after exclusion of HIV coinfection.

xxvii

FIG.3. Summary algorithm for diagnosis, treatment and monitoring1 of chronic HCV infection ANTI-HCV ANTIBODY Single RDT or laboratory-based immunoassay2

SEROLOGICAL TESTING

Anti-HCV + (reactive) Report positive Compatible with current or past HCV infection

Anti-HCV – (non-reactive) Report negative No serological evidence of HCV infection

CONFIRMATION OF VIRAEMIC INFECTION

HCV RNA NUCLEIC ACID TEST (NAT) (qualitative or quantitative) or HCV core antigen (cAg)

HCV RNA test or cAg + Report detected (with viral load if available) Compatible with viraemic HCV infection

HCV RNA test or cAg Report not detected No current viraemic HCV

ASSESSMENT OF STAGE OF LIVER DISEASE (using clinical criteria3 and non-invasive tests (NITs), i.e. APRI score4 >2 or based on TE5)

TREATMENT ASSESSMENT

OTHER CONSIDERATIONS FOR TREATMENT (e.g. comorbidities,

HCV genotyping, pregnancy and potential drug-drug interactions)

FACTORS TO BE CONSIDERED IN PRIORITIZING TREATMENT 1. Increased risk of death (e.g. advanced fibrosis and cirrhosis, post-liver transplantation) 2. Risk of accelerated fibrosis (e.g. HIV or HBV coinfection, metabolic syndrome, high level of alcohol use) 3. Extrahepatic manifestations and evidence of end-organ damage (e.g. debilitating fatigue, vasculitis and lymphoproliferative disorders) 4. Significant psychosocial morbidity (e.g. due to stigma, discrimination, fear of transmission to others) 5. Maximizing reduction in incidence (e.g. in PWID, MSM, prisoners, sex workers, women of childbearing age, health-care workers)

SELECT DIRECT-ACTING ANTIVIRAL (DAA) REGIMEN1,6 Daclatasvir/sofosbuvir or ledipasvir/sofosbuvir ± ribavirin for 12 or 24 weeks (depending on genotype and presence of cirrhosis)

MONITORING

ASSESSMENT OF CURE (sustained virological response (SVR) at 12 weeks (i.e. SVR12) after the end of treatment) HCV RNA NAT (qualitative or quantitative) DETECTION OF HCC Ultrasound and AFP in persons with cirrhosis (every 6 months)

Abbreviations: RDT: rapid diagnostic test; APRI: aspartase aminotransferase-to-platelet ratio index, TE: transient elastography; PWID: people who inject drugs; MSM: men who have sex with men; HCC: hepatocellular carcinoma; AFP: alpha fetoprotein 1 Guidelines for the screening, care and treatment of persons with chronic hepatitis C infection. Updated version, April 2016. Geneva: World Health Organization; 2016. 2 Laboratory-based immunoassays include enzyme immunoassay (EIA), chemoluminescence immunoassay (CLIA), and electrochemoluminescence assay (ECL). 3 Decompensated cirrhosis is defined by the development of portal hypertension (ascites, variceal haemorrhage and hepatic encephalopathy), coagulopathy, or liver insufficiency (jaundice). Other clinical features of advanced liver disease/cirrhosis may include: hepatomegaly, splenomegaly, pruritus, fatigue, arthralgia, palmar erythema, and oedema. 4 Aspartate aminotransferase (AST)-to-platelet ratio index (APRI) is a simple index for estimating hepatic fibrosis based on a formula derived from AST and platelet concentrations. The formula for calculating the APRI score is: APRI = (AST/AST ULN) x 100) /platelet count (109/L). Most recommend using 40 IU/L as the value for AST upper limit of normal (ULN). An online calculator can be found at: http://www.hepatitisc.uw.edu/page/clinical-calculators/apri 5 Transient elastography (Fibroscan) is a technique to measure liver stiffness (as a surrogate for fibrosis). 6 Caution: there is a potential but uncertain risk of HBV reactivation during or after HCV clearance. Prior to starting DAA therapy, test for HBV infection (HBsAg, HBeAg, and HBV DNA) to assess indication for HBV treatment. Continue careful monitoring after completion of DAA therapy, including for HCC.

xxviii

TABLE 1. SUMMARY OF RECOMMENDATIONS ON TESTING FOR CHRONIC HEPATITIS B AND C VIRUS INFECTION WHO TO TEST FOR CHRONIC HBV INFECTION Testing approach and population General population testing Recommendations* 1. In settings with a ≥2% or ≥5%1 HBsAg seroprevalence in the general population, it is recommended that all adults have routine access to and be offered HBsAg serological testing with linkage to prevention, care and treatment services. General population testing approaches should make use of existing communityor health facility-based testing opportunities or programmes such as at antenatal clinics, HIV or TB clinics. Conditional recommendation, low quality of evidence 2. In settings with a ≥2% or ≥5%%1 HBsAg seroprevalence in the general population, it is recommended that HBsAg serological testing be routinely offered to all pregnant women in antenatal clinics2, with linkage to prevention, care and treatment services. Couples and partners in antenatal care settings should be offered HBV testing services. Strong recommendation, low quality of evidence 3. In all settings (and regardless of whether delivered through facility- or communitybased testing), it is recommended that HBsAg serological testing and linkage to care and treatment services be offered to the following individuals: • Adults and adolescents from populations most affected by HBV infection3 (i.e. who are either part of a population with high HBV seroprevalence or who have a history of exposure and/or high-risk behaviours for HBV infection); • Adults, adolescents and children with a clinical suspicion of chronic viral hepatitis4 (i.e. symptoms, signs, laboratory markers); • Sexual partners, children and other family members, and close household contacts of those with HBV infection5; • Health-care workers: in all settings, it is recommended that HBsAg serological testing be offered and hepatitis B vaccination given to all health-care workers who have not been vaccinated previously (adapted from existing guidance on hepatitis B vaccination6) Strong recommendation, low quality of evidence 4. In all settings, screening of blood donors should be mandatory with linkage to care, counselling and treatment for those who test positive.

Routine testing in pregnant women

Focused testing in most affected populations

Blood donors Adapted from existing 2010 WHO guidance (Screening donated blood for transfusion transmissible infections7)

Abbreviations: HBsAg: hepatitis B surface antigen; PWID: people who inject drugs; MSM: men who have sex with men *The GRADE system (Grading of Recommendations, Assessment, Development and Evaluation) was used to categorize the strength of recommendations as strong or conditional (based on consideration of the quality of evidence, balance of benefits and harms, acceptability, resource use and programmatic feasibility) and the quality of evidence as high, moderate, low or very low. A threshold of ≥2% or ≥5% seroprevalence was based on several published thresholds of intermediate or high seroprevalence. The threshold used will depend on other country considerations and epidemiological context. Many countries have chosen to adopt routine testing in all pregnant women, regardless of seroprevalence in the general population, and particularly where seroprevalence ≥2%. A full vaccination schedule including birth dose should be completed in all infants, in accordance with the WHO position paper on hepatitis B vaccines 2009.6 3 Includes those who are either part of a population with higher seroprevalence (e.g. some mobile/migrant populations from high/intermediate endemic countries, and certain indigenous populations) or who have a history of exposure or high-risk behaviours for HBV infection (e.g. PWID, people in prisons and other closed settings, MSM and sex workers, HIV-infected persons, partners, family members and children of HBV-infected persons). 4 Features that may indicate underlying chronic HBV infection include clinical evidence of existing liver disease, such as cirrhosis or hepatocellular carcinoma (HCC), or where there is unexplained liver disease, including abnormal liver function tests or liver ultrasound. 5 In all settings, it is recommended that HBsAg serological testing with hepatitis B vaccination of those who are HBsAg negative and not previously vaccinated be offered to all children with parents or siblings diagnosed with HBV infection or with clinical suspicion of hepatitis, through community- or facility-based testing. 6 WHO position paper. Hepatitis B vaccines. Weekly Epidemiological Record. 2009;4 (84):405–20. 7 Screening donated blood for transfusion transmissible infections. Geneva: World Health Organization; 2010. 1 2

xxix

WHO TO TEST FOR CHRONIC HCV INFECTION Testing approach and population Focused testing in most affected populations Recommendations* 1. In all settings (and regardless of whether delivered through facility- or communitybased testing), it is recommended that serological testing for HCV antibody (antiHCV)1 be offered with linkage to prevention, care and treatment services to the following individuals: • Adults and adolescents from populations most affected by HCV infection2 (i.e. who are either part of a population with high HCV seroprevalence or who have a history of exposure and/or high-risk behaviours for HCV infection); • Adults, adolescents and children with a clinical suspicion of chronic viral hepatitis3 (i.e. symptoms, signs, laboratory markers). Strong recommendation, low quality of evidence Note: Periodic re-testing using HCV NAT should be considered for those with ongoing risk of acquisition or reinfection. 2. In settings with a ≥2% or ≥5%4 HCV antibody seroprevalence in the general population, it is recommended that all adults have access to and be offered HCV serological testing with linkage to prevention, care and treatment services. General population testing approaches should make use of existing community- or facility-based testing opportunities or programmes such as HIV or TB clinics, drug treatment services and antenatal clinics5. Conditional recommendation, low quality of evidence 3. This approach may be applied to specific identified birth cohorts of older persons at higher risk of infection6 and morbidity within populations that have an overall lower general prevalence. Conditional recommendation, low quality of evidence

General population testing

Birth cohort testing

Abbreviations: NAT: nucleic acid test; anti-HCV: HCV antibody; PWID: people who inject drugs; MSM: men who have sex with men *The GRADE system (Grading of Recommendations, Assessment, Development and Evaluation) was used to categorize the strength of recommendations as strong or conditional (based on consideration of the quality of evidence, balance of benefits and harms, acceptability, resource use and programmatic feasibility) and the quality of evidence as high, moderate, low or very low. This may include fourth-generation combined antibody/antigen assays Includes those who are either part of a population with higher seroprevalence (e.g. some mobile/migrant populations from high/intermediate endemic countries, and certain indigenous populations) or who have a history of exposure or high-risk behaviours for HCV infection (e.g. PWID, people in prisons and other closed settings, MSM and sex workers, and HIV-infected persons, children of mothers with chronic HCV infection especially if HIV-coinfected). 3 Features that may indicate underlying chronic HCV infection include clinical evidence of existing liver disease, such as cirrhosis or hepatocellular carcinoma (HCC), or where there is unexplained liver disease, including abnormal liver function tests or liver ultrasound. 4 A threshold of ≥2% or ≥5% seroprevalence was based on several published thresholds of intermediate and high seroprevalence. The threshold used will depend on other country considerations and epidemiological context. 5 Routine testing of pregnant women for HCV infection is currently not recommended. 6 Because of historical exposure to unscreened or inadequately screened blood products and/or poor injection safety. 1 2

xxx

HOW TO TEST FOR CHRONIC HBV INFECTION AND MONITOR TREATMENT RESPONSE Topic Which serological assays to use Recommendations* • For the diagnosis of chronic HBV infection in adults, adolescents and children (>12 months of age1), a serological assay (in either RDT or laboratory-based immunoassay format2) that meets minimum quality, safety and performance standards3 (with regard to both analytical and clinical sensitivity and specificity) is recommended to detect hepatitis B surface antigen (HBsAg). - In settings where existing laboratory testing is already available and accessible, laboratory-based immunoassays are recommended as the preferred assay format. - In settings where there is limited access to laboratory testing and/or in populations where access to rapid testing would facilitate linkage to care and treatment, use of RDTs is recommended to improve access. Strong recommendation, low/moderate quality of evidence In settings or populations with an HBsAg seroprevalence of ≥0.4%4, a single serological assay for detection of HBsAg is recommended, prior to further evaluation for HBV DNA and staging of liver disease. In settings or populations with a low HBsAg seroprevalence of <0.4%4, confirmation of HBsAg positivity on the same immunoassay with a neutralization step or a second different RDT assay for detection of HBsAg may be considered5. Conditional recommendation, low quality of evidence Directly following a positive HBsAg serological test, the use of quantitative or qualitative nucleic acid testing (NAT) for detection of HBV DNA is recommended as the preferred strategy and to guide who to treat or not treat. Strong recommendation, moderate/low quality of evidence

Serological testing strategies

•

Detection of HBV DNA – assessment for treatment Adapted from existing guidance (WHO HBV 2015 guidelines6) Monitoring for HBV treatment response and disease progression Existing guidance (WHO HBV 2015 guidelines6)

•

It is recommended that the following be monitored at least annually: - ALT levels (and AST for APRI), HBsAg7, HBeAg8, and HBV DNA levels (where HBV DNA testing is available) - Non-invasive tests (APRI score or transient elastography) to assess for presence of cirrhosis in those without cirrhosis at baseline; - If on treatment, adherence should be monitored regularly and at each visit. Strong recommendation, moderate quality of evidence

More frequent monitoring is recommended: • In persons on treatment or following treatment discontinuation: more frequent ontreatment monitoring (at least every 3 months for the first year) is indicated in: persons with more advanced disease (compensated or decompensated cirrhosis9); during the first year of treatment to assess treatment response and adherence; where treatment adherence is a concern; in HIV-coinfected persons; and in persons after discontinuation of treatment. Conditional recommendation, very low quality of evidence • In persons who do not yet meet the criteria for antiviral therapy: i.e. persons who have intermittently abnormal ALT levels or HBV DNA levels that fluctuate between 2000 IU/mL and 20 000 IU/mL (where HBV DNA testing is available) and in HIVcoinfected persons7. Conditional recommendation, low quality of evidence Abbreviations: ALT: alanine aminotransferase; AST: aspartate aminotransferase; APRI: aspartate-to-platelet ratio index; HBeAg: HBV e antigen; HBsAg: HBV surface antigen; NAT: nucleic acid test; RDT: rapid diagnostic test 1 A full vaccination schedule including birth dose should be completed in all infants in accordance with the WHO position paper on Hepatitis B vaccines, 2009. Testing of exposed infants is problematic within the first six months of life as HBsAg and hepatitis B DNA may be inconsistently detectable in infected infants. Exposed infants should be tested for HBsAg between 6 and 12 months of age to screen for evidence of hepatitis B infection. In all age groups, acute HBV infection can be confirmed by the presence of HBsAg and IgM anti-HBc. CHB is diagnosed if there is persistence of HBsAg for six months or more. 2 Laboratory-based immunoassays include enzyme immunoassay (EIA), chemoluminescence immunoassay (CLIA), and electrochemoluminescence assay (ECL). 3 Assays should meet minimum acceptance criteria of either WHO prequalification of in vitro diagnostics (IVDs) or a stringent regulatory review for IVDs. All IVDs should be used in accordance with manufacturers’ instructions for use and where possible at testing sites enrolled in a national or international external quality assessment scheme. 4 Based on results of predictive modelling of positive predictive values according to different thresholds of seroprevalence in populations to be tested, and assay diagnostic performance. 5 A repeat HBsAg assay after 6 months is also a common approach used to confirm chronicity of HBV infection. 6 For further details, see Chapter 5: Who to treat and who not to treat. Guidelines for the prevention, care and treatment of persons with chronic hepatitis B infection: World Health Organization; 2015. 7 In persons on treatment, monitor for HBsAg loss (although this occurs rarely), and for seroreversion to HBsAg positivity after discontinuation of treatment. 8 Monitoring of HBeAg/anti-HBe mainly applies to those who are initially HBeAg positive. However, those who have already achieved HBeAg seroconversion and are HBeAg negative and anti-HBe positive may serorevert. 9 Decompensated cirrhosis is defined by the development of portal hypertension (ascites, variceal haemorrhage and hepatic encephalopathy), coagulopathy, or liver insufficiency (jaundice). Other clinical features of advanced liver disease/cirrhosis may include: hepatomegaly, splenomegaly, pruritus, fatigue, arthralgia, palmar erythema and oedema.

xxxi

HOW TO TEST FOR CHRONIC HCV INFECTION AND MONITOR TREATMENT RESPONSE Topic Which serological assays to use Recommendations* • To test for serological evidence of past or present infection in adults, adolescents and children (>18 months of age1), an HCV serological assay (antibody or antibody/antigen) using either RDT or laboratory-based immunoassay formats2 that meet minimum safety, quality and performance standards3 (with regard to both analytical and clinical sensitivity and specificity) is recommended. - In settings where there is limited access to laboratory infrastructure and testing, and/or in populations where access to rapid testing would facilitate linkage to care and treatment, RDTs are recommended. Strong recommendation, low/moderate quality of evidence

Serological testing strategies

In adults and children older than 18 months1, a single serological assay for initial detection of serological evidence of past or present infection is recommended prior to supplementary nucleic acid testing (NAT) for evidence of viraemic infection. Conditional recommendation, low quality of evidence • Directly following a reactive HCV antibody serological test result, the use of quantitative or qualitative NAT for detection of HCV RNA is recommended as the preferred strategy to diagnose viraemic infection. Strong recommendation, moderate/low quality of evidence An assay to detect HCV core (p22) antigen, which has comparable clinical sensitivity to NAT, is an alternative to NAT to diagnose viraemic infection 4. Conditional recommendation, moderate quality of evidence Nucleic acid testing for qualitative or quantitative detection of HCV RNA should be used as test of cure at 12 or 24 weeks (i.e. sustained virological response (SVR12 or SVR24)) after completion of antiviral treatment. Conditional recommendation, moderate/low quality of evidence

Detection of viraemic infection

•

Assessment of HCV treatment response

•

Abbreviations: DBS: dried blood spot; IVD: in vitro diagnostics; NAT: nucleic acid test; RDT: rapid diagnostic test *The GRADE system (Grading of Recommendations, Assessment, Development and Evaluation) was used to categorize the strength of recommendations as strong or conditional (based on consideration of the quality of evidence, balance of benefits and harms, acceptability, resource use and programmatic feasibility) and the quality of evidence as high, moderate, low or very low. HCV infection can be confirmed in children under 18 months only by virological assays to detect HCV RNA, because transplacental maternal antibodies remain in the child’s bloodstream up until 18 months of age, making test results from serology assays ambiguous. 2 Laboratory-based immunoassays include enzyme immunoassay (EIA), chemoluminescence immunoassay (CLIA), and electrochemoluminescence assay (ECL). 3 Assays should meet minimum acceptance criteria of either WHO prequalification of IVDs or a stringent regulatory review for IVDs. All IVDs should be used in accordance with manufacturers’ instructions, and where possible at testing sites enrolled in a national or international external quality assessment scheme. 4 A lower level of analytical sensitivity can be considered, if an assay is able to improve access (i.e. an assay that can be used at the point of care or suitable for dried blood spot [DBS] specimens) and/or affordability. An assay with a limit of detection of 3000 IU/mL or lower would be acceptable and would identify 95% of those with viraemic infection, based on available data. 1

xxxii

INTERVENTIONS TO PROMOTE UPTAKE OF HEPATITIS TESTING AND LINKAGE TO CARE USE OF DRIED BLOOD SPOT (DBS) SPECIMENS FOR SEROLOGY AND NUCLEIC ACID TESTING Topic Serological testing Recommendations* • The use of DBS specimens for HBsAg and HCV antibody serology testing1 may be considered in settings where: - there are no facilities or expertise to take venous whole blood specimens; or - RDTs are not available or their use is not feasible; or - there are persons with poor venous access (e.g. in drug treatment programmes, prisons). Conditional recommendation, moderate (HBV)/low (HCV) quality of evidence The use of DBS specimens to test for HBV DNA and HCV RNA for diagnosis of HBV and HCV viraemia1, respectively, may be considered in settings where: - there is a lack of access to sites or nearby laboratory facilities for NAT, or provision for timely delivery of specimens to a laboratory; or - there are persons with poor venous access (e.g. in drug treatment programmes, prisons). Conditional recommendation, low (HBV)/moderate (HCV) quality of evidence

Detection of viraemia (nucleic acid testing)

•

OTHER INTERVENTIONS TO IMPROVE UPTAKE OF TESTING AND LINKAGE TO CARE Topic Uptake of testing and linkage to care Recommendations* • All facility- and community-based hepatitis testing services should adopt and implement strategies to enhance uptake of testing and linkage to care. Strong recommendation, moderate quality of evidence The following evidence-based interventions should be considered to promote uptake of hepatitis testing and linkage to care and treatment initiation: (Conditional recommendations) - Peer and lay health worker support in community-based settings (moderate quality of evidence). - Clinician reminders to prompt provider-initiated, facility-based HBV and HCV testing in settings that have electronic records or analogous reminder systems (very low quality of evidence). - Provision of hepatitis testing as part of integrated services within mental health/substance use services (very low quality of evidence).

•

*The GRADE system (Grading of Recommendations, Assessment, Development and Evaluation) was used to categorize the strength of recommendations as strong or conditional (based on consideration of the quality of evidence, balance of benefits and harms, acceptability, resource use and programmatic feasibility) and the quality of evidence as high, moderate, low or very low. 1 Well-functioning laboratory specimens referral network and system for return of results should be in place to maximize the impact of DBS specimens. There are currently few assays where the manufacturer’s instructions state that DBS specimens are validated for use. Therefore, currently use of DBS specimens would be considered “off-label”.

PART 1: BACKGROUND • Introductory chapters on objectives, scope and methodology of the guidelines Background to epidemiology, natural history, and serological and other markers of hepatitis B and C infection Background to diagnostics used to test for hepatitis B and C infection

•

1. INTRODUCTION

1.1. Current challenges in viral hepatitis testing Globally, hepatitis B virus (HBV) and hepatitis C virus (HCV) infection are major causes of acute and chronic liver disease (e.g. cirrhosis and hepatocellular carcinoma [HCC]), resulting in an estimated 1.4 million deaths annually (1). It is estimated that 248 million people are living with chronic HBV infection (CHB) (2), and that 110 million persons are HCV-antibody positive and 80 million have chronic viraemic HCV infection (3). Worldwide, it is estimated that a similar proportion of the total liver cancer mortality can be attributed to HCV (34 500) and HBV (30 000), with a smaller fraction due to alcohol (1). The burden of HBV and HCV remains disproportionately high in low- and middle-income countries (LMICs). Approximately 60% of the world’s population live in areas where HBV infection is highly endemic, particularly Asia and Africa. Additionally, even in low-prevalence areas, certain subpopulations have high levels of HCV and HBV infection, such as men who have sex with men (MSM), persons who inject drugs (PWID), people with HIV, as well as indigenous communities and migrants. The development of highly effective, well-tolerated, oral direct-acting antiviral (DAA) treatment regimens with high rates of cure has revolutionized the treatment of chronic HCV infection (4), although the high prices of the new medicines remain a major barrier to access in many countries (5). For people with chronic HBV infection, effective long-term suppressive treatment with tenofovir or entecavir is available (6). Despite the high global burden of disease due to chronic hepatitis B and C infection, and the advances and opportunities for treatment, most people infected with HBV and/or HCV remain unaware of their infection and therefore frequently present with advanced disease. The extent of this hidden burden is poorly documented, and largely based on limited data from higher-income settings (7–10). However, in low-income settings, it is estimated that less than 5% are aware of their diagnosis. This contrasts with the considerable recent progress in HIV testing coverage, whereby now more than half of all people living with HIV globally are aware of their status (11). Early identification of persons with chronic HBV or HCV infection would enable infected persons to receive the necessary care and treatment to prevent or delay the onset of liver disease and, in addition, prevent transmission by HBV vaccination of non-immune household contacts and sex partners.

There are several key reasons for this current low rate of hepatitis testing in LMICs. These include the limited facilities or services for hepatitis testing, lack of effective testing policies or national standards due to weak or non-existent hepatitis surveillance programmes to inform regional epidemiology and testing policies, costly and complex diagnostic assays and algorithms, poor laboratory capacity and infrastructure, and use of poor-quality test kits and reagents. In addition, in LMICs, HBV and HCV treatment remains unaffordable for those most in need, even if they have been diagnosed.

1.2. Goals of viral hepatitis testing Testing and diagnosis of HBV and HCV infection is the gateway for access to both prevention as well as care and treatment services (Fig. 1.1), and is a crucial component of an effective response to the hepatitis epidemic. The primary goals of testing are 1. to identify and link infected individuals, their partners and families to appropriate care and treatment services, and reduce hepatitis-related mortality by providing treatment to those in need through the use of directacting curative antiviral therapy for chronic hepatitis C and lifelong antiviral therapy for chronic hepatitis B infection; 2. to provide a link to preventive interventions to reduce transmission. For hepatitis, this includes provision of hepatitis B vaccination, and implementing individual- or facility-level prevention measures to reduce further transmission; 3. to monitor response to antiviral treatment. Testing is also undertaken for other reasons that are not within the scope of these guidelines. These include: surveillance for both acute hepatitis (to detect outbreaks, monitor trends in incidence and identify risk factors for new incident infections) and chronic hepatitis (to estimate the prevalence of chronic infection and monitor trends in sentinel groups) (12); and screening by blood transfusion services for hepatitis B and hepatitis C infection to exclude blood donations at risk of transmitting infections from donors to recipients.

FIG. 1.1. Cascade of viral hepatitis prevention, diagnosis, care and treatment VIRAL HEPATITIS CASCADE

ALL PEOPLE

PEOPLE REACHED BY PREVENTION ACTIVITIES

PEOPLE TESTED

AWARE OF STATUS

ENROLLED IN CARE

TREATMENT

RETAINED ON TREATMENT (HBV)

CURED (HCV)

ACCESSING CHRONIC CARE

CONTINUUM OF SERVICES

PREVENTION

TESTING

LINK TO CARE

TREATMENT

CHRONIC CARE

Source: Global health sector strategy on viral hepatitis 2016–2021. Geneva, World Health Organization; 2016 (16).

1.3. Why are testing guidelines needed? In 2010 and 2014, World Health Assembly resolutions WHA63.18 (13) and WHA67 (14) recognized viral hepatitis as a global public health problem. It directed WHO to develop and implement both a comprehensive strategy to address viral hepatitis, as well as provide clear guidance to Member States on the diagnosis and management of HBV and HCV infection. Recent WHO guidelines on treatment for HCV (5) and HBV (6) did not include comprehensive guidance on who to test and how to test for diagnosis. The Global Health Sector Strategy on Viral Hepatitis 2016–2021 (16) is the first global strategy on viral hepatitis, and covers the first six years of the Agenda for Sustainable Development. The Strategy outlines a set of global targets (see Web annex 1), including targets on diagnosis of chronic hepatitis B and hepatis C infection, and describes a set of priority actions for countries to achieve these hepatitis targets The Strategy is designed to contribute to the attainment of the 2030 Agenda for Sustainable Development and, specifically, to health-related Goal 3 (target 3.3). “By 2030, end the epidemics of AIDS, tuberculosis, malaria and neglected tropical diseases and combat hepatitis, water-borne diseases and other communicable diseases.”

1.4. Goals and objectives of the guidelines The overall objective of these guidelines is to provide the first WHO evidencebased guidance on testing for hepatitis B and C virus infection in adults, adolescents and children living, particularly in LMICs, where the burden of disease is highest and where access to treatment is becoming more readily available as treatment costs continue to decline. The guidelines are expected to provide the basis and rationale for the development of national guidelines for hepatitis testing, particularly in resource-limited settings, according to the local epidemiology of hepatitis B and C infection, health-care delivery system of the country, available resources and other determinants, with the overall aim of reducing the global burden of HBV and HCV infection. The specific objectives of the guidelines are • to provide recommendations in the area of who to screen for hepatitis B and hepatitis C infection, and which testing strategies and algorithms to use; • to provide evidence summaries, Grading of Recommendations Assessment, Development and Evaluation (GRADE) reviews, evaluation of the overall balance of benefits and harms, feasibility, costs and acceptability of the proposed recommendations; • to provide implementation guidance to support operationalization of the recommendations at country level, which includes a systematic approach to the selection and evaluation of assays, quality systems for all aspects of hepatitis testing, and a framework for planning the best mix of testing approaches; • to identify research gaps.

1.5. Scope of the guidelines The overall scope of these testing guidelines is the diagnosis, counselling and linkage to care of persons with chronic hepatitis B and hepatitis C infection. They are primarily aimed at resource-limited settings where hepatitis testing programmes are not yet well developed or where quality systems are lacking. The guidelines include the following components: • testing approaches – who to test for chronic hepatitis B and C infection • testing strategy – how to test for chronic hepatitis B and C infection • interventions to promote uptake of hepatitis testing and linkage to care • implementation issues with regard to product selection and procurement, validation of test kits, and quality assurance (QA).

Certain key topics were not included in the scope of work for these guidelines and are either addressed more fully in other WHO documents or guidelines, or will be included in future updates. These include: diagnosis and management of acute hepatitis B (6) and C infection (5); surveillance of acute and chronic hepatitis B and hepatitis C (12); treatment and side-effect monitoring of drugs for chronic hepatitis B and C (5, 6, 15); diagnosis and management of hepatitis A (17), hepatitis E (18) and hepatitis delta virus (19); use of HCV RNA or core antigen as a single test for the diagnosis of HCV infection; and recommendations and testing strategies for screening of donated blood (20).

1.6. Target audience These guidelines are primarily targeted at national hepatitis programme managers and other policy-makers in ministries of health, particularly in LMICs, who are responsible for the development of national hepatitis testing and treatment plans, policy and guidelines. These guidelines will also be useful for laboratory managers in ministries of health, reference laboratories and key hospital laboratories, who are responsible for validation of assays, development of national testing algorithms, and national procurement of assays and quality control (QC). Finally, the guidelines will serve as a reference for health-care providers who offer and implement hepatitis testing and care for persons with hepatitis B and hepatitis C infection, including those from community-based programmes.

1.7. Related WHO materials and guidelines These guidelines on testing for chronic hepatitis B and hepatitis C infection are intended to complement several existing WHO guidelines. These include the following: Guidelines for the prevention, care and treatment of persons with chronic hepatitis B infection (http://apps.who.int/iris/bitstream/10665/154590/1/9789241549059_ eng.pdf?ua=1&ua=1) (6) and for chronic hepatitis C infection (http://apps.who. int/iris/bitstream/10665/205035/1/9789241549615_eng.pdf) (5). These provide recommendations along the continuum of care, from diagnosis, initial assessment of stage of liver disease, initiation of treatment and monitoring. A summary of recommendations is provided in Web annexes 2 and 3, respectively. • Consolidated guidelines on HIV testing services (http://apps.who.int/iris/bitstre am/10665/251655/1/9789241549868-eng.pdf?ua=1) (11) and HIV self-testing supplement (http://apps.who.int/iris/bitstream/10665/251655/1/9789241549868eng.pdf?ua=1) (21). • Technical considerations and case definitions to improve surveillance for viral hepatitis (http://apps.who.int/iris/bitstream/10665/204501/1/9789241549547_

eng.pdf) (12) and Monitoring and evaluation for viral hepatitis B and C: recommended indicators and framework: technical report (http://apps.who. int/iris/bitstream/10665/204790/1/9789241510288_eng.pdf?ua=1) (22). • Consolidated guidelines on the use of antiretroviral drugs for treating and preventing HIV infection (http://apps.who.int/iris/bitstream/10665/208825/ 1/9789241549684_eng.pdf?ua=1) (23). • Hepatitis B control through immunization: a reference guide on prevention of perinatal and early childhood HBV infection through infant hepatitis B vaccination (24) (http://www.who.int/immunization/sage/ meetings/2015/october/8_WPRO_Hepatitis_B_Prevention_Through_ Immunization_Regional_Reference_Guide.pdf); as well as catchup vaccinations in key affected populations (http://apps.who.int/iris/ bitstream/10665/128048/1/9789241507431_eng.pdf?ua=1&ua=1 ns) (25), such as PWID, MSM (26) and sex workers (27) (http://apps.who.int/iris/ bitstream/10665/44619/1/9789241501750_eng.pdf?ua=1); (https://www. unfpa.org/sites/default/files/pub-pdf/9789241504744_eng.pdf). • Consolidated guidelines on HIV prevention, diagnosis, treatment and care for key populations (25) and Guidance on prevention of viral hepatitis B and C among people who inject drugs (http://apps.who.int/iris/ bitstream/10665/75357/1/9789241504041_eng.pdf?ua=1) (28). • Guidance on prevention of hepatitis infection in health-care settings (28–30) includes recommendations on hand hygiene, including surgical hand preparation, handwashing and use of gloves; safe handling and disposal of sharps and waste; safe cleaning of equipment; testing of donated blood and blood products; improved access to safe blood and blood products; and training of health personnel. There are also new WHO recommendations published in 2015 on the use of auto-disable syringes in immunization services, and safetyengineered injection devices, including reuse prevention (RUP) syringes and sharps injury prevention (SIP) devices for therapeutic injections (31) (http:// apps.who.int/iris/bitstream/10665/44102/1/9789241597906_eng.pdf); (http:// www.who.int/bloodsafety/publications/UniversalAccesstoSafeBT.pdf?ua=1); (http://www.euro.who.int/__data/assets/pdf_file/0005/268790/WHO-guidelineson-drawing-blood-best-practices-in-phlebotomy-Eng.pdf?ua=1).

2. GUIDING PRINCIPLES 2.1. Promoting human rights and equity in access to hepatitis testing Access to health care is a basic human right and applies equally to men, women and children, regardless of gender, race, sexual preference, socioeconomic status or behavioural practices, including drug use, and is in keeping with the United Nations Universal Declaration of Human Rights (32). The promotion of human rights and equity in access to hepatitis B and C testing, prevention, treatment and care are guiding principles central to these guidelines. Persons with hepatitis B and C infection may come from vulnerable groups because of low socioeconomic status with poor access to appropriate health care, or because they belong to groups that are marginalized or stigmatized such as PWID, MSM, migrants, indigenous peoples or prisoners. Hepatitis testing services need to ensure that testing is accessible to the populations most affected, and that these groups are offered testing in an environment that minimizes stigma and discrimination. Informed consent should always be obtained. Screening for viral hepatitis must not be used as a means to discriminate against those testing positive. The provision of adequate safeguards to ensure confidentiality, and a non-coercive approach are fundamental principles of good clinical practice.

2.2. The public health approach along the continuum of care In accordance with existing WHO guidance on HIV testing (11), use of antiretrovirals (ARVs) (23), and HBV and HCV treatment (6, 5), these guidelines are based on a public health approach to scaling up testing and treatment for hepatitis B and C across the entire continuum of care. The public health approach seeks to ensure the widest possible access to high-quality services at the population level, based on simplified and standardized approaches that can readily be taken to scale and decentralized, including in resource-limited settings. A public health approach aims to strike a balance between implementing the best-proven standard of care and what is feasible on a large scale in resourcelimited settings, and to achieve health equity, promote gender equality, engage communities, and leverage public and private sectors in the response.

2.3. The WHO “5 Cs” The WHO “5 Cs” are principles that apply to all models of hepatitis testing and in all settings: Consent, Confidentiality, Counselling, Correct test results and Connection (linkage to prevention, treatment and care services) (11). This means hepatitis testing for diagnosis must always be voluntary, and consent for testing informed by pre-test information. Testing should be linked to prevention, treatment, care and support services to maximize both individual and public health benefits. Mandatory, compulsory or coercive hepatitis testing is never appropriate, whether that coercion comes from a health-care provider, an employer, authorities (such as immigration services) or a partner or family member. All testing sites should ensure client confidentiality.

2.4. Accurate testing Patients have the right to accurate and high quality testing to ensure that those requiring treatment are identified and initiated, while those who are negative or not in need of treatment are not inappropriately treated. The foundation of accurate testing includes: (i) provision of reliable, high quality, regulatory approved test kits; (ii) qualified, trained, competent and supported testing personnel; and (iii) quality-assured testing environment that addresses quality (process) control, equipment management and maintenance, accurate recordkeeping and documentation (standard operating procedures (SOPs), and external quality assessment (EQA) schemes. Some countries will face significant challenges as they seek to implement testing for chronic hepatitis B and hepatitis C infection due to constraints in resources and health systems. Each country will need to plan its own approach to implementing quality hepatitis testing services. Such services should be informed by the local context, including national hepatitis B and C epidemiology, availability of appropriately trained individuals and suitable laboratory capacity with quality management systems in place. Other considerations are efficient supply systems for laboratory commodities, availability of financial resources, organization and capacity of the health system, anticipated cost–effectiveness of the various interventions, and fair and equitable expansion in access.

3. METHODOLOGY AND PROCESS OF DEVELOPING THE GUIDELINES 3.1. WHO guideline development process These WHO guidelines were developed following the recommendations for standard guidelines as described in the WHO Handbook for Guideline Development (33), and the GRADE framework (34–37) (Tables 3.1, 3.2 and Box 3.1). A Guidelines Development Group was formed with representation from different geographical regions as well as from a wide range of stakeholders, including researchers, clinicians and programme managers, advocacy groups and members of organizations that represent persons living with chronic hepatitis. There was an initial scoping and planning process to formulate questions most relevant to LMICs and patientimportant outcomes (see Web annex 4 for all PICO questions).

3.2. Systematic reviews and additional background work Systematic reviews on diagnostic performance. Systematic reviews and metaanalyses of the primary literature were commissioned externally to address the research questions and patient-important outcomes. For evaluation of HBV and HCV diagnostics and testing strategies, there was very limited or no evidence for patient-important outcomes. The Guidelines Development Group and PICO questions considered diagnostic accuracy (sensitivity, specificity, positive and negative predictive values) and in some cases analytical sensitivity (limit of detection) as surrogates for patient-important outcomes, assuming reasonable linkage and access to care. Search strategies and summaries of evidence are reported in Web annex 5. The glossary provides full definitions for diagnostic and analytical test performance. As part of the guidelines development process, WHO commissioned other work to provide additional data to support the recommendations. These are given below. • Existing systematic reviews on global and regional seroprevalence of HBsAg and HCV antibody in general population and specific high-risk populations (Table 4.1). • Review of the cost–effectiveness literature of different viral hepatitis testing approaches in different settings. The evidence base for different testing approaches remains very limited, especially for impact on patient-important outcomes and in LMICs, and largely relies on observational data and modelling.

The limited number of cost–effectiveness studies and the heterogeneity of study populations, testing approaches and outcomes measured precluded a formal systematic review and meta-analysis. A narrative review was therefore undertaken that included studies of: (i) focused or targeted testing of the highest-risk groups; (ii) routine testing among specific birth cohorts that are readily identified and have a high prevalence of HCV infection; and (iii) routine testing throughout the entire population, in different settings. • Predictive modelling of testing strategies (i.e. one- or two-test serological testing strategies). There were very few studies that directly compared different testing strategies for diagnostic accuracy and therefore a predictive modelling analysis was carried out to examine the accuracy of a testing strategy across a range of performance characteristics of the assays (sensitivity and specificity) based on the systematic reviews, and a hypothetical range of prevalence of the disease in the population (10%, 2%, 0.4%) representing high-, medium- and low-prevalence settings or populations (see Web annex 6). • Values and preferences survey of health-care workers and implementers for different testing strategies and approaches. A four-part online survey tool was undertaken in September 2015, which covered questions on current and preferences for future HBV and HCV testing practices, including a test of HCV cure. Respondents included clinicians, patient organizations, civil society representatives, programme managers, policy-makers and pharmaceutical industry employees. • Feasibility survey on programmatic experiences and reports of barriers/ challenges to HBV and/or HCV testing based on 22 interviewees across 13 LMICs conducted between June and September 2015. The 33-question semistructured questionnaire covered programme information (who is tested and where, what assays/algorithms are used, counselling and training, funding and costs of testing); protocol for hepatitis care and treatment; perceived barriers/ challenges and solutions; and provision of relevant epidemiological data. • Case examples of different models of hepatitis testing practices in different settings and populations were also solicited and identified through a hepatitis testing innovation contest, to illustrate effective and acceptable ways to deliver facility and community-based testing services, especially among most affected populations.

3.3. Grading of quality of evidence and strength of recommendations The quality of the evidence was assessed and either rated down or rated up based on criteria specified in GRADE methods, modified for diagnostic tests and test strategies (38, 39). Summaries of the quality of evidence to address each outcome were

entered in the GRADE profiler software (GRADE pro 3.6). The quality of evidence was categorized as high, moderate, low or very low (Box 3.1 and Table 3.1). Specific issues with rating quality of evidence for studies of diagnostic accuracy and strategies Diagnostic test accuracy. For evaluation of HBV and HCV diagnostics and testing strategies, there was very limited or no evidence on effects on patient-important outcomes. The Guidelines Development Group and PICO questions considered diagnostic accuracy (sensitivity, specificity, positive and negative predictive values) and in some cases analytical sensitivity (limit of detection) as surrogates for patientimportant outcomes, assuming reasonable linkage and access to care. Although observational studies of interventions start as low quality in GRADE, cross-sectional and cohort studies of diagnostic accuracy can provide reliable evidence (38), and were therefore initially categorized as high quality. Evidence was then rated down based on the presence of (i) risk of bias (using a tool designed for assessment of diagnostic accuracy studies, the QUADAS-2 tool) (40); (ii) inconsistency or heterogeneity; (iii) indirectness (addressing a different population than the one under consideration); or (iv) imprecision. However, evaluating inconsistency in studies of diagnostic accuracy is a challenge because methods to measure statistical heterogeneity are lacking and inconsistency is common, and therefore we did not downgrade for indirectness. Testing strategies. Clinical studies to evaluate comparisons of different testing strategies and approaches were generally not available. Therefore, the Guidelines Development Group considered instead predictive modelling to generate estimates of diagnostic performance of different testing strategies. This type of evidence was not formally graded but was considered low quality because it is very indirect.

BOX 3.1. Standard approach to rating the quality of evidence and strength of recommendations using the GRADE system The GRADE system separates the rating of the quality of evidence from the rating of the strength of the recommendation. The quality of evidence is defined as the confidence that the reported estimates of effect are adequate to support a specific recommendation. The GRADE system classifies the quality of evidence as high, moderate, low and very low (35, 37, 41–45). For studies of interventions, randomized controlled trials (RCTs) are initially rated as high-quality evidence but may be downgraded for several reasons, including the risk of bias, inconsistency of results across studies, indirectness of evidence, imprecision and publication bias. Observational studies of interventions are initially rated as low-quality evidence but may be upgraded if the magnitude of the treatment effect is very large, if multiple studies show the same effect, if evidence indicates a dose–response relationship, or if all plausible biases would underestimate the effect (41). The higher the quality of evidence, the more likely a strong recommendation can be made. The strength of a recommendation reflects the extent to which the Guidelines Development Group was confident that the desirable effects of following a recommendation outweigh the potential undesirable effects. The GRADE system classifies the strength of a recommendation in two ways: “strong” and “conditional” (37). The strength is influenced by the following factors: the quality of the evidence, balance of benefits and harms, values and preferences, resource use and the feasibility of carrying out the intervention (Table 3.2). A strong recommendation is one for which the Guidelines Development Group was confident that the desirable effects of adhering to the recommendation outweigh the undesirable effects. A conditional recommendation is one for which the Guidelines Development Group concluded that the desirable effects of adhering to the recommendation probably outweigh the undesirable effects but the Guidelines Development Group is not confident about these trade-offs. The implications of a conditional recommendation are that, although most people or settings would adopt the recommendation, many would not or would do so only under certain conditions. The reasons for making a conditional recommendation include the absence of high-quality evidence, imprecision in outcome estimates, uncertainty regarding how individuals value the outcomes, small benefits relative to harms, and benefits that may not be worth the costs (including the costs of implementing the recommendation).

TABLE 3.1. GRADE categories of the quality of evidence Level of evidence High Moderate Low Very low Rationale Further research is very unlikely to change our confidence in the estimate of effect. Further research is likely to have an important impact on our confidence in the effect. Further research is very likely to have an estimate of effect and is likely to change the estimate. Any estimate of effect is very uncertain.

TABLE 3.2. Key domains considered in determining the strength of recommendations Domain Benefits and risks/harms Values and preferences (acceptability) Costs and financial implications (resource use) Feasibility Rationale Desirable effects (benefits) need to be weighed against undesirable effects (risks/harms). The more the benefits outweigh the risks, the more likely a strong recommendation will be made. If the recommendation is likely to be widely accepted or highly valued, a strong recommendation will probably be made. If there are strong reasons that the recommended course of action is unlikely to be accepted, a conditional recommendation is more likely to be made. Lower costs (monetary, infrastructure, equipment or human resources) or greater cost–effectiveness will more likely result in a strong recommendation. If an intervention is achievable in a setting where the greatest impact is expected, a strong recommendation is more probable.

3.4. Formulation of recommendations At the September 2015 meeting of the Guidelines Development Group, for each of the PICO questions (see Web annex 4), the results of the systematic reviews and the evidence profiles (see Web annexes 5 and 6) were presented and reviewed. Commissioned surveys of diagnostic costs, values and preferences for different testing strategies of health-care workers and implementing partners, and a global survey of programmatic experience were also considered. Recommendations were then formulated based on the overall quality of the evidence, in addition to other considerations, including the balance between benefits and harms, values and preferences, feasibility and resource implications (Table 3.2). The strength of the recommendations was rated as either strong (the panel was confident that the benefits of the intervention outweighed the risks) or conditional (the panel considered that the benefits of the intervention outweighed the risks, but the balance of benefits to harms and burdens was small or uncertain). Recommendations were then formulated and the wording finalized by the entire Group. Implementation needs were subsequently evaluated, and areas and topics requiring further research identified.

For recommendations based on diagnostic accuracy, the Guidelines Development Group considered potential trade-offs between diagnostic accuracy and other factors. Although diagnostic accuracy was considered a critical outcome and a reasonable surrogate for patient outcomes, tests and testing strategies associated with slightly lower diagnostic accuracy could be recommended when associated with lower costs, increased testing access and linkage to care or greater feasibility.

3.5. Declaration and management of conflicts of interest In accordance with WHO policy, all members of the Guidelines Development Group and peer reviewers were required to complete and submit a WHO Declaration of Interest form (including participation in consulting and advisory panels, research support and financial investment) and, where appropriate, also provide a summary of research interests and activities. The WHO Secretariat then reviewed and assessed the declarations submitted by each member and, at the September 2015 meeting of the Guidelines Development Group, presented a summary to the Guidelines Development Group (see Web annex 7). The WHO Secretariat stated that there had been a transparent declaration of financial and academic interests, and concluded that there were no conflicts that required exclusion of any member from actively taking part in formulating the recommendations during the meeting. For the peer review group, the WHO Secretariat was also satisfied that no case necessitated exclusion from the review process.

3.6. Updating, disseminating and monitoring implementation of the guidelines The guidelines are accessible on the WHO website with links to other related websites, and translated into the official United Nations (UN) languages. WHO disseminates the guidelines to ministries of health in countries, as well as key international, regional and national collaborating partners (e.g. civil society, foundations, donors). Successful implementation of these guidelines will be assessed by the number of countries that incorporate the contents into national hepatitis plans and guidelines. The impact of the testing guidelines will be measured by monitoring the number of persons tested and treated for chronic hepatitis B and hepatitis C infection, in accordance with targets proposed in the WHO Global health sector strategy on viral hepatitis 2016–2021 (16) (see Web annex 1). The Guidelines Development Group recognized that the field of hepatitis diagnostics and testing is evolving rapidly, and it is anticipated that there will be a need for periodic updates.

4. BACKGROUND – EPIDEMIOLOGY AND NATURAL HISTORY An understanding of the global and regional epidemiology and burden of hepatitis B and C infection with respect to the main routes of transmission, most affected populations, and natural history and time course of serological markers is critical to inform strategies on both who to test and how to test. However, data are limited in many LMICs, particularly in the African region, due to weak surveillance systems with underreporting and therefore unreliable data. The nature of an epidemic within a specific country will determine the appropriate testing strategy and approaches. Table 4.1 provides an overview of the risk factors and primary routes of transmission for HBV and HCV infection in populations most affected by HBV and HCV, as well as data on seroprevalence from systematic reviews and other studies. TABLE 4.1. Overview of populations most affected by HBV and HCV infection with summary of risk factors, primary routes of transmission and seroprevalence rates Key and priority populations People who inject drugs (PWID) Hepatitis B Hepatitis C

High risk of infection through parenteral exposure, most commonly from sharing of needles and other injecting equipment. Prevalence rates of HBV infection among PWID similar to background population in HBV-endemic areas (1, 46, 47) Global prevalence estimated to be 67% among PWID in 77 countries and over 80% in 12 countries) (46). Prevalence is particularly high in settings where PWID are criminalized and lack access to harm reduction services. Non-injecting drug use, e.g. through intranasal drug use, has been associated with a small but increased risk of HCV infection (48).

People in prisons and closed settings

High risk of infection through parenteral exposure, most commonly from sharing of needles and razor blades and other injecting equipment, particularly when safe injecting equipment is not available (25, 49). Potential for increased risk of sexual transmission due to unsafe sex behaviours, lack of availability of prevention hardware such as condoms, and higher risk of experiencing men-on-men sexual violence (50) Estimated global prevalence ranges from 23% to 29%, with rates as high as 40% reported from some regions, including Australia, North America, western Europe, Central Asia, East and South-East Asia (51)

Key and priority populations Mobile or migrant populations

Hepatitis B Migrants from intermediate- and high-endemic HBV areas are at increased risk of chronic hepatitis B (CHB) (52–54). Displaced and marginalized populations may be at increased risk of sexual transmission of HBV due to increased vulnerability to sexual violence or coercion, or unsafe sex practices (53, 54).

Hepatitis C Migrant populations represent a heterogeneous group and HCV seroprevalence estimates vary widely (52–54).

Some marginalized mobile populations may be more likely to belong to other populations at high risk for HBV and HCV transmission, such as PWID or sex workers (53, 54). Indigenous populations Some indigenous populations may have higher rates of prevalence but poorer access to HBV vaccination or be more likely to belong to other high-risk populations, such as PWID (55, 56). Sex workers are at increased risk of sexual transmission of HBV due to exposure to multiple partners and poorer access to access safe sex materials such as condoms (56). Some indigenous populations may be more likely to belong to other high-risk populations, such as PWID (55, 56). Overall, the risk of sexual transmission of HCV is low. There may be a small, increased risk of transmission among persons with multiple sex partners.

Sex workers

Sex workers may be more likely to belong to other high-risk populations, such as PWID and persons in prisons or closed settings (56). Transgender people Men who have sex with men (MSM) Transgender people may be at increased risk for viral hepatitis through using unsafe injecting equipment for administration of hormones or through sexual transmission (57). MSM are at increased risk of sexual acquisition of HBV (58). Risk of sexual transmission of HCV is low among HIV-negative MSM. HIV-positive MSM are at significantly increased risk of sexual transmission of HCV, particularly those who engage in high-risk sex behaviours such as unprotected anal sex (59–63). In several outbreaks of HCV infection among MSM in Europe, Australia and the US, transmission has been linked to sexual exposure as well as potentially to underreported use of injecting and non-injecting recreational drugs (63–65).

Key and priority populations Health-care workers

Hepatitis B

Hepatitis C

The greatest proportion of occupational transmission of viral hepatitis is due to percutaneous injury via needles during vascular access. Transmission may also occur through exposure to blood and body fluids on skin lesions and mucous membranes (66). Multiple factors contribute to higher risk of occupational acquisition in LMICs. These include: working among populations with a higher prevalence of infection, higher rates of unnecessary injections in health-care settings, use of unsterilized needles and equipment lacking a needle-stick safety mechanism, lack of implementation of standard precautions, inadequate coverage of HBV vaccination (67). Among non-immune persons, the risk of HBV infection after percutaneous exposure ranges from less than 6% (if HBeAg negative) to 30% (if HBeAg posiitive) (68). Risk of HCV infection after percutaneous exposure estimated to be 1.8% (68).

Persons exposed in health-care settings

High risk of parenteral transmission in settings with a higher background seroprevalence of HBV and HCV and where infection control practices are inadequate (e.g. diagnostic and therapeutic procedures), and blood transfusions and other tissue donations are not screened for viral hepatitis (69–79). Persons who may have multiple exposures, such as patients with thalassaemia or haemophilia who receive multiple transfusions, and patients on haemodialysis, are at higher risk (80–82).

Persons exposed via other invasive procedures Persons living with HIV and those living with other sexually transmitted infections (STIs) Infants born to infected mothers

There is a small but increased risk of HBV and HCV transmission with other procedures where there is a risk blood-to-blood transmission via contaminated equipment, including cosmetic procedures (such as tattooing and body piercing), and traditional medicine procedures such as scarification and circumcision (80–82). Persons who have been exposed to HIV or other STIs via sexual transmission may be at increased risk of sexually acquired HBV infection (87). There is an increased risk of HCV infection among persons living with HIV (88–93).

Particularly in high HBV and HCV-prevalence settings, children who have been exposed to HIV through mother-to-child transmission (MTCT) are at increased risk of HBV and HCV infection (86). Perinatal or early childhood transmission is the main route of infection in many parts of the world, particularly in endemic countries, where 90% of CHB infections may be attributable to MTCT. HBV transmission in early life is associated with a much higher risk of developing chronic infection (90% in the perinatal period to 6 months of age) than acquisition later in childhood or adulthood) (94, 95). MTCT is the most common cause of HCV infection in young children. Risk of HCV transmission is 4–8% in the perinatal period, and 10%–25% among children born to mothers coinfected with HIV (96–99).

Key and priority populations Children

Hepatitis B Horizontal (household, intra-familial and child-to-child) transmission is an important route of infection. Up to 50% of childhood CHB infections cannot be accounted for by MTCT of HBV.

Hepatitis C Based on limited data, horizontal transmission does not appear to be a significant contributor to HCV transmission in children (100). High prevalence in some settings such as in children treated in hospital for malignancy, renal failure requiring haemodialysis, and those who have undergone surgical procedures likely reflects iatrogenic transmission (101). Overall, the risk of sexual transmission of HCV is low. However, there is a small increased risk among persons with multiple sex partners (59, 102).

Adolescents

There may also be adolescents who missed out on HBV vaccination, and were infected perinatally or in early childhood. Adolescents who engage in early sex, have multiple sex partners, or sex partners with CHB are at increased risk (25).

Vulnerable adolescents may be more likely to belong to other high-risk key populations, including PWID and sex workers, for example (25). Couples, partners and household contacts Persons who live in the same household as a person with CHB are at increased risk of horizontal acquisition of HBV infection (103). Overall, the risk of sexual transmission of HCV is low. However, the risk is increased among persons with multiple sex partners (59, 102, 104). There is no evidence to support transmission among household contacts who are not sexual partners (100). CHB: chronic hepatitis B; HBeAg: hepatitis B e antigen; HBsAg: hepatitis B surface antigen; HBV: hepatitis B virus; HCV: hepatitis C virus; insert MTCT: mother-to-child transmission; MSM: men who have sex with men; PWID: people who inject drugs; STI: sexually transmitted infection

4.1. Hepatitis B infection 4.1.1. Epidemiology of hepatitis B infection It is estimated that worldwide, 2 billion people have evidence of past or present infection with HBV, and 248 million are chronic carriers of HBV surface antigen (HBsAg), particularly in LMICs (2). Age-specific HBsAg seroprevalence varies markedly by geographical region, with the highest prevalence (>5%) in subSaharan Africa (SSA), east Asia, some parts of the Balkan region, the Pacific Islands and Amazon Basin of South America. Prevalence below 2% is seen in regions such as Central America, North America and Western Europe (2). Overall, almost half of the global population lives in areas of high and intermediate endemicity. The major complications of CHB are cirrhosis and HCC. Worldwide, it is estimated that around 686 000 people die each year from the complications

of CHB (1). Overall, HBV infection accounts for around 45% of cases of HCC and 30% of cirrhosis, with much higher proportions in LMICs (1, 105). In Asia and most other regions, the incidence of HCC and cirrhosis is low before the age of 35–40 years but then rises exponentially (1). However, in some parts of Africa, Alaska and the Amazon, the incidence of HCC is also high in infected children and young adult men (106). HIV and HBV. There is an estimated global HBsAg prevalence of 7·4% (IQR 5.0– 11.2%) in HIV-infected persons, and a burden of 2.73 million (IQR 1.8–3.9 million; IQR 1·3–4·4 million) HIV–HBsAg-coinfected persons (87). The highest burden for HIV–HBV coinfection is in sub-Saharan Africa (SSA) (71% of all cases; 1.96 million).

4.1.2. Transmission of hepatitis B infection Table 4.1 provides an overview of the risk factors and primary routes of transmission for HBV infection in populations most affected by hepatitis B. HBV is spread predominantly by percutaneous or mucosal exposure to infected blood and various body fluids, including saliva and menstrual, vaginal and seminal fluids. Perinatal transmission is the major route of HBV transmission in many parts of the world, and an important factor in maintaining the reservoir of the infection in some regions, particularly in China and South-East Asia (107, 108). Horizontal transmission, including household, interfamilial and especially child to child, is also important (103). Both sexual and oral transmission of hepatitis B may occur, particularly in unvaccinated MSM and heterosexual persons with multiple sex partners or contact with sex workers. Transmission of the virus may also result from accidental inoculation of minute amounts of blood or fluid during medical, surgical and dental procedures, or from razors and similar objects contaminated with infected blood; immunization with inadequately sterilized syringes and needles; injecting drug use; tattooing; body piercing; and acupuncture. Unvaccinated health-care workers are also at risk of accidental transmission of hepatitis B during handling contaminated sharps, body fluids and organs, and medical waste.

4.1.3. Natural history of HBV infection Hepatitis B virus is an enveloped DNA virus, and a member of the family Hepadnaviridae hepatotropic DNA viruses. Hepatitis B virus causes both acute and chronic infection that can range from asymptomatic infection or mild disease to severe or fulminant hepatitis. Acute hepatitis B is usually a self-limiting disease marked by acute inflammation and hepatocellular necrosis, with a case fatality rate of 0.5–1% (109). Chronic hepatitis B (CHB) encompasses a spectrum of disease, and is defined as persistent HBV infection (the presence of detectable HBsAg in the blood or serum for longer than six months), with or without associated active viral replication and evidence of hepatocellular injury and inflammation (109). Age is a key factor in determining the risk of chronic infection. Chronicity is common following acute infection in neonates (90% of neonates born to hepatitis B e antigen [HBeAg]-positive mothers) and in

young children under the age of 5 years (20–60%), but occurs less commonly (<5%) when infection is acquired in adulthood (94, 95) (Fig. 4.1). Worldwide, the majority of persons with CHB were infected at birth or in early childhood. FIG. 4.1 Outcomes of hepatitis B virus infection by age at infection 100 Chronic infection (%) 80 60 40 20 6 Birth 1–6 months 7–12 months 1–4 years Other children and adults

Symptomatic infections Chronic infections

Source: Guidelines for the prevention, care and treatment of persons with hepatitis B infection. Geneva: WHO; 2015 (http://www.who.int/hepatitis/publications/hepatitis-b-guidelines/en/, accessed 15 June 2016) (6).

The natural history of CHB is dynamic and complex, and progresses nonlinearly through several recognizable phases (6, 95). The phases are of variable duration, not necessarily sequential, and do not always relate directly to criteria and indications for antiviral therapy (47).

4.1.4. Time course and interpretation of serological markers of HBV infection A range of HBV markers other than HBsAg, such as anti-HBc total and anti-HBc IgM, HBeAg and antibodies to hepatitis B e and surface antigen (anti-HBe and anti-HBs) and HBV DNA can be used to further characterize HBV infection (see Table 4.2). When these markers are tested concurrently, a testing profile can be produced to differentiate acute from chronic infection, stage the disease and identify those who may benefit from treatment, monitor disease progression or response to antiviral treatment, as well as those who would benefit from HBV immunization or re-immunization. The appearance of HBsAg in the blood is followed by that of HBeAg, which is a marker of high levels of viral replication. In acute HBV infection that resolves by itself, HBeAg seroconverts relatively early to anti-HBe with the disappearance of HBsAg and HBeAg. But in chronic HBV infection, seroconversion to anti-HBe may be delayed for many years, HBeAg may persist, or neither anti-HBe nor HBeAg may be detectable in the presence of HBsAg. Antibodies to hepatitis B core antigen (anti-HBc) may occur relatively early in the infection, often within a week or two after the appearance of HBsAg, and is typified by a profound immunoglobulin (Ig)M anti-HBc response that wanes approximately 6 months later (Fig. 4.2 and 4.3.)

FIG. 4.2 Acute HBV infection with recovery Symptoms HBeAg anti-HBe

FIG. 4.3 Chronic HBV infection Acute (6 months) Chronic (years)

Titre

Total anti-HBc Titre

HBsAg

HBeAg

anti-HBe

IgM anti-HBc HBsAg anti-HBs

Total anti-HBc IgM anti-HBc IgM anti-HBc associated with flares (often with symptoms)

HBsAg levels may wane over time

8 12 16 20 24 28 32 36 Weeks after exposure

100

0 4 8 12 16 20 24 28 3236

Years

Weeks after exposure

CHB is defined as the persistence of HBsAg for more than 6 months. Previous HBV infection is characterized by the presence of antibodies (anti-HBs and anti-HBc). Immunity to HBV infection after vaccination is characterized by the presence of only anti-HBs. It also needs to be established whether the person is in the HBeAg-positive or HBeAg-negative phase of infection, though both require lifelong monitoring, as the condition may change over time. In persons with CHB, a positive HBeAg result suggests high-level HBV replication and high infectivity. Spontaneous improvement may occur following HBeAg-positive seroconversion (antiHBe), with a decline in HBV replication, and normalization of alanine aminotransferase (ALT) levels. This confers a good prognosis and does not require treatment. Further assessment of HBsAg-positive persons is needed to guide management and indicate the need for treatment (6). This generally includes assessment of additional serological markers of HBV infection (HBeAg), measuring aminotransferase levels to help determine liver inflammation, quantification of HBV DNA levels, and stage of liver fibrosis by non-invasive tests (NITs) such as transient elastography or serum biomarker-based tests such as aspartate aminotransferase (AST)-to-platelet ratio index (APRI), and fibrosis-4 (FIB-4). TABLE 4.2. Summary of markers of HBV infection Marker HBsAg Characteristics • • • • First serological marker of HBV infection to appear (Fig. 4.2 & 4.3) Window period between HBV infection and detection of HBsAg estimated to be around 38 days, but depends on analytical sensitivity of assay used, immunocompetence of host and individual virus kinetics Occult HBV infectiona has been observed, i.e. HBsAg is undetectable but HBV DNA can be detected in individuals not in the window period Quantification of HBsAgb is a potential alternative marker of viraemia and to monitor response to antiviral treatment

Marker Anti-HBc IgMc

Characteristics • • High levels present during acute infection but may remain detectable for up to 6 months Used to differentiate between acute and chronic HBV infection, but its reappearance during “flares” in chronic HBV infection make it an unreliable indicator of recent primary HBV infection (Fig. 4.3) Develops around 3 months after infection and most constant marker of infection Together with anti-HBs, indicates resolved infection Anti-HBc, with or without anti-HBs, also indicates individuals who may reactivate in the context of immunosuppression Present when the virus is actively replicating in the liver Associated with high levels of HBV viraemia and is therefore a marker of “high infectivity” Associated with progressive liver disease Represents host response to HBeAg and usually indicates decreasing HBV DNA and therefore infectivity Present in the immune-control and immune-escape phases May coexist with HBeAg during the period of seroconversion from e antigen to e antibody at the end of immune-tolerance phase Neutralizing antibody that confers protection from infection Present following spontaneous HBsAg clearance (with anti-HBc IgG) Generated by immunization and used to monitor post-immunization responses (anti-HBc absent) May coexist with HBsAg so presence cannot be used to exclude current infection Used as a more direct and accurate measure of active HBV viral replication, which correlates with disease progression Serum HBV DNA is measured in international units (IU)/mLd as the recognized international standard or copies/ml by nucleic acid testing (NAT) technologies Used to differentiate active from inactive HBeAg-negative, and to determine need for antiviral therapy in conjunction with ALT levels and degree of liver fibrosis Used to also monitor response to therapy (a rise may indicate inadequate adherence or the emergence of resistant variants) and as a marker of infectivity. May be detectable in early infection before HBsAg, and therefore useful in early diagnosis of at-risk individuals before HBsAg appears, but depends on sensitivity of the assay Also present at low levels in the absence of HBsAg in the context of occult infection

Anti-HBc (total)

• • •

HBeAg

• • •

Anti-HBe

• • •

Anti-HBs

• • • •

HBV DNA

• • • • • •

anti-HBc: antibody to hepatitis B core antigen; anti-HBs: antibody to hepatitis B surface antigen; HBeAg: hepatitis B e antigen; HBsAg: hepatitis B surface antigen; Ig: immunoglobulin a

Occult HBV infection: HBsAg is undetectable while HBV DNA can be detected in individuals who are not in the window period; mostly anti-HBc is also detectable. Evidence of onward transmission of occult HBV infection has been indicated in the literature, but impact on morbidity and/or mortality is less well described.

However, as most antivirals used to treat HBV block DNA replication pathways (by inhibiting reverse transcription) rather than transcription/translation HBsAg pathways, HBsAg levels are minimally impacted by antivirals. A potential consequence of misinterpreting a reactive IgM anti-HBc result is that a patient with chronic HBV infection experiencing flares in liver disease may not be offered timely antiviral treatment and, although they could be re-examined several months later to confirm the original diagnosis of acute HBV infection, a substantial proportion of such patients may be lost to follow up.

d 1 IU/mL = 5.3 copies/mL; 2000 IU/mL = 10 000 copies/mL; 20 000 IU/mL = 100 000 copies/mL; 200 000 IU/mL = 1 000 000 copies/mL

4.1.5. Preventing hepatitis B infection through vaccination Vaccination of infants and, in particular, delivery of hepatitis B vaccine within 24 hours of birth is 90–95% effective in preventing infection with HBV as well as in decreasing HBV transmission if followed by at least two other doses. WHO recommends universal hepatitis B vaccination for all infants, and giving the first dose as soon as possible after birth (24). This strategy has resulted in a dramatic decrease in the incidence and prevalence of CHB among young children in regions of the world where universal infant vaccination programmes have been implemented (110, 111). Target groups for catch-up vaccination as well as other preventive strategies include young adolescents, household and sexual contacts of persons who are HBsAg-positive, and persons at risk of acquiring HBV infection, such as PWID, MSM and persons with multiple sex partners.

4.1.6. Treatment of hepatitis B infection WHO recommends antiviral agents (tenofovir and entecavir) that are active against HBV infection and have been shown to effectively suppress HBV replication, prevent progression to cirrhosis, and reduce the risk of HCC and liver-related deaths (6, 112, 113). However, in the majority of patients, treatment with these drugs does not provide cure (i.e. the person continues to have replicating virus), necessitating potentially lifelong treatment.

4.2. Hepatitis C infection 4.2.1. Epidemiology of hepatitis C infection Recent analyses of the global prevalence of HCV indicate that there may be fewer persons living with hepatitis C infection than previously estimated. A recent systematic review estimated that 110 million persons have a history of HCV infection (i.e. are HCV-antibody positive) and 80 million have chronic viraemic infection (3). Regions estimated to have a high prevalence in the general population (>3.5%) are Central and east Asia, and North Africa/Middle East; those with a moderate prevalence (1.5–3.5%) include South and South-East Asia, Sub-Saharan Africa, Latin America (Andean, central, and southern regions), the Caribbean, Oceania, Australasia, and central, eastern and western Europe; whereas low-prevalence (<1.5%) regions include Asia–Pacific, Latin America, and North America (3). Updated estimates in Africa show a HCV prevalence of 2.98%, with a higher prevalence observed in west Africa and lower in south-east Africa (114). Despite the declining incidence, a large number of persons who were infected 30–60 years ago are now dying from HCV-related cirrhosis and liver cancer, as these complications often take decades to develop. According to estimates from the Global Burden of Disease study, the number of deaths due to hepatitis C increased from 333 000 in 1990 to 499 000 in 2010 and 704 000 in 2013 (1, 5, 115), and this increase is projected to continue for several more decades, unless treatment is scaled up considerably (116).

HIV and HCV have common routes of transmission, and persons with HIV infection, in particular PWID and MSM, are at increased risk of HCV infection (60, 62, 88–93, 117). In a recent comprehensive systematic review, it is estimated that, globally, 2.3 million persons are coinfected with these two viruses, of whom 1.2 million (interquartile range [IQR] 0.9–1.4 million) are PWID (88). With the widespread use of antiretroviral therapy (ART), which reduces the risk of HIV-associated opportunistic infections, HCV-related liver disease has started to overtake AIDS-defining illnesses as a leading cause of death among people living with HIV in some high-income countries (HICs) (118).

4.2.2. Transmission of hepatitis C infection Table 4.1 provides an overview of the primary routes of transmission for HCV infection and populations most affected. There are four main routes of transmission: healthcare-associated transmission, injecting drug use, mother-to-child transmission (MTCT), and sexual transmission. In LMICs, infection with HCV is most commonly associated with unsafe injection practices, and invasive procedures in health-care facilities with inadequate infection control practices, such as renal dialysis and unscreened (or inadequately screened) blood transfusions (70–74, 77, 78, 119). Persons who received untested blood products prior to the introduction of screening of blood for HCV in (HICs) are also at risk, and WHO reports suggest that there are still 39 countries that do not routinely screen blood transfusions for bloodborne viruses (120). In middle- and high-income countries, most HCV infections occur among people who use unsterile equipment to inject drugs. PWID have a high global prevalence of infection at around 67% (46). Of the estimated 16 million people in 148 countries who actively inject drugs, 10 million have serological evidence of HCV infection (46). There is a moderate risk of MTCT of HCV which is higher in HIV-coinfected mothers (10–20%) (96). The risk of sexual transmission of HCV is also greater in HIV-positive persons, particularly MSM (88), but is low among HIV-uninfected heterosexual couples (102, 121) and MSM (122, 123). Other routes of bloodborne transmission include acquisition by health-care workers, cosmetic procedures (such as tattooing and body piercing), scarification and circumcision (84, 85, 124), and intranasal drug use. As a result of these different routes of transmission, certain groups are at higher risk of HCV infection (Table 4.1). The relative importance of these risk groups varies substantially, depending on the geographical location and population studied. Persons at risk for HCV infection are also likely to be at risk for infection with other bloodborne viruses, including HBV and HIV. Generally, HCV epidemics around the world are heterogeneous and represent mixtures of three core epidemic components (Box 4.1). However, few countries have epidemics that fall into just one of these categories – most represent some combination of all components.

Box 4.1. Global epidemic patterns of HCV infection 1. Historic infection related to past generalized HCV exposures that have since been identified and removed, i.e. “birth cohort” epidemic. These exposures include blood transfusions and medical procedures prior to the identification of HCV, or prior to the availability of HCV diagnostic screening. Following introduction of HCV screening of the blood supply in the early 1990s, the incidence of HCV fell dramatically among the general population. However, there remains a burden of prevalent, chronic HCV infection among those exposed prior to the introduction of screening of the blood supply. This epidemic pattern, in which there is a high prevalence of HCV within a given older age group, is commonly referred to as a “birth cohort” epidemic (125). While typically identified as being the infection pattern in North America and Europe, it is likely to be a component of the HCV epidemic in many countries (126). In addition, some countries have other specific historical risks that reflect past medical practices or public health campaigns unique to that country, for example, the use of reusable syringes in the populationbased campaign to treat schistosomiasis in Egypt exposed a large proportion of the population to HCV. 2. Ongoing risk of HCV transmission reflecting current behaviours and practices a) Ongoing infection related to high-risk behaviours. In certain countries, HCV transmission occurs predominantly in high-risk populations, often via common routes of transmission. Among PWID, HCV prevalence is almost universally high (ranging from 30% to 75% (46), and in many HICs, PWID drive ongoing HCV transmission. Sex workers and prisoners also have increased prevalence (presumed to be related to both drug use and perhaps sexual transmission) (127, 128), as do MSM, especially those who are HIV infected (129). b) Ongoing infection and generalized population epidemic related to suboptimal infection control and injection safety procedures in clinical settings. This pattern is related to widespread exposure, often iatrogenic, which results in high prevalence (8–10%) across all age groups. An example of a generalized exposure is the common use of reusable hypodermic syringes and needles in medical settings without adequate sterilization between uses. The primary difference between a “birth cohort” pattern and a generalized pattern of infection is the duration of time that the generalized exposure has existed and whether it has been removed or mitigated.

FIG. 4.4 Global distribution of HCV genotypes

55.5 million 14 million 3.9 million 1.7 million 485 000

HCV genotype proportion

Source: Messina J.P, Humphreys I, Flaxman A, Brown A, Cooke GS, Pybus OG et al. Global distribution and prevalence of hepatitis C virus genotypes. Hepatology. 2015;77–87.

4.2.3. Natural history of hepatitis C infection HCV is a small, positive-stranded RNA-enveloped virus with multiple genotypes and subgenotypes, and their distribution varies substantially in different parts of the world (Fig. 4.4). The availability of pangenotypic DAA regimens will increasingly obviate the need for prior genotyping, which will help expand access to HCV treatment. Hepatitis C virus causes both acute and chronic infection. Acute HCV infection is defined as the presence of certain markers of HCV infection within six months of exposure to and infection with HCV, and is characterized by the appearance of HCV RNA, HCV core antigen (p22 Ag), and subsequently HCV antibodies, which may or may not be associated with viral clearance. Antibodies to HCV develop as part of acute infection and persist throughout life. Acute infection is usually clinically silent, and is only very rarely associated with life-threatening disease. Spontaneous clearance of acute HCV infection generally occurs within six months of infection in 15–45% of infected individuals in the absence of treatment, but this varies by region and population (130). Antibodies to HCV develop as part of acute infection and persist throughout life. Almost all the remaining 55–85% of persons who do not clear HCV within six months are defined as having chronic HCV infection. Left untreated, chronic HCV infection can cause liver cirrhosis, liver failure and HCC. Of those with chronic HCV infection, the risk of cirrhosis of the liver is 15–30% within 20 years (131–133). The risk of HCC in persons with cirrhosis is approximately 2–4% per year (Fig. 4.5) (134). Clearance of infection, whether spontaneous or as a result of antiviral treatment, does not provide lasting protection from reinfection. Diagnosis of HCV infection currently consists of initial screening for evidence of past or current HCV infection with a serological assay, followed by NAT for HCV RNA (either quantitative or qualitative) to confirm the presence of HCV viraemia, and therefore chronic HCV infection.

4.2.4. Time course of serological markers for HCV infection The exact time course of virological and immunological markers of HCV infection is not well defined, particularly during the first months of infection, due to differences in each host (patient) immune response, specific properties of the infecting virus, and sensitivity of assays used to determine the appearance of HCV markers. As illustrated in Fig. 4.5, following an initial eclipse phase of 1–2 weeks when no virological or serological markers of infection may be detected, the natural course of HCV infection is characterized by the appearance of HCV RNA, then HCV core p22 Ag in the absence of an antibody response for a further 6–10 weeks. During this serological window, it has been shown that free (i.e. not complexed with antibody) HCV core antigen (HCVcAg) can be detected in a proportion of individuals. Following the development of the antibody response, HCVcAg becomes complexed with these antibodies specific for HCV.

FIG. 4.5 Approximate Time course of virological and immunological markers of HCV infection with (A) Self-resolving HCV infection, and (B) Chronic HCV infection

A Eclipse phase

Serological window

Seroconversion

Virus clearance

Waning anti-HCV

Seroreversion

B Eclipse phase

Serological window

Seroconversion/acute phase

Chronic phase

HCV RNA

Anti-HCV Anti-HCV HCV RNA

HCV Ag

months

12 24

decades

months

12 24

decades

Window period. Assays designed solely to detect antibodies to HCV inevitably have a window period of infectivity in early infection, during which antibodies may be undetectable. This window period can be shortened by utilizing assays that also include direct detection of HCVcAg (50–60 days). HCV RNA is typically not used to determine exposure to HCV, in spite of its short window period (1–2 weeks after the onset of acute infection) primarily because of cost (135). There are increasing reports of occult HCV infection, i.e. HCV RNA detectable in the absence of any serological markers (i.e. HCV seronegative) (136–138) which may be due to underlying immunosuppression in, for example, HIV-infected populations.

4.2.5. Prevention of hepatitis C infection In the absence of a vaccine for hepatitis C, prevention of HCV infection depends upon reducing the risk of exposure to the virus. This is challenging because of the various routes of transmission and the different populations that are affected. Globally, most HCV infections occur in health-care settings as a result of inadequate infection control procedures. WHO has published guidelines with recommendations for preventing health-care-associated HCV infection, and for screening of blood products (20, 30, 139). Universal access to safe blood transfusion requires the implementation of key strategies to ensure access to a safe and sufficient blood supply, including 100% quality-assured testing of donated blood (139). Joint WHO–UNODC guidance recommends a comprehensive package of harm reduction interventions, which comprise nine harm reduction activities specifically for PWID, including the provision of sterile injecting equipment (140), alongside WHO guidance on prevention of viral hepatitis B and C transmission among PWID (28).

4.2.6. Treatment of hepatitis C infection A new class of medicines, called direct-acting antivirals (DAAs), have transformed the treatment of HCV, with regimens that can be administered for a short duration (as short as eight weeks), resulting in cure rates higher than 90%, but

are associated with fewer serious adverse events than the previous interferoncontaining regimens. WHO updated its hepatitis C treatment guidelines in 2016 to provide recommendations for the use of new DAAs (5) (see Web annex 3). There still remains some variation in recommended HCV treatment regimens and duration of therapy by genotype. This requirement to determine a patient’s genotype prior to treatment will soon change when antiviral agents that are active against all genotypes (referred to as pangenotypic) are licensed.

5. BACKGROUND – DIAGNOSTICS FOR TESTING FOR HEPATITIS B AND C INFECTION 5.1. Types of viral hepatitis assays Serological assays are typically used as the first line of the testing strategy to screen for exposure to a virus because of their relatively low cost (compared to NAT), and are therefore used to rule in all individuals who might potentially be infected with HCV or HBV. Serological assays detect the host immune response (antibodies to HCV) or a viral antigen (HBsAg, HCVcAg). They are based on the immunoassay principle, and are available in the form of rapid diagnostic tests (RDTs) or laboratory-based enzyme immunoassays (EIAs), chemoluminescence immunoassays (CLIAs) and electrochemoluminescence immunoassays (ECLs). In contrast, NAT technologies are typically used to detect the presence of the virus, determine if the infection is active and if the individual would benefit from antiviral treatment. NAT technologies are also used to determine when antiviral treatment should be discontinued (due to non-response or resistance) or to confirm virological cure (HCV) or effective suppression (HBV).

5.2 Serological assays 5.2.1. Rapid diagnostic tests Rapid diagnostic tests (RDTs) are single-use disposable assays that are provided in simple-to-use formats that generally require no additional reagents except those supplied in the test kit. They are read visually and can give a simple qualitative result in under 30 minutes. Due to their simplicity, cost and rapid turnaround time, they can be performed by trained lay providers or health-care workers, without the need for venepuncture for specimen collection. Quality-assured RDTs are therefore particularly useful in settings where conventional laboratory-based testing services are not available or accessible. They can also be used in outreach programmes (e.g. prison services, prevention and treatment services for people who use drugs). Most RDTs can be performed with capillary whole blood collected by a fingerstick procedure using a lancet, but many have also been developed for use with venous whole blood, serum or plasma. Certain ones have been validated for use with oral fluid specimens. It is critical to always refer to the manufacturer’s instructions for use for specific recommendations on specimen collection. Rapid tests are generally not suitable for testing large numbers of blood samples. The reading of results is dependent on subjective evaluation and no permanent record of the original test results can be kept.

5.2.2. Laboratory-based immunoassays Most laboratory-based serological immunoassays (EIAs, CLIAs and ECLs) detect antibodies, antigens or a combination of both and differ only in the mode of detection of immune complexes formed. A cut-off value, usually determined by the manufacturer of the assay, specifies the point at which the results are considered to be reactive, and therefore, EIA results are generally reported as optical density divided by the assay cut-off (OD/CO) values. These types of assays are best suited for and most cost–effective to perform in settings with a high throughput of specimens (in excess of 40 per day). They are meant for laboratory- or facilitybased testing rather than for use in the community, where infrastructure (electricity, cold storage, climate-controlled rooms) and skilled staff are consistently available, as cold-chain storage of test kits and the use of precision pipettes are usually required. These assays are typically used only with serum or plasma specimens, and therefore require phlebotomy to collect an appropriate specimen. These assays may be performed either manually or on non-dedicated automated assay or specific dedicated automated systems. Simple immunoanalysers automate a number of the processes and as such require less hands-on time than a manually run EIA. They can therefore be used in range of different situations from high-throughput laboratories for the screening of large numbers of samples with full automation, to medium-sized laboratories with semi-automation, to small laboratories, such as those in remote areas, which conduct a small number of tests manually.

5.2.3. Confirmatory assays For HBsAg – neutralization assays are used to confirm if observed antigen reactivity is neutralizable upon repeat testing with the same specimen using a neutralization step in the laboratory-based immunoassays, with a specific anti-HBs-containing reagent in the same assay. The result is confirmed when this neutralization reagent can abolish reactivity in the assay in comparison with a control reaction. For anti-HCV – line immunoassays or immunoblots are serological techniques to confirm the presence of antibodies to HCV that have already been detected by other serological assays. The use of confirmatory assays should be able to provide a definitive result, although these assays are more expensive than other assays and are prone to high rates of indeterminate results. These assays only confirm serostatus and cannot be used to diagnose viraemic active HCV infection.

5.3. Nucleic acid testing technologies These assays detect the presence of viral nucleic acid – DNA or RNA – through targeting a specific segment of the virus, which is then amplified. The amplification step enables the detection of low levels of the virus in the original specimen, which might not otherwise have been detectable. Laboratory-based technologies for NAT require

sophisticated equipment, rigorous laboratory conditions and specimen collection, and highly trained staff who can perform precision steps and avoid contamination. Not all NAT technologies detect all genotypes or subtypes equally well, unless they are optimized to do so. Newly developed NAT technologies that are simpler and more robust are intended for use at or near the point of care, and may avoid some of the logistical and technical disadvantages of laboratory-based NAT technologies. In addition to NAT assays that target a single virus, multiplex NAT screening assays have been developed, which can detect DNA or RNA from multiple viruses simultaneously.

5.4. Choice of serological assays Table 5.1 describes the advantages and disadvantages of RDTs and laboratorybased immunoassays. The choice of assay format will depend on a variety of factors, most importantly, performance characteristics (sensitivity and specificity), cost, ease of use and the characteristics of the testing site, such as storage facilities, infrastructure, and level of staff skills. Chapter 15 provides further details on how to set up laboratory services for viral hepatitis testing and selection of an assay, and how to assure the quality of testing.

5.5. Selection of one- or two-assay serological testing strategy A testing strategy defines the sequence of tests to be followed for a specific testing objective (i.e. to identify infected and non-infected individuals), taking into consideration the anticipated prevalence of HBsAg or HCV antibody in the population(s) to be tested. WHO recommends the use of standardized testing strategies to both maximize the accuracy of HBsAg or HCV antibody testing while simplifying the process through streamlining procurement and training (11). The choice between a one-assay versus two-assay serological testing strategy will depend on the seroprevalence in the population to be tested and diagnostic accuracy (sensitivity and specificity) of the assays used. In these guidelines, we refer to the use of testing strategy only in the context of serological testing and the use of a oneor two-serological assay testing strategy, though it is recognized that other sources refer to the use of a single HCV RNA NAT or core antigen as a one test strategy to replace the need for a two-step process of serological testing followed by NAT. One-assay serological testing strategy A one-assay serological testing strategy (Fig. 5.1) is when a single serological test is performed. If the test result is reactive, a “compatible with positive infection” status is reported. If the initial test result is non-reactive, a “negative infection” status is reported. This testing strategy efficiently rules out most uninfected individuals

correctly, and rules in those who are likely to be infected and therefore in need of further HBV DNA and HCV RNA NAT testing and staging of liver disease using NITs and clinical evaluation. This testing strategy is particularly suitable for highprevalence settings due to the relatively higher positive predictive values (PPVs), but needs a highly sensitive and specific assay to maintain acceptable predictive values. Two-assay serological testing strategy Two-assay serological testing strategy (Fig. 5.2) differs in that two different assays are used sequentially, to improve the PPV of the testing strategy, and so reduce the number of individuals inappropriately referred on to more specialist services. This can be achieved by either (i) repeating the serological test using a different assay of similar sensitivity, or (ii) in the case of HBsAg, performing a neutralization test using a specific anti-HBs-containing reagent in the same firstline assay after appropriate dilution of the specimen under test. If the first test result is non-reactive, a “negative infection” status is reported. If both test results are reactive, the status is reported as: “presumptive positive status infection for further diagnostic testing”. If the second test result is nonreactive, the status is reported as “infection inconclusive; requires additional testing”. If the second assay is less sensitive than the first, then it is likely that some true positives would be discarded if negative on the second test. Fig. 5.1. One-assay serological testing strategy Fig. 5.2. Two-assay serological testing strategy Assay 1 Assay 1

A1+ (Reactive) Report positive

A1(Non–reactive) Report negative

A1+ Assay 2

A1Report negative

Compatible with infection Further testing for viraemic infection

No evidence of infection

A1- A2+ Report positive

A1+ A2Report inconclusive

No evidence of infection

Compatible with infection Further testing for viraemic infection

Inconclusive result Further testing as appropriate

TABLE 5.1 Advantages and disadvantages of different assay formats Assay Laboratorybased immunoassays (EIA, CLIA, ECL) Advantages • Currently superior clinical/ diagnostic and analytical sensitivity/ specificity for HBsAg High throughput possible (>40 per day per operator) High throughput greater when using automated immunoanalysers Objective, automated reading of results, but not for line blots or simple assays Within-assay procedural quality control Accessible at the lowest level of the health-care system (including community settings) Does not specifically require laboratory facilities May be carried out by trained lay providers and health-care workers, as well as laboratory technicians Can be used with less invasive specimens that do not require venepuncture such as capillary whole blood or oral fluid If testing at or near to point of care, same-day results are possible, which may reduce number of individuals that are lost to follow up and therefore do not receive their test results Devices can be stored at 2–30 °C May be used at or near the point of care May be carried out by trained lay providers and health-care workers, as well as laboratory technicians Can be used with less invasive specimens that do not require venepuncture such as capillary whole blood Devices can be stored at 2–30 °C Disadvantages • Requires laboratory facilities, equipment, e.g. EIA plate washers, readers, incubators or immunoanalysers or randomaccess analysers. Requires trained laboratory technician Reagents require refrigeration Requires venepuncture to obtain specimen Time to result ~3 hours and generally batched as one run if manual EIA Lower clinical and analytical sensitivity/specificity for HBsAg Less sensitive in certain populations such as immunosuppressed, including HIV-positive individuals Ineffective within-assay quality control, i.e. most RDTs do not control for specimen addition Lack of test kit external control reagents for quality control with most RDTs, but some exceptions, e.g. Oraquick Stability at room temperature is impacted by environmental factors, e.g. heat, humidity, storage conditions Subjective reading and interpretation of results Requires manual transcription of testing results into laboratory logbook/testing register, partially mitigated by automated RDT readers Currently requires laboratory facilities and equipment, but this may not apply to future point-of-care options Requires trained laboratory technician Reagents require refrigeration Requires venepuncture to obtain specimen Time to result ~3 hours and generally batched as one

• • • •

Rapid diagnostic tests (RDTs)

•

• •

•

• •

•

Nucleic acid testing (NAT) technologies

• •

•

• • • •

•

PART 2: RECOMMENDATIONS • • Who to test for HBV and HCV infection How to test for chronic hepatitis B infection • • • serology and presence of viraemia monitoring of HBV treatment response

How to test for chronic hepatitis C infection • • serology and presence of viraemia monitoring of HCV treatment response

• •

Use of dried blood spot sampling Linkage to care and treatment

6. WHO TO TEST FOR CHRONIC HEPATITIS B OR C INFECTION – testing approaches and service delivery 6.1. Recommendations WHO TO TEST FOR CHRONIC HBV INFECTION Testing approach and population General population testing

Recommendations* 1. In settings with a ≥2% or ≥5%1 HBsAg seroprevalence in the general population, it is recommended that all adults have routine access to and be offered HBsAg serological testing with linkage to prevention, care and treatment services. General population testing approaches should make use of existing community- or health facility-based testing opportunities or programmes such as at antenatal clinics, HIV or TB clinics. Conditional recommendation, low quality of evidence 2. In settings with a ≥2% or ≥5%%1 HBsAg seroprevalence in the general population, it is recommended that HBsAg serological testing be routinely offered to all pregnant women in antenatal clinics2, with linkage to prevention, care and treatment services. Couples and partners in antenatal care settings should be offered HBV testing services. Strong recommendation, low quality of evidence 3. In all settings (and regardless of whether delivered through facility- or communitybased testing), it is recommended that HBsAg serological testing and linkage to care and treatment services be offered to the following individuals: • Adults and adolescents from populations most affected by HBV infection3 (i.e. who are either part of a population with high HBV seroprevalence or who have a history of exposure and/or high-risk behaviours for HBV infection); • Adults, adolescents and children with a clinical suspicion of chronic viral hepatitis4 (i.e. symptoms, signs, laboratory markers); • Sexual partners, children and other family members, and close household contacts of those with HBV infection5; • Health-care workers: in all settings, it is recommended that HBsAg serological testing be offered and hepatitis B vaccination given to all health-care workers who have not been vaccinated previously (adapted from existing guidance on hepatitis B vaccination6) Strong recommendation, low quality of evidence 4. In all settings, screening of blood donors should be mandatory with linkage to care, counselling and treatment for those who test positive.

Routine testing in pregnant women

Focused testing in most affected populations

Blood donors Adapted from existing 2010 WHO guidance (Screening donated blood for transfusion transmissible infections7)

Abbreviations: HBsAg: hepatitis B surface antigen; PWID: people who inject drugs; MSM: men who have sex with men *The GRADE system (Grading of Recommendations, Assessment, Development and Evaluation) was used to categorize the strength of recommendations as strong or conditional (based on consideration of the quality of evidence, balance of benefits and harms, acceptability, resource use and programmatic feasibility) and the quality of evidence as high, moderate, low or very low. A threshold of ≥2% or ≥5% seroprevalence was based on several published thresholds of intermediate or high seroprevalence. The threshold used will depend on other country considerations and epidemiological context. Many countries have chosen to adopt routine testing in all pregnant women, regardless of seroprevalence in the general population, and particularly where seroprevalence ≥2%. A full vaccination schedule including birth dose should be completed in all infants, in accordance with the WHO position paper on hepatitis B vaccines 2009.6 3 Includes those who are either part of a population with higher seroprevalence (e.g. some mobile/migrant populations from high/intermediate endemic countries, and certain indigenous populations) or who have a history of exposure or high-risk behaviours for HBV infection (e.g. PWID, people in prisons and other closed settings, MSM and sex workers, HIV-infected persons, partners, family members and children of HBV-infected persons). 4 Features that may indicate underlying chronic HBV infection include clinical evidence of existing liver disease, such as cirrhosis or hepatocellular carcinoma (HCC), or where there is unexplained liver disease, including abnormal liver function tests or liver ultrasound. 5 In all settings, it is recommended that HBsAg serological testing with hepatitis B vaccination of those who are HBsAg negative and not previously vaccinated be offered to all children with parents or siblings diagnosed with HBV infection or with clinical suspicion of hepatitis, through community- or facility-based testing. 6 WHO position paper. Hepatitis B vaccines. Wkly Epidemiol Rec. 2009;4 (84):405–20. 7 Screening donated blood for transfusion transmissible infections. Geneva: World Health Organization; 2010. 1 2

General population testing

Birth cohort testing

6.2. Background Viral hepatitis testing can be delivered to different populations and in different settings as part of general population testing, and/or a focused testing approach in most affected or high-risk populations, delivered through either health facilitybased or community-based testing. Chapter 17 provides additional details on the different facility- and community-based testing approaches available. Chapter 18 provides additional guidance on testing in specific populations.

Different hepatitis testing approaches There are several possible approaches to testing for HBV and HCV infection. 1. General population testing. This approach refers to routine testing throughout the entire population without attempting to identify high-risk behaviours or characteristics. It means that all members of the population should have potential access to the testing services. This approach might be indicated for those countries with an intermediate or high HBV or HCV seroprevalence. At present, only Japan recommends HCV testing for all individuals once in their lives regardless of demographics or specific behavioural risk. 2. Focused or targeted testing of specific high-risk groups. This approach refers to testing of specific populations who are most affected by hepatitis B or C infection, either because they are part of a population with high HBV or HCV seroprevalence (such as some migrant populations and some indigenous populations), or have a high risk of acquisition because of risk behaviours and/or exposures. This includes PWID, people in prisons and other closed settings, MSM and sex workers, HIV-infected persons, partners or family members of infected persons, and health-care workers. It may also involve testing on the basis of clinical suspicion of viral hepatitis (i.e. symptoms, signs or abnormal liver function tests or ultrasound scan). 3. Routine antenatal clinic (ANC) testing. This means routine testing of pregnant women especially in settings where there is an intermediate or high seroprevalence, to identify women in need of antiviral treatment for their own health and additional interventions to reduce MTCT of viral hepatitis. 4. “Birth cohort” testing. This approach means routine testing among easily identified age or demographic groups (i.e. specific “birth cohorts”) known to have high HCV prevalence due to past generalized exposures that have since been identified and removed. General one-time screening among this population avoids the need to identify risk behaviour. Most countries have at least some component of a “birth cohort” epidemic profile for HCV. Use of a birth cohort approach to HCV testing is currently recommended only in the United States. 5. Blood donor screening. WHO already recommends universal blood donor screening for viral hepatitis in order to prevent transmission of bloodborne viruses to the recipient

(20). However, at present, this is rarely accompanied by the HBsAg- or HCV antibody -positive donors being informed of this positive result, counselled and linked to care for clinical evaluation and treatment (141).

Service delivery of testing approaches (health facility- or community-based) The testing approaches described above can be offered and delivered using both healthfacility and/or community-based testing services. Health-facility-based testing includes primary care clinics, inpatient wards and outpatient clinics, including specialist dedicated clinics such as HIV, STI and TB clinics, in district and provincial or regional hospitals as well as their laboratories, and in private clinical services. Community-based testing can be offered and delivered using outreach (mobile) approaches in general and key populations; home-based testing (or door-to-door outreach); testing in workplaces, places of worship, parks, bars and other venues; in schools and other educational establishments; as well as through campaigns (e.g. screening for HIV or malaria alongside that for noncommunicable diseases such as diabetes and hypertension). Although many of these approaches were developed to increase the coverage and impact of HIV testing (11), they are equally applicable to the delivery of hepatitis testing.

6A TESTING APPROACHES TO DETECT CHRONIC HEPATITIS B 6.3. Summary of the evidence The evidence base for different HBV testing approaches (general population or focused testing) remains very limited, especially in LMICs, and relies largely on observational data and modelling. Although there are descriptive data showing that focused testing can increase the uptake of HBV testing, and detection rate of CHB cases, data showing impact on patient-important outcomes are limited (142). There is also a lack of evidence on and uncertainty regarding how successful focused testing is in reaching specific populations. For these reasons, a formal systematic review of the impact and cost–effectiveness of different testing approaches was precluded, and an updated narrative review of evidence was undertaken. The overall quality of evidence was therefore rated as low. There were 32 published studies of which nine studies met the inclusion criteria – all but one study were from HICs with low HBV prevalence (see Web annex 5.1). Two studies evaluated the cost–effectiveness of offering testing and treatment to the general population (one from the United States (143) and the other from west Africa (144)), and seven studies had examined targeted riskgroup testing in migrant populations (145–150) or “high-risk” groups (151). Several studies were based on modelling simulations using hypothetical data. Various outcome measures were used, including cost per quality-adjusted lifeyear (QALY) gained, cost per life-year (LY) saved and cost per case tested.

General population testing. The two studies performed in the United States and West Africa showed that offering HBsAg testing to the general population with provision of antiviral treatment in those eligible is cost–effective in both high-income (143) and low-income settings (144) , even down to a population prevalence as low as 0.3% and 1.5%, respectively. In addition, the feasibility of large-scale testing and treatment in sub-Saharan Africa based on real-world cost and effectiveness data was demonstrated by the PROLIFICA (Prevention of Liver Fibrosis and Liver Cancer in Africa) study in west Africa (152). This study screened almost 10 000 adults for HBsAg using an active outreach method at the community level in the Gambia and Senegal, followed by full clinical assessment of those found HBsAg positive, and provision of antiviral treatment if they met eligibility criteria. They showed this community-based screen-and-treat strategy was cost-effective compared to the status quo.

Focused risk-based testing. Testing and treatment of migrant or refugee populations in HICs was also found to be a cost-effective intervention in seven studies from Canada, the United States and Europe (145–150).

Pregnant women. Although the cost–effectiveness of HBsAg testing of pregnant women in ANC to reduce MTCT and benefits to the child has been addressed in several studies, there were none identified that considered interventions and antiviral treatment for the benefit of the mother to reduce her risk of progression of liver disease.

Drivers of cost–effectiveness. These analyses identified several key drivers of cost–effectiveness for countries to consider when planning testing approaches. These include: (i) Drug and testing costs. The key driver of the cost–effectiveness of a test-and-treat strategy reported is the cost of the antiviral drug (144, 146, 147), and to a lesser extent testing costs (145, 147). In the PROLIFICA study, despite an active community-based screening campaign, testing costs were low (US$ 7.43 per person offered screening) and the intervention remained costeffective even if there was a threefold increase in testing costs (144). (ii) Linkage to care and adherence. Adherence to treatment and linkage to care were reported as key drivers of cost–effectiveness in several studies (148), but not in the PROLIFICA study (152). (iii) Uptake of testing was not identified as a key driver of incremental cost-effectiveness ratio (ICER) in any of the studies. However, this does not imply that high participation levels in screening are not important. The implication of this result is that it is likely to be worthwhile performing screening and providing treatment, even if participation in screening may be low, in part because testing costs are low relative to the costs and health benefits of treatment for those who are infected. HBsAg prevalence also had a relatively small influence on cost–effectiveness across a wide range of prevalence levels examined.

6.4. Rationale for the recommendations on testing approaches for HBV infection In developing recommendations on which populations to test and what testing approaches to use, the Guidelines Development Group first considered the primary goals of testing (153): (i) to identify those in greatest need of treatment to reduce morbidity and mortality from HBV-related chronic liver disease; (ii) to reduce the risk of acquisition of disease, by vaccinating those who do not have HBV infection but remain at risk; and (iii) to reduce the risk of mother-to-child vertical transmission and so have benefits that extend beyond the person tested to others. These considerations were then balanced with the need for recommendations that are feasible and implementable by health programmes in LMICs. Overall, there was a very limited evidence base for the impact of different testing approaches (general population or focused high-risk) as well as for different settings (community- versus health facility-based). Therefore, recommendations were formulated based on consideration of evidence mainly from cost–effectiveness analyses together with data on HBsAg seroprevalence in different settings and populations, and in the general

population with considerations of feasibility and cost. The caveats of extrapolating costeffectiveness data from HICs to LMICs were recognized. The Guidelines Development Group recommended the use of three key testing approaches: routine testing in the general population; focused testing in most affected populations because of higher-risk behaviours or exposures; and routine ANC testing. These can be implemented both in health-care facilities and in the community, as appropriate to the local epidemiology and context.

Balance of benefits and harms General population testing. In settings where there is an HBsAg prevalence ≥2% in the general population, focused testing in higher-risk populations alone will be insufficient to identify many of those infected and in need of treatment. Additional general population testing approaches that use community- as well as health facility-based testing programmes are therefore needed to increase the coverage and impact of HBV testing. Although general population testing was estimated to be cost–effective down to prevalence levels <1%, the Guidelines Development Group proposed a higher threshold of ≥2% to reflect the wellaccepted thresholds for defining intermediate (≥2%)/high (≥5%) seroprevalence (154). The Guidelines Development Group recognized that the threshold used by countries will depend on other country considerations and epidemiological context. For this reason, a conditional recommendation was made. Focused risk-based testing in populations with high-risk behaviour or exposure to HBV infection. Certain populations are well recognized to be at high risk of acquisition and transmission of HBV infection (Table 4.1), and therefore should be prioritized for testing in all epidemic settings. These include people living with HIV, PWID, MSM, sex workers, people in prisons and other closed settings, some mobile/migrant populations from high/intermediate-endemic countries, some indigenous populations, children born to HBsAg-positive mothers, especially if they did not receive timely infant vaccination, and other family members, sexual partners and close household contacts of those with HBV infection; and health-care workers. In high-endemic settings, a clinically guided testing approach among adults and children with a clinical suspicion of chronic viral hepatitis (i.e. clinical symptoms or signs, or abnormal liver function tests or ultrasound scan) will identify a larger proportion of infected persons.

Key benefits of focused testing 1. Focused testing in health facilities can successfully increase the uptake of viral hepatitis testing, case detection rate, and referrals to specialist-level care and other important services. 2. Focused testing approaches can use existing opportunities and infrastructure for health facility-based testing (HIV, STI, and TB outpatient clinics, drug treatment programmes, primary care settings, inpatient and outpatient settings), as well as community-based testing.

3. Focused testing of high-prevalence populations or of those in settings where there is a large proportion of such persons (e.g. harm reduction and drug treatment services for PWID) or a clinically guided approach based on clinical suspicion is likely to be associated with higher rates of case–finding. This approach will generally be cost–effective compared to generalized testing, especially in low- and concentratedepidemic settings. 4. It is recognized that many high HBV-prevalence countries currently lack the resources to undertake general population screening, and therefore focused risk-based testing may be more readily feasible and cost–effective, particularly if it makes use of existing health-facility infrastructure and staff. Despite the limited formal evaluation of focused testing in high-risk groups and low quality of evidence, a strong recommendation was made because of the overall benefits of focused testing approaches.

Pregnant women – routine testing in antenatal clinics. The Guidelines Development Group strongly recommended routine HBsAg testing in ANC, despite limited or low-quality evidence, for several reasons. 1. To benefit their offspring through interventions to significantly reduce MTCT of HBV infection (6). This is because in high-prevalence, resource-limited settings, HBV is mainly transmitted through MTCT and early childhood horizontal transmissions. Infants born to HBV-infected mothers are at high risk for both acquisition of HBV infection and development of chronic infection (90%). Therefore, key interventions in this group could significantly reduce the burden of disease in the long term. To enable women to have knowledge of their HBV serostatus (together with their offspring and partners), allowing them to benefit for their own health through linkage to assessment and treatment services. Although a systematic review of cost–effectiveness studies on routine antenatal testing in LMICs was not undertaken, this would likely be cost-effective, since testing of mothers for HBV infection has benefits for both the child (reduced transmission) and mother (reduced morbidity). There is already universal HIV testing in ANC which has proved feasible and acceptable in many countries (6, 11, 153), and addition of HBV testing would be relatively low cost. Although many countries recommend routine screening of women for HBV infection in ANC, the proportion who are screened in many LMICs remains low (157).

Couples and partner testing in ANC. HIV testing of the partners of women attending ANC is now a focus in 21 priority countries aiming for elimination of MTCT (eMTCT) of HIV. Since these countries are also all highly endemic for HBV, this provides a unique

opportunity to integrate concurrent HBV testing for partners of women with CHB, or chronic HCV infection if risk factors are present, despite the lack of specific evidence for couples and partner testing for hepatitis in ANC.

Blood donor screening. WHO already recommends blood donor screening for HBV, HCV, HIV and syphilis in order to prevent transmission of bloodborne viruses to the recipient (20). However, this is rarely accompanied by the HBsAg or HCV antibody positive donor being informed of this positive result, counselled and linked into care for clinical evaluation and treatment (141). As part of the PROLIFICA study in west Africa, in addition to HBsAg testing and treatment in the community, blood donors who had tested HBsAg positive at the blood bank were linked to specialist care (152). A higher proportion of blood donors were HBsAg positive and requiring treatment, but had a lower rate of linkage to care. Although a formal cost–effectiveness analysis was not done, these factors are likely to make testing, linkage and treatment of blood donors even more cost-effective compared to community-based testing. However, as blood donors constitute only a small fraction of the population, this strategy is likely to be limited in its reach and population-level effectiveness, and probably should be seen as a complementary, rather than an alternative to a wider screening strategy.

Acceptability, values and preferences A values and preferences survey of 104 stakeholders from 43 (20 high-income, 23 lowand middle-income) countries provided additional strong support for testing of specific populations: blood donors (>85%), children born to HBV-infected mothers (75%), persons living with HIV (65%), pregnant women (78%), MSM (45%), sex workers (45%), prisoners (25%) and those chronically ill (around 25%). General population testing for HBV infection was supported by only one third of respondents.

6B TESTING APPROACHES TO DETECT CHRONIC HEPATITIS C 6.5. Summary of the evidence A systematic review and meta-analysis of the impact and cost–effectiveness of different HCV testing approaches (general population or focused testing) was precluded by the limited number of available studies and because of the heterogeneity of settings and populations studied and outcomes measured. Therefore, a narrative review was undertaken in different settings, alongside consideration of recent systematic reviews of HCV seroprevalence in different populations. The three main testing approaches evaluated were (i) routine testing throughout the entire population; (ii) focused or targeted testing of the highest-risk groups; and (iii) routine testing among specific birth cohorts. The overall quality of evidence was rated as low. Overall, there were 31 relevant studies based on a previously published systematic review (158) and 12 additional studies identified in an updated search (159–192) (see Web annex 5.2). The majority of studies were from Europe or the US, and very few from LMICs. Fourteen studies evaluated testing in the general population (125, 159–171); 13 in PWID and STD clinics (174–177, 179–181); three in recipients of blood transfusions (160, 161, 182); one among HIV-infected MSM (183); two among pregnant women (184, 185), and two in other populations (186, 187).

Focused testing. Focused testing of PWID, people in prisons or closed settings and HIV-infected MSM was shown to be cost-effective in all settings (159, 176, 180, 183, 188). This was the case among PWID even when the studies assumed poor follow-up rates, limited access to therapy (159, 180) and a high risk of reinfection. The higher the treatment rates, the greater the population impact, and the more cost–effective HCV case-finding becomes (189). Among prisoners, targeting testing to those prisoners with a history of injection drug use further improved cost– effectiveness (176). Among HIV-positive MSM population (183), cost–effectiveness was dependent on appropriate linkage to effective therapy and retention in care.

“Birth-cohort” testing. Most countries have at least some component of a “birth cohort” HCV epidemic (i.e. of easily identified age or demographic groups known to have a higher HCV prevalence), and several cost–effectiveness studies from the US and Portugal show that birth cohort testing is cost effective when compared to riskbased screening or current testing approaches (125, 166, 168, 190).

Routine testing in the general population. A major limitation of existing cost–effectiveness studies of testing in the general population is that they were conducted based on the use of interferon-based regimens and not using the new

DAA curative treatments, and in HICs (125, 151, 160, 162, 163, 165, 167, 169, 170). Only one cost–effectiveness study has been undertaken in a LMIC – in Egypt, which has a very high prevalence of disease. Routine testing was shown to be cost-effective even when treatment was based on use of pegylated interferon (PEG-IFN) and ribavirin (PEG-RBV) (191).

Drivers of cost–effectiveness. In all analyses, the cost–effectiveness of testing for HCV was most sensitive to variations in prevalence, treatment efficacy (i.e. the replacement of IFN/RBV with significantly improved efficacy of DAAs), progression rates from chronic HCV to cirrhosis, and levels of linkage to care and treatment (164, 165, 170). It was relatively insensitive to costs of screening and treatment. Based on this narrative review of heterogeneous studies of cost–effectiveness of testing approaches from HICs, the overall quality of evidence was rated as low.

6.6. Rationale for the recommendations on testing approaches for HCV infection The Guidelines Development Group recognized that HCV epidemics around the world are heterogeneous but are largely represented by mixtures of three main epidemic patterns for which a specific testing approach is appropriate. These are as follows: 1. Infection related to high-risk behaviours – requiring focused or targeted testing in the highest-risk groups; 2. Infection related to past generalized exposures that have since been identified and removed (i.e. “birth cohort epidemic”) – requiring routine testing among specific birth cohorts that are readily identified and that have a high prevalence of HCV infection; 3. Generalized population epidemic with high prevalence generally related to a widespread, often iatrogenic, exposure – requiring routine testing throughout the entire population. Few countries have epidemics that fall into one of the above three profiles. Rather, the majority have mixed epidemic profiles, with some combination of all these components. Determining the optimal strategic mix of HCV testing approaches to increase the diagnosis rate, and in particular, the approach to testing outside of high-risk risk groups will depend on a country’s unique HCV epidemic profile (see chapter 19). The lack of evidence from LMICs on evaluation of different testing approaches was noted. Testing in high-risk behaviour groups and in settings with a large proportion of patients such as PWID, MSM, prisoners, HIV-infected persons and commercial sex workers was cost–effective in all settings. The best approach to testing outside of high-risk risk groups depends on a country’s unique HCV epidemiology. Most countries have at least some component of a “birth cohort” epidemic, and “birth cohort” testing is likely to be cost–effective in most settings. In most epidemic settings, routine screening of the entire population may not be cost–effective.

Balance of benefits and harms Focused risk-based testing. The Guidelines Development Group considered that those specific populations at the highest risk of acquisition and transmission of HCV such as PWID, people in prisons and other closed settings, MSM and sex workers should be prioritized for testing, as this was both cost–effective and had a high yield of case-finding. In settings with a high prevalence, this also means focused testing of adults and children with a clinical suspicion of chronic viral hepatitis infection (i.e. clinical symptoms or signs of cirrhosis or HCC, or abnormal liver function tests or ultrasound scan). Other higher-risk groups for focused testing include persons who have had tattoos, body piercing or scarification, unsafe medical procedures, received blood products in countries where screening of blood is not carried out routinely, as well as partners and close contacts of people with HCV infection. The Guidelines Development Group recognized that the priority groups will differ across countries and settings, and that it will be important to ensure adequate linkage to care after diagnosis. Children. The Guidelines Development Group also considered that screening was indicated for children born to mothers with HCV infection (especially if also HIV infected) because of an increased risk for MTCT after 18 months of age.

Key benefits of focused testing 1. Focused testing in health facilities can successfully increase the uptake of viral hepatitis testing, case detection rate, and referral to specialist-level care and other key services. 2. Focused testing of these populations can be offered in high-prevalence settings such as harm reduction and drug treatment services for PWID. Other existing opportunities for health-facility -based testing can also be used (e.g. dedicated HIV, STI and TB outpatient clinics, and other primary care, outpatient and inpatient settings), as well as testing in the community. 3. A clinically guided testing approach is also likely to identify a larger proportion of people with HCV in highly endemic settings and therefore result in a lower cost per positive person found.

Risks of focused testing. Although ascertaining high-risk behaviours is a very effective way of identifying persons for testing, many people are unwilling to admit to stigmatizing behaviours, and health-care providers are also reluctant to ask (or have too limited time). As a result, medical records capture this information poorly, as the use of electronic medical records to flag high-risk persons for testing is limited.

Birth-cohort testing. The best approach to testing outside of groups with high risk behaviour or exposure depends on a country’s unique HCV epidemiology. For example, in many settings, unsafe injection practices will probably have more of an impact on HCV prevalence than illicit

injecting drug use. The Guidelines Development Group concluded that whenever there is an easily identified demographic group that has a high HCV prevalence (e.g. all individuals born in a certain time period), routine testing for HCV within that cohort, i.e. “birth cohort” testing will likely be cost–effective and should be considered. This will largely apply to those countries where routine screening of the blood supply for HCV in the 1990s and improvements in injection safety practices have since removed the exposure risk. A conditional recommendation was made mainly because of low quality of evidence.

Key benefits of birth cohort testing 1. Recent studies in the US showed birth cohort screening to be cost–effective when compared with risk-based screening. While typically identified as being the infection pattern in North America and Europe, many countries have at least some component of “birth cohort” epidemic in their HCV epidemiology, and therefore “birth cohort” testing is likely to be cost–effective in most settings. 2. A key advantage of this generally one-off screening approach is that it avoids the need to identify specific behavioural risks as the basis for screening, because providers may not be skilled at identifying high-risk behaviours, and individuals may not remember that they received a blood product, or report to previous risk-taking behaviour on direct questioning.

Key risks of birth cohort testing. More recent data suggest that a significant proportion of the HCV-infected population is not captured as part of birth cohort screening (192). A further challenge is that this approach requires reliable data on both the age distribution of the population and prevalence according to age, which is not available in most countries.

General population testing. Routine screening for HCV in the general population was generally not considered cost–effective outside specific settings with high general population prevalence. The application of a one-off birth cohort screening approach to testing the general population will be more widely applicable. Therefore, a conditional recommendation was made to support consideration of general population testing in intermediate- and high-prevalence settings.

Acceptability, values and preferences A values and preferences survey of 104 stakeholders from 43 (20 high-income, 23 low- and middle-income) countries identified the following target populations as priority for hepatitis C testing: blood donors (>85%), children born to HCV-infected mothers (55%), persons living with HIV (50%), pregnant women (40%), MSM (25%), prisoners (25%), sex workers (<10%), and those chronically ill (25%). General population testing for HCV infection was supported by 30% of respondents.

6C SERVICE DELIVERY OF HEPATITIS B AND C TESTING The Guidelines Development Group recommended the use of various health facility or community-based testing opportunities in the general population or focused on high-risk groups because of the overall benefits of these testing approaches. However, this was conditional because of the currently limited evidence base. Chapter 17 provides specific examples of the many different types of facility and community-based testing approaches.

6.7 Rationale for the recommendations on community-based testing Balance of benefits and harms Strategies for the delivery of general population or focused testing approaches include door-todoor/home-based testing and mobile outreach campaigns, and testing in workplaces, parks, bars, places of worship and educational establishments.

Benefits of community-based testing 1. There is some evidence that the offer of HBsAg testing in community settings may increase the acceptance and uptake of testing, and rates of early diagnosis (193). 2. The benefits of community-based testing to access the general population are that it can also reach first-time testers and people who seldom use clinical services or are unlikely to go to a facility. This particularly includes those from key and vulnerable populations in all settings (11, 194), but also those who are asymptomatic. 3. Community-based focused testing of high-risk groups. Innovative models of care have been developed and effectively implemented in many settings to provide integrated HIV (194) and hepatitis testing, and opioid substitution therapy (OST) services for PWID in community drug treatment services. Many of these programmes provide additional interventions, including education, harm reduction, mental health services, other general medical services, and referrals to care and treatment (195). These models can provide a framework for lowerincome countries to expand viral hepatitis testing and treatment for at-risk populations.

Risks of community-based testing. Key challenges encountered in delivering community-based HIV testing include ensuring the availability and accessibility of prevention, care and treatment services; and risks associated with potential lack of confidentiality in these settings, and associated stigmatization and discrimination. These will need to be addressed while delivering communitybased hepatitis testing.

Costs and cost–effectiveness Community-level screening with outreach components is one of the most active types of casefinding; and therefore also one of the most labour- and resource intensive (152). A US-based study reported that costs per person tested ranged from $40 up to $280 for programmes with more active outreach strategies (149). The only cost–effectiveness analysis of a community-based test-and-treat strategy from an LMIC setting in west Africa showed that a community-based test and treat approach (compared to the status quo) can be regarded as cost–effective (144).

Feasibility The feasibility in sub-Saharan Africa of large-scale community-level outreach testing, clinical assessment and antiviral treatment if meeting eligibility criteria, has been demonstrated by the PROLIFICA study in west Africa (152).

6.8 Rationale for the recommendations on facilitybased testing Balance of benefits and harms Testing for viral hepatitis in the populations most affected in health-care facilities or clinical settings provides a major opportunity to scale up hepatitis testing.

Benefits of facility-based testing 1. Facility-based testing for hepatitis B and C can be successfully integrated with other health services such as HIV testing and treatment, which can in turn provide an entry point for other treatment services. 2. Facility-based focused testing can increase the uptake of viral hepatitis testing, case detection rate, and referrals to specialist-level care and other important services. This approach will generally be associated with higher rates of case identification and be more cost–effective.

Risks of facility-based testing. In low-prevalence settings, routine health facility -based testing will most likely be cost–effective only if delivered to high-risk populations.

Feasibility Multiple studies have shown that HIV testing is feasible both in health-care and other clinical facilities such as harm reduction and drug treatment services for PWID (196), in TB and STI clinics (and other services), with a high uptake of testing, and has been widely implemented in developed countries. In a similar way to HIV testing, hepatitis testing can be integrated with other health services and offered as part of the package of care, either routinely to all those attending services or offered in particular clinical settings.

Costs and cost–effectiveness Integrating screening into existing clinical services was shown to be the least costly method of testing, but also reached the least people, whereas extending screening outside the clinical setting was more costly as it included costs of organizing events and volunteer time, but reached more people (149).

Acceptability Facility-based provider-initiated testing for HIV has been widely accepted in the context of testing in antenatal and TB clinic settings, as well as in drug treatment programmes (196). These clinics can in addition provide an entry point to services for viral hepatitis testing, prevention and care, which is likely to be acceptable to patients attending these clinics. It also can help normalize testing and remove the potential embarrassment of clients asking for a test, and saves time for clients attending clinical services for other reasons.

6.9. Implementation considerations for HBV and HCV testing approaches Determining the optimal strategic mix of HBV and HCV testing approaches to increase the diagnosis rate, and in particular, the approach to testing outside of high-risk risk groups will depend on a country’s epidemic profile. Chapter 19 provides a strategic framework to guide countries to make decisions on selecting testing approaches. Chapter 17 presents specific examples of the many different types of facility- and community-based service delivery approaches to viral hepatitis testing, and Chapter 18 details specific testing issues that need to be considered in many high-risk and other special populations.

Research gaps in HBV and HCV testing approaches Further evaluation and comparisons of different HBV and HCV testing approaches are needed (i.e. routine general population, focused risk-based, ANC, birth cohort testing) using different service delivery models (communityor health-facility -based). This can take the form of comparative trials, or largescale implementation studies in a range of epidemic settings and populations in LMICs. Key outcome measures should include impact (uptake, case detection and linkage to care and treatment); cost and cost–effectiveness (and key drivers of cost–effectiveness), and proportion of HBV- or HCV-infected individuals missed by a specific testing approach. Further research into the simplification of testing and care, and integration of hepatitis services with other health services (e.g HIV, TB services) is needed to guide how impact and cost–effectiveness can be improved.

7. HOW TO TEST FOR CHRONIC HEPATITIS B INFECTION – choice of serological assay and testing strategy 7.1. Recommendations HOW TO TEST FOR CHRONIC HBV INFECTION AND MONITOR TREATMENT RESPONSE Topic Which serological assays to use Recommendations • For the diagnosis of chronic HBV infection in adults, adolescents and children (>12 months of age1), a serological assay (in either RDT or laboratory-based immunoassay format2) that meets minimum quality, safety and performance standards3 (with regard to both analytical and clinical sensitivity and specificity) is recommended to detect hepatitis B surface antigen (HBsAg). - In settings where existing laboratory testing is already available and accessible, laboratory-based immunoassays are recommended as the preferred assay format. - In settings where there is limited access to laboratory testing and/or in populations where access to rapid testing would facilitate linkage to care and treatment, use of RDTs is recommended to improve access. Strong recommendation, low/moderate quality of evidence • In settings or populations with an HBsAg seroprevalence of ≥0.4%4, a single serological assay for detection of HBsAg is recommended, prior to further evaluation for HBV DNA and staging of liver disease. • In settings or populations with a low HBsAg seroprevalence of <0.4%4, confirmation of HBsAg positivity on the same immunoassay with a neutralization step or a second different RDT assay for detection of HBsAg may be considered5. Conditional recommendation, low quality of evidence

Serological testing strategies

Abbreviations: ALT: alanine aminotransferase; AST: aspartate aminotransferase; APRI: aspartate-to-platelet ratio index; HBeAg: HBV e antigen; HBsAg: HBV surface antigen; NAT: nucleic acid test; RDT: rapid diagnostic test 1 A full vaccination schedule including birth dose should be completed in all infants in accordance with the WHO position paper on Hepatitis B vaccines, 2009. Testing of exposed infants is problematic within the first six months of life as HBsAg and hepatitis B DNA may be inconsistently detectable in infected infants. Exposed infants should be tested for HBsAg between 6 and 12 months of age to screen for evidence of hepatitis B infection. In all age groups, acute HBV infection can be confirmed by the presence of HBsAg and IgM anti-HBc. CHB is diagnosed if there is persistence of HBsAg for six months or more. 2 Laboratory-based immunoassays include enzyme immunoassay (EIA), chemoluminescence immunoassay (CLIA), and electrochemoluminescence assay (ECL). 3 Assays should meet minimum acceptance criteria of either WHO prequalification of in vitro diagnostics (IVDs) or a stringent regulatory review for IVDs. All IVDs should be used in accordance with manufacturers’ instructions for use and where possible at testing sites enrolled in a national or international external quality assessment scheme. 4 Based on results of predictive modelling of positive predictive values according to different thresholds of seroprevalence in populations to be tested, and assay diagnostic performance. 5 A repeat HBsAg assay after 6 months is also a common approach used to confirm chronicity of HBV infection.

FIG. 7.1 WHO-recommended testing strategies for diagnosis of chronic HBV infection with (A) Single assay with HBsAg seroprevalence above 0.4%, and (B) Two assays with HBsAg seroprevalence below 0.4% A HBsAg (A1) B HBsAg (A1)

HBsAg (A1) + (Reactive) Report positive

HBsAg (A1) – (Non-reactive) Report negative

HBsAg (A1) + HBsAg (A2)

HBsAg (A1) – (Non-reactive) Report negative

Compatible with HBV infection Proceed to NAT testing for quantification of viraemic infection

No evidence of HBV infection Advise retesting ± immunization if ongoing risk of known exposure

HBsAg (A1) + HBsAg (A2) + Report positive

HBsAg (A1) + HBsAg (A2) Report negative

No evidence of HBV infection Advise retesting ± immunization if ongoing risk of known exposure

Compatible with HBV infection Proceed to NAT testing for quantification of viraemic infection

inconclusive result Further testing as appropriate

7.2. Background Testing to determine chronic HBV infection is conducted using serological assays, either RDTs or EIAs that detect HBsAg. Confirmation of the presence of HBsAg may be carried out by performing either a neutralization step in the same assay, or by repeating HBsAg testing using a different assay of similar sensitivity (i.e. two-assay serological testing strategy). The choice of which format of serological assays to use will depend on a variety of factors, such as the performance criteria of the test (sensitivity and specificity), cost, ease of use and the characteristics of the testing site, such as storage facilities, infrastructure, and level of staff skills. Chapter 5 provides a background to different IVDs and Table 5.1, summarizes the advantages and disadvantages of laboratory-based immunoassays and RDTs. WHO recommends the use of standardized testing strategies both to maximize the accuracy of HBsAg testing while minimizing cost and simplifying the process. A testing strategy describes a testing sequence for a specific testing objective,

taking into consideration the anticipated prevalence of HBsAg in the population. See section 5.1.6 for a background on one- and two-assay serological testing strategies. The choice between a one- versus two-assay serological testing strategy will depend on the HBsAg prevalence in the population as well as diagnostic accuracy (sensitivity and specificity) of the HBsAg assays used. A one-assay serological testing strategy (Fig. 7.1A) is when a single serological test is performed. If the test result is reactive, a “compatible with HBsAg-positive” status is reported. If the initial test result is non-reactive, an “HBsAg-negative” status is reported. The addition of a second serological test with a two-assay serological testing strategy (Fig. 7.1B) will generally improve the PPV (i.e. the proportion of individuals detected that actually have HBV infection), reduce the number of false-positive results and therefore the number of individuals inappropriately referred on to specialist services.

7.3. Summary of the evidence Which serological assay to use A systematic review (see Web annex 5.3) compared the diagnostic performance (sensitivity, specificity, positive and negative predictive values) of commercially available serological assays (RDTs and EIAs1) for the detection of HBsAg, when compared to a laboratory-based immunoassay reference standard (with or without a neutralization step). The review identified 30 studies (197–226) from 23 countries with varying prevalence of hepatitis B and evaluated 33 different RDTs. There were five studies of eight different EIAs against an immunoassay reference standard (214, 223, 227–229). A mixture of serum, plasma, capillary and venous whole blood specimens were used for RDTs, but only serum or plasma was used for EIAs. Seven studies assessed performance using capillary or venous whole blood (202, 206, 210, 215, 216, 218, 226). Sample size varied from 25 to 3928, and populations studied included healthy volunteers and blood donors, at-risk populations, pregnant women, incarcerated adults, and patients with confirmed hepatitis B.

RDTs. In 30 studies (197–226) of 33 different RDTs, the pooled clinical sensitivity of RDTs against different EIA reference standards was 90.0% (95% CI: 89.1–90.8) and pooled specificity was 99.5% (95% CI: 99.4–99.5) (Table 7.1). Brands: there was significant variation in performance between RDT brands and within the same brand of RDT, with sensitivity ranging from 50% to 100% and specificity from 69% to 100%. Specimen type: results for capillary whole blood specimens were comparable to serum but less heterogeneous. 1

CLIAs and ECLs were not included specifically in the research question. It is acknowledged that high-income settings are likely to be using these formats of immunoassays.

EIAs. In five studies (214, 223, 227–229) of eight EIAs there was wide variation in EIA performance, with sensitivity ranging from 74% to 100% and specificity from 88% to 100%. The pooled sensitivity was 88.9% (95% CI: 87–90.6) and pooled specificity was 98.4% (95% CI: 97.8–98.8). RDTs and EIAs in HIV-positive persons. Five studies (212, 214, 215, 218, 222) evaluated three different RDTs against different EIA reference standards. The pooled clinical sensitivity of RDTs was 72.3% (95% CI: 67.9–76.4), but specificity was 99.8% (95% CI: 99.5–99.9), compared to a pooled clinical sensitivity and specificity of 92.6% (95% CI: 89.8, 94.8) and 99.6% (95% CI: 99, 99.9), respectively, among HIV-negative persons. Possible explanations for this reduced sensitivity include an increased incidence of occult hepatitis B in HIV-positive persons (i.e. presence of HBV DNA with undetectable HBsAg levels, such that HBsAg might not be detected using the RDTs evaluated), and the use of tenofovir- or lamivudine-based antiretroviral regimens, which are active against HBV and may suppress HBV DNA and HBsAg levels. In the one study (214) that evaluated three EIAs against an EIA reference with neutralization, the overall pooled sensitivity in HIV-positive individuals was 97.9% (95% CI: 96.0–99.0) and specificity was 99.4% (95% CI: 99.0–99.7), suggesting that EIAs perform better in HIV-positive persons. Analytical sensitivity/limit of detection. The analytical sensitivity or limit of detection (LoD) is another important performance criteria, but there were insufficient data in the included studies to undertake a systematic comparison. However, no RDTs met the levels of analytical sensitivity (i.e. LoD of 0.130 IU/mL) required by the European Union through its Common Technical Specifications. Data from WHO prequalification assessment studies indicate that the LoD of EIAs for HBsAg was 50–100-fold better compared to RDTs (230). However, despite this difference in analytical sensitivity, clinical sensitivity is unlikely to be greatly reduced because the vast majority of chronic HBV infection is associated with blood HBsAg concentrations well over 10 IU/mL. This is important, as it has been suggested that false-negative RDTs for HBsAg are due to low HBsAg viral load levels, the presence of HBsAg mutants or specific genotypes, and the use of lamivudine- or tenofovir-based ART regimens (208, 214, 216, 230). The overall quality of the evidence for the recommendation of which serological assay to use was rated as low to moderate, with downgrading mainly due to serious risk of bias based on cross-sectional study design, and heterogeneity in results.

TABLE 7.1. Summary test accuracy of RDTs and EIAs for HBsAg (different assay formats and comparators, populations and specimen types) Comparison Pooled sensitivity (95% CI) Pooled specificity (95% CI)

Assay format and comparators RDT versus EIA (N=30) EIA versus another EIA (N=5) RDT versus NAT (N=3) RDT versus CMIA (N=5)

90.0 (89.1–90.8) 88.9 (87–90.6) 93.3 (91.3–94.9) 80.4 (77.9–82.6) 91.6 (90.1–92.9) 72.3 (67.9–76.4) 92.6 (89.8–94.8) 90.8 (88.9–92.4) 97.6 (96.2–98.6) 82.5 (77.5–86.7) 82.5 (77.5–86.7) 91.7 (89.1–93.9)

99.5 (99.4–99.5) 98.4 (97.8–98.8) 98.1 ( 97–98.9) 99.0 (99.6–99.3) 99.5 (99.3–99.7) 99.8 (99.5–99.9) 99.6 (99.0–99.9) 99.1 (98.9–99.4) 100 (99.7–100) 99.9 (99.8–100) 99.9 (99.8–100) 99.9 (99.8–99.9)

Population (RDT versus EIA) Blood donors (N=7) HIV positive (N =5) HIV negative (N=1)

RDT kit brand (RDT versus EIA) Determine HBsAg (N=10) BinaxNOW HBsAg (N=3) VIKIA HBsAg (N=3) Serodia HBsAg (N=3)

Specimen type (RDT) Capillary whole blood versus serum (N=8)

CMIA: chemiluminiscent microparticle immunoassay; EIA: enzyme immunoassay; RDT: rapid diagnostic test

Which testing strategy to use No studies were identified that directly compared the diagnostic accuracy of a one- versus two-assay serological testing strategy in high- and low-prevalence settings (see Web annex 5.5). A predictive modelling analysis was therefore undertaken, which examined diagnostic accuracy of a one- or two-assay strategy based on a hypothetical population of 1000 individuals across both a range of HBsAg seroprevalence levels (10%, 2%, 0.4% representing typical high-, medium- and low-seroprevalence settings or populations, respectively) and a range of assay performance characteristics (sensitivity of 98% and 90%,

and specificity of 99% and 98% derived from the systematic review pooled sensitivity and specificity for HBsAg RDTs. Prevalence had a strong impact on the PPV and the ratio of true-positive to false-positive results (see Web annex 6.1). The introduction of a second assay of similar sensitivity to be applied to all specimens reactive in the initial serological assay provides substantial potential gains in the PPV across all prevalence levels (>97%), but particularly at a low prevalence (0.4%) and with an assay that has a lower specificity. The overall quality of the evidence for the recommendation on use of a one- or two-assay serological testing strategy was rated as low, as this was based on predictive modelling simulation and hypothetical scenarios.

7.4. Rationale for the recommendations on which assay to use The Guidelines Development Group recognized the critical need to expand testing to identify as many persons as possible with chronic HBV infection who might benefit the most from antiviral treatment and other interventions, and therefore made strong recommendations for a simplified one-assay testing strategy using either EIA or RDTs. Overall, the selection of assay format (EIA3 or RDT) to test for HBsAg in a particular setting will depend first on the performance characteristics of the assay, but also on key operational considerations, such as accessibility, cost, ease of use in the intended-use setting i.e. technical complexity of test procedure and specimen collection methods. The most sensitive assay available, either RDT or EIA, in terms of clinical sensitivity, should be used.

Balance of benefits and harms Use of EIAs. In settings where existing laboratory testing infrastructure is available and there is good access to laboratory services, EIAs were recommended as the preferred testing method for several reasons: 1. Although RDTs and EIAs for HBsAg had similar clinical sensitivity and specificity when compared to an EIA reference standard, the sensitivity of different RDTs was highly variable, and some RDTs had suboptimal sensitivity. In HIV-infected individuals, clinical sensitivity of RDTs was poor (72.3%) and appears to be better for EIAs. The analytical sensitivity is much higher for EIAs (50- to 100-fold higher). The benefit of more analytically sensitive assays with better limits of detection is that it improves detection in persons with primary infection, and in individuals in whom HBsAg levels are extremely low.

It is assumed that CLIA and ECL would have similar performance principles as EIAs.

A confirmatory test using a neutralization step can be incorporated into laboratorybased EIAs. Testing using laboratory-based EIAs can be automated and may be more appropriate and cost–effective in settings where there are many tests being performed per day (>40 per day per operator).

Use of RDTs 1. The Guidelines Development Group recognized that despite the significant heterogeneity and suboptimal clinical and analytical sensitivity of certain RDTs for HBsAg, expanded use of quality-assured RDTs has a major potential to help scale up HBsAg testing in settings with poor access to or lack of existing laboratory infrastructure to conduct EIAs, such as in remote settings or with hard-to-reach populations. The use of RDTs may be also appropriate in high-income countries to increase the uptake of hepatitis testing in populations that may be reluctant to test or have poor access to health-care services (e.g. PWID) and in outreach programmes (e.g. prison services, harm reduction and drug treatment services). Key challenges to the use of RDTs include the limited availability of quality-assured RDTs for HBsAg detection, reduced analytical sensitivity compared to laboratorybased methods, and that very few HBsAg RDTs meet the analytical sensitivity (LoD 0.130 IU/mL) required by the European Union. However, overall, the Guidelines Development Group considered that the benefits of RDTs in terms of increased access would mitigate potential harms related to lower accuracy, especially if there was careful selection of RDTs that met minimum performance criteria.

In HIV-positive persons, RDTs had low clinical sensitivity (pooled sensitivity of 72.3%). Although this may be potentially explained by the impact of tenofovir- or lamivudinecontaining ART regimens, there is a need for caution in their use and interpretation in HIV-positive patients.

Minimum performance criteria for EIAs and RDTs. RDTs for HBsAg have reduced analytical sensitivity and LoD compared to EIAs, as well as wide variation in clinical sensitivity and specificity between assays, and between different studies of the same assay. However, clinical sensitivity is unlikely to be greatly reduced because the vast majority of chronic HBV infection is associated with blood HBsAg concentrations well over 10 IU/mL. However, careful consideration should be given to ensure that the assay chosen has minimal rates of false positivity (both analytical and clinical). The Guidelines Development Group decided against defining minimum performance characteristics for assays, but recommended that any assay used should meet the performance criteria of stringent (see chapter 15) regulatory authorities in terms of both analytical and clinical sensitivity and specificity.

The recommendations for use of either RDTs or EIAs/CLIAs/ECLs were based on the assumption that all HBsAg assays used should meet minimum performance criteria of either WHO prequalification of IVDs or a stringent regulatory review for IVDs. All IVDs should be used in accordance with manufacturers’ instructions for use.

7.5. Rationale for the recommendations on testing strategy Balance of benefits and harms When to use a one-assay strategy. A one-assay testing strategy is applicable to most testing settings in resource-limited countries based on simplicity and in populations where prevalence is ≥0.4%, and so PPVs are high. The Guidelines Development Group made an overall conditional recommendation for a oneassay serological testing strategy to diagnose chronic HBV infection based on low-quality evidence for the following reasons: 1. This approach will efficiently identify (rule in) most individuals likely to be infected and in need of further evaluation, and will rule out those who are uninfected. Although a one-assay serological testing strategy has a lower PPV than a twoassay serological testing strategy, particularly at lower levels of prevalence (0.4% and 2%), and will therefore generate more false-positive results, the Guidelines Development Group considered that the consequences of this would not be clinically significant. This is because all HBsAg-positive patients will have further evaluation with staging of liver disease and HBV DNA measurement to assess eligibility for treatment (i.e. presence of cirrhosis or evidence of raised HBV DNA levels). Therefore, no patient would be initiated on lifelong antiviral therapy on the basis of a single serological test. The Guidelines Development Group noted that it is also common practice in many settings to perform a second test after 6 months to confirm a diagnosis of CHB and so distinguish it from acute hepatitis B. This provides an additional approach to confirm a diagnosis of chronic hepatitis infection. If one uses a first test with high specificity then very few would require a second test. It would considerably simplify the process of testing and reduce costs, especially if delivered at the point of care.

4. 5.

More rapid reporting of test results (ideally same day) will help improve access and linkage to care.

When to use a two-assay strategy. The Guidelines Development Group made a conditional recommendation to consider a second serological assay in very low-prevalence settings (<0.4%) to improve the PPV. In low prevalence settings, there will be more false-positive than true-positive results with a single serological assay, even with a test of 99% specificity. Employing two assays with a specificity of around 99% increases the ratio of true-positive to false-positive diagnoses from 0.2 to 32–40. The recommendation is to confirm with a neutralization step if using laboratory-based immunoassays for detection of HBsAg, as per the assay manufacturer’s instructions. Where an RDT for HBsAg is used and no neutralization reagents are available or for EIAs with no neutralization reagents, a second different RDT assay may be used (231). However, there has been limited evaluation of the added value of a second RDT, and there are several challenges: (i) “different” RDT assays may fundamentally be the same, and therefore prone to similar inaccuracies and false-positive reactions; (ii) if the analytical or clinical sensitivity of the assay used is poor (high LoD), then a larger proportion of individuals who are truly HBsAg positive will not be identified, regardless of whether a one- or two-test strategy is used.

Acceptability, values and preferences In a values and preferences survey among 104 respondents from 43 (20 high-income, 23 low- and middle-income) countries, overall, there was strong support from patient groups for simplified testing strategies that would improve access to testing including for high-risk groups. Seventy-seven per cent expressed a strong preference for a one-serological assay testing strategy with same-day results using RDTs to reduce loss to follow up.

Feasibility In a survey of programmatic experience with hepatitis testing across 19 LMICs, implementing partners reported widespread use of RDTs in all settings, and use of a single HBsAg RDT assay by 68% of respondents.

Resource considerations The reagent costs for HBsAg assays are similar for RDTs (between US$ 0.95 and US$ 3.00) and EIAs (between US$ 0.40 and US$ 2.80). High-throughput EIAs require additional laboratory infrastructure and equipment, and precision and expertise in operation. In contrast, RDTs do not require capital investment in laboratory infrastructure, and so there is a concurrent reduction in maintenance costs for equipment.

8. HOW TO TEST FOR CURRENT OR PAST HCV INFECTION (HCV EXPOSURE) – choice of serological assay and testing strategy 8.1. Recommendations HOW TO TEST FOR CHRONIC HCV INFECTION AND MONITOR TREATMENT RESPONSE Topic Which serological assays to use Recommendations • To test for serological evidence of past or present infection in adults, adolescents and children (>18 months of age1), an HCV serological assay (antibody or antibody/antigen) using either RDT or laboratory-based immunoassay formats2 that meet minimum safety, quality and performance standards3 (with regard to both analytical and clinical sensitivity and specificity) is recommended. - In settings where there is limited access to laboratory infrastructure and testing, and/or in populations where access to rapid testing would facilitate linkage to care and treatment, RDTs are recommended. Strong recommendation, low/moderate quality of evidence

Serological testing strategies

Abbreviations: DBS: dried blood spot; IVD: in vitro diagnostics; NAT: nucleic acid test; RDT: rapid diagnostic test 1 HCV infection can be confirmed in children under 18 months only by virological assays to detect HCV RNA, because transplacental maternal antibodies remain in the child’s bloodstream up until 18 months of age, making test results from serology assays ambiguous. 2 Laboratory-based immunoassays include enzyme immunoassay (EIA), chemoluminescence immunoassay (CLIA), and electrochemoluminescence assay (ECL). 3 Assays should meet minimum acceptance criteria of either WHO prequalification of IVDs or a stringent regulatory review for IVDs. All IVDs should be used in accordance with manufacturers’ instructions, and where possible at testing sites enrolled in a national or international external quality assessment scheme. A lower level of analytical sensitivity can be considered, if an assay is able to improve access (i.e. an assay that can be used at the point of care or suitable for dried blood spot [DBS] specimens) and/or affordability. An assay with a limit of detection of 3000 IU/mL or lower would be acceptable and would identify 95% of those with viraemic infection, based on available data.

FIG. 8.1. WHO-recommended single-assay testing strategy for detection of HCV antibody, irrespective of prevalence Anti-HCV (A1)

Anti-HCV (A1) + (Reactive) Report positive

Anti-HCV (A1) – (Non-reactive) Report negative

Compatible with exposure to HCV Proceed to confirmatory NAT testing for viraemic infection

No serological evidence of HCV infection

8.2. Background The principal assays used to determine exposure to HCV infection and evidence of past or current HCV infection rely on detection of antibodies to HCV using relatively inexpensive serological assays. Such antibody-based assays are unable to detect infection soon after acquisition of HCV infection, as antibodies may not be detected for 2–3 months in an individual who has been recently infected (135). This diagnostic window period can be shortened by using assays that also directly detect HCV antigen. Assessing HCV exposure typically involves either a one- or twoserological assay testing strategy. The main rationale for the use of a second HCV antibody test is to minimize false-positive results and reduce the number of people referred for more costly NAT technologies to confirm viraemic HCV infection. If HCV antibody positivity is established consistent with past or current infection, testing for current viraemic HCV infection is performed to ascertain viral replication through the detection of HCV RNA or HCV (p22) core antigen (HCVcAg). Aside from blood and organ donation screening, HCV RNA is not currently used to determine exposure to HCV, in spite of the shorter window period (1–2 weeks after the onset of acute infection) primarily for reasons of access and cost (135).

8.3. Summary of the evidence Which serological assay to use A systematic review (see Web annex 5.4) compared the diagnostic performance (sensitivity, specificity, positive and negative predictive values) of commercially available serological assays (RDTs and EIAs) for the detection of HCV antibody, when compared to a laboratory-based immunoassay reference standard. Five studies evaluated RDTs compared to an EIA reference (201, 232–235), 13 studies compared RDT results to NAT or immunoblot (236–248) and 14 studies compared RDTs with a combination of EIA, immunoblot or NAT (197, 201, 232–235, 247, 249–255). Twelve studies compared RDTs using oral fluid to RDTs using whole blood (as well as serum or plasma) (235, 238, 240, 241, 246–248, 250, 251, 255–257). The studies were carried out on different source populations, including the general population, key populations and hospital patients. The sample sizes of the included studies ranged from 37 to 17 894. All studies used a cross-sectional or case–control design.

RDTs. Based on the five studies of RDTs compared to EIA-only reference standard, the pooled RDT sensitivity and specificity were, respectively, 99% (95% CI: 98–100) and 100% (95% CI: 100–100), but sensitivities in individual studies ranged from 83% to 100%, and specificities from 99% to 100%.

Brands. There was significant heterogeneity between studies and variable performance across RDT brands and even within the same brand. Although use

of NAT or immunoblot is not an appropriate reference to assess RDT diagnostic performance, the pooled sensitivity and specificity of RDTs were 93% (95% CI: 91–95%) and 98% (95% CI: 97–99%), respectively (Table 8.1).

Populations. A high sensitivity (>95%) and specificity (>99%) of RDTs for HCV antibody were observed across populations screened (general population, key populations, hospital patients) using different reference standards (EIA, immunoblot), but a patient selection bias was evident in around a third of studies.

Specimen type. RDTs using oral fluid showed a lower sensitivity but higher specificity compared to the reference standard, respectively, at 94% (95% CI: 93–96%) and 100% (95% CI: 100–100%). However, eight studies that examined OraQuick ADVANCE® HCV Rapid Antibody Test (OraSure Technologies, Inc.) had a higher sensitivity of 98% (95% CI: 97–98) compared to the other brands examined in six studies (pooled sensitivity of 88% [95% CI: 84–92]). There were insufficient data for other key brands, including the SD-Bioline which is now WHO prequalified.

RDTs and EIAs in HIV-positive persons. The number of studies was insufficient to undertake subanalyses based on HIV coinfection (256, 260–263). However, one recent study has reported that HCV EIAs may be associated with high rates of false positivity among HIV-infected persons in Africa (258). The overall quality of the evidence for the recommendation to use RDTs was rated from low to moderate with downgrading mainly due to a serious risk of bias based on cross-sectional study design, and heterogeneity of results.

TABLE 8.1. Summary diagnostic accuracy of HCV antibody tests (different assay format and comparators, populations, specimen type and oral kit brand) Comparison Pooled sensitivity (95% CI) Pooled specificity (95% CI)

Assay format and comparators RDT versus EIA only (N=5)

99 (98–100)

100 (100–100) 98 (97–99) 100 (100–100) 99 (98–100)

RDT versus NAT or or immunoblot 93 (91–95) (N=13) RDT versus EIA, NAT or immunoblot (N=14) Antibody and antigen combo testing (N=6) General population (N=17) Key populations (N=19) Hospital patients (N=16) Blood specimens (N=45) Oral fluid specimens (N=12) Oral RDT versus blood reference (N=12)

97 (96–98) 86 (79–94)

Populations (RDT versus EIA, NAT or immunoblot) 95 (94–96) 97 (96–98) 97 (96–98) 98 (97–98) 94 (93–96) 94 (93–96) 99 (98–99) 94 (94–95) 100 (100–100) 98 (98–99) 100 (100–100) 99.9 (99.8–100) 100 (100–100) 99 (99–100)

Specimen type (RDT vs EIA, NAT or immunoblot)

Oral kit brands OraQuick (N=8) Other brands (Chembio DPP, BIioeasy, ImmunoComb II (N=6)

98 (97–99) 88 (84–92)

EIA: enzyme immunoassay; NAT: nucleic acid test; RDT: rapid diagnostic test

Which testing strategy to use There was a single cost-effectiveness analysis that compared three testing strategies in a Brazilian population (259) (see Web annex 5.6). They found that a one-serological assay testing strategy for detection of anti-HCV followed by HCV RNA NAT to establish viraemic HCV infection was more cost–effective than a two-serological assay testing strategy. A predictive modelling study was also undertaken, which examined the diagnostic accuracy of a one- or two-assay HCV antibody testing strategy based on a hypothetical population of 1000 individuals across both a range of HCV antibody seroprevalence levels that reflect typical high prevalence rates among PWID [45%], intermediate prevalence among HIVinfected MSM [10%] and moderate [2%] to low [0.4%] endemicity in a general population), and across a range of assay performance characteristics (sensitivity of 98% and 90%, and specificity of 99% and 98% derived from the systematic review pooled sensitivity and specificity for HCV antibody RDTs, Table 8.1).

The outcomes show the strong influence of prevalence and assay specificity on PPV (see Web annex 6.2). The use of a highly sensitive and specific (98% and 99%) single assay yields a high PPV in excess of 90% at prevalences of 40% and 10%, and 67% at a 2% prevalence, and only a small number of false-positive diagnoses. Only at the lowest prevalence (e.g. 0.4%) does the PPV fall below 50%. Since overall the PPV is high at all prevalence levels with a single assay, the use of a second test would have a significant impact only in the lowest-prevalence populations (0.4%), especially if the initial assay was of lower performance.

8.4. Rationale for the recommendations on which assay to use Overall, the Guidelines Development Group made a strong recommendation for the use of serological assays, particularly RDTs, based on moderate/low-quality evidence for diagnostic performance. As for HBsAg, the selection of assay format (either EIA4 or RDT) to test for HCV antibody in a particular setting will depend first on the performance characteristics of the assay, cost and also on key operational considerations, such as accessibility and ease of use in the intended-use setting, such as a community-based drug treatment programme versus a hospital-based clinic.

Balance of benefits and harms Use of RDTs. In settings where access to laboratory services is limited, or where existing testing services do not have the capacity for conducting EIA, and for hard-to-reach and rural populations, the Guidelines Development Group recommended (as for HIV and HBsAg) the use of quality-assured RDTs rather than conventional laboratory-based EIAs. This was due mainly to their simplicity, relatively low cost and rapid turnaround time, and therefore their potential to substantially improve access to HCV testing, enhance linkage to care and reduce loss to follow up. Other reasons for the preferred use of RDTs include the following: 1. RDTs for the detection of antibodies to HCV have acceptable sensitivity and specificity compared to laboratory-based EIAs across a wide range of settings and different populations and for different brands. RDTs that use oral fluid are also available, which have adequate sensitivity and specificity, and may therefore be particularly useful where collection of venous or capillary whole blood is challenging. RDTs performed at the point of care, using less invasively collected specimens than venous whole blood, may allow for results to be available on the same day as testing, and so avoid the need for multiple follow-up appointments and reduce loss to follow up. For national programmes in resource-limited settings, expanded use of RDTs may mitigate the challenges of specimen collection, processing and transportation to laboratory services, and allow for the simplification and decentralization of testing.

It is assumed that CLIA and ECL would have similar performance principles to EIAs.

RDTs can also be used in outreach programmes (e.g. prison services, substance use/ treatment services) in HICs to increase the uptake of hepatitis screening. Well-trained community health workers can perform testing accurately and reliably.

Use of EIAs. In settings with existing laboratory infrastructure or where many tests are carried out per day, testing by laboratory-based methods, such as EIAs, may be cost– effective and appropriate. Although RDTs and EIAs had similar clinical sensitivity and specificity, testing using laboratory-based EIAs was recommended as the more appropriate and cost–effective assay in settings where suitable laboratory infrastructure is available, and where there is likely to be high-volume throughput, with many tests performed per day (>40 per day per operator), and in individuals who have good access to laboratory-based testing. It is important to note that the latest generation of assays designed to detect HCV antibody are also designed to detect HCVcAg in order to increase the sensitivity of the assay and reduce the diagnostic window period. However, these fourth-generation assays are not typically able to differentiate HCV exposure from chronic HCV infection. In HIV-positive persons. Insufficient studies were retrieved for the systematic review of persons with HIV/HCV coinfection for a formal evaluation of the diagnostic accuracy of RDTs to detect antibodies to HCV in persons who are HIV coinfected. Theoretically, the sensitivity of serological assays that detect antibodies only may be reduced if the patient is immunocompromised, e.g. persons with HIV infection, those undergoing immunosuppressive therapy or renal dialysis, and therefore exposure to HCV may not be detected in these individuals. It is estimated that this may occur in up to 6% of HIV-infected persons who undergo testing using an EIA for the detection of antibodies to HCV (260, 261) but may occur more often among persons with advanced immunosuppression due to HIV and during early HCV infection (262, 263). Conversely, there are also reports of a large proportion of false-positive HCV serological tests among HIV-infected persons, especially in SSA (258).

Minimum performance criteria for EIAs and RDTs. The Guidelines Development Group decided against defining minimum performance characteristics for assays, but recommended that any assay used should meet the performance criteria for stringent regulatory authorities (see Chapter 15). The Group also recognized that performance of RDTs in the field (i.e. setting of intended use) may vary and that certain RDTs are not validated by the manufacturer for use on capillary whole blood. The issues of analytical sensitivity and LoD for HCV antibody assays is less relevant than for HBsAg for several reasons. First, there are no WHO reference standard materials for anti-HCV antibody, and so IU/mL cannot be applied. Second, in contrast to the 40–100-fold difference in LoD for HBsAg detection, there is a minimal difference in end-point titres for anti-HCV antibody between RDTs and EIAs.

The recommendations for the use of either RDTs or EIAs/CLIAs/ECLs were based on the assumption that all HCV antibody assays used should meet minimum performance criteria of either WHO prequalification of IVDs or a stringent regulatory review for IVDs. All IVDs should be used in accordance with manufacturers’ instructions for use.

8.5. Rationale for the recommendation for a one-assay serological testing strategy Balance of benefits and harms The Guidelines Development Group made a conditional recommendation for a single test using an RDT or EIA, followed by HCV RNA NAT or core antigen on reactive specimens as the simplest and most feasible testing strategy in all settings based on low-quality evidence for several reasons. 1. Predictive modelling indicates that a one-assay testing strategy efficiently identifies all but a very few individuals likely to be infected and in need of NAT testing to confirm viraemic HCV infection, and similarly excludes nearly all HCV-uninfected individuals, even at low prevalence. The overall impact of the second assay on improving PPV was smaller compared to the situation with HBsAg because of the generally higher sensitivity and specificity of RDTs for HCV antibody. There are concerns about the cost implications and feasibility of implementing a second serological assay, particularly at the point of care and in resource-limited settings. In low-prevalence populations, a higher proportion of results would be false positive, and therefore individuals would undergo unnecessary and more expensive NAT to identify viraemic infection as a result of a single falsely reactive serological assay. In this situation, the two-assay serological testing strategy may be marginally cost saving. However, the Guidelines Development Group did not consider that the numbers of false- positive diagnoses were significant enough to justify a two-assay serological testing strategy. It was also recognized that many countries in SSA will fall into the low-seroprevalence category of less than 0.4%, and higher rates of false positivity have also been reported in these settings. As access to NAT remains very limited and costly at present, the use of a second serological test may be more cost–effective than performing NATs in multiple persons with false-positive results. The risks associated with a false positive HCV antibody result is minimal, as all individuals with a diagnosis of HCV exposure (HCV seropositive) will require supplemental testing to confirm viraemic HCV infection (by NAT to detect HCV RNA or serology to detect HCVcAg) before initiation of antiviral treatment.

Acceptability, values and preferences The values and preferences survey showed strong support by the majority for the use of RDTs delivered at the point of care to promote access, and a simplified one-serological assay testing strategy for HCV exposure, followed by supplementary testing to detect viraemic HCV infection. Providers and patients found RDTs that utilize oral fluid specimens to be more acceptable than capillary or venous whole blood specimens, especially in children.5

Resource use The cost of RDTs for HCV antibodies ranges from US$ 0.50 to US$ 2.00 for blood-based assays, and US$ 10 for oral fluid RDTs. The cost of EIAs ranges from US$ 0.50 to US$ 1.70, but EIAs require additional laboratory infrastructure and equipment, with precision and expertise required in their operation. RDTs do not require capital investment in laboratory infrastructure, and so there is a concurrent reduction in maintenance costs and reagents. Using a second different RDT assay would at least double the costs.

Feasibility A survey of hepatitis testing programmatic experience across 19 LMICs found that a oneserological assay testing strategy using mainly RDTs was being implemented in a range of hospital-based services, including blood donor screening, harm-reduction services, and HIV treatment and care clinics.

8.6. Implementation considerations for HBsAg and HCV antibody serological testing The most sensitive assay available, either RDT or EIA, in terms of both analytical sensitivity and clinical sensitivity, should be used. RDTs generally have a lower analytical sensitivity (IU/mL LoD) compared to EIAs. However, careful consideration should be given to ensure that the assay chosen has minimal rates of false positivity (both analytical and clinical). See also Chapter 15 for details on how to set up laboratory services for hepatitis testing and selection of an assay, and how to assure the quality of hepatitis B and hepatitis C testing. 1. Quality-assured and fit-for-purpose assays. Access to a range of well-performing quality-assured assays is critical to the success of any hepatitis testing programme. While a wide variety of EIAs (and CLIAs, ECLs) are commercially available, there is a lack of HBsAg RDTs that meet minimum performance criteria, as well as safety and quality standards. National regulatory authorities are responsible for approving RDTs for sale and use after assessment of their quality, safety and performance. 5

Most RDTs using oral fluid are yet to be validated and then evaluated in children.

2. Accurate testing. All hepatitis B and C testing should be performed in accordance with the assay manufacturer’s instructions, including an HBsAg neutralization step if it is being utilized. In addition, SOPs and job aids can help testing providers minimize testing and reporting errors, and thus improve the quality of the results. 3. Staff training and supervision. The testing environment should operate according to quality management systems (see Chapter 15), and have access to qualified, proficient and motivated laboratory staff, trained specifically in performance of the various assays, with adequate and supervisory support. Health-care workers should understand the strengths and limitations of any given testing strategy, counsel patients who are screened, and be able to act appropriately on the results, both positive and negative. Delivery of RDTs requires appropriate training of test providers in performing the test, reading the test result, storage of test kits and other supplies, and interpreting and reporting the results. 4. Linkage to care. As some preliminary results will be false positive, appropriate linkage will be needed to additional testing and clinical evaluation, especially where testing is conducted at the point of care in outreach programmes. Retesting strategies need to be implemented for those with a high risk of acquisition of HCV. 5. Provision of NAT or HCVcAg testing at the same site as serological testing would be optimal, enabling rapid turnaround, reduced loss to follow up, and reduced personal and health-care costs of referral to a distant centre. The Guidelines Development Group did not consider in these guidelines the potential future testing scenario of a single NAT or HCVcAg for both diagnosis and confirmation of active infection.

Research gaps for HBsAg and HCV antibody serological testing • The impact of HIV positivity (and of CD4 count, viral load and ART exposure, by regimen) on the diagnostic performance of RDTs for HBsAg and HCV antibody should be further evaluated. Evaluation should be done of the diagnostic performance, impact, cost and cost–effectiveness of a one- versus two-assay serological HBsAg or HCV testing strategy in diverse settings of both high and low HBsAg and HCV antibody prevalence. EIAs and RDTs assays should be validated using less invasive and simpler methods of sample collection, such as oral fluid and capillary whole blood and dried blood spots (DBS).

•

9. DETECTION OF VIRAEMIC HBV INFECTION – to guide who to treat or not treat 9.1. Recommendation Detection of HBV DNA – assessment for treatment Adapted from existing guidance (WHO HBV 2015 guidelines6) • Directly following a positive HBsAg serological test, the use of quantitative or qualitative nucleic acid testing (NAT) for detection of HBV DNA is recommended as the preferred strategy and to guide who to treat or not treat. Strong recommendation, moderate/low quality of evidence

Algorithm of WHO recommendations on the management of persons with chronic hepatitis B infectiona HBsAg positive CIRRHOSIS • Clinical criteriab • NITs (APRI score >2 in adults or FibroScan) Yes No

ASSESSMENT FOR TREATMENT

ALL AGESc >30 years (in particular) ALTd,e Persistently abnormal HBV DNA 2000– 20 000 IU/mL ALTd,e Intermittently abnormal ALTd,e Persistently normal

AGE ≤30 years ALTd,e Persistently normal

HBV DNA >20 000 IU/mL

HBV DNA <2000 IU/mL

INITIATE NA THERAPY AND MONITOR • Tenofovir or entecavir • Entecavir in children aged 2–11 years

DEFER TREATMENT AND MONITOR

Every 6 months

DETECTION OF HCCf (persons with cirrhosis or HCC family history)

MONITORING

Every 12 months

DISEASE PROGRESSION AND/OR TREATMENT RESPONSE IN ALLf

Baseline and every 12 months

TOXICITY MONITORING IN PERSONS ON TREATMENT Renal functiong and risk factors for renal dysfunction

STOPPING TREATMENT

CIRRHOSIS Lifelong treatment

NO CIRRHOSIS • and HBeAg loss and seroconversion to anti-HBe and after completion of at least one additional year of treatment • and persistently normal ALT • and persistently undetectable HBV DNA

NITs: non-invasive tests, ALT: alanine aminotransferase, APRI: aspartase aminotransferase-to-platelet ratio index a Defined as persistence of hepatitis B surface antigen (HBsAg) for six months or more. The algorithm does not capture all potential scenarios, but the main categories for treatment or monitoring. Recommendations for settings without access to HBV DNA testing are provided in the relevant chapters. b Clinical features of decompensated cirrhosis: portal hypertension (ascites, variceal haemorrhage and hepatic encephalopathy), coagulopathy, or liver insufficiency (jaundice). Other clinical features of advanced liver disease/cirrhosis may include: hepatomegaly, splenomegaly, pruritus, fatigue, arthralgia, palmar erythema, and oedema. c The age cut-off of >30 years is not absolute, and some persons with CHB less than 30 years may also meet criteria for antiviral treatment. d ALT levels fluctuate in persons with chronic hepatitis B and require longitudinal monitoring to determine the trend. Upper limits for normal ALT have been defined as below 30 U/L for men and 19 U/L for women, though local laboratory normal ranges should be applied. Persistently normal/abnormal may be defined as three ALT determinations below or above the upper limit of normal, made at unspecified intervals during a 6–12-month period or predefined intervals during 12-month period. e Where HBV DNA testing is not available, treatment may be considered based on persistently abnormal ALT levels, but other common causes of persistently raised ALT levels such as impaired glucose tolerance, dyslipidaemia and fatty liver should be excluded. f All persons with CHB should be monitored regularly for disease activity/progression and detection of HCC, and after stopping treatment for evidence of reactivation. More frequent monitoring maybe required in those with more advanced liver disease, during the first year of treatment or where adherence is a concern, and in those with abnormal ALT and HBV DNA levels >2000 IU/mL, not yet on treatment. g Before initiation, assessment should be done of renal function (serum creatinine level, estimated glomerular filtration rate, urine dipsticks for proteinuria and glycosuria, and risk factors for renal dysfunction [decompensated cirrhosis, CrCl <50 mL/min, poorly controlled hypertension, proteinuria, uncontrolled diabetes, active glomerulonephritis, concomitant nephrotoxic drugs, solid organ transplantation, older age, BMI <18.5 kg/m2 (or body weight <50 kg], concomitant use of nephrotoxic drugs or a boosted protease inhibitor [PI] for HIV). Monitoring should be more frequent in those at higher risk of renal dysfunction.

Existing recommendations on Who to treat and not treat (2015 WHO HBV guidelines) Who to treat • As a priority, all adults, adolescents and children with CHBa and clinical evidence of compensated or decompensated cirrhosisb (or cirrhosis based on APRI score >2 in adults) should be treated, regardless of ALT levels, HBeAg status or HBV DNA levels. (Strong recommendation, moderate quality of evidence) • Treatment is recommended for adults with CHBa who do not have clinical evidence of cirrhosis (or based on APRI score ≤2 in adults), but are aged more than 30 yearsc (in particular), and have persistently abnormal ALT levelsd,e and evidence of high-level HBV replication (HBV DNA >20 000 IU/mLf), regardless of HBeAg status. (Strong recommendation, moderate quality of evidence) › Where HBV DNA testing is not available: treatment may be considered based on persistently abnormal ALT levels alonee, regardless of HBeAg status. (Conditional recommendation, low quality of evidence)

Who not to treat but continue to monitor • Antiviral therapy is not recommended and can be deferred in persons without clinical evidence of cirrhosis (or based on APRI score ≤2 in adults), and with persistently normal ALT levelsd,e and low levels of HBV replication (HBV DNA <2000 IU/mLf), regardless of HBeAg status or age. (Strong recommendation, low quality of evidence) › Where HBV DNA testing is not available: Treatment can be deferred in HBeAg-positive persons aged 30 years or less and persistently normal ALT levels. (Conditional recommendation, low quality of evidence) • Continued monitoring is necessary in all persons with CHB, but in particular those who do not currently meet the above-recommended criteria for who to treat or not treat, to determine if antiviral therapy may be indicated in the future to prevent progressive liver disease. These include: - persons without cirrhosis aged 30 years or less, with HBV DNA levels >20 000 IU/ mLe but persistently normal ALT; - HBeAg-negative persons without cirrhosis aged 30 years or less, with HBV DNA levels that fluctuate between 2000 and 20 000 IU/mL, or who have intermittently abnormal ALT levelsd,e; › Where HBV DNA measurement is not available: persons without cirrhosis aged 30 years or less, with persistently normal or ALT levels, regardless of HBeAg status. a

Clinical features of decompensated cirrhosis: portal hypertension (ascites, variceal haemorrhage and hepatic encephalopathy), coagulopathy, or liver insufficiency (jaundice). Other clinical features of advanced liver disease/cirrhosis may include: hepatomegaly, splenomegaly, pruritus, fatigue, arthralgia, palmar erythema and oedema. Defined as persistence of hepatitis B surface antigen (HBsAg) for six months or more The age cut-off of >30 years is not absolute, and some persons with CHB aged less than 30 years may also meet the criteria for antiviral treatment.

b c d

ALT levels fluctuate in persons with CHB and require longitudinal monitoring to determine the trend. Upper limits for normal ALT have been defined as below 30 U/L for men and 19 U/L for women (based on greater sensitivity observed in hepatitis B for histological disease in the liver), though local laboratory normal ranges should be applied. Persistently normal/abnormal may be defined as three ALT determinations below or above the upper limit of normal, made at unspecified intervals during a 6–12-month period or predefined intervals during a 12-month period.

e Where HBV DNA testing is not available, other common causes of persistently raised ALT levels such as impaired glucose tolerance, dyslipidaemia and fatty liver should be excluded. f WHO has defined an international standard for expression of HBV DNA concentrations. Serum HBV DNA levels should be expressed in IU/mL to ensure comparability; the same assay should be used in the same patient to evaluate antiviral efficacy. All HBV DNA values in the recommendations are reported in IU/mL; values given as copies/mL were converted to IU/mL after dividing by a factor of 5 (10 000 copies/ mL = 2000 IU/mL; 100 000 copies/mL = 20 000 IU/mL; 1 million copies/mL =200 000 IU/mL).

Occasionally, extrahepatic manifestations of hepatitis B, including glomerulonephritis or vasculitis, may be indications for treatment.

9.2 Background The decision to initiate antiviral therapy is usually based on a combined assessment of the stage of liver disease (from clinical features, and now increasingly from blood or ultrasound-based NITs), together with levels of serum ALT and HBV DNA. Serum HBV DNA concentration quantified by real-time polymerase chain reaction (PCR) measures the extent of viral replication and correlates with disease progression (264–266). It is used to differentiate active HBeAg-negative disease from inactive chronic infection, and inform decisions on treatment and subsequent monitoring. The objective of treatment is to prevent the adverse outcomes of CHB. The decision to treat is usually clear in persons who present with life-threatening or advanced liver disease, such as acute liver failure, compensated or decompensated cirrhosis, and acute-on-chronic liver failure. However, in persons who have not yet progressed to cirrhosis, it is important that antiviral therapy is targeted to the stage of CHB when the risk of disease progression (fibrosis) is greatest while those persons with minimal fibrosis and low risk of CHB progression that do not require antiviral therapy are identified. Decisions are generally based on ALT and HBV DNA levels. However, not all persons will have raised ALT and HBV DNA levels. For example, during the immune-tolerant phase of disease, there will be high levels of HBV DNA but low or normal levels of ALT, and little liver inflammation or progression of fibrosis. Later on, during the immune-active phase, HBV DNA levels will be low, but ALT levels will be raised, with a much higher risk of disease progression to fibrosis.

9.3. Rationale for the recommendations on HBV DNA measurement (WHO 2015 HBV guidelines) The Guidelines Development Group recognized that access to HBV DNA measurement remains limited in LMICs. In the 2015 HBV guidelines (6), a strong recommendation was made for use of HBV DNA NAT (quantitative or qualitative) as the most important assay to guide decisions about who to treat or not treat. This was based on low/moderate-quality evidence from 22 observational studies (including four large population-based prospective cohort studies) to identify individuals with the highest and lowest risk of progression (i.e. cirrhosis and HCC) (Web appendix 2: SRs 5a and 5b – WHO HBV guidelines 2015, http://apps.who.int/iris/bitstream/10665/154590/1/9789241549059_ eng.pdf?ua=1&ua=1). There are caveats to the generalizability of the evidence. The majority of the studies were from Asia, and there were no data from cohorts in SSA or Latin America, and the data from the REVEAL study may not apply to those with adult-acquired HBV infection, those aged <30 or >65 years, and those infected with HBV genotypes non-B or C. There were also no studies in pregnant women, children or adolescents with CHB.

Balance of benefits and harms Who to treat In the 2015 HBV guidelines, it was recommended that antiviral therapy be prioritized for those with life-threatening liver disease (decompensated cirrhosis) and compensated cirrhosis, identified either clinically or using NITs (APRI score based on the single high cut-off >2 for cirrhosis in adults, or FibroScan®), regardless of ALT or HBV DNA levels. In persons who had not progressed to cirrhosis (APRI score ≤2 in adults), it was recommended to target treatment to those at highest risk of disease progression based on the detection of persistently abnormal ALT and HBV DNA levels >20 000 IU/mL (especially in those aged >30 years, regardless of HBeAg status). The recommended thresholds were derived from consistent evidence from large population-based cohort studies of increased risk of HCC and liver cirrhosis above these thresholds. It was recognized that there were uncertainties in the specific thresholds of age, HBV DNA and serum ALT levels for identifying significant fibrosis and/or necroinflammation.

Who not to treat Conversely, treatment was not recommended in persons with minimal liver disease or fibrosis, and at low risk of progression to cirrhosis and HCC on the basis of persistently normal ALT levels and low levels of HBV replication (<2000 IU/mL), and an APRI score ≤2, as the potential harms of long-term antiviral therapy outweigh the benefits. Long-term monitoring of these persons is required.

In settings where HBV DNA testing is not available The current limited access to HBV DNA testing in many LMICs is a significant impediment to the effective management of CHB in these settings, and this means that decisions to start treatment will be based on clinical features and serum ALT levels alone. Overall, there was a very limited evidence base to guide recommendations in the absence of HBV DNA levels, and two conditional recommendations were made in the 2015 WHO HBV guidelines based mainly on expert opinion. First, treatment should be initiated in persons with persistently abnormal ALT levels (regardless of HBeAg status), but where other common causes of persistently abnormal ALT such as impaired glucose tolerance, dyslipidaemia and fatty liver have been excluded. Conversely, treatment was not recommended in HBeAg-negative persons without cirrhosis aged less than 30 years with persistently normal ALT levels. It was recognized that there are several other categories of persons with CHB who do not meet the criteria for initiating or not initiating treatment, who would also require continued monitoring and observation.

9.4. Implementation considerations • • Access to HBV DNA testing is currently very limited in most LMICs, and is a significant impediment to the effective management of CHB in these settings. Serum HBV DNA levels should be expressed in IU/mL to ensure comparability; values given as copies/mL can be converted to IU/mL by dividing by a factor of 5 to approximate the conversion used in the most commonly used assays (i.e. 10 000 copies/mL = 2000 IU/mL; 100 000 copies/mL = 20 000 IU/mL; 1 million copies/mL = 200 000 IU/mL).

10. MONITORING FOR HBV TREATMENT RESPONSE AND DISEASE PROGRESSION 10.1. Recommendations Monitoring for HBV treatment response and disease progression Existing guidance (WHO HBV 2015 guidelines1) • It is recommended that the following be monitored at least annually: - ALT levels (and AST for APRI), HBsAg2, HBeAg3, and HBV DNA levels (where HBV DNA testing is available) - Non-invasive tests (APRI score or transient elastography) to assess for presence of cirrhosis in those without cirrhosis at baseline; - If on treatment, adherence should be monitored regularly and at each visit. Strong recommendation, moderate quality of evidence

More frequent monitoring is recommended: • In persons on treatment or following treatment discontinuation: more frequent on-treatment monitoring (at least every 3 months for the first year) is indicated in: persons with more advanced disease (compensated or decompensated cirrhosis4); during the first year of treatment to assess treatment response and adherence; where treatment adherence is a concern; in HIV-coinfected persons; and in persons after discontinuation of treatment. Conditional recommendation, very low quality of evidence • In persons who do not yet meet the criteria for antiviral therapy: i.e. persons who have intermittently abnormal ALT levels or HBV DNA levels that fluctuate between 2000 IU/mL and 20 000 IU/mL (where HBV DNA testing is available) and in HIVcoinfected persons7. Conditional recommendation, low quality of evidence Abbreviations: ALT: alanine aminotransferase; AST: aspartate aminotransferase; APRI: aspartate-to-platelet ratio index; HBeAg: HBV e antigen; HBsAg: HBV surface antigen; NAT: nucleic acid test; RDT: rapid diagnostic test 1 For further details, see Chapter 5: Who to treat and who not to treat. Guidelines for the prevention, care and treatment of persons with chronic hepatitis B infection: World Health Organization; 2015. 2 In persons on treatment, monitor for HBsAg loss (although this occurs rarely), and for seroreversion to HBsAg positivity after discontinuation of treatment. 3 Monitoring of HBeAg/anti-HBe mainly applies to those who are initially HBeAg positive. However, those who have already achieved HBeAg seroconversion and are HBeAg negative and anti-HBe positive may serorevert. 4 Decompensated cirrhosis is defined by the development of portal hypertension (ascites, variceal haemorrhage and hepatic encephalopathy), coagulopathy, or liver insufficiency (jaundice). Other clinical features of advanced liver disease/cirrhosis may include: hepatomegaly, splenomegaly, pruritus, fatigue, arthralgia, palmar erythema and oedema.

10.2. Background Prior to treatment The goal of monitoring HBV DNA and other markers is to identify progression of disease and when to initiate therapy. This can be ascertained by longitudinal monitoring of ALT, HBeAg and HBV DNA levels, where available. Fluctuations or persistently abnormal serum ALT and HBV DNA levels >20 000 IU/mL can indicate progressive disease and the need for treatment. Conversely, spontaneous improvement may occur with a decline in HBV replication, with normalization of ALT levels and seroconversion from HBeAg-positive to anti-HBe. This confers a good prognosis and does not require treatment. Similarly, persons with inactive disease, who are HBeAg-negative with normal ALT levels and low HBV DNA levels (previously called inactive HBsAg carriers), require regular monitoring of HBV DNA and ALT levels to ensure that they remain inactive carriers or, to determine the timing for treatment initiation, any increase in ALT or HBV DNA levels, or evidence of progression to cirrhosis.

During and after treatment Monitoring while on treatment is required to assess adherence, evaluate whether viral suppression is sustained (where HBV DNA can be measured), assess the potential for treatment discontinuation, and progression of liver disease, including development of HCC. Monitoring after treatment is important to check for reactivation early on and when to restart treatment.

10.3. Rationale for the recommendations (WHO 2015 HBV guidelines) The optimal timing and frequency of monitoring for serological markers (HBeAg, serum ALT) and HBV DNA to ascertain alterations in disease patterns prior to treatment, as well as assess treatment response have not been well established, and the evidence base is limited. Since no studies had directly compared different monitoring approaches and frequency of monitoring, and there was only indirect evidence from cohort studies, and imprecision due to few events, the quality of evidence was rated as low or very low (Web appendix 2: SR5a – WHO HBV guidelines 2015) http://apps.who.int/iris/bitstream/10665/154590/1/9789241549059_ eng.pdf?ua=1&ua=1) (6).

Balance of benefits and harms Monitoring prior to treatment The 2015 HBV guidelines (6) recommended at least annual monitoring of HBV DNA levels, HBeAg and serum ALT to determine any persistent abnormality in ALT or HBV DNA levels (i.e. HBV DNA threshold above >20 000 IU/mL and ALT levels) consistent with risk of disease progression, as well as for development of cirrhosis (based on clinical features or on NITs [APRI >2 in adults]), which would be an indication for antiviral therapy. Additional monitoring of HBeAg may be helpful for several reasons: it indicates the presence of active HBV replication and high infectivity, and spontaneous improvement may occur following HBeAgpositive seroconversion (anti-HBe), with a decline in HBV replication and normalization of ALT levels. This confers a good prognosis and does not require treatment. More frequent monitoring was recommended conditionally (based on limited evidence) in those who already have fluctuating raised ALT or HBV DNA levels (between 2000 IU/mL and 20 000 IU/mL) as they are at a higher risk of progression to active hepatitis and require treatment.

Monitoring during and after treatment The 2015 WHO HBV guidelines (6) recommended at least annual monitoring of ALT, HBeAg (for seroconversion to anti-HBe) and HBV DNA levels (where testing is available), and also NITs such as APRI to assess for progression to cirrhosis. HBV genotyping and resistance testing are not required to guide therapy. This was based on limited data from systematic reviews of multiple clinical trials and observational studies as the minimum and optimal frequency for monitoring treatment response during therapy have not been directly evaluated (Web appendix 2: SR5a – WHO HBV guidelines 2015). This shows that potent nucleos(t)ide

analogues (NAs) with a high barrier to resistance (i.e. tenofovir and entecavir) suppress HBV DNA replication to low or undetectable levels in the majority of persons (around 80% and 50– 70% in HBeAg-positive and -negative persons, respectively) by 24–48 weeks of treatment, with low rates of resistance. The data also suggest that if good adherence can be confirmed, monitoring can be relatively infrequent. However, there is limited success in achieving durable end-points, particularly loss of HBeAg in HBeAg-positive persons or loss of HBsAg. More frequent and careful monitoring was recommended conditionally based on limited evidence in: (i) those with more advanced disease (compensated or decompensated cirrhosis) because the risk of HCC is reduced but not eliminated with treatment, and their higher risk of adverse events; (ii) during the first year of treatment to assess treatment response; (iii) where adherence to therapy is a concern; and (iv) after stopping therapy, especially in the first year to detect severe exacerbations. Retreatment is recommended if there are consistent signs of reactivation (HBsAg or HBeAg becomes positive, ALT levels increase, or HBV DNA becomes detectable again).

10.4. Implementation considerations There are cost implications to regular ALT and DNA monitoring. Where there is limited access to HBV DNA assays, such as in LMICs (particularly rural areas), monitoring will require, at a minimum, serum ALT levels to establish the risk of progression. However, interpretation of disease stage and exacerbations of disease in HBeAg-positive and HBeAg-negative persons is enhanced by concomitant measurement of HBV DNA concentrations. Integrating routine monitoring for HCC alongside routine monitoring for disease progression provides a further opportunity to detect the development of cirrhosis and initiate antiviral therapy to prevent progression to HCC or liver failure.

11. DETECTION OF VIRAEMIC HCV INFECTION – to guide who to treat 11.1. Recommendations Detection of viraemic infection • Directly following a reactive HCV antibody serological test result, the use of quantitative or qualitative NAT for detection of HCV RNA is recommended as the preferred strategy to diagnose viraemic infection. Strong recommendation, moderate/low quality of evidence • An assay to detect HCV core (p22) antigen, which has comparable clinical sensitivity to NAT, is an alternative to NAT to diagnose viraemic infection1. Conditional recommendation, moderate quality of evidence

A lower level of analytical sensitivity can be considered if an assay is able to improve access (i.e. an assay that can be used at the point of care or suitable for dried blood spot [DBS] specimens) and/or affordability. An assay with a limit of detection of 3000 IU/mL or lower would be acceptable and would identify 95% of those with viraemic infection, based on available data.

11.2. Background Detection of antibodies to HCV is used to determine current or past HCV infection (i.e. exposure to HCV infection), and therefore to triage those who require further evaluation to determine if active viral replication is present. Approximately 15– 45% of persons who are infected with HCV will spontaneously clear the infection (267). These persons remain HCV antibody positive but are no longer infected with HCV. Diagnosis of viraemic HCV infection in those who are HCV antibody positive will distinguish persons with viraemic HCV infection and in need of treatment from those who have cleared the infection. This is generally done using NAT technologies to detect HCV RNA, but an alternative and potentially less costly option to NAT is to conduct testing to detect HCV core (p22) antigen.

Nucleic acid testing (NAT) Both quantitative and qualitative methods are available for the detection of viraemic HCV infection. Quantitative NAT has been widely used for measuring viral load and identifying those in need of treatment, as well as in assessing treatment response (5, 15). Qualitative NAT allows for rapid and sensitive detection of the virus as well as evidence of a decline in viral RNA level below a defined threshold. There are currently five quantitative HCV RNA (viral load) assays that are commercially available with another two in the pipeline (268). At present, there has been limited comparison of the two methods. Although NAT technologies are very sensitive and specific for the detection of viraemia, they require sophisticated laboratory equipment and skilled staff. Assays to detect HCV RNA that may be used at or near the point of care have recently become commercially available. A comprehensive review of the HCV diagnostics landscapes by UNITAID is available (268).

HCV core (p22) antigen testing In addition to NAT, it is possible to assess for viraemic infection by testing for HCVcAg – an HCV nucleocapsid peptide 22 (p22), which is released into the plasma during viral assembly and can be detected both early on and throughout the course of HCV infection (269). Serological methods that test for detection of HCVcAg have the potential to be less costly and centralized than NAT, but evaluation has been limited in low-resource settings. There are now several assays commercially available for stand-alone detection of HCVcAg (270). Detection of HCVcAg has also been used as an additional marker in a fourth-generation HCV Ag/Ab serological assay, because HCVcAg is detectable earlier than antibodies to HCV. However, the addition of core antigen was intended to increase sensitivity of the assay in early infection and not to differentiate seropositivity from active viraemic HCV infection.

11.3. Summary of the evidence Two systematic reviews were undertaken, which evaluated the diagnostic accuracy for detection of viraemic HCV infection of (i) qualitative versus quantitative NAT (see Web annex 5.7); and (ii) HCVcAg testing versus NAT (see Web annex 5.8) (271).

Diagnostic accuracy and limit of detection of HCV NAT assays The systematic review identified four eligible studies (272–275) that compared the performance of three quantitative HCV RNA NAT assays to a reference qualitative NAT (two assays used). Although early-generation qualitative NAT assays were able to detect the presence of HCV in plasma at concentrations a full log lower (i.e. about 10-fold less) than quantitative NAT assays, the lower limit of quantification of new versions of quantitative assays is now comparable to most commercial qualitative assays, i.e. 15 IU/mL.

Diagnostic accuracy and limits of detection of HCVcAg assays There were 50 studies that evaluated seven commercial HCVcAg assays. There was significant variation in performance between the different assay brands (Table 11.1) (271). The pooled sensitivity and specificity with 95% CI were: ARCHITECT 93.4% (95% CI: 88.7–96.2) and 98.7% (95% CI: 96.9–99.4); Ortho ELISA 93.2% (95% CI: 81.6–97.7) and 99.2% (95% CI: 87.9–100); and Hunan Jynda 59.5% (95% CI: 46–71.7) and 82.9% (95% CI: 58.6–94.3). The sensitivity for the Lumipulse was 95% (95% CI: 90.2– 99.8) in one study, but specificities could not be calculated. The estimates for the ARCHITECT assay were more homogeneous and precise as this assay has been the most extensively studied. A pooled quantitative analysis of data available from three studies demonstrated a close correlation between HCVcAg and HCV RNA at viral loads above 3000 IU/mL.

The LoD for the most sensitive assay is 3 fmol/L HCVcAg or 0.06 pg/mL, which equates to an LoD of about ~1000–3000 IU/mL by NAT, and is consistent with the analytical sensitivity (LoD) reported by the manufacturer. NAT assays are considered the reference standard for the detection of viraemia, but the quality of studies comparing quantitative versus qualitative assays for detection of viraemia was rated as low because of small numbers of studies and heterogeneity in populations. The overall quality of the evidence for the recommendation to use HCVcAg was rated as low to moderate because of inconsistency and imprecision. TABLE 11.1. Summary of diagnostic accuracy of HCV core antigen assays compared to NAT Index test Sample size (range) Diagnostic accuracy (95% CI) Sensitivity Specificity

Abbott Diagnostics GmbH, ARCHITECT HCV Ag Assay Ortho-Clinical Diagnostics, Ortho ELISA-Ag Bio-RAD Monolisa HCV Ag-Ab ULTRA EIKEN Lumispot HCV Ag Fujirebio Lumipulse Ortho HCV Ag Hunan Jynda HCV Core Ag ELISA DiaSorin S.A. Murex HCV Ag/Ab

20 (11–820) 5 (1–177) 5 (525) 2 (235) 1 (80) 4 (524) 4 (730)

93.4% (88.7–96.2) 93.2% (81.6–97.7) 28.6–95%a 97.5–98.1%a 95% (90.2–99.8)b 59.5% (46–71.7) 50–100%a

98.7% (96.9– 99.4) 99.2% (87.9–99.9) 94.9% (89.9–99.8)b ND ND 82.9% (58.6–94.3) 83.8–100%a

CI: confidence interval; ND = no data. a Meta-analysis not possible. Range of results seen across studies reported. b Result from one study only.

11.4. Rationale for the recommendations Balance of benefits and harms Use of quantitative or qualitative NAT assays for detection of HCV RNA The Guidelines Development Group made a strong recommendation for the use of a NAT assay (either qualitative or quantitative) as the preferred strategy for diagnosis of viraemic HCV infection based on moderate-/low-quality evidence for several reasons: 1. The new generation of quantitative and qualitative assays have the same LoD, which is around 15 IU/mL. However, quantitative assays are a reproducible method to detect and quantify HCV RNA in plasma or serum. A supplementary review of the literature showed that that 95% of those with chronic infection have a viral load >10 000 IU/mL except, temporarily, a minority with partial viral control between 5 and 12 months post infection. Therefore, the range of clinically observed HCV viral loads is rarely below the lower range of the limit of quantification (LoQ) of quantitative assays, and most NAT assays (quantitative or qualitative) will capture the majority of viraemic infections. Although quantitative RNA assays are considered the gold standard assays for the diagnosis and monitoring of HCV, the high cost of these assays and laboratory requirements means that they are not readily available in resource-limited settings. However, new NATs for use at or near the point of care for quantitation of HCV RNA are already available. These devices are easier to use than the laboratory-based NAT assays and can potentially improve access to diagnosis of viraemic HCV infection.

Use of HCV core antigen for detection of HCV RNA The Guidelines Development Group recognized that there is limited access to NAT assays in resource-limited settings and that this represents an important barrier to antiviral treatment. The Group made a conditional recommendation to consider use of HCVcAg assays as an alternative to NAT to diagnose viraemic HCV infection, based on moderate-quality evidence, for several reasons: 1. HCVcAg assays can utilize existing serological testing platforms and are potentially lower-cost options than NAT. They could serve as a more affordable replacement to NAT for HCV detection in the future. Although HCVcAg testing is currently limited to only a few platforms and even those with the highest performance do not reach the sensitivity of NAT, some well-performing HCVcAg assays have high sensitivity (up to 93.4% for certain commercial assays and high specificity (>98%), and good correlation with HCV RNA to an LoD of roughly 3000 IU/mL, which will detect over 95% of chronic HCV infections. However, it was noted that there was wide variation in sensitivity/specificity between assays and also within the same brand of assay for all but the Abbott ARCHITECT.

HCVcAg tests also offer the potential in the future to be applied as a one-step screening test as HCVcAg appears earlier than HCV antibodies (1–2 days after HCV RNA appears), has a high specificity, and so does not require any further confirmatory testing. However, such a strategy would be cost–effective only in very high-prevalence settings.

The risk in the use of HCVcAg is of potentially missing cases due to reduced clinical sensitivity. A further consideration is that it is preferable to select an assay that can be used for both diagnosis of viraemia and for test of cure to simplify the diagnostic pathway. On the basis of the limited current evidence, the HCVcAg assay cannot be recommended as a monitoring test.

Acceptability, values and preferences The values and preferences survey identified preferences for future HCV testing strategies among respondents. Key preferences were for a single-step HCV diagnostic strategy with a low-cost point-of-care test for confirming viraemic infection (48% of respondents). Of these, 52% opted for an HCV RNA test because of its high sensitivity, and 35% for an HCVcAg assay because of its lower cost and ease of use. More than half the respondents were prepared to compromise on sensitivity down to 95% in order to gain a reduction in the price of the test. Forty-seven per cent of respondents also indicated a preference for a test that uses capillary blood and therefore could be more easily performed in point-of-care settings, even at the expense of test sensitivity. A short turnaround time (at least same day) was identified as another key consideration to reduce loss to follow up, cost of transportation, and enable providers to see more patients within a day.

Feasibility The survey of hepatitis testing experience in 19 LMICs found that NAT for HCV RNA is available at a third of the sites but 40% of respondent countries do not have access to NAT for HCV diagnosis in their countries. The HCVcAg assay was not available at any site.

Resource considerations The resources required for quantitative NAT were considered to be substantial, with the cost per test ranging from US$ 30 to US$ 200. Furthermore, the laboratory equipment is expensive and requires technicians with specialized training. The cost of testing for HCVcAg is currently US$ 25–50 (MSF data), which is comparable to qualitative NAT (US$ 43–51), but this is still a major barrier to its use.

11.5. Implementation considerations 1. Near patient or point-of-care (POC) technologies. The development of reliable and affordable POC NAT and HCVcAg tests able to diagnose viraemic infection in field settings will be crucial for expanding hepatitis testing services (see Chapter 17.3.5 – diagnostic innovations). These devices offer the possibility of a same-day diagnosis of viraemic infection either alone or when combined with

an HCV antibody RDT and for test of cure. Since they are also more affordable than the laboratory-based assays, they can potentially improve access to early diagnosis, monitoring and linkage to care and treatment services, as well as reduce loss to follow up. 2. Immediate NAT directly after a positive serological result. The value of prompt testing for viraemia after a positive antibody result was highlighted, as patients with resolved HCV infection following spontaneous clearance can be reassured and those with viraemic infection could be promptly referred for care and treatment. 3. Genotyping. In most countries, there is a mix of HCV genotypes among persons with chronic HCV infection. The 2016 Hepatitis C treatment guidelines (5) provide recommendations on the preferred and alternative DAA regimens by HCV genotype. Therefore, knowing a patient’s genotype is still important for determining the most appropriate treatment regimen. Genotype determination, however, is expensive and not available in all settings. Where genotype information is unavailable, pragmatic decision-making may be required, taking into account the common genotypes circulating in the affected population. However, as pangenotypic regimens become available over the next year, this will no longer be required.

Research gaps • Establish the proportion of patients with chronic HCV infection with low viral loads that may be missed by HCV RNA or cAg assays that have a higher limit of detection (i.e. 3000 IU/mL). Evaluate the diagnostic accuracy, cost, cost–effectiveness and impact of HCVcAg or HCV RNA assays as a one-step diagnostic strategy. Assess the impact of HIV or HBV coinfection or genotype (particularly genotypes 4, 5 and 6, on which there are limited data) on detection of viraemia.

•

12. ASSESSMENT OF HCV TREATMENT RESPONSE - test of cure 12.1. Recommendation Assessment of HCV treatment response • Nucleic acid testing for qualitative or quantitative detection of HCV RNA should be used as test of cure at 12 or 24 weeks (i.e sustained virological response (SVR12 or SVR24)) after completion of antiviral treatment. Conditional recommendation, moderate/low quality of evidence

12.2. Background Detection of HCV viraemia is important to assess the response to treatment (276– 278). Prior to the introduction of curative oral DAA treatment regimens, treatment with interferon (IFN)-based regimens required frequent monitoring of HCV viral load levels during therapy to decide whether treatment should be stopped, or treatment duration could be shortened. Previously, these multiple measurements included a viral load measurement at week 4 of therapy to help predict the efficacy of therapy, and then again at week 12 (early viral response, EVR), and finally at 12 and 24 weeks after completion of therapy to test for cure (sustained viral response, SVR). These multiple assessments are now no longer relevant with the newer DAAs because of the relative infrequency of viral breakthrough and because the rate of viral load decline does not correlate with SVR. In fact, in most persons treated with DAAs, the viral load is undetectable 4 weeks after treatment initiation. In view of the high cost and relative unavailability of NAT testing for HCV RNA, this provides an important opportunity to reduce the frequency of on-treatment laboratory monitoring. HCVcAg testing has also been proposed as an alternative to HCV RNA for the diagnosis of viraemic HCV infection. However, there remains debate about whether HCVcAg can also be used as a tool for assessing the response to HCV antiviral treatment and to test for cure.

12.3. Summary of the evidence The accuracy of HCVcAg for treatment monitoring and to confirm successful viral clearance (test of cure) was assessed by descriptive analysis of five studies (279– 283) of two HCVcAg assays in comparison with HCV RNA NAT (qualitative and/or quantitative) (see Web annex 5.8). All studies were based on patients with mainly genotype 1b infection and on IFN-based therapy. The sensitivity of the HCVcAg assay in EVR ranged from 74% to 100% and specificity from 70% to 100%. SVR was assessed in only two studies with 100% sensitivity and specificity ranging from 94% to 100%. There were only three studies that evaluated the same assay – the Abbott ARCHITECT HCV Ag assay. There were no studies that evaluated the use of HCVcAg assay for monitoring treatment response using DAA IFN-free treatment regimens.

12.4. Rationale for the recommendation Balance of benefits and harms Use of qualitative or quantitative HCV RNA as a test of cure The Guidelines Development Group recommended the use of either qualitative or quantitative NAT detection of HCV RNA as a test of cure at 12 weeks (or 24 weeks if 12 weeks is not possible) after completion of treatment. As shown in Chapter 11, these assays have a broad dynamic range from 12 to 7 700 000 IU/mL, and the reviews showed analytical sensitivity as low as 5 IU/mL for qualitative HCV RNA by NAT. Although either assay was recommended, the lower cost of qualitative assays for HCV RNA makes them preferable to quantitative NAT as a test of cure at 12 weeks (284–287).

Use of HCVcAg as a test of cure The Guidelines Development Group recognized that dependence on detection of HCV RNA by NAT to assess response to HCV antiviral treatment and test of cure, especially in remote settings, could be a barrier to the setting up of hepatitis C treatment and testing services. However, the data on HCVcAg in treatment monitoring and assessment of test of cure (SVR) was considered to be too limited to recommend its use as a substitute for HCV RNA.

Timing of test of cure The Guidelines Development Group recognized that in the new era of treatment with curative DAA regimens, monitoring of viraemia during therapy with HCV RNA by NAT may no longer be necessary (288), and that a single negative test of viral load at 12 weeks after completion of therapy (SVR12) is now the benchmark for assessing treatment outcome and cure used in all clinical studies of DAA-based regimens.

Acceptability, values and preferences In the values and preferences survey of implementers and users of hepatitis testing services, almost half of the survey respondents expressed a preference for the test of cure to be performed 12 weeks after completion of HCV therapy because this was the earliest time point to reliably establish cure. However, one third expressed a preference for this to be performed more promptly after completion of treatment – at 4 weeks (20% of respondents) and 8 weeks (16%).

Feasibility In the values and preferences survey, HCVcAg assay was reported as not available at any of the sites, and 40% of respondents also reported that they did not have access to HCV NAT in their countries.

Resource considerations The availability of validated POC NAT assays, and further reduction in costs of both qualitative and quantitative NAT, will be critical to improve access to diagnosis and monitoring in LMICs.

12.5. Implementation considerations 1. Re-infection. The possibility of reinfection with HCV after successful treatment should be considered, and persons treated but who are still at active risk (e.g. current PWID) should be advised to retest annually for HCV RNA. 2. Timing of test of cure. A test of cure at 24 weeks (SVR24) after completion of treatment may be considered as an alternative SVR time-point, if SVR12 is not possible. Similarly, in populations for which there are limited data on the correlation between SVR12 and SVR24, e.g. patients with cirrhosis, HIV/HCV coinfection and other immunocompromised states, SVR24 may be considered. 3. Impact of co-morbidities. Clinical judgement based on the patient’s clinical circumstances, such as presence of HIV coinfection, cirrhosis or renal impairment, potential drug interactions and clinical well-being during treatment, may necessitate more frequent monitoring for side-effects and disease progression.

Research gaps • The impact of HIV or HBV coinfection and genotype on diagnostic accuracy of HCVcAg and quantitative/qualitative HCV RNA NAT as a test of cure should be assessed. The kinetics of HCVcAg with DAA treatment should be evaluated, and an optimal time-point identified to test for cure with DAA regimens using HCVcAg. The distribution of HCV viral load in the setting of viral rebound should be assessed to inform optimization of HCVcAg detection. Specific situations where quantitative NAT assay may be indicated, i.e. shortened DAA treatment course to 8 weeks, should be evaluated in those with lower baseline HCV RNA levels. The correlation between SVR12 and SVR24 should be evaluated in populations where there are more limited data, e.g. patients with cirrhosis, HIV/HCV coinfection and other immunocompromised states.

•

• •

•

13. USE OF DRIED BLOOD SPOT SPECIMENS FOR SEROLOGICAL AND VIROLOGICAL TESTING 13.1. Recommendations Topic Serological testing Recommendations • The use of DBS specimens for HBsAg and HCV antibody serology testing1 may be considered in settings where: - there are no facilities or expertise to take venous whole blood specimens; or - RDTs are not available or their use is not feasible; or - there are persons with poor venous access (e.g. in drug treatment programmes, prisons). Conditional recommendation, moderate (HBV)/low (HCV) quality of evidence The use of DBS specimens to test for HBV DNA and HCV RNA for diagnosis of HBV and HCV viraemia1, respectively, may be considered in settings where: - there is a lack of access to sites or nearby laboratory facilities for NAT, or provision for timely delivery of specimens to a laboratory; or - there are persons with poor venous access (e.g. in drug treatment programmes, prisons). Conditional recommendation, low (HBV)/moderate (HCV) quality of evidence

Detection of viraemia (nucleic acid testing)

•

Well-functioning laboratory specimens referral network and system for return of results should be in place to maximize the impact of DBS specimens. There are currently few assays where the manufacturer’s instructions state that DBS specimens are validated for use. Therefore, currently use of DBS specimens would be considered “off-label”.

13.2. Background Significant scale up in access to hepatitis testing and treatment will require further simplification of the process of diagnosis and monitoring, and methods to facilitate access to testing, especially in decentralized settings, and among vulnerable populations worldwide, such as PWID and people in prison. DBS is an alternative specimen collection method that does not require venepuncture, and is being increasingly used to facilitate access to serological testing and NAT for HIV, hepatitis B and C, and other infectious diseases (289–292), particularly in remote and underresourced regions with poor access to laboratory services, as well as for large epidemiological surveillance studies. DBS sampling involves obtaining a whole blood specimen, usually by capillary finger-stick (or heel-prick in infants), and embedding the drops of blood onto filter paper, or by pipetting venous blood onto filter paper. DBS specimens can then be transported from remote areas to a laboratory by standard means, e.g. posted to a laboratory, where testing would take place. The simplicity and relative ease of specimen collection, preparation, transport and storage make DBS specimens a potential option for serological testing and NAT in low-resource settings (289).

An increasing number of studies have been undertaken to validate the use of DBS specimens to test for HBsAg and antibodies to HCV, and NAT for HBV DNA and HCV RNA (291–293), including systematic reviews on HCV RNA detection using DBS specimens (294), and on the uptake of HCV screening (295).

13.3 Summary of the evidence Four updated systematic reviews and meta-analyses were undertaken to evaluate the diagnostic accuracy and impact of using DBS specimens compared to venous blood specimens for hepatitis B and C serological testing and NAT (see Web annex 5.9). The reviews evaluated the pooled sensitivity, specificity, positive and negative likelihood ratios, as well as the impact and duration of different storage conditions. Ten studies were included in the meta-analysis for HBsAg (290, 293, 296–303), 17 studies for HCV antibody (256, 293, 298, 299, 304–317), 10 studies for HBV DNA (293, 301, 318–325), and 9 studies for HCV RNA (305, 311, 312, 326–331) based on an update of one study in an existing review (294). The summary of results is shown in Table 13.1.

Serology The pooled sensitivity for detection of HBsAg in DBS specimens compared to plasma or serum specimens was 92.9% (95% CI: 86.2–96.5%) and specificity 99.9% (95% CI: 96.2–99.7%), and for antibodies to HCV the sensitivity was 98% (95% CI: 94–99%) and specificity 99% (95% CI: 97–100%). Impact of storage conditions. Most studies used storage conditions not applicable to typical field conditions (i.e. storage in a freezer or refrigerator). Those studies that investigated variation of results after storage of specimens in different conditions found that specimens could become false-positive with longer exposure at ambient temperatures for HCV antibody (305, 326, 332) and HBsAg (302, 303).

Nucleic acid testing Overall, studies and data were more limited for NAT than for serology testing, especially for HBV DNA, and had smaller sample sizes and were of lower quality. Nine studies contributed to the quantitative analysis of the diagnostic accuracy of HBV DNA measurement in DBS specimens compared to serum samples. Pooled sensitivity for HBV DNA measurement in DBS specimens was 96% (95% CI: 90–98%) and specificity 99% (95% CI: 55–100%), and for HCV RNA, 96.0% (95% CI: 93.4–97.6%) and 97.7% (95% CI: 94.7–99.0%). The descriptive review also shows a good correlation and strong association between quantitative values for HBV DNA on DBS specimens and in serum samples. Impact of storage conditions. No study reported on storage conditions longer than 24 hours at room temperature for DBS specimens to test for HBV, but

several studies that varied storage conditions for individual specimens found no effect on the qualitative result of these assays (301, 323, 324). Several HCV RNA studies stored DBS specimens at ambient room temperature. While these storage conditions did not affect accuracy, and RNA positivity could still be detected with DBS specimens, quantitative signals decreased over time in two studies (305, 332). The quality of evidence for recommendations to use DBS for HBV and HCV serology was rated as low to moderate for HBV and low for HCV, and for HBV and HCV NAT, it was low for HBV and moderate for HCV. Table 13.1. Summary of diagnostic performance of DBS specimens for serological and NAT testing DBS for HBsAg DBS for anti-HCV DBS for HBV DNA1 DBS for HCV RNA

No. of included studies Total sample size

10 (SR), 9 (meta-anal) 2481

18 (SR), 14 (meta-anal) 4524 98% [94–99] 99% [97–100] Storage at –20 °C associated with less variation compared to RT

10 (SR), 9 (meta-anal) 608 96% [90–98] 99 [55–100] Not possible to calculate because all accuracy studies conducted at –20 °C No effect on accuracy if 4–37 °C for ≤7 days (2 studies)

9 (SR & meta-anal) 1250 96.0% [93.4–97.6] 97.7% [94.7–99.0] Better result at –20 °C compared to RT; conflicting results re deterioration of sample at RT Conflicting results re deterioration over time

Overall pooled 92.9% sensitivity (95% CI) [86.2–96.5] Overall pooled 99.0% specificity (95% CI) [96.2–99.7] Impact of storage Cold chain: SN 78.7% [70–85] SP: 98.6% [68–100] ≥RT: SN: 96.1% [92–98] SP: 99.7% [98–100] Accuracy not affected if RT for ≤15 days (1 study) or 63 days (another study)

Impact of duration of storage

Accuracy not affected if RT for ≤3 days (1 study) or ≤6 days (another study) or ≤60 days (another study)

DBS: dried blood spot; meta-anal: meta-analysis; RT: room temperature; SN: sensitivity; SP specificity; SR: systematic review 1 HBV DNA testing is not recommended for ruling out HBV infection if HBsAg is positive. HBV DNA detection can be used to explore occult HBV infection in persons testing negative for HBsAg. A large proportion of HBV-infected persons have a low HBV replication level (inactive carriers).

13.4. Rationale for the recommendations Balance of benefits and harms Benefits of use of DBS. (Table 13.2) 1. The Guidelines Development Group recognized that the principal benefit of DBS specimens is their potential to facilitate greater access to testing in settings where venepuncture and laboratory facilities are not easy to access. This is largely because of the relative ease of specimen collection with avoidance of venepuncture, easier handling that does not require high skill, transportation with a lower biohazard risk, and easier storage options. The systematic reviews also showed generally high diagnostic accuracy of DBS specimens for both serology testing and NAT (although the evidence was more limited for evaluation of DBS for NAT, especially HBV DNA), and good precision based on low- to moderatequality evidence. For HBsAg, sensitivity was 92.9% (range 86.2–96.5%), which is below the WHO prequalification standard for RDTs, and may lead to cases being missed. Although data are still limited, DBS specimens are generally stable over time and maintain good accuracy in conditions with higher temperatures or humidity.

Risks of use of DBS. The Guidelines Development Group also recognized that several key caveats remain with the use of DBS specimens, as summarized in Table 13.2. 1. The main disadvantage is that the assay manufacturers have not yet validated their existing commercial assays with DBS specimens, which is required for regulatory approval for use of this specimen. The minimum performance criteria of a DBS (whole blood) specimen for HBV and HCV serology and NAT are not well established. Acceptable storage conditions need to be determined and validated by the manufacturer of the assay. The evidence review was not able to support the use of certain commercial assays over others for DBS testing, or to suggest minimum performance criteria that should be retained for HBV and HCV screening on DBS specimens. There is also conflicting evidence on whether DBS increases the uptake of hepatitis testing among different vulnerable populations, as some studies have not confirmed this trend (333–335).

Overall, despite these caveats, the Guidelines Development Group considered that the procedural advantages make DBS specimens a good option for HBV and HCV testing in remote settings or specific populations. A conditional recommendation was therefore made to consider the use of DBS specimens as an option for both HBV and HCV serology and/or NAT, especially in specific settings, where there are either no facilities or expertise to take venous

blood specimens, or there are persons with poor venous access (e.g. PWID and people in prisons). Use of DBS specimens for serological diagnosis was also recommended when RDTs are not available or their use is not feasible.

Assessing response to antiviral treatment The Guidelines Development Group did not make a recommendation for the use of DBS specimens to assess response to antiviral treatment because of the lack of studies that have specifically addressed this question. Preliminary evidence shows that patients failing DAAs have high viral loads at 12 or 24 weeks after treatment completion, suggesting that testing with DBS may be feasible.

Limits of detection For HBV DNA, existing WHO guidance (6) defines the HBV DNA threshold for initiation of treatment as ≥20 000 IU/mL, and does not recommend treating those with persistently normal ALT levels and low levels of HBV DNA replication (HBV DNA <2000 IU/mL) (6). Therefore, assays for use in the field would not need to detect HBV DNA below 2000 IU/mL. Most individual studies suggest that the sensitivity of HBV DNA detection above 2000 IU/ mL is good, and the LoD in DBS specimens is 900–4000 IU/mL, which means they would therefore be able to identify the majority of patients who require antiviral treatment. For HCV RNA, it is estimated that the majority of people with viraemic HCV infection will have high viral loads >10 000 IU/mL. Although the thresholds at which HCV RNA can be detected using DBS specimens are not well characterized, the evidence suggests that qualitative detection of HCV RNA using DBS specimens is possible and accurate within this range.

Acceptability, values and preferences The majority of respondents (implementers and users of hepatitis testing services) to the values and preferences survey from LMICs expressed a preference for DBS sampling because of the potential enhanced access to testing. They also considered that it was equally important for serology and NAT.

Feasibility and programmatic experience There is limited programmatic experience with the use of DBS specimens for hepatitis B and C testing, but DBS has been incorporated into several screening programmes and has been used to scale up hepatitis C testing in certain populations in the UK (336, 337) and France (338). It has also been used for testing of at-risk children for HBsAg at 12 months in the UK, and incorporated into research studies (see Box 13.1). These pilot programmes have already used DBS for antibodies to HCV on existing serological platforms without validation by the assay’s manufacturer (337, 339, 340).

Resource use DBS sampling may reduce costs associated with sample collection, storage and transportation, potential for batch testing in a centralized laboratory, in addition to staff costs by facilitating task-shifting to lay workers.

Table 13.2. Summary of key benefits and challenges of DBS specimens Benefits Challenges and concerns

Ease of specimen collection. DBS specimen collection involves pricking a finger or a heel without the need for venepuncture, and so avoids the need for a trained health worker.

Storage conditions. Few studies have systematically examined the effects of storage and transport conditions on the accuracy of results from DBS specimens. However, some data suggest that there may be instability of results when stored for prolonged durations (more than 14 days) at high temperatures (room temperature and above) and humidity, particular for serological testing. Manufacturers’ validation. Most manufacturers have not validated the use of their commercial assays with DBS specimens. In particular, procedures for pre-analytical treatment of specimens is not standardized, such as the type of filter paper, volume of capillary whole blood to be used, type of elution buffer and elution procedure. The impact on assay performance is uncertain but may result in potentially lower sensitivity/specificity. Lack of assays with regulatory approval. At present, there are few if any assays that have regulatory approval for the use of DBS as a specimen type for any HBsAg, anti-HCV, HBV DNA or HCV RNA assay.

Ease of sample transport. Specimen transport and logistics are simplified as, apart from the advantage of avoiding venepuncture, DBS specimen handling does not require high skill and the biohazard risk is thought to be less (292).

Minimal training required. When DBS specimens are used for serological testing, the need for training in how to use and interpret RDTs is eliminated. Lay providers can be trained to take capillary whole blood samples without having to be trained in the use and reading of RDTs. Facility to allow multiplex testing. Multiplex testing of multiple diseases may be possible, such as HIV/HBV/HCV in combination using the same DBS specimen card. Both serological testing and NAT can be conducted from multiple spots on the same DBS card at the same laboratory, thereby avoiding the need for collecting a second sample from the patient.

Assay cut-offs. The use of a DBS specimen may require adjustment of the assay cut-off to determine test positivity for serological screening, as DBS specimens use a small volume of blood.

Laboratory capacity. The actual laboratory work can be more difficult with a DBS specimen compared to a serum/plasma specimen, because it involves manual specimen processing, and requires a laboratory experienced in and competent at handling and processing these specimens. It also involves the need to maintain quality assurance of testing a specimen potentially off-label from the manufacturers’ validated specimen types (293).

Box 13.1. Examples of programmatic use of DBS • A voluntary counselling and testing service in France (The CheckPoint-Paris from the Kiosque) has offered rapid tests for screening and DBS for HCV RNA to confirm active infection since 2010. Hepatitis testing using DBS has also been adopted by associations such as the UK-based Hepatitis C Trust, le Réseau Hépatites LR in France, and by community pharmacists in the UK. • In 2012, the UK National Institute for Health and Care Excellence (NICE) recommended the use of DBS in certain settings for people with poor venous access and where there may be no facilities or expertise to take venous blood samples, such as prison and drug treatment services. • In France, guidelines from the Haute Autorité Sanitaire and AFEF-ANRS also recommended the use of DBS tests as an alternative to venous blood tests. However, the lack of standardization has limited the adoption of DBS, and as a result, there has been no clear recommendation for expansion of DBS testing. • In Scotland, 26% of new hepatitis C diagnoses during 2009–2013 were made in specialist drug services where DBS testing was introduced. • In a study of DBS sampling from substance misuse settings and prisons in Wales, less than 50% of those who were positive for anti-HCV returned for follow-up RNA testing, suggesting a low retention rate in care for those screened in such settings. Experience from DBS testing programmes in addiction centres in France has also observed a low rate of returning for results and linking to specialist care and treatment.

13.5. Implementation considerations 1. Settings for DBS implementation. The choice of whether to use DBS sampling for hepatitis B and C serological testing or NAT or both will depend on the health-care setting and infrastructure, and epidemiological context. Different programmes may opt for varying combinations: (i) DBS serology + DBS NAT (remote settings, hard-to-reach populations and those with poor venous access); (ii) RDT serology + DBS NAT (clinics, e.g. antenatal services); or (iii) EIA serology + plasma-based NAT (urban settings or larger hospitals). Therefore, if good-quality RDTs are available that can be performed using capillary blood, then the focus may be more on prioritizing DBS for NAT testing of HBV DNA and HCV RNA. However, if RDTs are not available and there are no facilities or expertise to take venous blood samples, then DBS testing may be equally important to increase access to serological testing as well as NAT and, conveniently, both could be performed from the same specimen if multiple spots are taken. Use of DBS may also be useful when large numbers of individuals are being tested at the same point in time, e.g.

drug treatment centres, prisons, or where polyvalent screening for multiple diseases, such as HIV/HBV/HCV, is being undertaken, but where multiplex RDTs for this purpose are not available or are more costly (298). 2. Validation of DBS with manufacturers’ assays. The use of DBS specimens has not yet been validated by assay manufacturers with their commercial assays, and under different storage and transport conditions. Addressing this is a priority for implementation, together with access to appropriate laboratory facilities and experience. 3. Laboratory QA/QC. The adoption of DBS sampling in a hepatitis testing programme requires the availability of a centralized laboratory experienced in and competent at handling and processing this sample type, as well as maintaining QA of testing a specimen potentially off-label from the manufacturers’ validated specimen types. 4. Training. Lay workers will need to be trained to perform finger-prick DBS, and systems put in place for timely and efficient communication of results. 5. Reducing loss to follow up. The mobility and instability of some vulnerable populations has raised concerns that DBS may be associated with a low rate of returning for results and linkage to care and treatment. This is suggested by reports of low rate of return and linkage to care and treatment in drug treatment programme centres in France, particularly with DBS testing.

Research gaps A major constraint to implementation of DBS sampling is the limited programmatic experience with its use for hepatitis testing in different settings. Priority areas for research and development include development of manufacturers’ guidance and regulatory approval of commercial assays using DBS specimens, and the establishment of large-scale demonstration projects of DBS for hepatitis serological testing and NAT in different settings. Specific research areas include the following: • Larger diagnostic accuracy and validation studies should be conducted on the use of DBS specimens for serology and NAT with optimal assay cut-offs, and across a range of storage and transport conditions common in the field (i.e. no cold chain, high and low humidity). This should also include additional studies in HIV-coinfected patients. The optimal preparation of DBS specimens prior to analysis should be studied. This includes differences between capillary versus venous blood, most appropriate volume of capillary whole blood and best type of filter paper to be used.

•

The use of DBS should be validated for monitoring of treatment response and HCV test of cure (SVR at 12 or 24 weeks) post DAA therapy, including threshold for detection. This includes validation of the rate of degradation of HCV RNA and detectability when stored at ambient temperatures and high humidity for different time periods. The diagnostic performance and impact on linkage to care of one DBS specimen card for serology and NAT should be compared with POC HCV RNA NAT or HCVcAg technologies in different settings, including mobile and outreach testing programmes and in prisons.

•

14. IMPROVING THE UPTAKE OF TESTING AND LINKAGE TO CARE AND PREVENTION 14.1. Recommendations Topic Uptake of testing and linkage to care Recommendations • All facility- and community-based hepatitis testing services should adopt and implement strategies to enhance uptake of testing and linkage to care. Strong recommendation, moderate quality of evidence The following evidence-based interventions should be considered to promote uptake of hepatitis testing and linkage to care and treatment initiation: (Conditional recommendations) - Peer and lay health worker support in community-based settings (moderate quality of evidence). - Clinician reminders to prompt provider-initiated, facility-based HBV and HCV testing in settings that have electronic records or analogous reminder systems (very low quality of evidence). - Provision of hepatitis testing as part of integrated services within mental health/substance use services (very low quality of evidence).

•

14.2. Background Uptake of testing and linkage to care are both essential initial components of the hepatitis B and C care continuum (Fig. 1.1). However, currently, levels of uptake of testing for hepatitis B and C are very low (341, 342), and a large proportion of people living with viral hepatitis B and C are unaware of their infection, especially in LMICs, and those from vulnerable populations, such as PWID, sex workers or migrants. Poor linkage to care and loss to follow up after receiving a diagnosis of hepatitis B or C is a further challenge, contributing to delayed treatment initiation. Population-level data on the care continuum for viral hepatitis is limited, but even in high-income countries, only a small fraction of the estimated population living with HBV or HCV is ultimately treated and achieves viral suppression (341, 342). Those who test negative, if at continuing high risk, as well as those who test positive, need linkage to prevention services and HBV vaccination. Without linkage to prevention, treatment and care, testing and learning one’s hepatitis B or C status has limited value. Suboptimal linkage to prevention, care and treatment results in avoidable morbidity and mortality, poorer treatment outcomes, increased cost of care, and preventable transmission Multiple factors may hinder the successful uptake of testing and linkage to care and prevention. These include patient-level factors (such as depression, lack of social or family support, and fear of disclosure), as well as structural or economic factors such as stigma and discrimination, distance from care sites, lack of or cost of transportation,

and long waiting times at the facility) (343). Hepatitis C also disproportionally affects individuals with comorbid mental health or substance use issues. Traditionally, services for hepatitis, mental health and substance use have been provided by separate clinicians or teams often located in different health facilities, which may contribute to HCV treatment dropout and/or treatment failure (344). Optimizing the impact of effective treatments and prevention will require interventions to both expand uptake of testing and improve linkage to care and retention across the care continuum, from initial screening to treatment initiation and viral suppression (HBV) or cure (HCV). Such interventions may vary based on the local context, including the health-care delivery system, geography and target population. There are several well-established evidence-based interventions that improve linkage to care and treatment of people who have received an HIVpositive diagnosis, and were included as recommendations in the WHO 2015 HIV consolidated testing guidelines (11) and the 2016 ARV consolidated guidelines (23), which may also apply to viral hepatitis care and prevention.

14.3. Summary of the evidence A systematic review was undertaken to assess the impact of different interventions to enhance five key steps along the continuum of care for chronic viral hepatitis – screening, linkage to care, treatment uptake, treatment adherence, and ultimately viral suppression. Fifty-four studies were included in the review, of which 37 studies addressed interventions and outcomes across the HCV care continuum, 15 across the HBV care continuum, and two across both (334, 345–397) (see Web annex 5.11) (398). Thirty-three studies were included in a meta-analysis that generated pooled effect size estimates for different outcomes. Interventions to improve retention along the HBV continuum of care were limited to promoting testing and linkage to care, while interventions along the HCV continuum of care addressed all five steps. Interventions to address adherence, viral suppression, and uptake of HCV testing were the best studied, but there were few methodologically rigorous studies for promoting linkage to care, and particularly few studies on HBV. All studies except one (370) were from high-income countries. Most existing studies were rated as being of low or very low methodological quality, because of risk of bias due to study design issues, and a high degree of heterogeneity across studies.

Promotion of HBV testing by lay health workers Nearly half (7/15) of the interventions to improve the uptake of HBV testing involved lay health worker interventions (348, 354, 368, 376, 388–390). The majority of these were one-time activities that delivered educational content tailored to a particular community’s cultural and social context, mainly Asian communities in the United States or Canada. Pooled meta-analysis from six studies showed that a single HBV test promotion intervention by lay health

workers increased HBV testing rates compared to groups that received no or unrelated educational interventions (relative risk [RR] = 2.68 [1.82–3.93]). The quality of evidence was rated as moderate.

Clinician reminders to prompt HCV testing during clinical visits Unlike interventions to improve HBV testing, which were primarily delivered in community settings, all 11 of the interventions to improve HCV testing either targeted health-care providers or took place at established health-care facilities (333, 334, 358, 360, 362, 363, 371, 374, 379, 385, 386). Providers were prompted by reminders to either order HCV tests if patients belonged to a high-risk birth cohort (371), reported risk behaviour (360) or both (374), using reminder stickers attached to patient charts or in an electronic medical records system (371). These studies found that clinician reminders to prompt HCV screening during clinical visits substantially increased HCV testing rates compared to no clinician reminders (RR = 3.70 [95% CI: 1.81–7. 57]). The quality of evidence was rated as very low.

Integrated care between mental health and HCV treatment specialists Several studies evaluated interventions providing “coordinated”, “integrated”, or “multidisciplinary” care to improve treatment adherence and viral suppression in patients with mental health issues. (345, 352, 359, 365, 369, 381). Three RCTs demonstrated that interventions facilitating referral and scheduling to specialist sites increased patient attendance at HCV specialist visits (RR = 1.57 [95% CI: 1.03– 2.41], moderate-quality evidence). Individually tailored mental health counselling and motivational therapy to treat mental health and/or substance use issues also increased the number of patients who were regarded as eligible for treatment compared to usual care (OR = 3.43, 95% CI: 1.81–6.49). Coordinated care between mental health and treatment specialists along with psychological therapy and counselling for patients with mental health and/or substance use comorbidities increased HCV treatment initiation (OR = 3.03 [95% CI: 1.24–7.37]), improved treatment completion (RR = 1.22 [95% CI: 1.05–1.41]), and increased SVR (RR = 1.21 [95% CI: 1.07–1.38]) compared to usual care. Nurse-led therapeutic educational interventions also improved treatment completion and increased SVR). The quality of evidence was rated as low to very low.

Interventions to promote linkage to care for HIV There are several well-established evidence-based interventions that improve linkage to care and treatment of people who have received an HIV-positive diagnosis, and were included as recommendations in the WHO 2015 HIV consolidated guidelines on HIV testing services (11) and the 2016 ARV consolidated guidelines (23). Box 14.1 outlines some of these approaches, which may also be applicable to viral hepatitis care and prevention.

Box 14.1. Good practices for promoting linkage to care from HIV testing services • • • • • • Comprehensive home-based testing, which includes offering home assessment and homebased treatment initiation; Integrated services, where testing, prevention, treatment and care, TB and STI screening, and other relevant services are provided together at a single facility or site; Providing on-site or immediate testing with same-day results; Providing assistance with transport, such as transportation vouchers, if the treatment site is far from the testing service site; Decentralized treatment provision and community-based distribution of treatment; Support and involvement of trained lay providers who are peers and act as peer navigators, expert patients/clients, and community outreach workers to provide support, and identify and reach people lost to follow up; • • Intensified post-test counselling by community health workers; Using communication technologies, such as mobile phones and text messaging, which may help with disclosure, adherence and retention; • Providing brief strengths-based case management, which emphasizes people’s selfdetermination and strengths, is client-led and focuses on future outcomes, helps clients set and accomplish goals, establishes good working relationships among the client, health worker and other sources of support in the community, and provides services outside of office settings; • Promoting partner testing may increase rates of testing and linkage to care. Source: Consolidated guidelines on HIV testing services. Geneva: WHO; 2015 (11).

14.4. Rationale for the recommendations Balance of benefits and harms The Guidelines Development Group recognized that poor uptake of viral hepatitis testing and linkage to care is a major barrier to access to care and treatment. To expand access to testing and treatment, programmes need to not only make use of multiple testing approaches at the facility and community levels, but also adopt interventions to promote optimal linkage to prevention and treatment. The Guidelines Development Group made a strong recommendation for the general adoption and implementation of a series of relatively simple, low-cost but effective strategies (promotion of testing by lay health workers, clinician reminders and coordinated care between hepatitis and mental health specialists) to enhance uptake of hepatitis testing and linkage to care, based on generally low-quality evidence from the systematic review.

Specifically, promotion of HBV screening activities by lay health workers increased HBV testing uptake, while clinician reminders to prompt HCV screening during clinical visits increased HCV testing rates. Coordinated care between hepatitis and mental health specialists along with psychological therapy and counselling for patients with mental health and/or substance use comorbidities also increased HCV treatment initiation and treatment completion, and resulted in higher SVR rates. Integration of services among certain populations of individuals with HCV such as PWID may also be useful. The Guidelines Development Group also considered evidence from recent systematic reviews on interventions to improve linkage to care following HIV testing, which was considered relevant to hepatitis care and treatment services.

Feasibility, acceptability, and resource use Education and support for peer and lay health-care workers. The findings are also consistent with the growing body of evidence demonstrating that lay health workers effectively perform a range of interventions that would otherwise be undertaken by trained medical personnel, strengthen service delivery capacity in a variety of clinical settings in LMICs (399–402), and are critical to supporting decentralization of services and nonfacility-based testing. Evidence supports such peer-led interventions as being feasible and acceptable to both those individuals screened and lay health workers themselves (403). The low-cost nature of this intervention could facilitate its use in resource-limited settings. The lay health workers in the seven studies received training in order to help tailor the educational intervention; this training component was relatively simple and of low cost. Clinician reminders. Clinician reminders are consistent with the broader shift towards standardizing clinical practice, including provider-initiated screening and systems-based approaches to improving clinical outcomes. Implementation is relatively easy and similar systems have demonstrated effectiveness in multiple disease modalities, such as breast (404) and colorectal cancer screening (405).

Integrated care. Integrating HCV screening and treatment with mental health and substance use services is feasible and acceptable to targeted clients (195, 406). Chapter 17 on service delivery provides a range of examples of integrated care in different settings promoting linkage to hepatitis care. While the interventions addressing multidisciplinary or integrated care in the evidence review were diverse, a likely key contributor to improved outcomes was co-location and coordination of services.

Costs and cost–effectiveness None of the studies identified in the systematic review reported estimates of the direct cost or cost–effectiveness of interventions. However, effective linkage to hepatitis care and treatment following a positive diagnosis would be expected to improve programme effectiveness, support earlier treatment initiation and reduce loss to follow up before treatment initiation, thus resulting in potential cost savings along the continuum of care.

100

14.5. Implementation considerations 1. Policies on linkage to care. Proactive linkage approaches are a critical component of comprehensive hepatitis testing services. Countries should ensure that they have specific policies and strategies to improve and prioritize linkages between hepatitis testing and prevention, treatment and care services. Interventions that impact on multiple steps along the care continuum will generally be more resource efficient. The effectiveness of linkage will vary for different testing approaches. 2. Linkage to prevention services. As for HIV (11), a range of prevention services should be available for those diagnosed with hepatitis, as well as for those who test negative. Linkage to prevention services for people who test HBV or HCV negative is not well documented or studied. Supporting linkage to prevention services is particularly important for those with high ongoing risk, such as PWID and serodiscordant couples. 3. Monitoring and evaluation. Monitoring people’s linkage following hepatitis testing is critical to strengthening the treatment and prevention cascades. The success of linkage should be measured by enrolment in care and not by intermediary process indicators such as the number of referrals issued, and areas identified for improvement. Without strategies that ensure linkage and enrolment in care, the effect of hepatitis testing in reducing HBV or HCV transmission, morbidity and mortality cannot be fully realized.

Research gaps Most existing studies were rated as being of low or very low methodological quality, and there has been a lack of methodologically rigorous studies, in particular, on interventions to promote uptake of HBV and HCV testing, linkage to HBV treatment uptake, and HCV treatment in the era of DAA therapy. There is a need for studies evaluating the effectiveness, costs and cost–effectiveness of packages of different interventions and combinations of interventions to optimize engagement and retention, and treatment outcomes for people living with HBV and HCV along the continuum of care, especially in LMICs.

101

PART 3: IMPLEMENTATION Laboratory testing: • How to organize laboratory testing services for viral hepatitis

Service delivery: • • • • Pre- and post-test counselling Service delivery approaches to hepatitis testing Testing issues in priority populations Strategic planning for implementing testing services and approaches

102

15. IMPLEMENTING LABORATORY TESTING SERVICES FOR VIRAL HEPATITIS 15.1. Key elements for national testing services The efficient coordination of testing services at the national level is important for an effective and sustainable national hepatitis testing programme. Table 15.1 summarizes the key elements for countries to consider while planning or further expanding testing for viral hepatitis4 to ensure the quality and accuracy of testing. Table 15.1. Key elements to consider while planning and expanding hepatitis testing services Section 15.2 National framework for viral hepatitis testing National testing policy National hepatitis strategic plan (including testing) National regulatory mechanisms (for diagnostic products, staff and services) National reference laboratory Section 15.3 Building capacity for testing services Human resource management Inventory management (procurement and supply chain) Storage and transportation Equipment management Laboratory information management systems Section 15.4 Product selection Monitoring testing algorithms Post-marketing surveillance of diagnostics Section 15.5 Assuring the quality of testing services Quality management system Quality control Personnel and training Section 15.6 Assuring the safety of testing services Facilities and safety

Adapted from Asia Pacific strategy for strengthening health laboratory services (2010–2015). Manila/Delhi: WHO; 2010 (http://www.searo.who.int/about/administration_structure/cds/BCT_Asia_Pacific_Strategy10-15.pdf, accessed 06 February 2017). Development of national health laboratory policy and plan. Manila/Delhi: WHO; 2011 (http://www. wpro.who.int/health_technology/documents/docs/Nationalhealthlab2_0F38.pdf?ua=1, accessed 06 February 2017).

103

A national technical working group should be developed to discuss and agree on each of the different laboratory aspects. It should comprise representatives from patient groups and civil society, public and private testing providers, national governmental agencies (national regulatory authorities, directorate of laboratory services, viral hepatitis programmes), laboratory specialists, programme experts, other implementing partners and nongovernmental agencies.

15.2. National framework for viral hepatitis testing A framework approach is useful for guiding national authorities on how to arrange testing services that allow access to all populations that would benefit most from testing and linkage to prevention and antiviral treatment.

15.2.1. National hepatitis testing policy A national testing policy for viral hepatitis is a statement of intent to provide testing services for viral hepatitis. It should set out the goals and objectives of the national testing services and define who is responsible for testing services at each level of the tiered testing network (see Fig. 15.2). It should be complementary to the wider national health policy, with a link to the justice system, given that many key populations at risk of acquiring viral hepatitis live in prisons and other closed settings. The national testing policy may be used to engage and build consensus through active participation with stakeholders for its development. This allows for all implementing partners to understand the national priorities for testing in all settings to reach equity in testing coverage. This document will contribute to the development of a national strategic plan for viral hepatitis that includes testing.

15.2.2. National hepatitis strategic plans (including testing) A national strategic plan for viral hepatitis (16) that includes testing should describe how practically testing services will be established and delivered to support different testing objectives, i.e. diagnosis, prevention, surveillance and treatment. It allows for planning of public and private testing services in the context of the different tiers of the testing network (level 0 through to level 4; see Fig. 15.2). Development of national and regional plans should include all relevant stakeholders involved in organizing the structure and network of facilities for hepatitis testing. Strategic planning for testing services requires first a situation analysis (or needs assessment) to identify and map all existing facilities (public and private; facilityand community-based) that have the capacity to undertake viral hepatitis testing, with an assessment of their organizational structure, infrastructure, technical and human resources, and financing. A minimum package of testing services to be provided at each of the four levels of the health-care system (see Fig. 15.2), as well as at the community level, should be agreed upon and articulated in the plan. Each testing facility will require a specific and sufficient budget,

104

a suitable infrastructure, with some facilities requiring additional infrastructure (such as reliable water and equipment). Plans should also include a monitoring and evaluation mechanism, with a baseline, targets and indicators in order to measure progress and impact.

15.2.3. National regulatory mechanisms (for diagnostics, staff and services) Each national programme should ensure that there are regulatory mechanisms that can perform oversight functions for the various activities of the viral hepatitis testing programme. These may be carried out by the health authority and health product regulatory authorities or through a designated governmental agency. The scope of regulation should encompass regulatory controls (pre-market and post-market) for IVDs available for sale and use within the country, certification of competency of testing staff, and accreditation of testing services. An effective oversight system gives confidence in the testing services to all stakeholders. For countries that do not currently possess capacity and/or competency to regulate IVDs and other laboratory equipment and items, the WHO prequalification assessment serves as a mechanism to provide an independent assessment of the quality, safety and performance of IVDs that are intended for sale and use in resource-limited settings. A number of guidance documents are available to direct nascent regulators in their capacity-building efforts. WHO conducts the prequalification assessment of IVDs using a standardized procedure to determine if the product meets WHO prequalification requirements. The assessment consists of three key components: • • • review of the safety, quality and performance of the assay presented in a product dossier prepared by the manufacturer; desk review of the quality management systems applied during production, followed by a site inspection; independent performance evaluation of performance and operational characteristics.

15.2.4. National reference laboratory for oversight At least one laboratory with the relevant expertise and experience could be designated as a national reference laboratory for viral hepatitis (level 4; see Fig. 15.2). A national public health or disease-specific reference laboratory is generally suitable for this task. The role of the reference laboratory may include: • • • provision of QC specimens; organization of quality assessment schemes; training and supervisory support.

105

15.3. Building capacity for testing services Building or expanding existing capacity for viral hepatitis testing services should be considered in the wider framework of expanding access to testing services for a range of related diseases. The rise in use of multi-disease platforms for testing and the need to service coinfected individuals (such as HIV/HCV-coinfected patients) means that development of a trained and motivated health workforce should be considered across disease control programmes.

15.3.1. Human resource management A range of personnel may be required for the different roles in testing services, including phlebotomists, test operators (laboratory technicians, POC test providers), data clerks and other auxiliary staff. The national strategic plan should detail how testing staff will be trained and certified, with their roles and responsibilities made clear. All staff should have appropriate qualifications, such as certifications according to national guidelines, and demonstrated proficiency in performing the tasks within their scope of work. Supervisory support to staff and regular site visits as part of a quality assessment system provides an opportunity for troubleshooting and feedback to higher management.

15.3.2. Inventory management (procurement and supply chain of assays and reagents) Continuity in the supply of test kits, reagents and other consumables required for testing depends on reliable and responsive procurement and supply systems. Stockouts of test kits or essential consumables can contribute to poor testing services. The testing programme should ensure that procurement procedures (either through a centralized medical store or direct from the supplier) are conducted in accordance with best global practice for procurement. Importantly, any bidding or evaluation of bidding should be conducted in a fair and transparent manner. To support inventory management, testing services at user level should have appropriate systems in place to monitor stocks and expiry dates of test kits and reagents, as well a method to track consumption and wastage. Further reading Forthcoming second edition of Guidance for procurement of in vitro diagnostics and related laboratory items and equipment. Geneva: WHO; 2013. First edition available on the WHO website (http://www.who.int/diagnostics_laboratory/procurement/131024_procurement_of_diagnostics_ finalversion.pdf?ua=1, accessed 06 February 2017).

15.3.3. Storage and transportation All test kits and reagents should be transported and stored under controlled conditions, according the manufacturer’s instructions for use of the product. Testing services should ensure that so-called room temperature-stable test kits such as RDTs are stored according to their labelling.

106

15.3.4. Equipment management When purchasing diagnostics that require instrumentation, such as analysers (either closed or open platform, polyvalent or otherwise), ancillary equipment (e.g. refrigerator, freezer, incubators), and other equipment that requires installation and validation (e.g. autoclaves, water purification systems), it is necessary to ensure that these are maintained. This means calibration upon installation, as well as preventive and corrective maintenance, which should be foreseen as part of financial planning and procurement procedures. Further reading Maintenance manual for laboratory equipment, 2nd edition. Geneva: WHO; 2008 (http://apps.who. int/iris/bitstream/10665/43835/2/9789241596350_eng_low.pdf, accessed 06 February 2017).

15.3.5. Laboratory information management systems Information management consists of paper-based and electronic systems for storing records and documents, including laboratory information management systems and mobile mechanisms that provide testing results or reminders to health-care facility staff or clients. It is closely linked to documentation and record-keeping. To assure the quality and integrity of the test status given to a client, the testing service must minimize the risk of transcription errors. Assigning patient identification numbers and specimen identification numbers to each subsequent specimen received from the same individual will serve to reduce the possibility of transcription errors. It will also protect the confidentiality of people undergoing testing.

15.4 Product selection While a testing strategy provides a generic approach to how many assays should be used and how many tests should be conducted on each assay, a testing algorithm defines the specific products (assay by brand name) to be used in a given testing strategy. The design of a testing algorithm will be determined by the specific disease marker to be tested for, and operational aspects such as the required expertise of the users, infrastructure and testing conditions, and assay characteristics. Selection should include some consideration of products that have been approved by (i.e. conforms to requirements of) stringent regulatory assessment, such as by the WHO prequalification programme or any of the founding members of the Global Harmonization Task Force (GHTF)5 (Table 15.2).

Founding members of the Global Harmonization Task Force are Australia, Canada, European Union, Japan, USA.

107

TABLE 15.2. Examples of stringent regulatory assessments Regulatory jurisdiction Risk class Documentary evidence

WHO prequalification European Union

All classes Annex II, List A

WHO prequalification public report EC Full Quality Assurance Certificate EC Production Quality Assurance Certificate EC Type-Examination Certificate

US Food and Drug Administration Health Canada

Class III Class IV

PMA letter or BLA license Medical Device Licence and summary report for a Class IV IVD CMDCAS-issued ISO 13485 Certificate

Therapeutic Goods Class 4 Administration (TGA), Australia

TGA Licence for Manufacture TGA Issued ISO 13485 Certificate AUST R Number TGA Full Quality Assurance Certificate TGA Type-Examination Certificate TGA Production Quality Assurance Certificate

Japan Ministry of Health, Labour and Welfare (JMHLW)

Class III

JMHLW Minister’s Approval JMHLW License for Manufacturer JMHLW Recognised Foreign Manufacturer

BLA: Biologics License Application; CMDCAS: Canadian Medical Devices Conformity Assessment System; EC: European Commission; PMA: pre-market approval

If there are no products available that meet these quality criteria (WHO prequalified or stringent review by GHTF founding member), then efforts should be made to review any other existing quality certification held by the manufacturer for the product of intended supply. This might include a request from the supplier to provide • • a list of all quality reviews conducted on the product; all internationally recognized standards such as ISO 13485 or equivalent.

The performance of potential candidate assays may have been published as independent performance evaluations in the peer-reviewed literature. The Standards for Reporting Diagnostic Accuracy (STARD) criteria should be used to exclude low-quality studies (407).

108

15.4.1. Monitoring testing algorithms A periodic assessment of the programmatic performance of diagnostics should be conducted. For example, when testing for HCV, if there are low rates of HCV viraemia in those that are HCV-antibody seropositive, this might indicate that there are high rates of false-positive results (poor specificity) for the serological assay(s). Conversely, if seroprevalence rates are low but a higher proportion are found to have detectable HCV RNA, then the possibility of false-negative results should be considered. Either scenario should trigger further revalidation of the testing algorithm to facilitate the selection of assays with increased sensitivity and/or specificity. The following quality indicators should be monitored, as appropriate: • • • • • rate of defective consumables, e.g. specimen transfer pipettes, lancets; rate of invalid test devices (if single-use devices such as RDTs); rate of invalid runs (disaggregated by error codes); rate of equipment breakdown and respective down-time rate of out-of-range QC results; and rate of discrepant results within a testing algorithm consisting of two or more assays.

15.4.2. Post-market surveillance of diagnostics Once a product is placed on the market, its quality, safety and performance must be monitored to ensure that diagnostics continue to meet standards. WHO has established a system for post-market surveillance of diagnostics that supplements the obligations of manufacturers, who must also conduct their own post-market evaluation activities. In this context, post-market surveillance consists of the following: • proactive post-market surveillance (to identify any problem before use) through in-country lot verification testing, both before and after distribution of test kits to testing sites; and reactive post-market surveillance (when a problem has been identified during the use of the diagnostic) through reporting and evaluation of complaints, including reports of adverse events, and any required actions to correct the problem and prevent recurrence.

•

Lot verification testing conducted independently of the manufacturer is particularly useful where manufacturing quality has not been adequately assured, and to verify that an assay has minimal lot-to-lot variation (408).

109

15.5. Assuring the quality of testing services 15.5.1. Quality management systems, irrespective of the testing setting Effective quality management systems are essential for the overall effectiveness of a hepatitis testing programme. It should encompass all activities of the testing programme and not be limited to laboratories only, but include testing in healthand community-based facilities. Any site conducting hepatitis testing should implement a quality management system that incorporates the 12 interconnected components summarized in Fig. 15.1. Many of these components have already been described in the context of the national framework for organizing testing services for viral hepatitis (409). FIG. 15.1. The 12 components of quality management systems Organization Personnel Equipment

Purchasing & inventory

Process control

Information management

Documents & records

Organizational management

Assessment

Process improvement

Customer service

Facilities & safety

Source: Laboratory quality management system: handbook. Geneva: WHO: 2011 (http://www.who.int/ihr/publications/ lqms/en/, accessed 06 February 2017).

Further reading Laboratory quality management system: handbook. Geneva: World Health Organization; 2011 (http://www.who.int/ihr/publications/lqms/en/, accessed 06 February 2017). Improving the quality of HIV-related point-of-care testing: ensuring reliability and accuracy of test results. Geneva: World Health Organization; 2015 (http://www.who.int/hiv/pub/toolkits/handbook-point-of-care-testing/en/, accessed 06 February 2017).

Quality assurance (QA) should be seen as an integral part of the continuing roles and responsibilities of each and every staff member. Through this QA framework, countries can plan, implement, evaluate, improve and sustain QA

110

activities. Such frameworks and provisions apply not only to test accuracy but also to ensuring the quality of pre-test information and post-test counselling.

15.5.2. Quality control Quality control (QC), also known as process control, refers to processes and activities to ensure that testing procedures are performed correctly, that environmental conditions are suitable and that the assay works as expected. The intention of QC is to detect, evaluate and correct errors due to assay failure, environmental conditions or operator performance before results are reported. Hence, QC is a multistep process with certain checkpoints throughout the testing process. Before testing (pre-analytical) • Ensure the appropriate sample type and/or volume has been used. • Check the expiry of test kits and required consumables. While testing (analytical) • Ensure that any QC specimens have been run (e.g. test kit controls and/ or external QC specimen) and that the results are within QC acceptance criteria. • Ensure that a test result is read correctly. After testing (post-analytical) • Double-check the report of test status to the client. Internal QC refers to processes within the assay that check whether the test procedure is working; the appearance of a control line for HBsAg or anti-HCV RDTs is an example of internal QC. As an addition to the test kit controls, external quality control specimens may be produced. These are prepared and validated for the assay by the specimen provider, usually the national reference laboratory or commercial entity, separately from the manufacturer. Many errors occur due to incorrect transcription of testing results and reporting of the status. All necessary steps should be taken to mitigate these errors such as rechecking the reports and re-reading visually read assays independently by a second individual (409).

15.5.3. Personnel and training All testing services must employ an adequate number of trained, certified and supported personnel to conduct each of the elements of hepatitis B and C testing for the expected number of tests conducted and the number of people being served. To assess and manage human resource planning, tools such as the WHO Workload indicators for staffing need (WISN) (http://www.who. int/hrh/resources/wisn_user_manual/en/, accessed 06 February 2017) can be useful for calculating the number of health workers and lay providers needed to provide adequate viral hepatitis testing services.

111

All personnel must be trained adequately, including those taking specimens, conducting testing, providing test reports, and data clerks and other auxiliary staff. All staff should have appropriate qualifications, such as certifications according to the national guidelines, and demonstrated proficiency in performing the tasks within their scope of work. Both pre-service and inservice training, including periodic refresher training, should be part of the training requirements for all testing services. In addition, regular supportive supervision and ongoing mentoring of all staff are essential.

15.6 Assuring the safety of testing services 15.6.1. Facilities and safety It is critical that testing facilities are well designed and maintained. The testing site, including where counselling takes place, where specimens are taken and where the test is performed, should be clean and comfortable, with adequate lighting (for reading visually read assays) and free of any potential hazards. It is critical to guard against harm to any client, testing provider or other person at the testing site. This means that a safe working environment must be maintained by and for all staff, with necessary procedures in place. These procedures include universal precautions (assume that all specimens are potentially infectious), prevention of and/or response to needle-stick injuries or other occupational exposures, chemical and biological safety, spill containment, waste disposal and use of personal protective equipment. It is imperative to follow the assay manufacturer’s recommendations for the control of room temperature of areas where testing is performed. Where possible, testing should take place in climate-controlled areas. There must be proper waste disposal for biological (infectious and non-infectious), chemical and paper waste, and sharps.

15.7. Other practical considerations for testing 15.7.1. Testing at different levels of the health-care system Testing for serological markers of HBV and HCV infection may take place at any level of the health-care system and for virological tests at levels 2 to 4. Fig. 15.2 depicts how testing services are typically organized, with the different assays formats that could feasibly be available at each of the levels when their operational characteristics and other factors such as need for phlebotomy are considered. The degree of physical infrastructure required for each assay format, such as the need for reliable electricity to store reagents and climate-controlled testing rooms to run tests, as well as the staff skills and competencies required, will determine how complex the assay can be for a given testing setting. With further expanded use of RDTs, more people could access testing at the primary care level (level 1).

112

FIG. 15.2. A tiered testing service, with test format menu and staff qualifications Lab-NAT Lab-IA (EIA/ECL/CLIA) Lab-NAT/POC-NAT Lab-IA (EIA/ECL/CLIA) POC-NAT and HCVcAG Lab-IA (EIA/ECL/RDT) RDT 1 RDT 4 3 2 National Reference Centre Senior Laboratory Specialist Provincial/Regional hospital Senior Laboratory Specialist/technicians District hospital Laboratory technicians/ Health-care workers Primary care Health-care workers Lay providers Community/Outreach Community health workers, Lay providers

Facility-based testing

Non-facilitybased testing

CLIA: chemiluminescence immunoassay; ECL: electrochemiluminescence immunoassay; EIA: enzyme immunoassay; RDT: rapid diagnostic test; Lab-NAT: laboratory-based nucleic acid testing; POC-NAT: nucleic acid testing at point of care Source: Consolidated guidelines on HIV testing services. Geneva: WHO; 2015 (11).

15.7.2. Specimen types and collection methods Specimen integrity is critical to the accuracy of testing. Table 15.3 provides a broad summary of specimen types and processing requirements, but each manufacturer specifies in the instructions for use the recommended specimen collection procedures, the storage requirements and specimen stability after collection, and these instructions should always take precedence. Where the instructions for use do not include a certain specimen type within the intended use, it indicates that the assay manufacturer has not yet validated that specimen type for use with their assay. Serum/plasma specimens are most commonly used for testing of HBV and HCV, both serological testing and NAT. However, taking whole blood specimens using venepuncture requires technical skill and proficiency, with the need for additional processing steps to generate serum/plasma from the venous whole blood, which requires a centrifuge and refrigerated storage facilities. Collection of oral fluid and capillary whole blood is less invasive than venepuncture. However, oral fluid testing is currently limited to serological tests, and may have lower sensitivity than testing performed on capillary whole blood or serum/ plasma specimens.

113

TABLE 15.3. Specimen types and processing requirements Specimen type Time to processing/storage/time to testing

Venous whole blood • Fresh whole blood collected by venepuncture Serum Freshly collected whole blood is allowed to coagulate, and serum fraction is collected away from the clotted red blood cells. • • •

Use the specimen immediately. Collect whole blood, mix by hand 4–5 times immediately and let stand for the clot to form. Process within 30 minutes of collection. Store at 2–8 °C. Test within 5 days or as specified by the instructions for the assay to be used. Collect whole blood, mix by hand 8–10 times immediately and centrifuge for up to 10 minutes. Process within 6 hours of collection. Store at 2–8 °C. Test within 5 days or as specified by the instructions for the assay to be used. Use the specimen immediately, with the specimen transfer device recommended by the instructions for use. Note that the specimen transfer device may or may not include an anticoagulant. An anticoagulant contributes to accuracy. Use the specimen immediately, with the specimen transfer device recommended in the instructions for use. Store at 4 °C for up to 3 months, or at –20 °C for longer. Use of specific assays with DBS should be validated by the manufacturer. If the manufacturer has not validated their assay for DBS, the use of DBS is considered “off-label”, or unauthorized for returning medical results.

Plasma Freshly collected whole blood is added to recommended anticoagulant, such as EDTA, heparin or citrate. After centrifugation, plasma is separated.

• • •

Capillary whole blood Capillary (finger-stick) whole blood is collected using a lancet and a specimen transfer device.

• •

Oral fluid Oral mucosal transudate (not saliva) is collected from the gums using a collection device. Dried blood spot (DBS) Venous or capillary whole blood is applied to a filter paper by hanging drop or microcapillary action. Whole blood is later eluted from the filter paper and used for the test procedure.

•

• •

114

16. PRE-TEST AND POST-TEST COUNSELLING This chapter discusses essential counselling and services prior to HBV and HCV testing. It also discusses post-test counselling for individuals who are diagnosed with chronic HBV or HCV infection, as well as those who test negative or who have an inconclusive result.

Key points • The 5 “Cs” are essential for all hepatitis testing services: consent, confidentiality, counselling, correct test results and connection to hepatitis prevention, treatment and care. Verbal consent is usually adequate, but all individuals should have an opportunity to refuse testing. Mandatory testing is never warranted. Viral hepatitis testing services must ensure that all test results and client information are confidential. Although disclosure to supportive family members and health workers is often beneficial, this must be done only with the consent of the person being tested. Everyone who is diagnosed positive for hepatitis B or hepatitis C should receive post-test counselling. People who test negative for hepatitis B or C will usually need only brief health information about how to prevent acquisition of viral hepatitis in the future, where and how to link to prevention services, as appropriate, and be offered HBV vaccination. People with significant ongoing risk such as PWID may need more active support and linkage to harm reduction services. Connection or linkage to prevention, treatment and care is an essential component of viral hepatitis testing. Chapter 14 provides recommendations on approaches to improving linkages.

•

16.1. Promoting testing awareness Depending on the current levels of knowledge and awareness of hepatitis in different countries and settings, general promotion and awareness campaigns for viral hepatitis testing and where it is available may be necessary. This may include promotion through the mass media, including radio, television, billboards and posters, the Internet and electronic social media. In other countries,

115

promotional activities may need to focus on specific populations in whom viral hepatitis testing rates remain suboptimal, such as PWID. There is also a need for clear signs, printed information, posters that direct clients to where testing is available in health facilities (such as ANC, STI and TB clinics), the community or through mobile services and social media.

16.2. Creating an enabling environment Critical enablers are strategies, activities and approaches (generally outside the purview of the health sector) that are key to the success of health sector interventions. Addressing such critical enablers as reducing stigma and discrimination, empowering the community and reviewing certain national laws, policies and practice can help strengthen interventions to support the uptake of viral hepatitis testing and linkages to prevention, care and treatment. In particular, it can improve the accessibility, acceptability, uptake, equitable coverage, quality, effectiveness and efficiency of viral hepatitis interventions, especially among populations that are reluctant to use or have limited access to current hepatitis testing, such as PWID.

16.3. The WHO 5 “Cs” The WHO 5 “Cs” are principles that apply to all models of HIV and hepatitis testing and in all settings (Box 16.1) (11). Box 16.1 “The 5 Cs” for hepatitis testing services (11) • Consent. People being tested for hepatitis B or C must give informed consent to be tested and counselled. Verbal consent is sufficient, and they should be informed of the process for testing and of their right to decline. Provision of information about testing and the need for consent can be delivered in a group setting, such as group health education, but clients should give consent in an individual and private manner. Health workers should carefully explain how a client can decline testing and ensure that no one coerces clients into being tested, and each person has a private opportunity to opt out of testing. • Confidentiality – ensuring a confidential setting and preserving confidentiality. Testing must be confidential, meaning that what the provider and the client discuss will not be disclosed to anyone else without the expressed consent of the person being tested. Confidentiality applies not only to the test results and report of hepatitis status but also to any personal information, such as information concerning sexual behaviour and the use of illegal drugs. Hepatitis testing services should avoid practices that can inadvertently reveal a client’s test results to others in the waiting room or in the health facility. Experiences with HIV testing services have shown that a lack of confidentiality discourages people from using testing services. Health workers and others who will provide testing may need special training and sensitization regarding the confidentiality of medical records.

116

Although confidentiality should always be respected, it should not be allowed to reinforce secrecy, stigma or shame. Counsellors should discuss, among other issues, whom the person may wish to inform and how they would like this to be done. Shared confidentiality with a partner or family members and health-care providers is often highly beneficial. • Counselling. Pre-test information can be provided in a group setting, but all people should have the opportunity to ask questions in a private setting if they request it. All hepatitis testing must be accompanied by appropriate post-test counselling, based on the specific hepatitis test result and hepatitis status reported. QA mechanisms as well as supportive supervision and mentoring systems should be in place to ensure the provision of high-quality counselling. • Correct. Providers of hepatitis testing should strive to provide high-quality testing services, and QA mechanisms should ensure that people receive a correct diagnosis. QA may include both internal and external measures, including support from the national reference laboratory. All people who receive a positive serological diagnosis of HBV or HCV should have a NAT to confirm the presence of viraemic infection and assess their need for care and treatment before starting antiviral therapy (Chapter 15). • Connection. Linkage to prevention, treatment and care services should include effective and appropriate follow up, including long-term prevention and treatment support. Providing viral hepatitis testing where there is no access to care, or poor linkage to care and treatment, has limited benefit for those with hepatitis (Chapter 14).

16.4. Providing pre-test information With the increasing availability of RDTs, many people will receive their initial serology test results on the same day as testing. Therefore, intensive pre-test counselling is not needed and may create barriers to service delivery. Depending on local conditions and resources, programmes may provide pre-test information through individual or group information sessions and through media such as posters, brochures, websites and short video clips shown in waiting rooms. When testing children and adolescents, information should be presented in an age-appropriate way to ensure comprehension. Offering or recommending viral hepatitis testing to a client or a group of clients includes providing clear and concise information on: • viral hepatitis and the benefits of testing for hepatitis B or C; and the meaning of a positive and negative test result; a brief description of prevention options; the confidentiality of the test result, as well as any information shared by the client; the potential negative consequences of testing to the client in settings

• •

•

117

where certain sexual or injecting drug use behaviour is stigmatized or even criminalized, or where a positive test could result in discrimination, for example, with regard to employment or with insurance policies where there may be financial consequences of either taking a test or of a positive result. In addition, the practical implications of a positive test result should be explained, including when there is no treatment currently available.

16.5. Post-test counselling and services 16.5.1. For those who test positive Health workers, professional counsellors, social workers and trained lay providers can provide counselling. The information and counselling that health workers or others should provide to HBV- or HCV-positive clients is listed below. However, counselling should always be responsive to and tailored to the unique situation of each individual. • • Explain the test results and diagnosis. Provide clear information on further tests required to confirm viraemic infection and stage of liver disease, indications for treatment for both HBV and HCV and its benefits, as well as where and how to obtain the appropriate care and treatment (and advice if treatment is not currently available). Make an active referral for viral hepatitis clinical care for a specific time and date, i.e. tester makes an appointment or if services are co-located, accompanies the client to an appointment. Provide information on how to prevent transmission of infection. Preventive measures include HBV vaccination of non-immune clients, family members (including children), and sexual partners. Counselling on lifestyle. This includes assessment of alcohol consumption and advice on alcohol reduction (the WHO ASSIST package includes Alcohol, Smoking and Substance Involvement Screening Test)(410), diet and physical activity. Discuss possible disclosure of the result and the risks and benefits of disclosure, particularly among couples and partners. Offer couples counselling to support mutual disclosure. Encourage and offer HBV and HCV testing for family members, including children, and sexual partners. This can be done individually, through couples testing or partner notification. Provide additional referrals for prevention, counselling, support and

•

118

other services, as appropriate. These could include, for example, HIV, TB, STI diagnosis and treatment, contraception, antenatal care, reducing alcohol use, OST, access to sterile needles and syringes, and brief safe sex counselling. • Considerations in special populations. In certain populations, such as PWID or those with mental health problems, intensified post-test counselling combined with follow-up counselling by referral to community health workers and to other services such as OST should be included in post-test counselling. A peer counsellor may particularly help people understand the diagnosis and support linkage to care and treatment by serving as a “peer navigator”, who assists with finding, choosing and obtaining a full range of services, and can potentially increase the proportion of people who start treatment. Chapter 18 addresses considerations for testing in specific populations, e.g. pregnant and postpartum women, adolescents, and children.

16.5.2. For those who test negative Individuals who test negative for HBV or HCV infection should receive brief health information about their test results. In general, a lengthy counselling session is not necessary and may divert counselling resources that are needed by those who test positive. Counselling for those who test negative should include the following, particularly in high-prevalence settings: • • an explanation of the negative test result; an offer of HBV vaccination and education on methods to prevent acquisition, and referral to harm reduction prevention services, as appropriate; repeat testing for HCV based on the client’s level of recent exposure and/or ongoing risk of exposure. The majority of individuals do not require retesting to verify a negative test, particularly in the absence of any ongoing risk. However, certain individuals who test negative warrant retesting because of an ongoing risk, especially for HCV, but also for HBV if they have not yet been vaccinated. These include the following: persons from high-risk populations, such as PWID, sex workers and MSM; persons with a known HBsAg- or HCV RNA-positive partner or family member; pregnant women in high-prevalence settings (at each pregnancy); individuals seen for a diagnosis or treatment of HIV or STIs; encouraging the client to return for a further test to confirm the diagnosis when the hepatitis status is inconclusive

•

119

17. SERVICE DELIVERY APPROACHES FOR VIRAL HEPATITIS TESTING – examples from the field This chapter summarizes the different facility- and community-based testing approaches available, i.e. where to test, and supports implementation of the recommendations on who to test for viral hepatitis in Chapters 6 and 7, with examples of their use in the field in different populations and settings. Chapter 19 provides a strategic framework to guide countries’ decision-making on selecting testing approaches.

Key points • Viral hepatitis testing can be delivered in different populations and different settings through both health-care facility-based testing and community-based testing. Many of these approaches have successfully increased the coverage and impact of HIV testing, and can be applied to the delivery of viral hepatitis testing. Health-care facilities for testing are primary care clinics and outpatient clinics that include specialist clinics such as HIV, STI and TB clinics, antenatal clinics, OST services, as well as inpatient wards in district, provincial and regional hospitals, and private clinical services. Community-based testing can be offered through outreach/ mobile, home-based or door-to-door approaches, in schools and other educational establishments, and in workplaces, places of worship, parks, bars and other venues. Effective health system programme practices that may be appropriate for increasing access to hepatitis testing in some settings include integration with other health services (e.g. HIV), decentralization of testing to primary care facilities and outside the health system (e.g. workplaces, schools, places of worship), and task-sharing of testing responsibilities to other health workers, including trained lay providers. Countries need to identify the most strategic mix of facility- and community-based testing opportunities (as well as the use of integration, decentralization and task-sharing) to best reach those with undiagnosed infection and populations at high risk (see Chapter 19).

•

120

17.1. Health-care facility-based testing and provider-initiated testing and counselling Health facility-based viral hepatitis testing refers to testing provided in a health facility or laboratory setting. There are several approaches to facility-based testing. Provider or practitioner-initiated testing and counselling (PITC) denotes testing that is routinely offered at a health facility (411), as well as for persons who request testing or who exhibit clinical signs, symptoms or laboratory results that could indicate HBV or HCV infection. It includes provision of pre-test information and obtaining consent, with the option for individuals to decline testing. Although voluntary counselling and testing (VCT) in stand-alone facilities was an early model for delivering HIV testing, it was recognized that offering testing in clinical sites as part of general medical care through PITC (411) resulted in increased HIV testing uptake, coverage and case detection. It also helped normalize testing by removing the potential reluctance of clients to request a test (11, 411). PITC for hepatitis can be implemented and integrated in a number of clinical settings, as summarized below, and these represent major opportunities for scaling up viral hepatitis testing. HIV clinics. In many populations and high-risk groups, prevalence of HIV and HBV or HCV is high, and there are also high rates of HIV/HBV or HIV/ HCV coinfection (88). Existing HIV and ART programmes provide an important opportunity to integrate testing for viral hepatitis with that for HIV (see Box 17.1). Box 17.1. Integrated HIV/hepatitis testing Médecins Sans Frontières (MSF) India. The MSF team in collaboration with the National AIDS Control Organization in India provided integrated HIV/TB/HCV services. Counsellors experienced in HIV testing and adherence counselling have also been trained to provide HCV pre-test counselling, HCV viral load testing, genotyping and FibroScan for staging of liver fibrosis. Of the 1367 HIVinfected persons who were tested, 383 (28%) were HCV antibody seropositive. Source: Hepatitis Testing Innovation Contest, 2016

TB clinics. Some populations who are at high risk for HBV and HCV infection are often also at risk for TB, e.g. PWID, prisoners, migrants and persons coinfected with HIV. WHO already recommends routine HIV testing for all TB patients (both active and presumptive cases) (412), and this has proven highly acceptable (413). Integrating HBV and HCV testing as part of a comprehensive package of care for TB patients should be both feasible and acceptable, particularly in settings and populations where the prevalence of TB and viral hepatitis is high. STI clinics. HBV and HIV, and to a lesser extent HCV, are all sexually transmissible infections, and services providing care for STIs are therefore a key entry point for

121

both HIV and viral hepatitis prevention and treatment services. New acquisition of STIs, such as gonorrhoea and syphilis, indicate recent unprotected sex and can help identify people at a heightened risk of acquisition of HIV and viral hepatitis. WHO already recommends routinely offering HIV testing for persons diagnosed with other STIs (414), with a high uptake of testing (415, 416). Extension to include targeted HBV and HCV testing is also likely to be feasible and acceptable. Drug treatment and harm reduction services. Many innovative models of care to provide integrated hepatitis and OST services for PWID in community drug treatment services have been developed and effectively implemented, mostly in developed countries (196, 362, 377, 417–419). Many of these programmes provide additional interventions, including education, harm reduction, mental health services, other general medical services, and direct provision of referrals to care and treatment. These models can provide a framework for lower-income countries to expand viral hepatitis testing and treatment for at-risk populations. Inpatient and outpatient hospital settings present a further opportunity for testing in patients with symptoms or laboratory test findings, such as unexplained abnormal liver function tests that may be indicative of viral hepatitis infection. Testing in hospitals, particularly in low- or concentrated-epidemic settings, has proven effective in HIV case-finding in Europe (420). Testing in hospital emergency departments has also been recently piloted in Europe and the United States (see Box 17.2) (421). Box 17.2. Using emergency departments to promote testing The Barts Health NHS Trust, London, United Kingdom, “Going Viral” campaign brought together health authorities, pharmaceutical companies, and national media to promote a testing initiative for bloodborne viruses (HIV/HBV/HCV) in nine emergency departments. In addition to promotion in the emergency department, social media celebrity endorsements and television coverage were used. Of 7800 individuals having blood drawn in the emergency department, 2118 (27%) agreed to be tested, and 39 individuals with HCV and 15 with HBV were identified. Approximately half were previously unaware of their diagnosis, and 41 (76%) of these were linked to care, and attended at least one follow-up clinic. (http://bartshealth.nhs.uk/). Source: Hepatitis Testing Innovation Contest, 2016

Primary-care settings may be more accessible and less stigmatizing than hospitalbased clinics, particularly for high-risk and vulnerable populations such as PWID. Targeted case-finding of people with a history of injecting drug use can increase the number of people who are offered and accept HCV testing (422). Studies also show that multidisciplinary care with integration of other services (e.g. drug and alcohol support, psychiatric services) at the same primary-care setting are acceptable and particularly effective for these populations, who often have multiple

122

health comorbidities and complex needs (417, 423–425). Peer-led models, or provider-led models with peer support, can be particularly effective in enabling integration of services in one place in a way that is acceptable to certain high-risk groups such as PWID. Electronic medical records (EMR) have been successfully used to identify and flag higher-risk patients for viral hepatitis testing in several primary-care clinic and hospital-based programmes (see Box 17.3). Box 17.3. Use of electronic medical records (EMRs) to flag higher-risk primary care patients for testing National Nurse Care Consortium, Philadelphia, United States. Five Philadelphia primary health centres have integrated HCV testing and linkage to care within community primary-care services. The EMRs were used to identify testing eligibility and expedite laboratory requisitions. A total of 9225 HCV tests were performed between October 2012 and January 2016. Of these, 1114 (12.1%) were HCV antibody positive and 1057 (95%) also had HCV RNA testing, of whom 765 (72%) were positive. Of these 765, 512 (67%) had a follow-up HCV medical evaluation and 110 (22%) received treatment. (www.nncc.us) Source: Hepatitis Testing Innovation Contest, 2016

Paediatric and adolescent clinics may be important settings for identifying cases of previously undiagnosed hepatitis B or C infections, particularly in highprevalence countries. Offering testing to all children whose mother or father has either HBV or HCV infection, and to those with symptoms or laboratory findings that could be indicative of viral hepatitis infection may identify many infections. This could be integrated in clinics where PITC for HIV is already provided. Routine testing in antenatal clinics (ANC) is a key opportunity to reduce the global burden of HBV disease, which is primarily propagated through ongoing MTCT in high-prevalence resource-limited settings. Testing for HBsAg enables women to have knowledge of their HBV serostatus for their own health, and for their offspring to benefit from interventions to prevent MTCT, including birth dose and infant HBV vaccination, use of hepatitis B immune globulin (HBIG), and antiviral therapy. PITC for HBV offered routinely in ANC has proven feasible and acceptable in several settings. However, although many countries recommend routine screening, the proportion actually screened in many high-burden LMICs remains low (157). The additional cost of also testing pregnant women for HCV alongside HIV and HBV is likely to be low (see Box 17.4).

123

Box 17.4. Antenatal clinic testing for hepatitis B infection Antenatal testing is a key opportunity to prevent MTCT with neonatal vaccination and use of HBIG and antiviral therapy. Yunnan AIDS Initiative in China. This nongovernmental organization implemented a combined HIV, HBV and syphilis testing campaign in ANC, labour and delivery units throughout the province. An opt-out testing model for the women was used and partners were also offered testing. The Chinese government has recently adopted routine HIV, HBV and syphilis testing in ANC in over 1000 counties nationwide after the success of the Yunnan AIDS Initiative and several other demonstration projects. Source: Hepatitis Testing Innovation Contest, 2016

17.2. Community-based testing Community-based testing can complement facility-based approaches, which may fail to reach certain high-risk populations, especially PWID, who are often marginalized because of stigma and discrimination or legal sanctions, as well as those in remote or rural areas, including pregnant women with limited access to facility-based testing. There is some evidence that offering hepatitis testing in community settings may increase testing acceptance and uptake, achieve earlier diagnosis, reach first-time testers and people who seldom use clinical services (194, 422). However, the same barriers encountered in ensuring linkage to HIV prevention, care and treatment services will need to be addressed. Such outreach methods aimed at HIV prevention have been shown to be particularly effective in engaging with hard-to-reach PWID populations, and decreasing injection and sexual risk behaviours (426) (see Box 17.5). Mobile/Outreach testing approaches include outreach to community sites through mobile vans or tents, at community sites such as churches, mosques or other faith settings, in places of entertainment such as bars and clubs, at cruising sites. Such services may be offered on a regular schedule, at night (“moonlight testing”), or as a one-time or occasional promoted event, linked to public events, such as sports events, music performances, theatre, agricultural fairs and holiday festivals. Door-to-door/home-based testing takes place in the home. There are two main models: (i) testing that is offered door to door and provided to all consenting individuals, couples or families in a geographical area; and (ii) testing that is offered to households with an index patient (i.e. persons known to have HIV, viral hepatitis or active or presumptive TB), with consent obtained from the index patient before the home visit. Door-to-door testing during the daytime may reach only people who are not working and younger children, while services during the evening or on weekends may increase uptake among others, such as men.

124

Box 17.5 Reaching people who inject drugs Tailored HCV testing and linkage to care services for PWID are a critical part of delivering HCV services. Integration of HCV testing into PWID services (including drug dependence treatment services, needle and syringe services, PWID community health services) is effective in a range of settings. Care & Cure Service Center (CCSC) Kerinchi in Kuala Lumpur, Malaysia. A methadone clinic screened and counselled individuals for HCV. Their programme collaborated with public hospitals, facilitating service integration after testing. Of 544 methadone clinic attendees 304 (55%) have been evaluated and tested for HCV antibody. Of those, 235 (77%) tested positive and 81 (34.5%) were referred for clinical care. (www.ceria.um.edu.my/) Testing camps: Community Network for Empowerment (CONE) in Manipur, India. As part of a province-wide campaign to identify more cases of viral hepatitis, a community organization partnered with provincial government and pharmaceutical companies to establish testing camps. They provided free screening and HCV RNA confirmatory tests together with liver scans for staging of liver fibrosis. Of 1011 individuals (including PWID, people living with HIV, and general clinic attendees) tested for HCV, 463 were confirmed to be HCV RNA positive. Linkage to treatment is now under way since DAAs have become available. (www.conemanipur.net) Source: Hepatitis Testing Innovation Contest, 2016

National testing campaigns are nationwide efforts to increase access to and uptake of testing. Some have focused on testing in facilities while others have used a community-based approach or a combination of the two. Outcomes have varied with regard to coverage of different population groups, linkage and cost– effectiveness. A national or regional hepatitis testing campaign has the potential to reach a significant proportion of the population, which includes both those known to be at risk as well as those not at risk for HBV or HCV infection (see Box 17.6). However, experience with national HIV campaigns has shown that they can be expensive, and that a substantial number of people with HIV remain undiagnosed. In addition, linkage to care and treatment from campaigns has been problematic.

125

Box 17.6 Reaching general populations through community-based and primary clinic testing Egyptian Liver Research Institute, Mansoura, Egypt. In El Othmanya village in northern Egypt, a social marketing and community mobilization campaign was implemented to promote household testing for hepatitis B and C among all adolescents and adults. The majority (98%, 3500/3573) of household members were tested and 270 (7.7%) were confirmed HCV RNA positive and 8 (0.22%) HBV DNA positive. Treatment-eligible cases were linked to care at the Egyptian Liver Hospital and treatment costs were covered through community fund-raising or the government health insurance system. The model is being scaled up in 30 other villages. (http://www.nrc.sci.eg/) Ishikawa Prefecture and Kanazawa University Hospital, Japan. Since 2001, the government has provided free hepatitis testing at five-year intervals for all citizens aged between 40 and 70 years. A total of 240 180 individuals, or 38% of the target population in Ishikawa prefecture, were tested. The programme also provides at least annual follow up of all diagnosed patients for liver fibrosis staging and treatment. (http://www.m-kanazawa.jp/english/index.html) Prevention of Liver Fibrosis and Cancer in Africa (PROLIFICA), Gambia. Is the first community- and facility-based screen and treat programme for HBV infection in sub-Saharan Africa and offers testing to all rural and urban inhabitants aged 30 years or older in the western part of Gambia. HBsAg screening was accepted by 68.9% of 8170 adults and 81.4% of 6832 blood donors, and was positive in 495 (8.8%) of individuals in the community and 721 (13%) of blood donors. All individuals who tested HBsAg positive were referred for comprehensive outpatient assessment to determine eligibility for treatment. Linkage to care was high in the community (81.3%) but lower (41.6%) among blood donors, and treatment eligibility was 4.4% and 9.7%, respectively. (https:// www.prolifica.org.uk/) Source: Hepatitis Testing Innovation Contest, 2016

Mass media and social media. Knowledge of hepatitis testing and availability is limited in many countries, and there is a need for promotion and awareness campaigns in the general population. Some countries and programmes promote viral hepatitis testing and education through the mass media, including radio, television, billboards and posters, the Internet and electronic social media (see Box 17.7). This approach has also been used to facilitate more targeted and efficient screening in regions of low prevalence. This also applies to testing in health facilities, in the community and through mobile services (see Box 17.7).

126

Box 17.7 Use of Internet and social media to promote testing in the general population Public Health Service of Amsterdam, Netherlands. The campaign goal was to motivate at-risk groups to independently assess their HCV risk, using an online risk assessment tool developed by the Amsterdam Public Health Service. Participants received a laboratory form for anonymous HCV testing via the website if the tool rated them as being at high risk. Approximately 9700 individuals took the online risk assessment, 1500 were offered testing, and 28% of these used the organization’s testing facility. The HCV antibody positivity rate was 3.6%. Test results were also available online along with invitations for confirmatory testing. (www.ggd.amsterdam.nl) Source: Hepatitis Testing Innovation Contest, 2016

Workplace testing provides employed men and women access to testing, who otherwise might have limited access to clinical services because they need to take time off work to seek health care. Concerns with workplace testing include the potential for coercion, breaches in confidentiality, and weak linkages to services, and care must be taken that this approach is not promoted where it is likely to be abused. For example, 60% of HIV testing in the Middle East and North Africa region is undertaken through workplace testing and work visa procedures, and is generally mandatory (427). It should not therefore be considered an effective model for scale up of hepatitis testing. However, workplace testing for HIV and TB with onward linkage to HIV and TB services has been successfully implemented in several high-burden settings (428–430). Many workplace health programmes do not include hepatitis programmes, creating an opportunity for expanding workplace testing (see Box 17.8). Box 17.8 Workplace testing Asian Liver Center, Stanford University, United States. This project partnered with 42 corporations to increase awareness of HBV and HCV testing in several countries. An online tool (www.hepbhra.org) was developed, which allowed individuals in the workplace to assess their risk of hepatitis B infection and make decisions about being tested. (http://liver.stanford.edu/). Source: Hepatitis Testing Innovation Contest, 2016

Testing in schools, colleges or other educational establishments can facilitate access to testing among sexually active young persons by bringing services to students who may find it challenging to seek HIV or hepatitis testing during school hours, and be otherwise hard to reach, as they do not use health services or community services. The service may also provide sexual health education and counselling on risk reduction. In South Africa, a national campaign provides HIV testing to students aged 12 years and older in schools (431). However, school-based HIV testing remains controversial, and few countries have established such programmes. Further evaluation is needed to understand

127

issues of confidentiality, linkage to care and adolescents’ experiences with and expectations of school-based testing for both viral hepatitis and HIV, as well as the impact and acceptability of testing among university students. Testing in prisons and other correctional system settings is a potentially effective way to expand testing uptake among high-risk populations, as many prisoners are at increased risk of acquiring hepatitis B and C infection. There are also additional ethical and regulatory procedures involved in establishing testing programmes in prison settings. Several effective case studies demonstrate how hepatitis testing can be undertaken in prison and justice system settings (see Box 17.9). Box 17.9 Prison and correctional hepatitis testing St Vincent’s Hospital Melbourne and Department of Justice and Regulation, Victorian State Government, Australia. HCV prevalence among prisoners in Victoria state is high at 25% and new transmissions also occur within prisons. A state-wide programme was initiated with the goal of eliminating transmission of viral hepatitis within its thirteen prisons through the assessment, education and management of prisoners with chronic viral hepatitis. All prisoners are screened for viral hepatitis on prison entry or transfer. Prisoners who are seropositive are referred to the Victorian State-wide Prison Hepatitis Program for further assessment and initiation of antiviral therapy supervised by trained clinical nurse consultants during visits every 2–4 weeks. This programme is integrated into the prison primary care system and uses telemedicine for consultations with two part-time hepatologists, and provision of DAAs since March 2015. (http://www. svhm.org.au/) Source: Hepatitis Testing Innovation Contest, 2016

17.3. Good practices for delivery of effective viral hepatitis testing services 17.3.1. Effective health system programming practices Delivery models for viral hepatitis testing, care and treatment can be informed and strengthened by experience from the global scale up of HIV testing and treatment. The WHO-recommended effective health programming practices of integration with other health services; decentralization to primary health-care facilities as well as outside the health system (e.g. workplaces, schools, places of worship); and taskshifting of responsibilities to increase the role of trained lay providers were originally developed to improve the delivery of HIV testing (11) (see Glossary). Inclusion of one or more of these practices may improve the accessibility of hepatitis testing, and onward linkage to services and support in some settings. With the availability of simplified viral hepatitis diagnostic tests and treatment regimens, decentralization and task-shifting or -sharing in particular can be increasingly used in service delivery models to scale up hepatitis testing and treatment, especially in settings where there is limited access to hospital facilities and laboratory services (432).

128

17.3.2. Integration of viral hepatitis testing with other services Integration involves not only providing related services in a single setting, but also linking recording and reporting systems to share information and referrals between settings and providers. There is already a range of clinical services for which WHO recommends the integration of HIV testing, and this may also apply to hepatitis testing in some settings (see Table 17.1). These include clinical services for TB, HIV, maternal and child health, sexual and reproductive health (STI clinics), mental health and harm reduction programmes for PWID, migrant and refugee services, and persons in prisons (5, 6, 25, 28, 432). Integration with HIV testing and treatment services will be particularly appropriate in HBV and HCV epidemic settings where the HIV prevalence is also high. Table 17.1. Potential populations and programmes for integration to promote hepatitis testing Disease High-risk groups and potential programme integration

Hepatitis B

• • • • •

Infants of infected mothers (delivery units, maternal and child health [under-5 and immunization] clinics) Children in endemic regions (maternal and child health [under-5 and immunization] clinics) Sexual transmission in adults (STI and HIV clinics) People who inject drugs (harm reduction and drug treatment services) Health-care workers (occupational health) People who have received unsafe therapeutic injections/blood products (health promotion) People who inject drugs (harm reduction and drug treatment services) Men who have sex with men (STI and HIV clinics) Health-care workers (occupational health)

Hepatitis C

• • • •

Source: Adapted from: Mihigo R, Nshimirimana D, Hall A, Kew M, Wiersma S, Clements CJ. Control of viral hepatitis infection in Africa: are we dreaming? Vaccine. 2013;31 (2):341–6.

The primary purpose of such integration is to make HBV, HCV and HIV testing more convenient for people coming to health facilities for other reasons, and so expand the reach and uptake of viral hepatitis testing. For the patient, integration of hepatitis testing into other health services may facilitate addressing other health needs at the same time, saving time and money. For the health system, integration may reduce duplication of services and improve coordination, for example, in stock management, overall efficiency and cost–effectiveness. The goal of programme collaboration is to create integrated delivery systems that best facilitate access to and increase the impact of hepatitis testing, treatment and other health services. Aspects of coordination across programmes that need consideration include: mobilizing, allocating and sharing resources (including multitasking and task-shifting of human resources to increase the availability of highly skilled workers); training, mentoring and supervising health workers; procuring and managing medicines, test kits and other medical supplies; maintaining the quality of testing; and reducing stigma and discrimination (433).

129

17.3.3. Decentralization of hepatitis testing services Decentralization of services refers to delivery of services provided in peripheral health facilities, community-based venues and locations beyond urban hospital sites, nearer to patients’ homes. This may reduce transportation costs and waiting times experienced in central hospitals and, therefore, improve uptake of testing. Decentralization of HIV treatment services, in high-burden LMICs was a key component of the global scale up of HIV services and successfully improved uptake of testing and reduced loss to follow up (402). To date, delivery of viral hepatitis testing and treatment has in general relied on specialist-led centralized models of care in hospital settings (432). Currently, there are only a few successful models of decentralized viral hepatitis testing and treatment for hard-to-reach populations and general populations at high risk (see Box 17.10). With the development of simpler diagnostic tests for HBV and HCV, and simpler and more effective treatment regimens for HCV, decentralization has the potential to also increase the uptake of hepatitis testing. Box 17.10 Decentralization In Taiwan, a community-based outreach model delivering free testing for HBV and HCV and targeting the general population has been successfully implemented since 1996, and identified a high overall seroprevalence of both HBV and HCV (17.3% and 4.4%, respectively) and significant geographical variations in prevalence (434). Source: Hepatitis Testing Innovation Contest, 2016

Decentralization of services, however, may not always be appropriate for or acceptable to potential users. In some settings, centralized viral hepatitis services can provide greater anonymity than neighbourhood services for highrisk populations or others who fear stigma and discrimination. Also, in some low-prevalence settings, decentralizing hepatitis testing may be inefficient and costly. Context, needs, access to laboratory infrastructure and tests, and overall costs and benefits should inform decisions about where hepatitis testing should be decentralized. Decentralization of testing services will also require access to quality-assured RDTs, and DBS specimen collection and analysis.

17.3.4. Task-shifting or -sharing in delivery of hepatitis testing Many countries, including those affected by HBV, HCV and HIV epidemics, continue to face shortages of trained health workers. Task-shifting is a pragmatic response to health workforce shortages. It seeks to increase the effectiveness and efficiency of available personnel and so enable the existing workforce to provide testing services to more people. Several systematic reviews from different areas of health care support the general conclusion that good health outcomes can be achieved by devolving tasks

130

to nurses and lay or community health workers (435–438), with appropriate training and supervision. Task-shifting has been adopted for over a decade to expand HIV testing across the Americas (441), Europe (442, 443), sub-Saharan Africa (444–449) and Asia (450), especially in resource-limited settings where there is a shortage of health-care professionals (438). WHO now recommends that lay providers who are trained and supervised can independently perform HIV counselling and testing using RDTs (11, 439). HIV testing in pregnancy can also be promoted through prescription of ART by nurses (440). In a similar way, task-shifting may also be important for scale up of hepatitis testing, particularly in settings with high HBV or HCV prevalence in the general population or subpopulations (see Box 17.11). Incorporation of viral hepatitis testing into existing task-shifting models of care providing HIV services could be an effective and cost–effective means of fulfilling these objectives. Peerled interventions have also been effective in increasing viral hepatitis testing, care and treatment for marginalized groups of PWID. In addition to providing services, peers can act as role models and offer non-judgemental and respectful support that may contribute to reducing stigma, facilitating access to services and improving their acceptability (25). However, increasing task-shifting and broadening the scope of responsibilities of trained lay providers will not alone fully rectify staff shortages and poor-quality services. Box 17.11 Task-shifting There is emerging evidence from high-income countries that task-shifting can help deliver effective HCVrelated services to vulnerable key populations with outcomes comparable to specialist-level care (346, 451). In British Columbia, Canada, a nurse-coordinated but specialist-supported model of care with specific training and clear protocols resulted in good HCV treatment outcomes in rural and small urban centres (452), and in prisons in Australia (451). Nurses were responsible for patient assessments and education as well as making referrals to other appropriate services according to patients’ needs. In the ECHO programme, New Mexico, United States, primary health-care providers successfully provided high-quality HCV treatment to patients in rural areas and prisons (346). Source: Hepatitis Testing Innovation Contest, 2016

17.4. Diagnostic innovations to promote access to testing Advances in hepatitis virus detection technology have created new opportunities for enhancing hepatitis testing, as well as monitoring the response to treatment. Future directions and innovations in testing include simplified single virological assay testing algorithms, near patient or POC assays for NAT and core antigen, DBS sampling (Chapter 13), multiplex/polyvalent platforms, and self-testing.

131

Simplified testing algorithms. Simplifying testing algorithms will be critical to ensuring affordability and the success of scaling up testing. Potential future testing approaches for HCV infection are the adoption of a less expensive and more manageable single virological test for both diagnosis and confirmation of viraemia (453). However, this may only ever be cost–effective in high-prevalence settings and high-risk populations. Near patient or POC testing. The development of reliable, accurate, practical and affordable near patient tests will be crucial for expanding hepatitis testing services, especially in community-based settings. POC technologies for viral hepatitis include molecular NAT-based tests for diagnosis and treatment monitoring. These emerging POC devices are able to perform conventional laboratory molecular testing (qualitative and quantitative) in field settings; are easier to use than the laboratory-based NAT assays, as they require minimum training and hands-on time; can be operated on battery or conventional power source; do not require phlebotomy; and provide a result within 2 hours. They include cartridge-based HCV RNA assays, which can be used with existing diagnostic platforms developed for TB or HIV early infant diagnosis and viral load monitoring, but HCVcAg POC platforms are also in development. They offer the possibility of a same-day diagnosis of viraemic infection, either alone or when combined with an HCV antibody RDT, as well as test of cure. Multiplex and multi-disease analysers. Multiplex or multi-disease analysers allow for integrated testing of hepatitis B and C alongside other pathogens, e.g. HIV and syphilis, and can leverage technology developed for other infectious disease programmes. Key advantages include the requirement for lower specimen volume, improved client flow with results for multiple pathogens available at the same time, and so fewer patient visits and transport costs. Multiplex RDTs are in development for anti-HIV/anti-HCV, anti-HIV/syphilis/anti-HCV, anti-HIV/syphilis/HBsAg and anti-HIV/ anti-HCV/HBsAg. Data on their diagnostic accuracy and impact on patient-important outcomes are required before adoption. Self-testing. Self-testing is a process in which an individual, who wants to know his or her status collects a specimen, performs a test and interprets the result themselves, often in private. HIV self-testing (HIVST) is now being conducted in many settings. Most studies report that HIVST is highly acceptable across a variety of populations (454–456), and has increased uptake of testing among people not reached by other existing HIV testing services, many of whom are first-time testers (21, 457). The experience with hepatitis self-testing is currently very limited, but it represents a potentially important approach to expand access to testing in the future.

132

18. TESTING ISSUES IN SPECIFIC POPULATIONS This chapter addresses special considerations that apply to viral hepatitis testing in certain priority and high-risk or key populations. This includes PWID; persons in prisons or closed settings; MSM; sex workers; transgender people; persons living with HIV; TB-infected populations; migrant and mobile populations; healthcare workers; couples, partners and household contacts; pregnant women; children; and adolescents.

18.1. Principles for testing in all populations • • Hepatitis testing must emphasize the WHO 5 “Cs”. Mandatory, compulsory or coercive testing is never appropriate (see Chapter 16). All sites that provide hepatitis testing should have SOPs and ethical codes of conduct. They should protect client information and confidentiality, and should employ trained and supervised health workers (including lay providers). All HBV and HCV testing should follow WHO testing strategies and a validated national testing algorithm. Hepatitis testing should have appropriate QA and quality improvement (QI) mechanisms in place. Testing should be part of a care pathway that includes access to prevention, treatment and vaccination services. All persons who test positive for HBV and HCV should be linked to hepatitis care and treatment services. Priority for testing is to diagnose the undiagnosed as well as to identify those both in greatest need of treatment and at greatest risk of transmitting infection.

•

18.2. Principles for testing in key and high-risk populations In some countries, HIV, HBV and HCV infections occur predominantly in certain key or high-risk populations, often via common routes of transmission. Key populations include PWID, MSM, people in prisons and other closed settings, sex workers and transgender people. These populations not only have an increased risk of infection, but their behaviours are often stigmatized, discriminated and criminalized. In almost all countries and settings, hepatitis testing for these key and priority populations is inadequate, and access to prevention, care and treatment services remains low.

133

•

Promotion of health equity and human rights in hepatitis B and C testing is critical, as many of the affected populations such as PWID, prisoners, MSM, and sex workers are those who are systematically excluded from access to testing, treatment and care. Expanded testing and access should be fair, equitable and voluntary, and provided in a supportive environment free of stigma and discrimination. Essential strategies to create an enabling environment for access to hepatitis testing and treatment in these populations include: supportive legislation, policy and financial commitment, such as decriminalization of behaviours of key populations; addressing stigma and discrimination and violence against people from key populations; and community empowerment. In prisons, this can also include addressing additional systemic barriers contributing to transmission of viral hepatitis and other infectious diseases, such as confined unhygienic living spaces, lack of access to clean drinking water and adequate nutrition (458). Testing in prisons. Prisons provide an opportunity to offer testing and treatment to marginalized populations that otherwise might have difficulties accessing care. However, there is a need to guard against the negative consequences of testing in prisons such as mandatory or coercive testing and segregation of prisoners. There are also often major challenges to continuity of care between prisons and the community. All people who test positive need to be linked to viral hepatitis care and treatment services on discharge. Provision of a comprehensive package of prevention and treatment interventions. The high prevalence of comorbidities (e.g. viral hepatitis/ HIV coinfection, TB, mental health issues and polydrug use) in PWID and other high-risk populations means that the provision of comprehensive prevention, treatment, care and social services is important. WHO has outlined a comprehensive set of interventions and approaches for PWID (140), prisoners (459), MSM and sex workers (27, 57). These include provision of condoms, STI screening, HBV vaccination, OST provision and needle–syringe programme (NSP), and referral for ART and antiviral therapy. Full HBV vaccination or adoption of a catch-up vaccination programme is recommended for certain populations at increased risk of HBV, including PWID, MSM (26), sex workers and prisoners, without the need for prior HBsAg testing. Provision of accessible testing and treatment services. ° Integration of testing and service delivery. To facilitate access, testing in certain populations such as PWID should be integrated, where possible, with delivery of other harm-reduction or drug dependency services and HIV testing (460).

•

134

° Training of health-care workers. In many settings, health-care workers lack experience or training on how to provide inclusive and nonjudgemental testing, and there are reports of discrimination against high-risk populations. Countries should prioritize the training of health workers so that they can provide acceptable services, better understand the needs of these populations, and be familiar with local support and prevention services. Similarly, services for transgender persons should be welcoming, with staff who are respectful and sensitive to transgender issues, and are knowledgeable about transgender medical concerns, such as the integration of hormone therapy and hepatitis care. • Testing and repeat testing. Testing should be offered to not only current injecting drug users but to all persons who have ever injected drugs. Repeat screening is required in PWID and other groups such as MSM at ongoing risk of infection with a negative test. The possibility of reinfection after spontaneous clearance or successful treatment should also be considered. Those who have been previously infected should be retested using RNA testing, as the antibody remains positive after the first infection.

Further reading * Consolidated guidelines on HIV prevention, diagnosis, treatment and care for key populations. Geneva: WHO; 2014 and update 2016 (25) describes essential services for key populations and interventions to reduce barriers to testing and linkage to care after testing. (http://www.who. int/hiv/pub/guidelines/keypopulations/en/) * Integrating collaborative TB and HIV services within a comprehensive package of care for people who inject drugs. Geneva: WHO; 2016 (461). (http://apps.who.int/iris/ bitstream/10665/204484/1/9789241510226_eng.pdf?ua=1) * Guidance on prevention of viral hepatitis B and C among people who inject drugs. Geneva: WHO 2012 (http://apps.who.int/iris/bitstream/10665/75357/1/9789241504041_eng.pdf) (28) * WHO, UNODC, UNAIDS Technical guide for countries to set targets for universal access to HIV prevention, treatment and care for injecting drug users –2012 revision. Geneva: WHO 2013 (http://www.who.int/hiv/pub/idu/targets_universal_access/en/, accessed 08 July 2016) (140). *UNODC, ILO, UNDP, WHO, UNAIDS. HIV prevention, treatment and care in prisons and other closed settings: a comprehensive package of interventions. Vienna: UNODC; 2013 (http://www. who.int/hiv/pub/prisons/interventions_package/en/, accessed 08 July 2016) (459). *Prevention and treatment of HIV and other sexually transmitted infections for sex workers in low- and middle-income countries . Geneva: WHO; 2012. (https://www.unfpa.org/sites/default/ files/pub-pdf/9789241504744_eng.pdf) (27). *Implementing comprehensive HIV/STI programmes with sex workers: practical approaches from collaborative interventions. Geneva: World Health Organization; 2013 (http://apps.who.int/ iris/bitstream/10665/90000/1/9789241506182_eng.pdf) (57).

135

18.3 Persons living with HIV Concurrent infection with HIV usually results in more severe and progressive liver disease, and a higher incidence of cirrhosis, HCC and mortality (462–465). HIV-infected persons are therefore a priority group for early diagnosis of viral hepatitis coinfection, and provision of both ART and specific antiviral therapy.

Implementation considerations • Comparable outcomes of DAA therapy have been seen in persons with HIV coinfection as for those with HCV monoinfection, with cure rates higher than 95%, even for those with prior HCV treatment failure or advanced fibrosis (5). Therefore, there is no longer a need to consider HIV/HCV-coinfected patients as a special, difficult-to-treat patient population. HBV vaccination. The risk of HBV infection may be higher in HIV-infected adults, and therefore all persons newly diagnosed with HIV should be screened for HBsAg and anti-HBs to identify those with CHB, and vaccinated if non-immune. Response to HBV vaccine may be lower in HIVinfected persons especially those with a low CD4 count. A schedule using four double (40 µg) doses of the vaccine provides a higher protective antiHBs titre than the regular three 20 µg dose schedule (466).

•

18.4 Tuberculosis-infected populations Certain groups, such as PWID and people in prisons who at increased risk of HCV and HBV infection, are also at risk of infection with TB, largely because they live in regions and/or settings (e.g. prisons or regions of the world) that are endemic for these infections (467, 468).

Implementation considerations • Supporting intensified tuberculosis case-finding at testing facilities. Screening for active TB should be part of the clinical evaluation of patients being considered for HBV and/or HCV and HIV treatment. WHO recommends a four-symptom screening algorithm to rule out active TB (412). In the absence of a cough, weight loss, fever and night sweats, active TB can be confidently ruled out. In the presence of these symptoms, further investigations for TB would be recommended. Drug interactions. Drug-induced liver injury is three- to sixfold higher in persons coinfected with HBV, HCV or HIV who are receiving antituberculosis drugs. All existing DAA combination regimens interact with rifampicin, but there are no serious interactions anticipated between sofosbuvir or daclatasvir and multidrug-resistant (MDR) or extensively drug-resistant (XDR)-TB regimens (469).

•

136

18.5. Migrant and mobile populations In some low-prevalence HBV and HCV regions, such as North America, Europe and Australia, the prevalence of viral hepatitis infection among persons born in high- and intermediate-endemic countries is higher, and reflects that in their country of origin. In other settings, minority ethnic groups and other mobile populations such as migrant workers, refugees, asylum seekers, fisher folk and lorry drivers, are particularly vulnerable to HBV, HCV and HIV infection. All these groups can be hard to reach and have difficulty in accessing health care for HIV or hepatitis testing services because of stigma, language differences, discrimination and legal barriers (53). Displacement of populations through human trafficking may further complicate the provision of testing services (53).

Implementation considerations • Knowledge of the underlying prevalence of viral hepatitis as well as other important diseases of public health significance in migrants and refugees is key for an effective country programme. Barriers to testing uptake among migrant groups, such as language and cultural barriers, need to be addressed in order to increase uptake of testing (193). There is evidence that provision of information and education on hepatitis B to migrant populations may improve knowledge about risk, screening and prevention (470), but not necessarily lead to increased uptake of testing. Persons who have travelled to high-prevalence countries and had an invasive procedure, including tattoos, acupuncture, body piercings, with equipment that may not have been properly sterilized, or those who may have engaged in high-risk sexual behaviours or injecting drug use should also be considered for targeted testing.

•

18.6. Health-care workers Due to the risks associated with occupational exposure to blood and body fluids, health-care workers are a population at risk for acquisition of both hepatitis B and C infection. Exposure to blood and bodily fluids can occur through needle-stick and other sharps injuries, contact with blood and bodily fluids through scratches, abrasions or burns on the skin as well as mucosal surfaces of the eyes, nose, or mouth through accidental splashes (68). However, the largest proportion of occupational transmission of viral hepatitis is due to percutaneous injury via needles during vascular access (66). The risk of HBV transmission with such exposure is estimated to be 6–30% and for HCV transmission around 1.8% (68).

137

Implementation considerations In all settings, testing for hepatitis B (and in many settings for hepatitis C) and the offer of HBV vaccination to health-care workers who are non-immune should be standard practice, but this is currently not widely implemented in LMICs. • Infection control and injection safety. In settings where infection control practices and occupational health and safety standards are inadequate, testing initiatives should take place alongside improvements in safety standards and procedures to protect health-care workers against possible exposure. Post-exposure prophylaxis. In the event of exposure to HBV, post-exposure prophylaxis with HBV vaccine and HBIG should be made available for health-care workers exposed to HBV where the worker has not received vaccination or where the antibody response to HBV vaccination is unknown. Early diagnosis and management of chronic hepatitis B and C infection should be available to all health-care workers where occupational transmission of HBV or HCV has occurred. Those who are HBsAg positive and undertake exposureprone procedures, such as surgeons, gynaecologists, nurses, phlebotomists, personal care attendants and dentists, should be considered for HBV antiviral therapy to reduce direct transmission to others, and DAA therapy for HCV.

•

18.7. Couples, partners, family members and household contacts Testing of couples and partners, family members and household contacts of persons with CHB infection, may be an efficient and effective way of identifying additional people with HBV infection who can also benefit from treatment and monitoring. This may also enable adoption of prevention strategies by the couple or family members (e.g. HBV vaccination, condom use, safe injecting practices) (471). Although the risk of HCV transmission to household contacts and sexual partners among heterosexual and HIV-negative MSM partners is low, there is a small but increased risk among sexual partners of PWID and MSM who engage in high-risk sexual behaviours or are HIV positive. An increasing number of countries offer couples and partner HIV testing (471) in various settings, including ANC, community-based TB services, and HIV/ART clinics, and this can also inform the service delivery of partner testing for viral hepatitis.

Implementation considerations • Couples counselling requires additional training and enhanced counselling skills. Providers must be aware of the potential for intimate partner-based violence and should accept people’s decisions not to test with their partners.

138

Testing for couples who ask to be tested together promotes mutual disclosure of status and increases adoption of prevention measures, especially in the case of discordant couples. Further reading Guidance on couples HIV testing and counselling – including antiretroviral therapy for treatment and prevention in serodiscordant couples: recommendations for a public health approach. Geneva: WHO; 2012 (http://apps.who.int/iris/bitstream/10665/44646/1/9789241501972_ eng.pdf?ua=1) (471).

18.8. Pregnant women Hepatitis B infection. Universal HBV testing in pregnant women already occurs in many parts of the world, but remains suboptimal in resource-limited settings (157). Box 18.1 summarizes the existing WHO guidelines on HBV infection prevention in newborns (6), but the most important preventive strategy is to deliver the first dose of hepatitis B vaccine as soon as possible after birth, preferably within 24 hours followed by at least two timely subsequent doses. Recent studies have suggested that there may also be a role for antiviral therapy in the third trimester in HBV-infected pregnant women to further reduce the risk of MTCT (157, 472, 473). Hepatitis C infection. Although the risk of MTCT of HCV infection is much lower than that of HBV infection, perinatal transmission of HCV occurs in between 4% and 8% of births, but the risk is two to three times higher if the mother is coinfected with HIV (96). Although the costs of implementing HCV testing alongside HIV and HBV is likely to be low, there is currently no effective public health intervention to decrease the risk of MTCT of HCV infection. However, identifying pregnant women who are HCV positive allows avoidance of procedures that promote mixing of fetal and maternal blood (e.g. use of scalp electrodes, amniocentesis), and may thus decrease transmission risk (98). It can also help promote testing of the child at 18 months. Identifying and treating women of reproductive age before they become pregnant preferable, but if DAAs are found to be safe and effective for use in pregnancy, they will also contribute to the prevention of MTCT.

Implementation considerations • Integration with HIV testing. WHO now recommends HIV testing for all pregnant women (11). The offer of HBV testing alongside existing HIV testing and PMTCT interventions is an effective and efficient mechanism of scaling up HBV testing for pregnant women and their partners. Information on risk factors for HCV infection should be communicated to pregnant women and, if present, or in high-endemic settings, testing for HCV should also be considered alongside testing for HIV and HBV.

139

•

Timing of testing. Testing should be done as early as possible during pregnancy to enable pregnant women to benefit most from prevention, treatment and care, and to reduce the risk of transmission to their infants. It can also be performed late in pregnancy, in labour or, if that is not feasible, as soon as possible after delivery. Pre- and post-test counselling. Pre-test information for women who are or may become pregnant or who are postpartum should include: the benefits of early diagnosis of HBV or HCV infection for their own health, as well as to reduce the risk of HBV or HCV transmission to the infant; and importance of testing also for HIV and syphilis. Post-test counselling should include: use of antiviral therapy for the mother’s health as appropriate; measures to reduce the risk of transmitting HBV or HCV infection to the infant; encouragement for partner testing; advice on childbirth plans and infant-feeding options with an encouragement to deliver in a health facility to ensure access to PMTCT services; and HBV and HCV testing for the infant. Linkage to care. There is a significant loss to follow up of pregnant women testing HBV- or HCV-positive who need to be linked to care to assess the need for antiviral treatment and ongoing monitoring. Pregnant women without any serological markers for HBV can be offered HBV vaccination. Follow up should continue through the breastfeeding period to ensure that infants born to mothers with CHB receive the recommended three doses of vaccine, especially if they did not receive the HBV birth-dose vaccination.

•

18.9. Children There are significant gaps and missed opportunities for diagnosis and documenting the HBV and HCV status of children of HBV-positive parents or HCV-positive mothers. Hepatitis B infection. In endemic countries, HBV- infection is predominantly transmitted perinatally or in early childhood. In some settings, up to 50% of childhood infections may be attributable to horizontal intrafamilial transmission. In non-endemic settings, most children with CHB are migrants or children of migrants from endemic countries. Box 18.1 summarizes the existing WHO guidelines on HBV infection prevention in newborns. Although 70–90% of children who are exposed perinatally will become chronically infected, HBV-related morbidity is low during childhood as they are generally in the immune-tolerant phase. Since there are also low curative rates with both long-term NA and IFN treatment, and concerns over long-term safety and risk of drug resistance, a conservative approach to antiviral therapy is indicated, unless there are other criteria for treatment, such as cirrhosis or evidence of severe ongoing necroinflammatory disease (6). Tenofovir is approved for use in adolescents and children above the age of 12 years for HBV treatment (and 3 years or older for HIV treatment), and entecavir above 2 years of age.

140

Box 18.1. WHO recommendations on HBV prevention in newborns and children All infants should receive their first dose of hepatitis B vaccine as soon as possible after birth, preferably within 24 hours, followed by two or three doses. HBIG prophylaxis, in conjunction with HBV vaccination, may be of additional benefit for the following: newborn infants whose mothers are HBsAg positive, particularly if they are also HBeAg positive. In full-term neonates born to mothers who are HBsAg positive but HBeAg negative, protection against perinatally acquired infection achieved by immediate vaccination against HBV (given within 24 hours) may not be significantly improved by the addition of HBIG. Source: Guidelines for the prevention, care and treatment of persons with chronic hepatitis B infection. Geneva: WHO; 2015 ((http:// apps.who.int/iris/bitstream/10665/154590/1/9789241549059_eng.pdf?ua=1&ua=1) (6).

Hepatitis C infection. In countries where adults have a high prevalence of HCV infection, an increased prevalence in children is often observed. This rate is particularly high in those exposed to medical interventions and treated in hospitals (101). Children born to mothers with HCV infection, especially those who are HIV-coinfected, are also at risk (96–99), and MTCT is the most common cause of HCV infection in young children. As with HBV infection, the progression of HCV liver disease is usually slow in infected children. None of the DAAs have yet been approved for use among children (data from ongoing clinical trials will provide the necessary safety and efficacy data for paediatric regulatory approval), and so the only approved treatment remains PEG-IFN/ribavirin. However, as DAAs offer the potential for curative treatment at an early stage before progression of liver disease in children, earlier HCV testing in infants and children will also become more important.

Implementation considerations • Service delivery approaches to delivering testing to infants and children. Box 18.2 shows potential testing approaches to improve hepatitis case-finding among infants and children. Infants whose mothers have been diagnosed with HBV or HCV should be followed up and routinely offered testing, and those diagnosed with either should be regularly monitored for signs of liver disease so that treatment can be offered when necessary. In highprevalence settings, testing of HBV- and HCV-exposed infants could be available through a variety of services – child health services, immunization clinics, under-5 clinics, malnutrition services, well-child services, services for hospitalized and all sick children, TB clinics, and services for orphans and vulnerable children. Follow up through the breastfeeding period is also important to be able to offer HBV testing and vaccination for infants born to mothers with CHB who did not receive the HBV birth-dose vaccination, and to ensure that all children are followed up to receive the recommended

141

three doses of vaccine. However, many infants are lost to follow up, which makes additional paediatric case-finding important. • Testing in infants and children under 18 months. Hepatitis B. Testing of exposed infants is problematic within the first six months of life, as HBsAg and HBV DNA may be inconsistently detectable in infected infants. Exposed infants should be tested for HBsAg at 12 months of age – CHB is diagnosed if there is persistence of HBsAg for six months or more (95). Hepatitis C. HCV infection in children under 18 months can be confirmed only by virological assays to detect HCV RNA, because transplacental maternal antibodies remain in the child’s bloodstream up until 18 months of age, making test results from serology assays ambiguous.

Box 18.2 Potential testing approaches to improve hepatitis case-finding among infants and children • • Prioritize testing children of all HBV- or HCV-positive mothers (especially if the mother is HCV/HIV-coinfected) through home- or facility-based testing. Offer testing to all children and adolescents presenting with signs and symptoms that suggest acute viral hepatitis, including anorexia, nausea, jaundice, right upper quadrant discomfort and abnormal liver function tests. Consider offering viral hepatitis testing to all children and adolescents attending HIV services, STI clinics and TB clinics. Offer viral hepatitis testing or retesting to mothers or infants in immunization clinics or under-5 clinics. Target HCV testing to children who have had medical interventions or received blood products in countries where screening of blood is not routine or where medical equipment is inadequately sterilized.

• • •

18.10. Adolescents In high HBV- or HCV-prevalence settings, two groups of adolescents (defined as 10–19 years of age) are at potential risk of HBV or HCV infection and may need access to testing. These include the following: (1) undiagnosed adolescents who were HBV exposed perinatally or in early childhood in highly endemic HBV settings, and who missed out on HBV vaccination. These adolescents need to be diagnosed and started on antiviral treatment if and when this is clinically indicated, or if negative, vaccinated for HBV. (2) Adolescents who acquire HBV or HCV sexually or through injecting drug use through sex with multiple partners, or with MSM. It is important that these adolescents receive targeted interventions to increase access to HIV and hepatitis testing (474).

142

Implementation considerations • Service delivery – delivering adolescent-friendly services. Engaging adolescents in testing for both HIV and viral hepatitis, either within the health services or community, should be based on adolescent-friendly principles to ensure that psychological as well as physical needs are addressed. Services need to be convenient and available, offer flexible opening hours and/or walk-in or sameday appointments. Separate hours and special events for adolescents may help overcome concerns that they will be seen attending viral hepatitis/HIV services by relatives or neighbours. Disclosure. Adolescents may particularly need support with when and to whom to disclose a positive status (474). When appropriate, and only with the adolescent’s specific permission, health-care personnel should engage the support of adults – family members, teachers, community members. Vulnerable adolescents. Special considerations are needed for particularly vulnerable adolescents, such as those living on the streets, orphans, boys who have sex with men, adolescents in child-headed households, girls engaged in sex with older men, in multiple or concurrent sexual partnerships, or those who are sexually exploited (25). Specific campaigns, use of social media or other web-based approaches, and involving adolescents in identifying appropriate language may help to reach this group in some settings. Age of consent. The age of consent for HIV testing varies from country to country, and this can pose barriers to adolescents’ access to HIV and viral hepatitis testing (474). Testing services should be aware of laws and policies governing the age of consent, and develop appropriate procedures based on this legal framework to ensure that children and adolescents have access to testing. WHO also recommends that children and adolescents themselves be involved in the testing decision as much as possible (474).

•

Further reading HIV and adolescents: guidance for HIV testing and counselling and care for adolescents living with HIV. Geneva: WHO; 2013 (474).

143

19. STRATEGIC PLANNING FOR IMPLEMENTING TESTING SERVICES AND APPROACHES This chapter provides a strategic framework to guide countries’ decision-making on selecting testing approaches, and summarizes the key steps for assessing and improving the selection of hepatitis testing approaches. This includes setting targets, reviewing the effectiveness of existing testing activities and identifying gaps, and then adjusting programme activities.

Key points • There are many facility- and community-based opportunities for and approaches to delivering viral hepatitis testing (see Chapter 17). Countries need to consider a strategic mix of these testing approaches to reach different populations, identify people who are unaware that they are infected in the early stages of infection, and support the timely linkage to prevention, care and treatment services for those who test positive or negative. The selection and mix of testing approaches and application of effective programming practices should be based on a situational assessment that includes: national context and epidemiology (prevalence, populations affected and undiagnosed burden); existing health-care and testing infrastructure; current testing uptake and coverage (number and proportion ever tested by population); programme costs and cost– effectiveness of different testing approaches at national and subnational levels; available financial and human resources; and preferences of the populations to be served. All available epidemiological data from surveillance, surveys and programmes should be used to guide geographical, population, facility and service prioritization. Programmes should monitor data from testing services and in general favour the testing approaches that result in the highest proportion of positive diagnoses in priority populations.

•

144

Key steps for assessing and selecting hepatitis testing approaches Box 19.1 and Fig. 19.1 summarize the key steps for assessing and improving the selection of hepatitis testing approaches, which include setting targets, reviewing the effectiveness of existing testing activities and identifying gaps, and then adjusting programme activities. The final selection and mix of testing approaches with the greatest public health benefit and impact should be based on a situational assessment. This assessment should consider prevalence, unmet need (the estimated number of people who remain undiagnosed), priority populations for the country and the anticipated proportion testing positive, gaps in coverage in geographical areas with undiagnosed HBV and/or HCV infection, the available financial and human resources, and cost–effectiveness. Overall, a mix of hepatitis testing approaches that are focused on populations and/or geographical locations with high HBV or HCV prevalence, and that maximize linkage will have the greatest impact and likely be most cost–effective. FIG. 19.1. Steps to assess, select and evaluate hepatitis testing approaches

Set targets Based on: • Treatment targets • Epidemiology • Current coverage

Review effectiveness and identify gaps • Proportion HBV- or HCV-positive by different testing approaches • Cost per new case identified • Linkage rate • By population group and locale

Adjust programme Identify areas for focus/re-focus • Population • Geography • Sites and setting • Clinical services and conditions

145

Box 19.1. Key steps for assessing and selecting hepatitis testing approaches 1. Review national and subnational epidemiology (prevalence, populations most affected and undiagnosed burden). 2. Set testing (and treatment) coverage targets. 3. Review the effectiveness of existing testing services, and identify gaps. This involves the following: • mapping current services, including availability, uptake (by sex, age and population), coverage rate, funding source and location of all current testing settings and sites; • analysing data from the testing services to assess current testing activities and coverage using different approaches in various sites and locations (e.g. number and proportion of people tested by population, age and sex, new cases diagnosed and enrolled in care); analysing and identifying gaps in current testing coverage in relation to burden, by geographical location and population, focusing on areas of highest prevalence or incidence, which are not being reached by available services; assessing barriers to testing, including social, cultural and geographical factors, psychosocial and behavioural factors, stigma and discrimination, gender and legal factors (including age-of-consent requirements), and structural and health system factors that may impede access; assessing linkage between testing and existing care and treatment programmes following a positive diagnosis; assessing commodity and human resource needs; assessing available human and financial resources.

•

• • •

4. Assess costs and cost–effectiveness of different testing approaches. 5. Monitor, evaluate and adjust testing programme activities.

146

Step 1: Review national and subnational epidemiology In order to devise successful testing services, it is important for countries to understand which populations and settings have the highest prevalence and incidence of HBV and HCV, the estimated number of people affected in the population, and where the greatest burden of undiagnosed infection exists geographically, and by age, sex and population group. Although it is difficult to know the exact number of people with chronic hepatitis B or C infection or the number of new infections in a given area, this can be estimated through the analysis of all available epidemiological data from multiple sources, including surveillance, surveys and programmes. As populationbased household surveys seldom reach or identify high-risk populations and marginalized vulnerable groups, additional studies may be required. A summary of the epidemiological situation would include the following, and the information collated can be summarized in Table 19.1. • Estimates of HBsAg and HCV antibody prevalence in the general and specific high-risk populations ° An estimate of HBsAg and HCV antibody prevalence in the general population stratified by place (if relevant and available) and age group (for general population to identify which ages are at highest risk), as well as in pregnant women attending ANC from national population-based household surveys and surveillance data among pregnant women; ° For each high-risk population group identified with a higher prevalence: - the prevalence of chronic infection in that population group - an estimation of the proportion of the infected population that belongs to that population group - an estimation of the size of that population group. Once the information on the prevalence has been summarized for the general population and specific groups, an analysis of the situation may guide the selection of groups to target with testing services. Testing population groups with a higher prevalence may have a higher yield but lead to the identification of a lower proportion of those living with infection. Testing the general population (or a defined age or birth cohort) may have a lower yield but leads to the identification of a larger proportion of those living with infection. • Hepatitis testing uptake, by different populations and testing approaches;

147

and proportion of those tested who are positive, by population, testing approach or facility; • Number and proportion of people who are aware of their HBV and HCV status Depending on the data available, this may be the proportion of individuals who have ever been tested for HBV or HCV, or of people who were tested in the past 12 months and received their results. These data may be disaggregated by sex, age, geographical region, population type, testing approach and facility. Proportion of people who tested positive and who have been enrolled in hepatitis care and treatment services.

•

Step 2: Set testing (and treatment) coverage targets For each type of testing service, a target may be set in terms of the number of persons to test and to refer for care and treatment or prevention if they are not infected but are at high ongoing risk. An additional target may include the proportion of persons living with viral hepatitis who are diagnosed. The consolidation of targets of all services considered will lead to an overall target for the number of persons for testing and treatment. Coordinating testing with treatment scale up. As the primary reason for diagnosing people with chronic hepatitis B and C is so that they can benefit from treatment, it is important to directly link testing and treatment targets. Plans for major scale up of treatment services will not succeed without testing. Similarly, major scale up of testing, which will create a demand for treatment, will have limited benefit without concurrently expanding treatment capacity.

Step 3: Review the effectiveness of existing testing services and identify gaps Following an epidemiological analysis, an assessment and mapping of current hepatitis testing activities and coverage can determine how well existing services are covering populations in need. This exercise could include the following and the information can be summarized in Table 19.1. • Mapping of existing services, including location of all current testing settings and sites, uptake and coverage rate (by sex, age and population), and funding source. This may include facility-based testing in ANC, TB, STI clinics as well as in harm reduction, outpatient and inpatient services; outreach testing for key populations, community-based and mobile testing, testing within the workplace or educational institutions; and testing by private health-care providers.

148

A detailed situational assessment should also be undertaken with regard to HIV testing services, as in many settings the same populations may be affected, providing the opportunity to integrate hepatitis testing into existing HIV testing delivery models. • Analysing hepatitis testing services data to see what is being achieved by specific approaches in various sites and locations, in terms of the number and proportion of people tested, new cases diagnosed and enrolled in care; • Analysing and identifying gaps in current hepatitis testing coverage in relation to burden, by geographical location and population, focusing on areas of highest prevalence or incidence, which are not being reached by available services; • Assessing the strengths and weaknesses of these testing services, including preferences for testing approaches through key informant interviews with clients and health-care workers; • Assessing barriers to testing, including social, cultural and geographical factors, psychosocial and behavioural factors, stigma and discrimination, gender and legal factors (including age-of-consent requirements), and structural and health system factors that may impede access; • Assessing the linkage between hepatitis testing and existing care and treatment programmes, in particular, following a positive diagnosis; • Assessing laboratory site performance, including the quality of test performance; • Assessing commodity and human resource needs, their availability, and policies to identify barriers to and opportunities for expanding or shifting the focus of programmes (e.g. availability of rapid test kits or trained lay providers and policies regarding task-sharing), and what education, training and certification are required for those conducting tests. The initial assessment should be followed by an inventory of the resources needed and available for testing services. These include (i) equipment (e.g. testing devices) and supplies (e.g. testing kits); (ii) financial resources; and (iii) human resources. • Assessing available financial resources for hepatitis testing, including investments by the government and funding partners.

149

Step 4: Assess costs and cost–effectiveness of different testing approaches • Assessing costs. Comparing the costs associated with a given testing approach between countries can be challenging. Costs for similar services often differ significantly between countries and by testing approach within a country, due to both general cost differences between countries and to differences in the specific services provided (e.g. referral to a clinic for those testing positive versus enhanced linkage support), cadre of staff employed (e.g. nurses versus community health workers), and the ease of reaching different populations. Direct cost comparisons of different testing approaches are easier to interpret when they use the same costing inputs. A common approach to estimating costs involves identifying costs incurred in the following broad categories: personnel (e.g. staff salaries and allowances); recurrent costs (e.g. test kits and commodities, printed materials, office supplies); and capital expenses, often totalled over their useful life and discounted annually at 3% (e.g. office space, vehicles, equipment). These costs can be added to compute the total expected cost of an intervention per year. • Estimating cost–effectiveness. Cost–effectiveness analyses compare the costs and health impacts of different interventions to identify those that provide good value for money, and are useful for optimizing the allocation of public health resources. Health outcomes used in cost–effectiveness analyses of hepatitis testing services include: number of people tested; number of hepatitis B or C cases identified; number of infections averted (when linked to vaccination and prevention of MTCT); number of disability-adjusted life-years (DALYs) lost or number of quality-adjusted life-years (QALYs) gained (dependent not only on being diagnosed but linked to treatment). The health benefits associated with testing are not derived from the test itself, but rather from the treatment and prevention interventions that occur subsequently, including the effectiveness of linkage from testing to treatment. The cost of a programme and its relative cost–effectiveness also depends greatly on the specifics of the programme itself. For example, a programme designed to reach PWID by running mobile camps at various locations can have significantly different costs from providing testing in a fixed location, such as a drug treatment programme. Still, both testing approaches may be necessary to reach this key population. Assessing which testing approaches make the most efficient use of resources requires a detailed understanding of the approaches themselves, including how and to whom they are delivered. Different approaches may be cost–effective for different populations.

150

Step 5: Monitor, evaluate and adjust programme activities Ensuring that hepatitis testing programmes are reaching their intended populations and identifying previously undiagnosed positive persons will require continued monitoring and evaluation. For long-term success, the impact of different hepatitis testing approaches on uptake, the proportion that tests positive, costs, and changes in the prevalence of hepatitis B or C in different population groups must be evaluated and measured regularly, and programmes must be adjusted appropriately. Other activities include the following: • Revisit and revise national targets for and approaches to hepatitis testing so as to better reach those who are undiagnosed, taking into account linkage and enrolment in treatment. • Develop and follow a national consensus plan for expanding and refocusing hepatitis testing in line with the treatment plan. • Evaluate implemented programmes through routine programme monitoring, programme-specific evaluations, surveillance and population-based surveys. • Testing services also require their own monitoring and evaluation framework. In 2016, WHO published a monitoring and evaluation framework for hepatitis B and C (Monitoring and evaluation for viral hepatitis B and C: recommended indicators and framework. Geneva: WHO, 2016) that proposes ten core indicators (22), and includes the proportion of persons living with HBV or HCV infection diagnosed.

151

Table 19.1. A simplified sample template of results of a baseline assessment of testing services Proport io charac n of populati teristic on with the Proport io is aware n of the pop ulation of its s that tatus Estima te popula d proportion tion th of infec at te popula tion gro belongs to th d e up Prevale n group ce of infectio n in th e

Facility or community service access points

d size o f popula ti

Numbe r annuall of tests cond ucte y

s of po pulati

ons

Access point

Strength s

Estima te

General population Pregnant women PWID

XX% 100%

XXXXXX 100%

XXXX XX%

Primary care Hospital ANC clinics OST services Outreach Dialysis centres

XX% XX% XX% XX%

XXXXXX XX% XXXXXX XX% XXXXXX XX% XXXXXX XX%

XXXX XX% XXXX XX% XXXX XX% XXXX XX%

MSM XX% XX% Persons XX% XX% undergoing haemodialysis

Weakn es

Examp le

ses

152

REFERENCES 1. GBD 2013 Mortality and Causes of Death Collaborators. Global, regional, and national age-sex specific all-cause and cause-specific mortality for 240 causes of death, 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet. 2015;385(9963):117–71. Schweitzer A, Horn J, Mikolajczyk RT, Krause G, Ott JJ. Estimations of worldwide prevalence of chronic hepatitis B virus infection: a systematic review of data published between 1965 and 2013. Lancet. 2015;386(10003):1546–55. Gower E, Estes C, Blach S, Razavi-Shearer K, Razavi H. Global epidemiology and genotype distribution of the hepatitis C virus infection. J Hepatol. 2014;61(1 Suppl):S45–S57. Schinazi R, Halfon P, Marcellin P, Asselah T. HCV direct-acting antiviral agents: the best interferon-free combinations. Liver Int. 2014;34 Suppl 1:69–78. Guidelines for the screening, care and treatment of persons with chronic hepatitis C infection. Updated version, April 2016. Geneva: World Health Organization; 2016 (http://apps.who.int/iris/ bitstream/10665/205035/1/9789241549615_eng.pdf?ua=1, accessed 6 February 2017). Guidelines for the prevention, care and treatment of persons with chronic hepatitis B infection. Geneva: World Health Organization; 2015 (http://apps.who.int/iris/bitstream/10665/154590/1/9789241549059_ eng.pdf?ua=1&ua=1, 6 February 2017). Mathurin P. HCV burden in Europe and the possible impact of current treatment. Dig Liver Dis. 2013;45(Suppl 5):S314–S317. Mitchell AE, Colvin HM, Palmer Beasley R. Institute of Medicine recommendations for the prevention and control of hepatitis B and C. Hepatology. 2010;51(3):729–33. Papatheodoridis G, Sypsa V, Kantzanou M, Nikolakopoulos I, Hatzakis A. Estimating the treatment cascade of chronic hepatitis B and C in Greece using a telephone survey. J Viral Hepat. 2015;22(4):409–15. 2. 3. 4. 5.

7. 8. 9.

10. Hatzakis A, Wait S, Bruix J, Buti M, Carballo M, Cavaleri M et al. The state of hepatitis B and C in Europe: report from the hepatitis B and C summit conference. J Viral Hepat. 2011;18 (Suppl 1):1–16. 11. Consolidated guidelines on HIV testing services. Geneva: World Health Organization; 2015 (http://apps. who.int/iris/bitstream/10665/179870/1/9789241508926_eng.pdf?ua=1&ua=1, 6 February 2017). 12. Technical considerations and case definitions to improve surveillance for viral hepatitis. Geneva: World Health Organization; 2016 (http://www.who.int/hepatitis/publications/hep-surveillance-guide-pub/en/, accessed 6 February 2017). 13. Resolution WHA63.18. Viral hepatitis. In: Sixty-third World Health Assembly. Geneva: World Health Organization; 2010 (http://apps.who.int/gb/ebwha/pdf_files/WHA63-REC1/WHA63_REC1-en.pdf accessed 6 February 2017). 14. Resolution WHA67.6. Hepatitis. In: Sixty-seventh World Health Assembly. Geneva: World Health Organization; 2014 (http://apps.who.int/gb/ebwha/pdf_files/WHA67/A67_R6-en.pdf accessed 6 February 2017). 15. Guidelines for the screening, care and treatment of persons with hepatitis C infection. Geneva: World Health Organization; 2014 (http://apps.who.int/iris/bitstream/10665/111747/1/9789241548755_eng. pdf?ua=1&ua=1, accessed 6 February 2017).

16. WHO global health sector strategy on viral hepatitis. Geneva: World Health Organization; 2016 (http:// www.who.int/hepatitis/strategy2016-2021/Draft_global_health_sector_strategy_viral_hepatitis_13nov. pdf?ua=1, accessed 16 February 2017). 17. Hepatitis A. Fact sheet. Geneva: World Health Organization; July 2016 (http://www.who.int/mediacentre/ factsheets/fs328/en/, accessed 05 February 2017). 18. Waterborne outbreaks of hepatitis E: recognition, investigation and control. Geneva: World Health Organization; 2014 (http://apps.who.int/iris/bitstream/10665/129448/1/9789241507608_eng. pdf?ua=1&ua=1, accessed 6 February 2017). 19. Hepatitis delta. Fact sheet. Geneva: World Health Organization; July 2016 (http://www.who.int/ mediacentre/factsheets/hepatitis-d/en/, 05 February 2017). 20. Screening donated blood for transfusion transmissible infections. Geneva: World Health Organization; 2010 (http://www.who.int/bloodsafety/ScreeningDonatedBloodforTransfusion.pdf, accessed 6 February 2017). 21. HIV self-testing and partner notification. Geneva: World Health Organization; 2016 (http://apps.who.int/ iris/bitstream/10665/251655/1/9789241549868-eng.pdf?ua=1, accessed 19 January 2017). 22. Monitoring and evaluation for viral hepatitis B and C: recommended indicators and framework: technical report. Geneva: World Health Organization; 2016 (http://apps.who.int/iris/ bitstream/10665/204790/1/9789241510288_eng.pdf, accessed 6 February 2017). 23. Consolidated guidelines on the use of antiretrovirals for treating and preventing HIV infection. Recommendations for a public health approach. Geneva: World Health Organization; 2016 (http://apps. who.int/iris/bitstream/10665/208825/1/9789241549684_eng.pdf?ua=1, accessed 6 February 2017). 24. Hepatitis B vaccines. Wkly Epidemiol Rec. 2009;84:405–20. 25. Consolidated guidelines on HIV prevention, diagnosis, treatment and care for key populations. Geneva: World Health Organization; 2014 (http://apps.who.int/iris/bitstream/10665/128048/1/9789241507431_ eng.pdf?ua=1&ua=1, accessed 6 February 2017).

153

26. Prevention and treatment of HIV and other sexually transmitted infections among men who have sex with men and transgender people. Geneva: World Health Organization; 2011 (http://apps.who.int/iris/ bitstream/10665/44619/1/9789241501750_eng.pdf, accessed 6 February 2017). 27. Prevention and treatment of HIV and other sexually transmitted infections for sex workers in lowand middle-income countries: recommendations for a public health approach. Geneva: World Health Organization; 2012 (https://www.unfpa.org/sites/default/files/pub-pdf/9789241504744_eng.pdf, accessed 6 February 2017). 28. Guidance on prevention of viral hepatitis B and C among people who inject drugs. Geneva: World Health Organization; 2012 (http://apps.who.int/iris/bitstream/10665/75357/1/9789241504041_eng. pdf?ua=1, accessed 18 December 2016). 29. WHO guidelines on hand hygiene in health care. Geneva: World Health Organization; 2009 (http:// apps.who.int/iris/bitstream/10665/44102/1/9789241597906_eng.pdf, accessed 6 February 2017). 30. Universal access to safe blood transfusion. Geneva: World Health Organization; 2008 (http://www.who. int/bloodsafety/publications/UniversalAccesstoSafeBT.pdf?ua=1, accessed 6 February 2017). 31. WHO guideline on the use of safety-engineered syringes for intramuscular, intradermal and subcutaneous injections in health care settings. Geneva: World Health Organization; 2016. 32. The Universal Declaration of Human Rights. Geneva: United Nations; 1948 (http://www.un.org/en/ documents/udhr/index.shtml, accessed 6 February 2017). 33. Handbook for guideline development, second edition. Geneva: World Health Organization; 2014 (http:// apps.who.int/iris/bitstream/10665/145714/1/9789241548960_eng.pdf, accessed 6 February 2017). 34. Guyatt G, Oxman AD, Akl EA, Kunz R, Vist G, Brozek J et al. GRADE guidelines: 1. Introduction – GRADE evidence profiles and summary of findings tables. J Clin Epidemiol. 2011;64(4):383–94. 35. Balshem H, Helfand M, Schunemann HJ, Oxman AD, Kunz R, Brozek J et al. GRADE guidelines: 3. Rating the quality of evidence. J Clin Epidemiol. 2011;64(4):401–6. 36. Guyatt GH, Oxman AD, Kunz R, Atkins D, Brozek J, Vist G et al. GRADE guidelines: 2. Framing the question and deciding on important outcomes. J Clin Epidemiol. 2011;64(4):395–400. 37. Andrews J, Guyatt G, Oxman AD, Alderson P, Dahm P, Falck-Ytter Y et al. GRADE guidelines: 14. Going from evidence to recommendations: the significance and presentation of recommendations. J Clin Epidemiol. 2013;66(7):719–25. 38. Schunemann HJ, Oxman AD, Brozek J, Glasziou P, Jaeschke R, Vist GE et al. Grading quality of evidence and strength of recommendations for diagnostic tests and strategies. BMJ. 2008;336(7653):1106–10. 39. Gopalakrishna G, Mustafa RA, Davenport C, Scholten RJ, Hyde C, Brozek J et al. Applying Grading of Recommendations Assessment, Development and Evaluation (GRADE) to diagnostic tests was challenging but doable. J Clin Epidemiol. 2014;67(7):760–8. 40. Whiting PF, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155(8):529–36. 41. Guyatt GH, Oxman AD, Vist G, Kunz R, Brozek J, Alonso-Coello P et al. GRADE guidelines: 4. Rating the quality of evidence--study limitations (risk of bias). J Clin Epidemiol. 2011;64(4):407–15. 42. Guyatt GH, Oxman AD, Montori V, Vist G, Kunz R, Brozek J et al. GRADE guidelines: 5. Rating the quality of evidence – publication bias. J Clin Epidemiol. 2011;64(12):1277–82. 43. Guyatt GH, Oxman AD, Kunz R, Brozek J, Alonso-Coello P, Rind D et al. GRADE guidelines 6. Rating the quality of evidence – imprecision. J Clin Epidemiol. 2011;64(12):1283–93. 44. Guyatt GH, Oxman AD, Kunz R, Woodcock J, Brozek J, Helfand M et al. GRADE guidelines: 7. Rating the quality of evidence – inconsistency. J Clin Epidemiol. 2011;64(12):1294–302. 45. Guyatt GH, Oxman AD, Kunz R, Woodcock J, Brozek J, Helfand M et al. GRADE guidelines: 8. Rating the quality of evidence – indirectness. J Clin Epidemiol. 2011;64(12):1303–10. 46. Nelson PK, Mathers BM, Cowie B, Hagan H, Des Jarlais D, Horyniak D et al. Global epidemiology of hepatitis B and hepatitis C in people who inject drugs: results of systematic reviews. Lancet. 2011;378(9791):571–83. 47. Lok AS, McMahon BJ. Chronic hepatitis B. Hepatology. 2007;45(2):507–39. 48. Scheinmann R, Hagan H, Lelutiu-Weinberger C, Stern R, Des Jarlais DC, Flom PL et al. Non-injection drug use and hepatitis C virus: a systematic review. Drug Alcohol Depen. 2007;89(1):1–12. 49. El Maerrawi I, Carvalho HB. Prevalence and risk factors associated with HIV infection, hepatitis and syphilis in a state prison of Sao Paulo. Int J STD AIDS. 2015;26(2):120–7. 50. Evidence for action technical papers: effectiveness of interventions to address HIV in prisons. Geneva: World Health Organization; 2007 (http://apps.who.int/iris/bitstream/10665/43806/1/9789241596190_ eng.pdf, accessed 6 February 2017). 51. Larney S, Kopinski H, Beckwith CG, Zaller ND, Jarlais DD, Hagan H et al. Incidence and prevalence of hepatitis C in prisons and other closed settings: results of a systematic review and meta-analysis. Hepatology. 2013;58(4):1215–24. 52. Padovese V, Egidi AM, Melillo TF, Farrugia B, Carabot P, Didero D et al. Prevalence of latent tuberculosis, syphilis, hepatitis B and C among asylum seekers in Malta. J Public Health (Oxf). 2014;36(1):22–7. 53. Policy statement on HIV testing and counselling for refugees and other persons of concern to UNHCR. Geneva: United Nations High Commissioner for Refugees; 2014 (http://www.unhcr.org/53a816729. pdf, accessed 6 February 2017).

154

54. Hahne SJ, Veldhuijzen IK, Wiessing L, Lim TA, Salminen M, Laar M. Infection with hepatitis B and C virus in Europe: a systematic review of prevalence and cost-effectiveness of screening. BMC Infect Dis. 2013;13:181. 55. Bloodborne viral and sexually transmissible infections in Aboriginal and Torres Strait Islander people: annual surveillance report 2015. Sydney, Australia: The Kirby Institute for infection and immunity in society 2015 (https://kirby.unsw.edu.au/sites/default/files/hiv/resources/atsip2015_v3.pdf, accessed 8 February 2017). 56. Implementing comprehensive HIV/STI programmes with sex workers: practical approaches from collaborative interventions. Geneva: World Health Organization; 2013 (http://apps.who.int/iris/ bitstream/10665/90000/1/9789241506182_eng.pdf, accessed 6 February 2017). 57. Policy brief: transgender people and HIV. Geneva: World Helath Organization; 2015 (http://apps.who. int/iris/bitstream/10665/179517/1/WHO_HIV_2015.17_eng.pdf?ua=1, accessed 5 February 2017). 58. Diamond C, Thiede H, Perdue T, Secura GM, Valleroy L, Mackellar D et al. Viral hepatitis among young men who have sex with men: prevalence of infection, risk behaviors, and vaccination. Sex Transm Dis. 2003;30(5):425–32. 59. Tohme RA, Holmberg SD. Is sexual contact a major mode of hepatitis C virus transmission? Hepatology. 2010;52(4):1497–505. 60. Yaphe S, Bozinoff N, Kyle R, Shivkumar S, Pai NP, Klein M. Incidence of acute hepatitis C virus infection among men who have sex with men with and without HIV infection: a systematic review. Sex Transm Infect. 2012;88(7):558–64. 61. Bradshaw D, Matthews G, Danta M. Sexually transmitted hepatitis C infection: the new epidemic in MSM? Curr Opin Infect Dis. 2013;26(1):66–72. 62. Rauch A, Rickenbach M, Weber R, Hirschel B, Tarr PE, Bucher HC et al. Unsafe sex and increased incidence of hepatitis C virus infection among HIV-infected men who have sex with men: The Swiss HIV cohort study. Clin Infect Dis. 2005;41(3):395–402. 63. van de Laar T, Pybus O, Bruisten S, Brown D, Nelson M, Bhagani S et al. Evidence of a large, international network of HCV transmission in HIV-positive men who have sex with men. Gastroenterology. 2009;136(5):1609–17. 64. Fierer DS, Uriel AJ, Carriero DC, Klepper A, Dieterich DT, Mullen MP et al. Liver fibrosis during an outbreak of acute hepatitis c virus infection in HIV-infected men: a prospective cohort study. J Infect Dis. 2008;198(5):683–6. 65. Danta M, Brown D, Bhagani S, Pybus OG, Sabin CA, Nelson M et al. Recent epidemic of acute hepatitis C virus in HIV-positive men who have sex with men linked to high-risk sexual behaviours. AIDS. 2007;21(8):983–91. 66. Deuffic-Burban S, Delarocque-Astagneau E, Abiteboul D, Bouvet E, Yazdanpanah Y. Blood-borne viruses in health care workers: prevention and management. J Clin Virol. 2011;52(1):4–10. 67. Lee R. Occupational transmission of bloodborne diseases to healthcare workers in developing countries: meeting the challenges. J Hosp Infect. 2009;72(4):285–91. 68. Beltrami EM, Williams IT, Shapiro CN, Chamberland ME. Risk and management of blood-borne infections in health care workers. Clin Microbiol Rev. 2000;13(3):385–407. 69. Global database on blood safety. In: Blood transfusion safety [webpage]. Geneva: World Health Organization; 2011 (http://www.who.int/bloodsafety/global_database/en/, accessed 6 February 2017). 70. Shepard CW, Finelli L, Alter MJ. Global epidemiology of hepatitis C virus infection. Lancet Inf Dis. 2005;5(9):558–67. 71. Frank C, Mohamed MK, Strickland GT, Lavanchy D, Arthur RR, Magder LS et al. The role of parenteral antischistosomal therapy in the spread of hepatitis C virus in Egypt. Lancet. 2000;355(9207):887–91. 72. Singh S, Dwivedi SN, Sood R, Wali JP. Hepatitis B, C and human immunodeficiency virus infections in multiply-injected kala-azar patients in Delhi. Scand J Infect Dis. 2000;32(1):3–6. 73. Marx MA, Murugavel KG, Sivaram S, Balakrishnan P, Steinhoff M, Anand S et al. The association of health-care use and hepatitis C virus infection in a random sample of urban slum community residents in southern India. Am J Trop Med Hyg. 2003;68(2):258–62. 74. Wang CS, Chang TT, Chou P. Differences in risk factors for being either a hepatitis B carrier or antihepatitis C+ in a hepatoma-hyperendemic area in rural Taiwan. J Clin Epidemiol. 1998;51(9):733–8. 75. Ho MS, Hsu CP, Yuh Y, King CC, Tsai JF. High rate of hepatitis C virus infection in an isolated community: persistent hyperendemicity or period-related phenomena? J Med Virol. 1997;52(4):370–6. 76. Lin CC, Hwang SJ, Chiou ST, Kuan CL, Chen LW, Lee TC et al. The prevalence and risk factors analysis of serum antibody to hepatitis C virus in the elders in northeast Taiwan. J Chin Med Assoc. 2003;66(2):103–8. 77. Saxena R, Thakur V, Sood B, Guptan RC, Gururaja S, Sarin SK. Transfusion-associated hepatitis in a tertiary referral hospital in India. A prospective study. Vox Sang. 1999;77(1):6–10. 78. Candotti D, Sarkodie F, Allain JP. Residual risk of transfusion in Ghana. Br J Haematol. 2001;113(1):37–9. 79. Ministry of Health and Population [Egypt], El-Zanaty and Associates [Egypt], ICF International. Egypt Demographic and Health Survey 2014. Cairo, Egypt and Rockville, Maryland, USA: Ministry of Health and Population and ICF International; 2015.

155

80. Omar N, Salama K, Adolf S, El-Saeed GS, Abdel Ghaffar N, Ezzat N. Major risk of blood transfusion in hemolytic anemia patients. Blood Coagul Fibrinolysis. 2011;22(4):280–4. 81. Ghosh K, Joshi SH, Shetty S, Pawar A, Chipkar S, Pujari V et al. Transfusion transmitted diseases in haemophilics from western India. Indian J Med Res. 2000;112:61–4. 82. Arababadi MK, Nasiri Ahmadabadi B, Yousefi Daredor H, Kennedy D. Epidemiology of occult hepatitis B infection among thalassemic, hemophilia, and hemodialysis patients. Hepat Mon. 2012;12(5):315–9. 83. Nishioka Sde A, Gyorkos TW, Joseph L, Collet JP, Maclean JD. Tattooing and risk for transfusiontransmitted diseases: the role of the type, number and design of the tattoos, and the conditions in which they were performed. Epidemiol Infect. 2002;128(1):63–71. 84. Jafari S, Copes R, Baharlou S, Etminan M, Buxton J. Tattooing and the risk of transmission of hepatitis C: a systematic review and meta-analysis. Int J Infect Dis. 2010;14(11):E928–E940. 85. Karmochkine M, Carrat F, Dos Santos O, Cacoub P, Raguin G. A case control study of risk factors for hepatitis C infection in patients with unexplained routes of infection. J Viral Hepat. 2006;13(11):775–82. 86. Ezechi OC, Kalejaiye OO, Gab-Okafor CV, Oladele DA, Oke BO, Musa ZA et al. Sero-prevalence and factors associated with hepatitis B and C co-infection in pregnant Nigerian women living with HIV infection. Pan Afr Med J. 2014;17:197. 87. Easterbrook PJ, Platt L, Gower E, McDonald B, Sabin K, McGowan C et al. Global systematic review and meta-analysis of the seroprevalence of HBV and HCV infection in HIV-infected persons. 8th IAS Conference on HIV Pathogenesis, Treatment and Prevention. 19–22 July 2015. [Abstract No TU PEB254]. 88. Platt L, Easterbrook P, Gower E, McDonald B, Sabin K, McGowan C et al. Prevalence and burden of HCV co-infection in people living with HIV: a global systematic review and meta-analysis. Lancet Infect Dis. 2016;16(7):797–808. 89. Karuru JW, Lule GN, Joshi M, Anzala O. Prevalence of HCV and HCV/HIV co-infection among inpatients at the Kenyatta National Hospital. East Afr Med J. 2005;82(4):170–2. 90. Quaranta JF, Delaney SR, Alleman S, Cassuto JP, Dellamonica P, Allain JP. Prevalence of antibody to hepatitis C virus (HCV) in HIV-1-infected patients (nice SEROCO cohort). J Med Virol. 1994;42(1):29–32. 91. Sherman KE, Rouster SD, Chung RT, Rajicic N. Hepatitis C virus prevalence among patients infected with human immunodeficiency virus: a cross-sectional analysis of the US adult AIDS Clinical Trials Group. Clin Infect Dis. 2002;34(6):831–7. 92. D’Oliveira A, Jr., Voirin N, Allard R, Peyramond D, Chidiac C, Touraine JL et al. Prevalence and sexual risk of hepatitis C virus infection when human immunodeficiency virus was acquired through sexual intercourse among patients of the Lyon University Hospitals, France, 1992–2002. J Viral Hepat. 2005;12(3):330–2. 93. Taylor LE, Swan T, Mayer KH. HIV coinfection with hepatitis C virus: evolving epidemiology and treatment paradigms. Clin Infect Dis. 2012;55 (Suppl 1):S33–S42. 94. Shimakawa Y, Toure-Kane C, Mendy M, Thursz M, Lemoine M. Mother-to-child transmission of hepatitis B in sub-Saharan Africa. Lancet Infect Dis. 2016;16(1):19–20. 95. McMahon BJ. The natural history of chronic hepatitis B virus infection. Semin Liver Dis. 2004;24 (Suppl 1):17–21. 96. Thomas DL, Villano SA, Riester KA, Hershow R, Mofenson LM, Landesman SH et al. Perinatal transmission of hepatitis C virus from human immunodeficiency virus type 1-infected mothers. Women and Infants Transmission Study. J Infect Dis. 1998;177(6):1480–8. 97. Benova L, Mohamoud YA, Calvert C, Abu-Raddad LJ. Vertical transmission of hepatitis C virus: systematic review and meta-analysis. Clin Infect Dis. 2014;59(6):765–73. 98. Mast EE, Hwang LY, Seto DS, Nolte FS, Nainan OV, Wurtzel H et al. Risk factors for perinatal transmission of hepatitis C virus (HCV) and the natural history of HCV infection acquired in infancy. J Infect Dis. 2005;192(11):1880–9. 99. Floreani A. Hepatitis C and pregnancy. World J Gastroenterol. 2013;19(40):6714–20. 100. Camarero C, Martos I, Delgado R, Suarez L, Escobar H, Mateos M. Horizontal transmission of hepatitis C virus in households of infected children. J Pediatr. 1993;123(1):98–9. 101. Thursz M, Fontanet A. HCV transmission in industrialized countries and resource-constrained areas. Nat Rev Gastroenterol Hepatol. 2014;11(1):28–35. 102. Terrault NA, Dodge JL, Murphy EL, Tavis JE, Kiss A, Levin TR et al. Sexual transmission of hepatitis C virus among monogamous heterosexual couples: the HCV partners study. Hepatology. 2013;57(3):881–9. 103. Gupta S, Gupta R, Joshi YK, Singh S. Role of horizontal transmission in hepatitis B virus spread among household contacts in north India. Intervirology. 2008;51(1):7–13. 104. Vandelli C, Renzo F, Romano L, Tisminetzky S, De Palma M, Stroffolini T et al. Lack of evidence of sexual transmission of hepatitis C among monogamous couples: results of a 10-year prospective followup study. Am J Gastroenterol. 2004;99(5):855–9. 105. Perz JF, Armstrong GL, Farrington LA, Hutin YJ, Bell BP. The contributions of hepatitis B virus and hepatitis C virus infections to cirrhosis and primary liver cancer worldwide. J Hepatol. 2006;45(4):529–38. 106. Goldstein ST, Zhou F, Hadler SC, Bell BP, Mast EE, Margolis HS. A mathematical model to estimate global hepatitis B disease burden and vaccination impact. Int J Epidemiol. 2005;34(6):1329–39. 107. Yi P, Chen R, Huang Y, Zhou RR, Fan XG. Management of mother-to-child transmission of hepatitis B virus: propositions and challenges. J Clin Virol. 2016;77:32–9.

156

108. Sarin SK, Kumar M, Lau GK, Abbas Z, Chan HL, Chen CJ et al. Asian-Pacific clinical practice guidelines on the management of hepatitis B: a 2015 update. Hepatol Int. 2016;10(1):1–98. 109. Lavanchy D. Hepatitis B virus epidemiology, disease burden, treatment, and current and emerging prevention and control measures. J Viral Hepat. 2004;11(2):97–107. 110. Yonghao G, Jin X, Jun L, Pumei D, Ying Y, Xiuhong F et al. An epidemiological serosurvey of hepatitis B virus shows evidence of declining prevalence due to hepatitis B vaccination in central China. Int J Infect Dis. 2015;40:75–80. 111. Wasley A, Kruszon-Moran D, Kuhnert W, Simard EP, Finelli L, McQuillan G et al. The prevalence of hepatitis B virus infection in the United States in the era of vaccination. J Infect Dis. 2010;202(2):192–201. 112. Head-to-head comparison of two years efficacy of entecavir and tenofovir in patients with treatment naive chronic hepatitis B – the real life data. Hepatogastroenterology. 2015;62(140):982–6. 113. Hadziyannis SJ. Update on hepatitis B virus infection: focus on treatment. J Clin Transl Hepatol. 2014;2(4):285–91. 114. Rao VB, Johari N, du Cros P, Messina J, Ford N, Cooke GS. Hepatitis C seroprevalence and HIV co-infection in sub-Saharan Africa: a systematic review and meta-analysis. Lancet Infect Dis. 2015;15(7):819–24. 115. Lozano R, Naghavi M, Foreman K, Lim S, Shibuya K, Aboyans V et al. Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380(9859):2095–128. 116. Razavi H, Waked I, Sarrazin C, Myers RP, Idilman R, Calinas F et al. The present and future disease burden of hepatitis C virus (HCV) infection with today’s treatment paradigm. J Viral Hepat. 2014;21:34–59. 117. Eyster ME, Alter HJ, Aledort LM, Quan S, Hatzakis A, Goedert JJ. Heterosexual co-transmission of hepatitis C virus (HCV) and human immunodeficiency virus (HIV). Ann Intern Med. 1991;115(10):764–8. 118. Bica I, McGovern B, Dhar R, Stone D, McGowan K, Scheib R et al. Increasing mortality due to end-stage liver disease in patients with human immunodeficiency virus infection. Clin Infect Dis. 2001;32(3):492–7. 119. Ministry of Health and Population (Egypt), El-Zanaty and Associates (Egypt), ICF International. Egypt health issues survey 2015. Cairo, Egypt and Rockville, Maryland, USA: Ministry of Health and Population and ICF International; 2015. 120. Global database on blood safety. Summary report 2011. Geneva: World Health Organization; 2011 (http://www.who.int/bloodsafety/global_database/GDBS_Summary_Report_2011.pdf?ua=1, accessed 6 February 2017). 121. Marincovich B, Castilla J, del Romero J, Garcia S, Hernando V, Raposo M et al. Absence of hepatitis C virus transmission in a prospective cohort of heterosexual serodiscordant couples. Sex Transm Infect. 2003;79(2):160–2. 122. Tseng YT, Sun HY, Chang SY, Wu CH, Liu WC, Wu PY et al. Seroprevalence of hepatitis virus infection in men who have sex with men aged 18–40 years in Taiwan. J Formos Med Assoc. 2012;111(8):431–8. 123. Price H, Gilson R, Mercey D, Copas A, Parry J, Nardone A et al. Hepatitis C in men who have sex with men in London – a community survey. HIV Med. 2013;14(9):578–80. 124. Metwally A, Mohsen A, Saleh R, Foaud W, Ibrahim N, Rabaah T et al. Prioritizing high-risk practices and exploring new emerging ones associated with hepatitis C virus infection in Egypt. Iranian J Publ Health. 2014;43(10):1385–94. 125. Rein DB, Smith BD, Wittenborn JS, Lesesne SB, Wagner LD, Roblin DW et al. The cost-effectiveness of birth-cohort screening for hepatitis C antibody in US primary care settings. Ann Intern Med. 2012;156(4):263–70. 126. Mohd Hanafiah K, Groeger J, Flaxman AD, Wiersma ST. Global epidemiology of hepatitis C virus infection: new estimates of age-specific antibody to HCV seroprevalence. Hepatology. 2013;57(4):1333–42. 127. Alonso M, Gutzman A, Mazin R, Pinzon CE, Reveiz L, Ghidinelli M. Hepatitis C in key populations in Latin America and the Caribbean: systematic review and meta-analysis. Int J Public Health. 2015;60(7):789–98. 128. Chen Y, Shen Z, Morano JP, Khoshnood K, Wu Z, Lan G et al. Bridging the epidemic: a comprehensive analysis of prevalence and correlates of HIV, hepatitis C, and syphilis, and infection among female sex workers in Guangxi Province, China. PLoS One. 2015;10(2):e0115311. 129. Hagan H, Jordan AE, Neurer J, Cleland CM. Incidence of sexually transmitted hepatitis C virus infection in HIV-positive men who have sex with men. AIDS. 2015;29(17):2335–45. 130. Grebely J, Page K, Sacks-Davis R, van der Loeff MS, Rice TM, Bruneau J et al. The effects of female sex, viral genotype, and IL28B genotype on spontaneous clearance of acute hepatitis C virus infection. Hepatology. 2014;59(1):109–20. 131. Tong MJ, Elfarra NS, Reikes AR, Co RL. Clinical outcomes after transfusion-associated hepatitis-C. N Engl J Med. 1995;332(22):1463–6. 132. Tremolada F, Casarin C, Alberti A, Drago C, Tagger A, Ribero ML et al. Long-term follow-up of non-A, non-B (type C) post-transfusion hepatitis. J Hepatol. 1992;16(3):273–81. 133. Thein HH, Yi Q, Dore GJ, Krahn MD. Estimation of stage-specific fibrosis progression rates in chronic hepatitis C virus infection: a meta-analysis and meta-regression. Hepatology. 2008;48(2):418–31.

157

134. El-Serag HB, Rudolph KL. Hepatocellular carcinoma: epidemiology and molecular carcinogenesis. Gastroenterology. 2007;132(7):2557–76. 135. Busch MP. Insights into the epidemiology, natural history and pathogenesis of hepatitis C virus infection from studies of infected donors and blood product recipients. Transfus Clin Biol. 2001;8(3):200–6. 136. Abdelrahim SS, Khiry RM, Esmail MA, Ragab M, Abdel-Hamid M, Abdelwahab SF. Occult hepatitis C virus infection among Egyptian hemodialysis patients. J Med Virol. 2016;88(8):1388–93. 137. El-Shishtawy S, Sherif N, Abdallh E, Kamel L, Shemis M, Saleem AA et al. Occult hepatitis C virus infection in hemodialysis patients; single center study. Electron Physician. 2015;7(8):1619–25. 138. Rezaee-Zavareh MS, Hadi R, Karimi-Sari H, Hossein Khosravi M, Ajudani R, Dolatimehr F et al. Occult HCV infection: the current state of knowledge. Iran Red Crescent Med J. 2015;17(11):e34181. 139. Blood donor selection: guidelines on assessing donor suitability for blood donation. Geneva: World Health Organization; 2012 (http://www.ncbi.nlm.nih.gov/pubmed/23700651, accessed 6 February 2017). 140. WHO, UNODC, UNAIDS technical guide for countries to set targets for universal access to HIV prevention, treatment and care for injecting drug users. Geneva: World Health Organization; 2012 revision (http://apps.who.int/iris/bitstream/10665/77969/1/9789241504379_eng.pdf?ua=1, accessed 6 February 2017). 141. Allain JP, Opare-Sem O, Sarkodie F, Rahman R, Owusu-Ofori S. Deferred donor care in a regional hospital blood center in Ghana. Transfusion. 2009;49(4):669–75. 142. Pollack H, Wang S, Wyatt L, Peng CH, Wan K, Trinh-Shevrin C et al. A comprehensive screening and treatment model for reducing disparities in hepatitis B. Health Aff (Millwood). 2011;30(10):1974–83. 143. Eckman MH, Kaiser TE, Sherman KE. The cost-effectiveness of screening for chronic hepatitis B infection in the United States. Clin Infect Dis. 2011;52(11):1294–306. 144. Nayagam S, Conteh L, Sicuri E, Shimakawa Y, Suso P, Tamba S et al. Cost-effectiveness of communitybased screening and treatment for chronic hepatitis B in The Gambia: an economic modelling analysis. Lancet Glob Health. 2016;4(8):e568–78. 145. Wong WW, Woo G, Jenny Heathcote E, Krahn M. Cost effectiveness of screening immigrants for hepatitis B. Liver Int. 2011;31(8):1179–90. 146. Rossi C, Schwartzman K, Oxlade O, Klein MB, Greenaway C. Hepatitis B screening and vaccination strategies for newly arrived adult Canadian immigrants and refugees: a cost-effectiveness analysis. PLoS One. 2013;8(10):e78548. 147. Hutton DW, Tan D, So SK, Brandeau ML. Cost-effectiveness of screening and vaccinating Asian and Pacific Islander adults for hepatitis B. Ann Intern Med. 2007;147(7):460–9. 148. Veldhuijzen IK, Toy M, Hahne SJ, De Wit GA, Schalm SW, de Man RA et al. Screening and early treatment of migrants for chronic hepatitis B virus infection is cost-effective. Gastroenterology. 2010;138(2):522–30. 149. Rein DB, Lesesne SB, Smith BD, Weinbaum CM. Models of community-based hepatitis B surface antigen screening programs in the US and their estimated outcomes and costs. Public Health Rep. 2011;126(4):560–7. 150. Jazwa A, Coleman MS, Gazmararian J, Wingate LT, Maskery B, Mitchell T et al. Cost-benefit comparison of two proposed overseas programs for reducing chronic Hepatitis B infection among refugees: is screening essential? Vaccine. 2015;33(11):1393–9. 151. Ruggeri M, Cicchetti A, Gasbarrini A. The cost-effectiveness of alternative strategies against HBV in Italy. Health Policy. 2011;102(1):72–80. 152. Lemoine M, Shimakawa Y, Nije R, Taal M, Ndow G, Chemin I et al., on behalf of the PROLIFICA investigators. Acceptability and feasibility of a screen-and-treat programme for hepatitis B virus infection in The Gambia: the Prevention of Liver Fibrosis and Cancer in Africa (PROLIFICA) study. Lancet Glob Health. 2016;4(8):e559–e567. 153. Easterbrook P, Johnson C, Figueroa C, Baggaley R. HIV and hepatitis testing: global progress, challenges and future directions. AIDS Rev. 2016;18:3–14. 154. Ott JJ, Stevens GA, Groeger J, Wiersma ST. Global epidemiology of hepatitis B virus infection: new estimates of age-specific HBsAg seroprevalence and endemicity. Vaccine. 2012;30(12):2212–9. 155. Qvist T, Cowan SA, Graugaard C, Helleberg M. High linkage to care in a community-based rapid HIV testing and counseling project among men who have sex with men in Copenhagen. Sex Transm Dis. 2014;41(3):209–14. 156. Hensen B, Baggaley R, Wong VJ, Grabbe KL, Shaffer N, Lo YR et al. Universal voluntary HIV testing in antenatal care settings: a review of the contribution of provider-initiated testing & counselling. Trop Med Int Health. 2012;17(1):59–70. 157. Thio CL, Guo N, Xie C, Nelson KE, Ehrhardt S. Global elimination of mother-to-child transmission of hepatitis B: revisiting the current strategy. Lancet Infect Dis. 2015;15(8):981–5. 158. Geue C, Wu O, Xin Y, Heggie R, Hutchinson S, Martin NK et al. Cost-effectiveness of HBV and HCV screening strategies – a systematic review of existing modelling techniques. PLoS One. 2015;10(12):e0145022. 159. Castelnuovo E, Thompson-Coon J, Pitt M, Cramp M, Siebert U, Price A et al. The cost-effectiveness of testing for hepatitis C in former injecting drug users. Health Technol Assess. 2006;10(32):iii–iv, ix–xii, 1–93.

158

160. Loubiere S, Rotily M, Moatti JP. Prevention could be less cost-effective than cure: the case of hepatitis C screening policies in France. Int J Technol Assess Health Care. 2003;19(4):632–45. 161. Loubiere S, Rotily M, Moatti JP. [Medico-economic assessment of the therapeutic management of patients with hepatitis C]. Gastroenterol Clin Biol. 2000;24(11):1047–51. 162. Nakamura J, Terajima K, Aoyagi Y, Akazawa K. Cost-effectiveness of the national screening program for hepatitis C virus in the general population and the high-risk groups. Tohoku J Exp Med. 2008;215(1):33–42. 163. Singer ME, Younossi ZM. Cost effectiveness of screening for hepatitis C virus in asymptomatic, averagerisk adults. Am J Med. 2001;111(8):614–21. 164. Stein K, Dalziel K, Walker A, Jenkins B, Round A, Royle P. Screening for hepatitis C in genito-urinary medicine clinics: a cost utility analysis. J Hepatol. 2003;39(5):814–25. 165. Coffin PO, Scott JD, Golden MR, Sullivan SD. Cost-effectiveness and population outcomes of general population screening for hepatitis C. Clin Infect Dis. 2012;54(9):1259–71. 166. McGarry LJ, Pawar VS, Panchmatia HR, Rubin JL, Davis GL, Younossi ZM et al. Economic model of a birth cohort screening program for hepatitis C virus. Hepatology. 2012;55(5):1344–55. 167. Eckman MH, Talal AH, Gordon SC, Schiff E, Sherman KE. Cost-effectiveness of screening for chronic hepatitis C infection in the United States. Clin Infect Dis. 2013;56(10):1382–93. 168. Liu S, Cipriano LE, Holodniy M, Goldhaber-Fiebert JD. Cost-effectiveness analysis of risk-factor guided and birth-cohort screening for chronic hepatitis C infection in the United States. PLoS One. 2013;8(3):e58975. 169. Lapane KL, Jakiche AF, Sugano D, Weng CS, Carey WD. Hepatitis C infection risk analysis: who should be screened? Comparison of multiple screening strategies based on the National Hepatitis Surveillance Program. Am J Gastroenterol. 1998;93(4):591–6. 170. Wong WW, Tu HA, Feld JJ, Wong T, Krahn M. Cost-effectiveness of screening for hepatitis C in Canada. CMAJ. 2015;187(3):E110–21. 171. Ruggeri M, Coretti S, Gasbarrini A, Cicchetti A. Economic assessment of an anti-HCV screening program in Italy. Value Health. 2013;16(6):965–72. 172. Honeycutt AA, Harris JL, Khavjou O, Buffington J, Jones TS, Rein DB. The costs and impacts of testing for hepatitis C virus antibody in public STD clinics. Public Health Rep. 2007;122 (Suppl 2):55–62. 173. Josset V, Torre JP, Tavolacci MP, Van Rossem-Magnani V, Anselme K, Merle V et al. Efficiency of hepatitis C virus screening strategies in general practice. Gastroenterol Clin Biol. 2004;28(4):351–7. 174. Leal P, Stein K, Rosenberg W. What is the cost utility of screening for hepatitis C virus (HCV) in intravenous drug users? J Med Screen. 1999;6(3):124–31. 175. Stein K, Dalziel K, Walker A, Jenkins B, Round A, Royle P. Screening for hepatitis C in injecting drug users: a cost utility analysis. J Public Health (Oxf). 2004;26(1):61–71. 176. Sutton AJ, Edmunds WJ, Gill ON. Estimating the cost-effectiveness of detecting cases of chronic hepatitis C infection on reception into prison. BMC Public Health. 2006;6:170. 177. Sutton AJ, Edmunds WJ, Sweeting MJ, Gill ON. The cost-effectiveness of screening and treatment for hepatitis C in prisons in England and Wales: a cost-utility analysis. J Viral Hepat. 2008;15(11):797–808. 178. Thompson Coon J, Castelnuovo E, Pitt M, Cramp M, Siebert U, Stein K. Case finding for hepatitis C in primary care: a cost utility analysis. Fam Pract. 2006;23(4):393–406. 179. Tramarin A, Gennaro N, Compostella FA, Gallo C, Wendelaar Bonga LJ, Postma MJ. HCV screening to enable early treatment of hepatitis C: a mathematical model to analyse costs and outcomes in two populations. Curr Pharm Des. 2008;14(17):1655–60. 180. Schackman BR, Leff JA, Barter DM, DiLorenzo MA, Feaster DJ, Metsch LR et al. Cost-effectiveness of rapid hepatitis C virus (HCV) testing and simultaneous rapid HCV and HIV testing in substance abuse treatment programs. Addiction. 2015;110(1):129–43. 181. Cipriano LE, Zaric GS, Holodniy M, Bendavid E, Owens DK, Brandeau ML. Cost effectiveness of screening strategies for early identification of HIV and HCV infection in injection drug users. PLoS One. 2012;7(9):e45176. 182. Jusot JF, Colin C. Cost-effectiveness analysis of strategies for hepatitis C screening in French blood recipients. Eur J Public Health. 2001;11(4):373–9. 183. Linas BP, Wong AY, Schackman BR, Kim AY, Freedberg KA. Cost-effective screening for acute hepatitis C virus infection in HIV-infected men who have sex with men. Clin Infect Dis. 2012;55(2):279–90. 184. Plunkett BA, Grobman WA. Routine hepatitis C virus screening in pregnancy: a cost-effectiveness analysis. Am J Obstet Gynecol. 2005;192(4):1153–61. 185. Urbanus AT, van Keep M, Matser AA, Rozenbaum MH, Weegink CJ, van den Hoek A et al. Is adding HCV screening to the antenatal national screening program in Amsterdam, the Netherlands, costeffective? PLoS One. 2013;8(8):e70319. 186. Deuffic-Burban S, Abiteboul D, Lot F, Branger M, Bouvet E, Yazdanpanah Y. Costs and costeffectiveness of different follow-up schedules for detection of occupational hepatitis C virus infection. Gut. 2009;58(1):105–10. 187. Miners AH, Martin NK, Ghosh A, Hickman M, Vickerman P. Assessing the cost-effectiveness of finding cases of hepatitis C infection in UK migrant populations and the value of further research. J Viral Hepat. 2014;21(9):616–23.

159

188. Martin NK, Hickman M, Miners A, Hutchinson SJ, Taylor A, Vickerman P. Cost-effectiveness of HCV case-finding for people who inject drugs via dried blood spot testing in specialist addiction services and prisons. BMJ Open. 2013;3(8). 189. Martin NK, Vickerman P, Foster GR, Hutchinson SJ, Goldberg DJ, Hickman M. Can antiviral therapy for hepatitis C reduce the prevalence of HCV among injecting drug user populations? A modeling analysis of its prevention utility. J Hepatol. 2011;54(6):1137–44. 190. Brogueira P, Costa A, Miranda A, Peres S, Baptista T, Aldir I et al. Improve screening of HCV infection by targeting high prevalence aged groups: analysis of a cohort of HCV and HIV co-infected patients. J Int AIDS Soc. 2014;17(4 Suppl 3):19601. 191. Kim DD, Hutton DW, Raouf AA, Salama M, Hablas A, Seifeldin IA et al. Cost-effectiveness model for hepatitis C screening and treatment: implications for Egypt and other countries with high prevalence. Glob Public Health. 2015;10(3):296–317. 192. Asrani SK, Davis GL. Impact of birth cohort screening for hepatitis C. Curr Gastroenterol Rep. 2014;16(4):381. 193. Jones L, Bates G, McCoy E, Beynon C, McVeigh J, Bellis M. A systematic review of the effectiveness and cost-effectiveness of interventions aimed at raising awareness and engaging with groups who are at an increased risk of hepatitis B and C infection – final report. Liverpool: Centre for Public Health, Faculty of Health and Applied Social Sciences, Liverpool John Moores University; 2012. 194. Suthar AB, Ford N, Bachanas PJ, Wong VJ, Rajan JS, Saltzman AK et al. Towards universal voluntary HIV testing and counselling: a systematic review and meta-analysis of community-based approaches. PLoS Med. 2013;10(8):e1001496. 195. Sylvestre DL, Loftis JM, Hauser P, Genser S, Cesari H, Borek N et al. Co-occurring hepatitis C, substance use, and psychiatric illness: treatment issues and developing integrated models of care. J Urban Health. 2004;81(4):719–34. 196. Gunn RA, Lee MA, Callahan DB, Gonzales P, Murray PJ, Margolis HS. Integrating hepatitis, STD, and HIV services into a drug rehabilitation program. Am J Prev Med. 2005;29(1):27–33. 197. Mvere D, Constantine NT, Katsawde E, Tobaiwa O, Dambire S, Corcoran P. Rapid and simple hepatitis assays: encouraging results from a blood donor population in Zimbabwe. Bull World Health Organ. 1996;74(1):19–24. 198. Sato K, Ichiyama S, Iinuma Y, Nada T, Shimokata K, Nakashima N. Evaluation of immunochromatographic assay systems for rapid detection of hepatitis B surface antigen and antibody, Dainascreen HBsAg and Dainascreen Ausab. J Clin Microbiol. 1996;34(6):1420–2. 199. Abraham P, Sujatha R, Raghuraman S, Subramaniam T, Sridharan G. Evaluation of two immunochromatographic assays in relation to `RAPID’ screening of HBsAG. Indian J Med Microbiol. 1998;16(1):23–5. 200. Oh J, Kim T, Yoon H, Min H, Lee H, Choi T. Evaluation of Genedia® HBsAg rapid and Genedia® antiHBs rapid for the screening of HBsAg and anti-HBs. Korean J Clin Pathol. 1999;19(1):114–7. 201. Kaur H, Dhanao J, Oberoi A. Evaluation of rapid kits for detection of HIV, HBsAg and HCV infections. Indian J Med Sci. 2000;54(10):432–4. 202. Lien TX, Tien NT, Chanpong GF, Cuc CT, Yen VT, Soderquist R et al. Evaluation of rapid diagnostic tests for the detection of human immunodeficiency virus types 1 and 2, hepatitis B surface antigen, and syphilis in Ho Chi Minh City, Vietnam. Am J Trop Med Hyg. 2000;62(2):301–9. 203. Raj AA, Subramaniam T, Raghuraman S, Abraham P. Evaluation of an indigenously manufactured rapid immunochromatographic test for detection of HBsAg. Indian J Pathol Microbiol. 2001;44(4):413–4. 204. Clement F, Dewint P, Leroux-Roels G. Evaluation of a new rapid test for the combined detection of hepatitis B virus surface antigen and hepatitis B virus e antigen. J Clin Microbiol. 2002;40(12):4603–6. 205. Lau DT, Ma H, Lemon SM, Doo E, Ghany MG, Miskovsky E et al. A rapid immunochromatographic assay for hepatitis B virus screening. J Viral Hepat. 2003;10(4):331–4. 206. Akanmu AS, Esan OA, Adewuyi JO, Davies AO, Okany CC, Olatunji RO et al. Evaluation of a rapid test kit for detection of HBsAg/eAg in whole blood: a possible method for pre-donation testing. Afr J Med Med Sci. 2006;35(1):5–8. 207. Nyirenda M, Beadsworth MB, Stephany P, Hart CA, Hart IJ, Munthali C et al. Prevalence of infection with hepatitis B and C virus and coinfection with HIV in medical inpatients in Malawi. J Infect. 2008;57(1):72–7. 208. Lin YH, Wang Y, Loua A, Day GJ, Qiu Y, Nadala ECJ et al. Evaluation of a new hepatitis B virus surface antigen rapid test with improved sensitivity. J Clin Microbiol. 2008;46(10):3319–24. 209. Randrianirina F, Carod JF, Ratsima E, Chretien JB, Richard V, Talarmin A. Evaluation of the performance of four rapid tests for detection of hepatitis B surface antigen in Antananarivo, Madagascar. J Virol Methods. 2008;151(2):294–7. 210. Ola SO, Otegbayo JA, Yakubu A, Aje AO, Odaibo GN, Shokunbi W. Pitfalls in diagnosis of hepatitis B virus infection among adult Nigerians. Niger J Clin Pract. 2009;12(4):350–4. 211. Khan J, Lone D, Hameed A, Munim R, Bhatti M, Khattak A et al. Evaluation of the performance of two rapid immunochromatographic tests for detection of hepatitis B surface antigen and anti HCV antibodies using ELISA tested samples. Ann King Edw Med Univ. 2010;16(1):84–7. 212. Davies J, van Oosterhout JJ, Nyirenda M, Bowden J, Moore E, Hart IJ et al. Reliability of rapid testing for hepatitis B in a region of high HIV endemicity. Trans R Soc Trop Med Hyg. 2010;104(2):162–4.

160

213. Bjoerkvoll B, Viet L, Ol HS, Lan NT, Sothy S, Hoel H et al. Screening test accuracy among potential blood donors of HBsAg, anti-HBc and anti-HCV to detect hepatitis B and C virus infection in rural Cambodia and Vietnam. Southeast Asian J Trop Med Public Health. 2010;41(5):1127–35. 214. Geretti AM, Patel M, Sarfo FS, Chadwick D, Verheyen J, Fraune M et al. Detection of highly prevalent hepatitis B virus coinfection among HIV-seropositive persons in Ghana. J Clin Microbiol. 2010;48(9):3223–30. 215. Hoffmann CJ, Dayal D, Cheyip M, McIntyre JA, Gray GE, Conway S et al. Prevalence and associations with hepatitis B and hepatitis C infection among HIV-infected adults in South Africa. Int J STD AIDS. 2012;23(10):e10–e3. 216. Bottero J, Boyd A, Gozlan J, Lemoine M, Carrat F, Collignon A et al. Performance of rapid tests for detection of HBsAg and anti-HBsAb in a large cohort, France. J Hepatol. 2013;58(3):473–8. 217. Chameera E, Noordeen F, Pandithasundara H, Abeykoon A. Diagnostic efficacy of rapid assays used for the detection of hepatitis B virus surface antigen. Sri Lankan Journal of Infectious Diseases. 2013;3(2):21–7. 218. Franzeck FC, Ngwale R, Msongole B, Hamisi M, Abdul O, Henning L et al. Viral hepatitis and rapid diagnostic test based screening for HBsAg in HIV-infected patients in rural Tanzania. PLoS One. 2013;8(3):e58468. 219. Chevaliez S, Challine D, Naija H, Luu TC, Laperche S, Nadala L et al. Performance of a new rapid test for the detection of hepatitis B surface antigen in various patient populations. J Clin Virol. 2014;59(2):89–93. 220. Erhabor O, Kwaifa I, Bayawa A, Isaac Z, Dorcas I, Sani I. Comparison of ELISA and rapid screening techniques for the detection of HBsAg among blood donors in Usmanu Danfodiyo University Teaching Hospital Sokoto, North Western Nigeria. J Blood Lymph. 2013;4(2):124. 221. Gish RG, Gutierrez JA, Navarro-Cazarez N, Giang K, Adler D, Tran B et al. A simple and inexpensive point-of-care test for hepatitis B surface antigen detection: serological and molecular evaluation. J Viral Hepat. 2014;21(12):905–8. 222. Honge B, Jespersen S, Medina C, Te D, da Silva Z, Ostergaard L et al. Hepatitis B virus surface antigen and anti-hepatitis C virus rapid tests underestimate hepatitis prevalence among HIV-infected patients. HIV Med. 2014;15(9):571–6. 223. Liu C, Chen T, Lin J, Chen H, Chen J, Lin S et al. Evaluation of the performance of four methods for detection of hepatitis B surface antigen and their application for testing 116,455 specimens. J Virol Methods. 2014;196:174–8. 224. Mutocheluh M, Owusu M, Kwofie TB, Akadigo T, Appau E, Narkwa PW. Risk factors associated with hepatitis B exposure and the reliability of five rapid kits commonly used for screening blood donors in Ghana. BMC Res Notes. 2014;7:873. 225. Upreti SR, Gurung S, Patel M, Dixit SM, Krause LK, Shakya G et al. Prevalence of chronic hepatitis B virus infection before and after implementation of a hepatitis B vaccination program among children in Nepal. Vaccine. 2014;32(34):4304–9. 226. Njai HF, Shimakawa Y, Sanneh B, Ferguson L, Ndow G, Mendy M et al. Validation of rapid point-of-care (POC) tests for detection of hepatitis B surface antigen in field and laboratory settings in the Gambia, Western Africa. J Clin Microbiol. 2015;53(4):1156–63. 227. Ol HS, Bjoerkvoll B, Sothy S, Van Heng Y, Hoel H, Husebekk A et al. Prevalence of hepatitis B and hepatitis C virus infections in potential blood donors in rural Cambodia. Southeast Asian J Trop Med Public Health. 2009;40(5):963–71. 228. Peng J, Cheng L, Yin B, Guan Q, Liu Y, Wu S et al. Development of an economic and efficient strategy to detect HBsAg: application of ‘gray-zones’ in ELISA and combined use of several detection assays. Clin Chim Acta. 2011;412(23–24):2046–51. 229. Viet L, Lan NT, Ty PX, Bjorkvoll B, Hoel H, Gutteberg T et al. Prevalence of hepatitis B & hepatitis C virus infections in potential blood donors in rural Vietnam. Indian J Med Res. 2012;136(1):74–81. 230. Scheiblauer H, El-Nageh M, Diaz S, Nick S, Zeichhardt H, Grunert HP et al. Performance evaluation of 70 hepatitis B virus (HBV) surface antigen (HBsAg) assays from around the world by a geographically diverse panel with an array of HBV genotypes and HBsAg subtypes. Vox Sang. 2010;98(3 Pt 2):403–14. 231. National HBV testing policy. Sydney: Australasian Society for HIV, Viral hepatitis and Sexual Health Medicine; 2012 (http://testingportal.ashm.org.au/images/HepB_TESTING_POLICY_MARCH2014_ V1.1_FOR%20PRINT.pdf, accessed 5 February 2017). 232. Gao F, Talbot EA, Loring CH, Power JJ, Dionne-Odom J, Alroy-Preis S et al. Performance of the OraQuick HCV rapid antibody test for screening exposed patients in a hepatitis C outbreak investigation. J Clin Microbiol. 2014;52(7):2650–2. 233. Hess KL, Fisher DG, Reynolds GL. Sensitivity and specificity of point-of-care rapid combination syphilisHIV-HCV tests. PLoS One. 2014;9(11):e112190. 234. Hui AY, Chan FK, Chan PK, Tam JS, Sung JJ. Evaluation of a new rapid whole-blood serological test for hepatitis c virus. Acta Virol. 2002;46(1):47–8. 235. Lee SR, Kardos KW, Schiff E, Berne CA, Mounzer K, Banks AT et al. Evaluation of a new, rapid test for detecting HCV infection, suitable for use with blood or oral fluid. J Virol Methods. 2011;172(1–2):27–31. 236. Al-Tahish G, El-Barrawy MA, Hashish MH, Heddaya Z. Effectiveness of three types of rapid tests for the detection of hepatitis C virus antibodies among blood donors in Alexandria, Egypt. J Virol Methods. 2013;189(2):370–4.

161

237. Buti M, Cotrina M, Chan H, Jardi R, Rodriguez F, Costa X et al. Rapid method for the detection of antiHCV antibodies in patients with chronic hepatitis C. Rev Esp Enferm Dig. 2000;92(3):140–6. 238. Cha YJ, Park Q, Kang ES, Yoo BC, Park KU, Kim JW et al. Performance evaluation of the OraQuick hepatitis C virus rapid antibody test. Ann Lab Med. 2013;33(3):184–9. 239. da Rosa L, Dantas-Correa EB, Narciso-Schiavon JL, Schiavon L de L. Diagnostic performance of two point-of-care tests for anti-HCV detection. Hepat Mon. 2013;13(9):e12274. 240. Ibrahim S, Al Attas SA, Mansour GA, Ouda S, Fallatah H. Accuracy of rapid oral HCV diagnostic test among a Saudi sample. Clin Oral Investig. 2015;19(2):475–80. 241. Jewett A, Smith BD, Garfein RS, Cuevas-Mota J, Teshale EH, Weinbaum CM. Field-based performance of three pre-market rapid hepatitis C virus antibody assays in STAHR (Study to Assess Hepatitis C Risk) among young adults who inject drugs in San Diego, CA. J Clin Virol. 2012;54(3):213–7. 242. Kosack CS, Nick S, Shanks L. Diagnostic accuracy evaluation of the ImmunoFlow HCV rapid immunochromatographic test for the detection of hepatitis C antibodies. J Virol Methods. 2014;204:6–10. 243. Kim MH, Kang SY, Lee WI. Evaluation of a new rapid test kit to detect hepatitis C virus infection. J Virol Methods. 2013;193(2):379–82. 244. Montebugnoli L, Borea G, Miniero R, Sprovieri G. A rapid test for the visual detection of anti-hepatitis C virus antibodies in whole blood. Clin Chim Acta. 1999;288(1–2):91–6. 245. Poovorawan Y, Theamboonlers A, Chumdermpadetsuk S, Thong CP. Comparative results in detection of HCV antibodies by using a rapid HCV test, ELISA and immunoblot. Southeast Asian J Trop Med Public Health. 1994;25(4):647–9. 246. Scalioni Lde P, Cruz HM, de Paula VS, Miguel JC, Marques VA, Villela-Nogueira CA et al. Performance of rapid hepatitis C virus antibody assays among high- and low-risk populations. J Clin Virol. 2014;60(3):200–5. 247. Smith BD, Drobeniuc J, Jewett A, Branson BM, Garfein RS, Teshale E et al. Evaluation of three rapid screening assays for detection of antibodies to hepatitis C virus. J Infect Dis. 2011;204(6):825–31. 248. Yaari A, Tovbin D, Zlotnick M, Mostoslavsky M, Shemer-Avni Y, Hanuka N et al. Detection of HCV salivary antibodies by a simple and rapid test. J Virol Methods. 2006;133(1):1–5. 249. Daniel HD, Abraham P, Raghuraman S, Vivekanandan P, Subramaniam T, Sridharan G. Evaluation of a rapid assay as an alternative to conventional enzyme immunoassays for detection of hepatitis C virusspecific antibodies. J Clin Microbiol. 2005;43(4):1977–8. 250. Drobnik A, Judd C, Banach D, Egger J, Konty K, Rude E. Public health implications of rapid hepatitis C screening with an oral swab for community-based organizations serving high-risk populations. Am J Public Health. 2011;101(11):2151–5. 251. Lee SR, Yearwood GD, Guillon GB, Kurtz LA, Fischl M, Friel T et al. Evaluation of a rapid, point-of-care test device for the diagnosis of hepatitis C infection. J Clin Virol. 2010;48(1):15–7. 252. Njouom R, Tejiokem MC, Zanga MC, Pouillot R, Ayouba A, Pasquier C et al. A cost-effective algorithm for the diagnosis of hepatitis C virus infection and prediction of HCV viremia in Cameroon. J Virol Methods. 2006;133(2):223–6. 253. O’Connell RJ, Gates RG, Bautista CT, Imbach M, Eggleston JC, Beardsley SG et al. Laboratory evaluation of rapid test kits to detect hepatitis C antibody for use in predonation screening in emergency settings. Transfusion. 2013;53(3):505–17. 254. Yuen MF, Hui CK, Yuen JC, Young JL, Lai CL. The accuracy of SM-HCV rapid test for the detection of antibody to hepatitis C virus. Am J Gastroenterol. 2001;96(3):838–41. 255. Lee S, Kardos K, Yearwood G, Kurtz L, Roehler M, Feiss G. Results of a multi-center evaluation of a new rapid test for detection of HCV infection using whole blood, serum, plasma and oral fluid. J Hepatol. 2010;52(Suppl. 1):S271. 256. Larrat S, Bourdon C, Baccard M, Garnaud C, Mathieu S, Quesada JL et al. Performance of an antigen-antibody combined assay for hepatitis C virus testing without venipuncture. J Clin Virol. 2012;55(3):220–5. 257. Smith BD, Teshale E, Jewett A, Weinbaum CM, Neaigus A, Hagan H et al. Performance of premarket rapid hepatitis C virus antibody assays in 4 national human immunodeficiency virus behavioral surveillance system sites. Clin Infect Dis. 2011;53(8):780–6. 258. Mullis CE, Laeyendecker O, Reynolds SJ, Ocama P, Quinn J, Boaz I et al. High frequency of falsepositive hepatitis C virus enzyme-linked immunosorbent assay in Rakai, Uganda. Clin Infect Dis. 2013;57(12):1747–50. 259. Barreto AM, Takei K, E CS, Bellesa MA, Salles NA, Barreto CC et al. Cost-effective analysis of different algorithms for the diagnosis of hepatitis C virus infection. Braz J Med Biol Res. 2008;41(2):126–34. 260. Bonacini M, Lin HJ, Hollinger FB. Effect of coexisting HIV-1 infection on the diagnosis and evaluation of hepatitis C virus. J Acquir Immune Defic Syndr. 2001;26(4):340–4. 261. George SL, Gebhardt J, Klinzman D, Foster MB, Patrick KD, Schmidt WN et al. Hepatitis C virus viremia in HIV-infected individuals with negative HCV antibody tests. J Acquir Immune Defic Syndr. 2002;31(2):154–62. 262. Thio CL, Nolt KR, Astemborski J, Vlahov D, Nelson KE, Thomas DL. Screening for hepatitis C virus in human immunodeficiency virus-infected individuals. J Clin Microbiol. 2000;38(2):575–7.

162

263. Thomson EC, Nastouli E, Main J, Karayiannis P, Eliahoo J, Muir D et al. Delayed anti-HCV antibody response in HIV-positive men acutely infected with HCV. AIDS. 2009;23(1):89–93. 264. Lok AS, McMahon BJ. Chronic hepatitis B: update 2009. Hepatology. 2009;50(3):661–2. 265. Brunetto MR, Oliveri F, Colombatto P, Moriconi F, Ciccorossi P, Coco B et al. Hepatitis B surface antigen serum levels help to distinguish active from inactive hepatitis B virus genotype D carriers. Gastroenterology. 2010;139(2):483–90. 266. Chen CJ, Iloeje UH, Yang HI. Long-term outcomes in hepatitis B: the REVEAL-HBV study. Clin Liver Dis. 2007;11(4):797–816. 267. Gerlach JT, Diepolder HM, Zachoval R, Gruener NH, Jung MC, Ulsenheimer A et al. Acute hepatitis C: high rate of both spontaneous and treatment-induced viral clearance. Gastroenterology. 2003;125(1):80–8. 268. UNITAID. Hepatitis C diagnostic technology landscape. Geneva: World Health Organization; 2015 (http://unitaid.org/images/marketdynamics/publications/UNITAID-HCV_Diagnostic_Landscape-1st_ edition.pdf, accessed 5 February 2017). 269. Tanaka E, Kiyosawa K, Matsumoto A, Kashiwakuma T, Hasegawa A, Mori H et al. Serum levels of hepatitis C virus core protein in patients with chronic hepatitis C treated with interferon alfa. Hepatology. 1996;23(6):1330–3. 270. Tillmann HL. Hepatitis C virus core antigen testing: role in diagnosis, disease monitoring and treatment. World J Gastroenterol. 2014;20(22):6701–6. 271. Freiman JM, Tran TM, Schumacher SG, White LF, Ongarello S, Cohn J et al. Hepatitis C core antigen testing for diagnosis of hepatitis C virus infection: a systematic review and meta-analysis. Ann Intern Med. 2016;165(5):345–55. 272. Lee SC, Antony A, Lee N, Leibow J, Yang JQ, Soviero S et al. Improved version 2.0 qualitative and quantitative AMPLICOR reverse transcription-PCR tests for hepatitis C virus RNA: calibration to international units, enhanced genotype reactivity, and performance characteristics. J Clin Microbiol. 2000;38(11):4171–9. 273. Yu ML, Chuang WL, Dai CY, Chen SC, Lin ZY, Hsieh MY et al. Clinical evaluation of the automated COBAS AMPLICOR HCV MONITOR test version 2.0 for quantifying serum hepatitis C virus RNA and comparison to the quantiplex HCV version 2.0 test. J Clin Microbiol. 2000;38(8):2933–9. 274. Sabato MF, Shiffman ML, Langley MR, Wilkinson DS, Ferreira-Gonzalez A. Comparison of performance characteristics of three real-time reverse transcription-PCR test systems for detection and quantification of hepatitis C virus. J Clin Microbiol. 2007;45(8):2529–36. 275. Vermehren J, Kau A, Gartner BC, Gobel R, Zeuzem S, Sarrazin C. Differences between two real-time PCR-based hepatitis C virus (HCV) assays (RealTime HCV and Cobas AmpliPrep/Cobas TaqMan) and one signal amplification assay (Versant HCV RNA 3.0) for RNA detection and quantification. J Clin Microbiol. 2008;46(12):3880–91. 276. Vermehren J, Susser S, Berger A, Perner D, Peiffer KH, Allwinn R et al. Clinical utility of the ARCHITECT HCV Ag assay for early treatment monitoring in patients with chronic hepatitis C genotype 1 infection. J Clin Virol. 2012;55(1):17–22. 277. European Association for the Study of the Liver. EASL Clinical Practice Guidelines: management of hepatitis C virus infection. J Hepatol. 2014;60(2):392–420. (http://www.easl.eu/medias/cpg/HEPCrevised-version/English-report.pdf, accessed 5 February 2017). 278. Ghany MG, Nelson DR, Strader DB, Thomas DL, Seeff LB. An update on treatment of genotype 1 chronic hepatitis C virus infection: 2011 practice guideline by the American Association for the Study of Liver Diseases. Hepatology. 2011;54(4):1433–44. 279. Loggi E, Cursaro C, Scuteri A, Grandini E, Panno AM, Galli S et al. Patterns of HCV-RNA and HCV core antigen in the early monitoring of standard treatment for chronic hepatitis C. J Clin Virol. 2013;56(3):207–11. 280. Fujino T, Nakamuta M, Aoyagi Y, Fukuizumi K, Takemoto R, Yoshimoto T et al. Early decline of the HCV core antigen can predict SVR in patients with HCV treated by pegylated interferon plus ribavirin combination therapy. J Dig Dis. 2009;10(1):21–5. 281. Takahashi M, Saito H, Higashimoto M, Atsukawa K, Ishii H. Benefit of hepatitis C virus core antigen assay in prediction of therapeutic response to interferon and ribavirin combination therapy. J Clin Microbiol. 2005;43(1):186–91. 282. Feng B, Yang RF, Xie Q, Shang J, Kong FY, Zhang HY et al. Hepatitis C virus core antigen, an earlier and stronger predictor on sustained virological response in patients with genotype 1 HCV infection. BMC Gastroenterol. 2014;14:47. 283. Moscato GA, Giannelli G, Grandi B, Pieri D, Marsi O, Guarducci I et al. Quantitative determination of hepatitis C core antigen in therapy monitoring for chronic hepatitis C. Intervirology. 2011;54(2):61–5. 284. Gu J, Yu T, Liang Z. Performances of HCV Ag or HCV RNA kits for screening of HCV-infected samples. Chinese Journal of Biologicals. 2014;27(9):1181–4. 285. van Helden J, Weiskirchen R. Hepatitis C diagnostics: clinical evaluation of the HCV-core antigen determination. Z Gastroenterol. 2014;52(10):1164–70. 286. Schnuriger A, Dominguez S, Valantin MA, Tubiana R, Duvivier C, Ghosn J et al. Early detection of hepatitis C virus infection using a new combined antigen-antibody detection assay: potential use in HIV co-infected individuals. Pathol Biol (Paris). 2007;54(10):578–86.

163

287. Okazaki K, Nishiyama Y, Saitou T, Shibata N, Yamamoto C, Oosaga J et al. Fundamental evaluation of HCV core antigen method comparison with Cobas Amplicor HCV monitor v2.0 (high range method). Rinsho Byori. 2008;56(2):95–100. 288. Sidharthan S, Kohli A, Sims Z, Nelson A, Osinusi A, Masur H et al. Utility of hepatitis C viral load monitoring on direct-acting antiviral therapy. Clin Infect Dis. 2015;60(12):1743–51. 289. Snijdewind IJ, van Kampen JJ, Fraaij PL, van der Ende ME, Osterhaus AD, Gruters RA. Current and future applications of dried blood spots in viral disease management. Antiviral Res. 2012;93(3):309–21. 290. Grüner N, Stambouli O, Ross RS. Dried blood spots – preparing and processing for use in immunoassays and in molecular techniques. J Vis Exp. 2015(97):52619. 291. McDade TW, Williams S, Snodgrass JJ. What a drop can do: dried blood spots as a minimally invasive method for integrating biomarkers into population-based research. Demography. 2007;44(4):899–925. 292. Hirtz C, Lehmann S. Blood sampling using “dried blood spot”: a clinical biology revolution underway? Ann Biol Clin (Paris). 2015;73(1):25–37. 293. Ross RS, Stambouli O, Gruner N, Marcus U, Cai W, Zhang W et al. Detection of infections with hepatitis B virus, hepatitis C virus, and human immunodeficiency virus by analyses of dried blood spots – performance characteristics of the ARCHITECT system and two commercial assays for nucleic acid amplification. Virol J. 2013;10:72. 294. Greenman J, Roberts T, Cohn J, Messac L. Dried blood spot in the genotyping, quantification and storage of HCV RNA: a systematic literature review. J Viral Hepat. 2015;22(4):353–61. 295. Jones L, Bates G, McCoy E, Beynon C, McVeigh J, Bellis MA. Effectiveness of interventions to increase hepatitis C testing uptake among high-risk groups: a systematic review. Eur J Public Health. 2014;24(5):781–8. 296. Boa-Sorte N, Purificacao A, Amorim T, Assuncao L, Reis A, Galvao-Castro B. Dried blood spot testing for the antenatal screening of HTLV, HIV, syphilis, toxoplasmosis and hepatitis B and C: prevalence, accuracy and operational aspects. Braz J Infect Dis. 2014;18(6):618–24. 297. Forbi JC, Obagu JO, Gyar SD, Pam CR, Pennap GR, Agwale SM. Application of dried blood spot in the sero-diagnosis of hepatitis B infection (HBV) in an HBV hyper-endemic nation. Ann Afr Med. 2010;9(1):44–5. 298. Kania D, Bekale AM, Nagot N, Mondain AM, Ottomani L, Meda N et al. Combining rapid diagnostic tests and dried blood spot assays for point-of-care testing of human immunodeficiency virus, hepatitis B and hepatitis C infections in Burkina Faso, West Africa. Clin Microbiol Infect. 2013;19(12):E533–41. 299. Lee CE, Sri Ponnampalavanar S, Syed Omar SF, Mahadeva S, Ong LY, Kamarulzaman A. Evaluation of the dried blood spot (DBS) collection method as a tool for detection of HIV Ag/Ab, HBsAg, anti-HBs and anti-HCV in a Malaysian tertiary referral hospital. Ann Acad Med Singapore. 2011;40(10):448–53. 300. Mendy M, Kirk GD, van der Sande M, Jeng-Barry A, Lesi OA, Hainaut P et al. Hepatitis B surface antigenaemia and alpha-foetoprotein detection from dried blood spots: applications to field-based studies and to clinical care in hepatitis B virus endemic areas. J Viral Hepat. 2005;12(6):642–7. 301. Mohamed S, Raimondo A, Penaranda G, Camus C, Ouzan D, Ravet S et al. Dried blood spot sampling for hepatitis B virus serology and molecular testing. PLoS One. 2013;8(4):e61077. 302. Villa E, Cartolari R, Bellentani S, Rivasi P, Casolo G, Manenti F. Hepatitis B virus markers on dried blood spots. A new tool for epidemiological research. J Clin Pathol. 1981;34(7):809–12. 303. Villar LM, de Oliveira JC, Cruz HM, Yoshida CF, Lampe E, Lewis-Ximenez LL. Assessment of dried blood spot samples as a simple method for detection of hepatitis B virus markers. J Med Virol. 2011;83(9):1522–9. 304. Brandao CP, Marques BL, Marques VA, Villela-Nogueira CA, Do OK, de Paula MT et al. Simultaneous detection of hepatitis C virus antigen and antibodies in dried blood spots. J Clin Virol. 2013;57(2):98–102. 305. Tuaillon E, Mondain AM, Meroueh F, Ottomani L, Picot MC, Nagot N et al. Dried blood spot for hepatitis C virus serology and molecular testing. Hepatology. 2010;51(3):752–8. 306. O’Brien J, Kruzel K, Wandell M, Vinogradov I, Sheagren J, Frank A. Detection of hepatitis C antibody with at-home collection kits using an innovative laboratory algorithm. Infect Disin Clin Pract. 2001;10(9):474–80. 307. Croom HA, Richards KM, Best SJ, Francis BH, Johnson EI, Dax EM et al. Commercial enzyme immunoassay adapted for the detection of antibodies to hepatitis C virus in dried blood spots. J Clin Virol. 2006;36(1):68–71. 308. Marques BL, Brandao CU, Silva EF, Marques VA, Villela-Nogueira CA, Do OK et al. Dried blood spot samples: optimization of commercial EIAs for hepatitis C antibody detection and stability under different storage conditions. J Med Virol. 2012;84(10):1600–7. 309. Waterboer T, Dondog B, Michael KM, Michel A, Schmitt M, Vaccarella S et al. Dried blood spot samples for seroepidemiology of infections with human papillomaviruses, Helicobacter pylori, hepatitis C virus, and JC virus. Cancer Epidemiol Biomarkers Prev. 2012;21(2):287–93. 310. Nandagopal P, Iqbal HS, Saravanan S, Solomon SS, Mehta S, Selvakumar M et al. Evaluation of dried blood spot as an alternative specimen for the diagnosis of anti-HCV in resource-limited settings. Indian J Med Microbiol. 2014;32(2):208–10. 311. Dokubo EK, Evans J, Winkelman V, Cyrus S, Tobler LH, Asher A et al. Comparison of hepatitis C virus RNA and antibody detection in dried blood spots and plasma specimens. J Clin Virol. 2014;59(4):223–7.

164

312. Tejada-Strop A, Drobeniuc J, Mixson-Hayden T, Forbi JC, Le NT, Li L et al. Disparate detection outcomes for anti-HCV IgG and HCV RNA in dried blood spots. J Virol Methods. 2015;212:66–70. 313. Chevaliez S, Soulier A, Poiteau L, Pawlotsky JM. Dried blood spots (DBS), a promising tool for largescale hepatitis C screening, diagnosis and treatment monitoring. International Liver Congress 2014. 49th Annual Meeting of the European Association for the Study of the Liver (EASL), 9–13 April 2014, London. J Hepatol. 2014;60 (1 Suppl):S325–S326. [Abstract P765] 314. Parker SP, Cubitt WD, Ades AE. A method for the detection and confirmation of antibodies to hepatitis C virus in dried blood spots. J Virol Methods. 1997;68(2):199–205. 315. Lukacs Z, Dietrich A, Ganschow R, Kohlschutter A, Kruithof R. Simultaneous determination of HIV antibodies, hepatitis C antibodies, and hepatitis B antigens in dried blood spots – a feasibility study using a multi-analyte immunoassay. Clin Chem Lab Med. 2005;43(2):141–5. 316. McCarron B, Fox R, Wilson K, Cameron S, McMenamin J, McGregor G et al. Hepatitis C antibody detection in dried blood spots. J Viral Hepat. 1999;6(6):453–6. 317. Shepherd SJ, Kean J, Hutchinson SJ, Cameron SO, Goldberg DJ, Carman WF et al. A hepatitis C avidity test for determining recent and past infections in both plasma and dried blood spots. J Clin Virol. 2013;57(1):29–35. 318. Alidjinou EK, Moukassa D, Sane F, Twagirimana Nyenyeli S, Akoko EC, Mountou MV et al. Detection of hepatitis B virus infection markers in dried plasma spots among patients in Congo-Brazzaville. Diagn Microbiol Infect Dis. 2014;78(3):229–31. 319. Alhusseini N, Abadeer M, El-Taher S. Hepatitis B virus DNA can be amplified directly from dried blood spot on filter paper. Am J Biochem Biotechnol. 2012;8(2):143–9. 320. Durgadevi S, Dhodapkar R, Parija S. Serological and molecular diagnosis of hepatitis B virus. BMC Infect Dis. 2012;12(Suppl 1):P31. 321. Gupta BP, Jayasuryan N, Jameel S. Direct detection of hepatitis B virus from dried blood spots by polymerase chain reaction amplification. J Clin Microbiol. 1992;30(8):1913–6. 322. Halfon P, Raimondo A, Ouzan D, Bourlière M, Khiri H, Cohen-Bacrie S et al. Dried blood spot for hepatitis B virus serology and molecular testing. International Liver Congress 2012. 47th Annual Meeting of the European Association for the Study of the Liver (EASL), 18–22 April 2012. Barcelona. J Hepatol. 2012;56 (Suppl 2):S62. [Abstract 142] 323. Jardi R, Rodriguez-Frias F, Buti M, Schaper M, Valdes A, Martinez M et al. Usefulness of dried blood samples for quantification and molecular characterization of HBV-DNA. Hepatology. 2004;40(1):133–9. 324. Lira R, Maldonado-Rodriguez A, Rojas-Montes O, Ruiz-Tachiquin M, Torres-Ibarra R, Cano-Dominguez C et al. Use of dried blood samples for monitoring hepatitis B virus infection. Virol J. 2009;6:153. 325. Vinikoor MJ, Zurcher S, Musukuma K, Kachuwaire O, Rauch A, Chi BH et al. Hepatitis B viral load in dried blood spots: a validation study in Zambia. J Clin Virol. 2015;72:20–4. 326. Abe K, Konomi N. Hepatitis C virus RNA in dried serum spotted onto filter paper is stable at room temperature. J Clin Microbiol. 1998;36(10):3070–2. 327. Bennett S, Gunson RN, McAllister GE, Hutchinson SJ, Goldberg DJ, Cameron SO et al. Detection of hepatitis C virus RNA in dried blood spots. J Clin Virol. 2012;54(2):106–9. 328. De Crignis E, Re MC, Cimatti L, Zecchi L, Gibellini D. HIV-1 and HCV detection in dried blood spots by SYBR Green multiplex real-time RT-PCR. J Virol Methods. 2010;165(1):51–6. 329. Soulier A, Poiteau L, Rosa I, Hezode C, Roudot-Thoraval F, Pawlotsky JM et al. Dried blood spots: a tool to ensure broad access to hepatitis C screening, diagnosis, and treatment monitoring. J Infect Dis. 2016;213(7):1087–95. 330. Santos C, Reis A, Dos Santos CV, Damas C, Silva MH, Viana MV et al. The use of real-time PCR to detect hepatitis C virus RNA in dried blood spots from Brazilian patients infected chronically. J Virol Methods. 2012;179(1):17–20. 331. Solmone M, Girardi E, Costa F, Pucillo L, Ippolito G, Capobianchi MR. Simple and reliable method for detection and genotyping of hepatitis C virus RNA in dried blood spots stored at room temperature. J Clin Microbiol. 2002;40(9):3512–4. 332. Solmone M, Girardi E, Costa F, Ippolito G, Capobianchi MR. Simple and reliable method for HCV-RNA detection/genotyping in dried blood spots. J Hepatol. 2002;36(Suppl 1):131. 333. Craine N, Whitaker R, Perrett S, Zou L, Hickman M, Lyons M. A stepped wedge cluster randomized control trial of dried blood spot testing to improve the uptake of hepatitis C antibody testing within UK prisons. Eur J Public Health. 2015;25(2):351–7. 334. Hickman M, McDonald T, Judd A, Nichols T, Hope V, Skidmore S et al. Increasing the uptake of hepatitis C virus testing among injecting drug users in specialist drug treatment and prison settings by using dried blood spots for diagnostic testing: a cluster randomized controlled trial. J Viral Hepat. 2008;15(4):250–4. 335. Hutchinson SJ, Dillon JF, Fox R, McDonald SA, Innes HA, Weir A et al. Expansion of HCV treatment access to people who have injected drugs through effective translation of research into public health policy: Scotland’s experience. Int J Drug Policy. 2015;26(11):1041–9. 336. Hepatitis B and C: ways to promote and offer testing to people at increased risk of infection. London: National Institute for Health and Clinical Excellence (NICE); 2012 (https:// https://www.nice.org.uk/ guidance/ph43/resources/hepatitis-b-and-c-ways-to-promote-and-offer-testing-draft-guidance2, accessed 6 February 2017).

165

337. McLeod A, Weir A, Aitken C, Gunson R, Templeton K, Molyneaux P et al. Rise in testing and diagnosis associated with Scotland’s Action Plan on Hepatitis C and introduction of dried blood spot testing. J Epidemiol Community Health. 2014;68(12):1182–8. 338. Place des tests rapides d’orientation diagnostique (TROD) dans la stratégie de dépistage de l’hépatite C. Saint-Denis La Plaine Cedex: Haute Autorité Sanitaire; 2014 (http://www.has-sante.fr/portail/upload/ docs/application/pdf/2014-05/place_des_trod_dans_la_strategie_de_depistage_de_vhc-_rapport.pdf, accessed 6 February 2017). 339. Bravo MJ, Vallejo F, Barrio G, Brugal MT, Molist G, Pulido J et al. HCV seroconversion among neverinjecting heroin users at baseline: no predictors identified other than starting injection. Int J Drug Policy. 2012;23(5):415–9. 340. Allen EJ, Palmateer NE, Hutchinson SJ, Cameron S, Goldberg DJ, Taylor A. Association between harm reduction intervention uptake and recent hepatitis C infection among people who inject drugs attending sites that provide sterile injecting equipment in Scotland. Int J Drug Policy. 2012;23(5):346–52. 341. Allard NL, MacLachlan JH, Cowie BC. The cascade of care for Australians living with chronic hepatitis B: measuring access to diagnosis, management and treatment. Aust N Z J Public Health. 2015;39(3):255–9. 342. Yehia BR, Schranz AJ, Umscheid CA, Lo Re V 3rd. The treatment cascade for chronic hepatitis C virus infection in the United States: a systematic review and meta-analysis. PLoS One. 2014;9(7):e101554. 343. Govindasamy D, Ford N, Kranzer K. Risk factors, barriers and facilitators for linkage to antiretroviral therapy care: a systematic review. AIDS. 2012;26(16):2059–67. 344. Willenbring ML. Integrating care for patients with infectious, psychiatric, and substance use disorders: concepts and approaches. AIDS. 2005;19 Suppl 3:S227–37. 345. Ahmed I, Habibi A, Iqbal J, Niaz Z, Naqvi A. Improving outcome in hepatitis C management: a need for dedicated multi-disciplinary service to improve compliance with treatment. J Gastroenterol Hepatol Res. 2013;2(8):737–9. 346. Arora S, Thornton K, Murata G, Deming P, Kalishman S, Dion D et al. Outcomes of treatment for hepatitis C virus infection by primary care providers. N Engl J Med. 2011;364(23):2199–207. 347. Asthana A, Choong J, Lubel J. Education does not improve hepatitis B screening uptake in those receiving cytotoxic chemotherapy – time for alternative strategies. J Gastroenterol Hepatol. 2012;27:162. 348. Bastani R, Glenn BA, Maxwell AE, Jo AM, Herrmann AK, Crespi CM et al. Cluster-randomized trial to increase hepatitis B testing among Koreans in Los Angeles. Cancer Epidemiol Biomarkers Prev. 2015;24(9):1341–9. 349. Bonkovsky HL, Tice AD, Yapp RG, Bodenheimer HC Jr, Monto A, Rossi SJ et al. Efficacy and safety of peginterferon alfa-2a/ribavirin in methadone maintenance patients: randomized comparison of direct observed therapy and self-administration. Am J Gastroenterol. 2008;103(11):2757–65. 350. Bruce RD, Eiserman J, Acosta A, Gote C, Lim JK, Altice FL. Developing a modified directly observed therapy intervention for hepatitis C treatment in a methadone maintenance program: implications for program replication. Am J Drug Alcohol Abuse. 2012;38(3):206–12. 351. Cacoub P, Ouzan D, Melin P, Lang JP, Rotily M, Fontanges T et al. Patient education improves adherence to peg-interferon and ribavirin in chronic genotype 2 or 3 hepatitis C virus infection: a prospective, real-life, observational study. World J Gastroenterol. 2008;14(40):6195–203. 352. Carrion JA, Gonzalez-Colominas E, Garcia-Retortillo M, Canete N, Cirera I, Coll S et al. A multidisciplinary support programme increases the efficiency of pegylated interferon alfa-2a and ribavirin in hepatitis C. J Hepatol. 2013;59(5):926–33. 353. Chakrabarty G, Rice P, Forton D. Randomized controlled trial of home-based self-administered dried blood spot testing versus written advice for community screening of hepatitis B contacts. Hepatology. 2013;58:616A. 354. Chen MS Jr, Fang DM, Stewart SL, Ly MY, Lee S, Dang JH et al. Increasing hepatitis B screening for hmong adults: results from a randomized controlled community-based study. Cancer Epidemiol Biomarkers Prev. 2013;22(5):782–91. 355. Chen JY, Feeney ER, Chung RT. HCV and HIV co-infection: mechanisms and management. Nat Rev Gastroenterol Hepatol. 2014;11(6):362–71. 356. Cioe PA, Stein MD, Promrat K, Friedmann PD. A comparison of modified directly observed therapy to standard care for chronic hepatitis C. J Community Health. 2013;38(4):679–84. 357. Craine N, Parry J, O’Toole J, D’Arcy S, Lyons M. Improving blood-borne viral diagnosis; clinical audit of the uptake of dried blood spot testing offered by a substance misuse service. J Viral Hepat. 2009;16(3):219–22. 358. Cullen W, Stanley J, Langton D, Kelly Y, Staines A, Bury G. Hepatitis C infection among injecting drug users in general practice: a cluster randomised controlled trial of clinical guidelines’ implementation. Br J Gen Pract. 2006;56(532):848–56. 359. Curcio F, Di Martino F, Capraro C, Angelucci F, Bulla F, Caprio N et al. Together ... to take care: multidisciplinary management of hepatitis C virus treatment in randomly selected drug users with chronic hepatitis. J Addict Med. 2010;4(4):223–32. 360. Drainoni ML, Litwin AH, Smith BD, Koppelman EA, McKee MD, Christiansen CL et al. Effectiveness of a risk screener in identifying hepatitis C virus in a primary care setting. Am J Public Health. 2012;102(11):e115–21.

166

361. Evon DM, Simpson K, Kixmiller S, Galanko J, Dougherty K, Golin C et al. A randomized controlled trial of an integrated care intervention to increase eligibility for chronic hepatitis C treatment. Am J Gastroenterol. 2011;106(10):1777–86. 362. Hagedorn H, Dieperink E, Dingmann D, Durfee J, Ho SB, Isenhart C et al. Integrating hepatitis prevention services into a substance use disorder clinic. J Subst Abuse Treat. 2007;32(4):391–8. 363. Helsper CW, van Essen GA, Bonten MJ, de Wit NJ. A support programme for primary care leads to substantial improvements in the effectiveness of a public hepatitis C campaign. Fam Pract. 2010;27(3):328–32. 364. Hirsch AA, Lawrence RH, Kern E, Falck-Ytter Y, Shumaker DT, Watts B. Implementation and evaluation of a multicomponent quality improvement intervention to improve efficiency of hepatitis C screening and diagnosis. Jt Comm J Qual Patient Saf. 2014;40(8):351–7. 365. Ho SB, Brau N, Cheung R, Liu L, Sanchez C, Sklar M et al. Integrated care increases treatment and improves outcomes of patients with chronic hepatitis C virus infection and psychiatric illness or substance abuse. Clin Gastroenterol Hepatol. 2015;13(11):2005–14.e3 366. Hsu L, Bowlus CL, Stewart SL, Nguyen TT, Dang J, Chan B et al. Electronic messages increase hepatitis B screening in at-risk Asian American patients: a randomized, controlled trial. Dig Dis Sci. 2013;58(3):807–14. 367. Hussein M, Benner JS, Lee D, Sesti AM, Battleman DS, Brock-Wood C. Propensity score matching in the evaluation of drug therapy management programs: an illustrative analysis of a program for patients with hepatitis C virus. Qual Manag Health Care. 2010;19(1):25–33. 368. Juon HS, Lee S, Strong C, Rimal R, Kirk GD, Bowie J. Effect of a liver cancer education program on hepatitis B screening among Asian Americans in the Baltimore-Washington metropolitan area, 2009– 2010. Prev Chronic Dis. 2014;11:130258. 369. Knott A, Dieperink E, Willenbring ML, Heit S, Durfee JM, Wingert M et al. Integrated psychiatric/ medical care in a chronic hepatitis C clinic: effect on antiviral treatment evaluation and outcomes. Am J Gastroenterol. 2006;101(10):2254–62. 370. Koruk I, Koruk S, Çopur AC, Simsek Z. A intervention study to improve HBsAg testing and preventive practices for hepatitis B in an obstetrics hospital. TAF Prev Med Bull. 2011;10(3):287–92. 371. Krauskopf K, Kil N, Sofianou A, Toribio W, Lyons J, Singer M et al. Evaluation of an electronic health record prompt for hepatitis c antibody screening of baby boomers in primary care – a cluster randomized control trial. J Gen Intern Med. 2014;29:S88–S9. 372. Larrey D, Salse A, Ribard D, Boutet O, Hyrailles-Blanc V, Niang B et al. Education by a nurse increases response of patients with chronic hepatitis C to therapy with peginterferon-alpha2a and ribavirin. Clin Gastroenterol Hepatol. 2011;9(9):781–5. 373. Lee R, Vu K, Bell CM, Hicks LK. Screening for hepatitis B surface antigen before chemotherapy: current practice and opportunities for improvement. Curr Oncol. 2010;17(6):32–8. 374. Litwin AH, Smith BD, Drainoni ML, McKee D, Gifford AL, Koppelman E et al. Primary care-based interventions are associated with increases in hepatitis C virus testing for patients at risk. Dig Liver Dis. 2012;44(6):497–503. 375. Lubega S, Agbim U, Surjadi M, Mahoney M, Khalili M. Formal hepatitis C education enhances HCV care coordination, expedites HCV treatment and improves antiviral response. Liver Int. 2013;33(7):999–1007. 376. Ma GX, Gao W, Tan Y, Chae WG, Rhee J. A community-based participatory approach to a hepatitis B intervention for Korean Americans. Prog Community Health Partnersh. 2012;6(1):7–16. 377. Masson CL, Delucchi KL, McKnight C, Hettema J, Khalili M, Min A et al. A randomized trial of a hepatitis care coordination model in methadone maintenance treatment. Am J Public Health. 2013;103(10):e81–e8. 378. Matthews H, McLeod M, Oakes K, McCurdy G, Zuckerman M, Carey I et al. Perinatal hepatitis B in a high prevalence inner city population: direct electronic referral improves care. Gut. 2012;61:A79–A80. 379. Merchant RC, Baird JR, Liu T, Taylor LE, Montague BT, Nirenberg TD. Brief intervention to increase emergency department uptake of combined rapid human immunodeficiency virus and hepatitis C screening among a drug misusing population. Acad Emerg Med. 2014;21(7):752–67. 380. Mostert MC, Richardus JH, de Man RA. Referral of chronic hepatitis B patients from primary to specialist care: making a simple guideline work. J Hepatol. 2004;41(6):1026–30. 381. Neri S, Bertino G, Petralia A, Giancarlo C, Rizzotto A, Calvagno GS et al. A multidisciplinary therapeutic approach for reducing the risk of psychiatric side effects in patients with chronic hepatitis C treated with pegylated interferon alpha and ribavirin. J Clin Gastroenterol. 2010;44(9):e210–7. 382. Ramsey SE, Engler PA, Stein MD, Brown RA, Cioe P, Kahler CW et al. Effect of CBT on depressive symptoms in methadone maintenance patients undergoing treatment for hepatitis C. J Addict Res Ther. 2011;2(2):2–10. 383. Reimer J, Schmidt CS, Schulte B, Gansefort D, Golz J, Gerken G et al. Psychoeducation improves hepatitis C virus treatment during opioid substitution therapy: a controlled, prospective multicenter trial. Clin Infect Dis. 2013;57 (Suppl 2):S97–104. 384. Rifai MA, Moles JK, Lehman LP, Van der Linden BJ. Hepatitis C screening and treatment outcomes in patients with substance use/dependence disorders. Psychosomatics. 2006;47(2):112–21. 385. Rosenberg SD, Goldberg RW, Dixon LB, Wolford GL, Slade EP, Himelhoch S et al. Assessing the STIRR model of best practices for blood-borne infections of clients with severe mental illness. Psychiatr Serv. 2010;61(9):885–91.

167

386. Sahajian F, Excler G, Bailly F, Caillat-Vallet E, Trepo C, Sepetjan M et al. Hepatitis C screening practices among private practitioners: impact of an information campaign. Gastroenterol Clin Biol. 2004;28(8–9):714–9. 387. Tait JM, McIntyre PG, McLeod S, Nathwani D, Dillon JF. The impact of a managed care network on attendance, follow-up and treatment at a hepatitis C specialist centre. J Viral Hepat. 2010;17(10):698–704. 388. Taylor VM, Bastani R, Burke N, Talbot J, Sos C, Liu Q et al. Evaluation of a hepatitis B lay health worker intervention for Cambodian Americans. J Community Health. 2013;38(3):546–53. 389. Taylor VM, Gregory Hislop T, Bajdik C, Teh C, Lam W, Acorda E et al. Hepatitis B ESL education for Asian immigrants. J Community Health. 2011;36(1):35–41. 390. Taylor VM, Hislop TG, Tu SP, Teh C, Acorda E, Yip MP et al. Evaluation of a hepatitis B lay health worker intervention for Chinese Americans and Canadians. J Community Health. 2009;34(3):165–72. 391. van der Veen YJ, van Empelen P, de Zwart O, Visser H, Mackenbach JP, Richardus JH. Cultural tailoring to promote hepatitis B screening in Turkish Dutch: a randomized control study. Health Promot Int. 2014;29(4):692–704. 392. Le Lan C, Guillygomarc’h A, Danielou H, Le Dreau G, Laine F, Vedeilhie C et al. A multi-disciplinary approach to treating hepatitis C with interferon and ribavirin in alcohol-dependent patients with ongoing abuse. J Hepatol. 2012;56(2):334–40. 393. Impact of Physician Directed Education on Patient Compliance With Hepatitis C Therapy (OPTIMAL). Chronic Liver Disease Foundation; 2014. (https://clinicaltrials.gov/ct2/show/NCT01405027, accessed 6 February 2017). 394. Compliance of HCV genotype 1 infected patients receiving pegintron/rebetol and a patient assistance program (Study P04671). Merck Sharp & Dohme Corp.; 2007. 395. Adherence in patients receiving pegintron pen/rebetol for hepatitis C in conjunction with a patient assistance program (Study P04281). Merck Sharp & Dohme Corp.; 2009. 396. Adherence in patients receiving pegintron/rebetol for hepatitis C in conjunction with a psychotherapy support program (Study P04252). Merck Sharp & Dohme Corp.; 2009. 397. Renou C LP, Pariente A. Impact of therapeutic education on the outcome of chronic hepatitis C treatment. Hepatology. 2009;50:729A. 398. Zhou K, Fitzpatrick T, Walsh N, Kim JY, Chou R, Lackey M et al. Interventions to optimise the care continuum for chronic viral hepatitis: a systematic review and meta-analyses. Lancet Infect Dis. 2016;16(12):1409–22. 399. Mwai GW, Mburu G, Torpey K, Frost P, Ford N, Seeley J. Role and outcomes of community health workers in HIV care in sub-Saharan Africa: a systematic review. J Int AIDS Soc. 2013;16:18586. 400. Patel V, Weiss HA, Chowdhary N, Naik S, Pednekar S, Chatterjee S et al. Lay health worker led intervention for depressive and anxiety disorders in India: impact on clinical and disability outcomes over 12 months. Br J Psychiatry. 2011;199(6):459–66. 401. Joshi R, Alim M, Kengne AP, Jan S, Maulik PK, Peiris D et al. Task shifting for non-communicable disease management in low and middle income countries – a systematic review. PLoS One. 2014;9(8):e103754. 402. Kredo T, Adeniyi FB, Bateganya M, Pienaar ED. Task shifting from doctors to non-doctors for initiation and maintenance of antiretroviral therapy. Cochrane Database Syst Rev. 2014;(7):CD007331. 403. Glenton C, Colvin CJ, Carlsen B, Swartz A, Lewin S, Noyes J et al. Barriers and facilitators to the implementation of lay health worker programmes to improve access to maternal and child health: qualitative evidence synthesis. Cochrane Database Syst Rev. 2013;(10):CD010414. 404. Mandelblatt JS, Gold K, O’Malley AS, Taylor K, Cagney K, Hopkins JS et al. Breast and cervix cancer screening among multiethnic women: role of age, health, and source of care. Prev Med. 1999;28(4):418–25. 405. Green BB, Wang CY, Anderson ML, Chubak J, Meenan RT, Vernon SW et al. An automated intervention with stepped increases in support to increase uptake of colorectal cancer screening: a randomized trial. Ann Intern Med. 2013;158(5 Pt 1):301–11. 406. Norman J, Walsh NM, Mugavin J, Stoove MA, Kelsall J, Austin K et al. The acceptability and feasibility of peer worker support role in community based HCV treatment for injecting drug users. Harm Reduct J. 2008;5:8. 407. Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig L et al., for the STARD Group. STARD 2015: an updated list of essential items for reporting diagnostic accuracy studies. Clin Chem. 2015;61(12):1446–52. 408. Guidance for post-market surveillance of in vitro diagnostics. Geneva: World Health Organization; 2015 (http://www.who.int/diagnostics_laboratory/postmarket/150210_pms_ivds_guidance.pdf?ua=1, accessed 02 July 2016, accessed 6 February 2017). 409. Laboratory quality management system: handbook. World Health Organization; 2011 (http://www.who. int/ihr/publications/lqms/en/, accessed 6 February 2017). 410. The Alcohol, Smoking and Substance Involvement Screening Test (ASSIST). Manual for use in primary care. Geneva: World Health Organization; 2011 (http://apps.who.int/iris/ bitstream/10665/44320/1/9789241599382_eng.pdf, accessed 23 January 2017). 411. Guidance on provider-initiated HIV testing and counselling in health facilities. Geneva: World Health Organization; 2007 (http://apps.who.int/iris/bitstream/10665/43688/1/9789241595568_eng.pdf, accessed 6 February 2017).

168

412. Systematic screening for active tuberculosis: principles and recommendations. Geneva: World Health Organization; 2013 (http://www.who.int/tb/tbscreening/en/, accessed 6 February 2017). 413. Corneli A, Jarrett NM, Sabue M, Duvall S, Bahati E, Behets F et al. Patient and provider perspectives on implementation models of HIV counseling and testing for patients with TB. Int J Tuberc Lung Dis. 2008;12(3 Suppl 1):79–84. 414. Guidelines for the management of sexually transmitted infections. Geneva: World Health Organization; 2003 (http://apps.who.int/iris/bitstream/10665/42782/1/9241546263_eng.pdf?ua=1, 6 February 2017). 415. Tucker JD, Bien CH, Peeling RW. Point-of-care testing for sexually transmitted infections: recent advances and implications for disease control. Curr Opin Infect Dis. 2013;26(1):73–9. 416. Dukers-Muijrers NH, Niekamp AM, Vergoossen MM, Hoebe CJ. Effectiveness of an opting-out strategy for HIV testing: evaluation of 4 years of standard HIV testing in a STI clinic. Sex Transm Infect. 2009;85(3):226–30. 417. Bruggmann P, Litwin AH. Models of care for the management of hepatitis C virus among people who inject drugs: one size does not fit all. Clin Infect Dis. 2013;57 (Suppl 2):S56–S61. 418. Islam MM, Topp L, Conigrave KM, White A, Reid SE, Grummett S et al. Linkage into specialist hepatitis C treatment services of injecting drug users attending a needle syringe program-based primary healthcare centre. J Subst Abuse Treat. 2012;43(4):440–5. 419. Kresina TF, Lubran R, Clark HW, McCance-Katz EF. Advancing service integration in opioid treatment progams for the care and treatment of hepatitis C infection. Int J Clin Med. 2014;5(3):118–25. 420. HIV indicator conditions: guidance for implementing HIV testing in adults in health care settings. Copenhagen: HIV in Europe; 2012 (http://hiveurope.eu/Portals/0/Guidance.pdf.pdf, accessed 6 February 2017). 421. d’Almeida KW, Kierzek G, de Truchis P, Le Vu S, Pateron D, Renaud B et al. Modest public health impact of nontargeted human immunodeficiency virus screening in 29 emergency departments. Arch Intern Med. 2012;172(1):12–20. 422. Jones L, Bates G, McCoy E, Beynon C, McVeigh J, Bellis M. A systematic review of the effectiveness and cost-effectiveness of interventions aimed at raising awareness and engaging with groups who are at an increased risk of hepatitis B and C infection – final report. Liverpool: NICE; 2012 (https://www.nice. org.uk/guidance/ph43/evidence/evidence-review-2-69062510, accessed 6 February 2017). 423. Jack K, Willott S, Manners J, Varnam MA, Thomson BJ. Clinical trial: a primary-care-based model for the delivery of anti-viral treatment to injecting drug users infected with hepatitis C. Aliment Pharmacol Ther. 2009;29(1):38–45. 424. Grebely J, Knight E, Genoway KA, Viljoen M, Khara M, Elliott D et al. Optimizing assessment and treatment for hepatitis C virus infection in illicit drug users: a novel model incorporating multidisciplinary care and peer support. Eur J Gastroenterol Hepatol. 2010;22(3):270–7. 425. Crawford S, Bath N. Peer support models for people with a history of injecting drug use undertaking assessment and treatment for hepatitis C virus infection. Clin Infect Dis. 2013;57 (Suppl 2):S75–9. 426. Evidence for action: effectiveness of community-based outreach in preventing HIV/AIDS among injecting drug use. Geneva: World Health Organization; 2004 (http://www.who.int/hiv/pub/prev_care/ en/evidenceforactionalcommunityfinal.pdf, accessed 6 February 2017). 427. Hermez J, Petrak J, Karkouri M, Riedner G. A review of HIV testing and counseling policies and practices in the Eastern Mediterranean Region. AIDS. 2010;24 (Suppl 2):S25–S32. 428. Corbett EL, Dauya E, Matambo R, Cheung YB, Makamure B, Bassett MT et al. Uptake of workplace HIV counselling and testing: a cluster-randomised trial in Zimbabwe. PLoS Med. 2006;3(7):e238. 429. Collier AC, Van der Borght SF, Rinke de Wit T, Richards SC, Feeley FG. A successful workplace program for voluntary counseling and testing and treatment of HIV/AIDS at Heineken, Rwanda. International journal of occupational and environmental health. 2007;13(1):99–106. 430. Charalambous S, Innes C, Muirhead D, Kumaranayake L, Fielding K, Pemba L et al. Evaluation of a workplace HIV treatment programme in South Africa. AIDS. 2007;21 (Suppl 3):S73–8. 431. Counselling and testing children for HIV in South Africa. Lancet. 2013;381(9865):424. 432. Ford N, Swan T, Beyer P, Hirnschall G, Easterbrook P, Wiktor S. Simplification of antiviral hepatitis C virus therapy to support expanded access in resource-limited settings. J Hepatol. 2014;61(1 Suppl):S132–S8. 433. Xia J, Rutherford S, Ma Y, Wu L, Gao S, Chen T et al. Obstacles to the coordination of delivering integrated prenatal HIV, syphilis and hepatitis B testing services in Guangdong: using a needs assessment approach. BMC Health Serv Res. 2015;15:117. 434. Chen CH, Yang PM, Huang GT, Lee HS, Sung JL, Sheu JC. Estimation of seroprevalence of hepatitis B virus and hepatitis C virus in Taiwan from a large-scale survey of free hepatitis screening participants. J Formos Med Assoc. 2007;106(2):148–55. 435. Community health workers: what do we know about them? The state of the evidence on programmes, activities, costs and impact on health outcomes of using community health workers. Geneva: World Health Organization; 2007 (http://www.who.int/hrh/documents/community_health_workers.pdf, accessed 6 February 2017). 436. Lewin SA, Dick J, Pond P, Zwarenstein M, Aja G, van Wyk B et al. Lay health workers in primary and community health care. Cochrane Database Syst Rev. 2005;(1):CD004015.

169

437. Laurant M, Reeves D, Hermens R, Braspenning J, Grol R, Sibbald B. Substitution of doctors by nurses in primary care. Cochrane Database Syst Rev. 2005;(2):CD001271. 438. Callaghan M, Ford N, Schneider H. A systematic review of task-shifting for HIV treatment and care in Africa. Hum Resour Health. 2010;8:8. 439. Task shifting: global recommendations and guidelines. Geneva: World Health Organization; 2007 (http://www.who.int/healthsystems/task_shifting/en/, accessed 6 February 2017). 440. Optimizing health worker roles to improve access to key maternal and newborn health interventions through task shifting. Geneva: World Health Organization; 2012 (http://apps.who.int/iris/ bitstream/10665/77764/1/9789241504843_eng.pdf?ua=1, accessed 6 February 2017). 441. Walensky RP, Reichmann WM, Arbelaez C, Wright E, Katz JN, Seage GR 3rd et al. Counselor- versus provider-based HIV screening in the emergency department: results from the universal screening for HIV infection in the emergency room (USHER) randomized controlled trial. Ann Emerg Med. 2011;58(1 Suppl 1):S126–S132. e4. 442. Champenois K, Le Gall JM, Jacquemin C, Jean S, Martin C, Rios L et al. ANRS-COM’TEST: description of a community-based HIV testing intervention in non-medical settings for men who have sex with men. BMJ Open. 2012;2(2):e000693. 443. Lorente N, Preau M, Vernay-Vaisse C, Mora M, Blanche J, Otis J et al. Expanding access to nonmedicalized community-based rapid testing to men who have sex with men: an urgent HIV prevention intervention (the ANRS-DRAG study). PLoS One. 2013;8(4):e61225. 444. Fylkesnes K, Sandoy IF, Jurgensen M, Chipimo PJ, Mwangala S, Michelo C. Strong effects of homebased voluntary HIV counselling and testing on acceptance and equity: a cluster randomised trial in Zambia. Soc Sci Med. 2013;86:9–16. 445. Molesworth AM, Ndhlovu R, Banda E, Saul J, Ngwira B, Glynn JR et al. High accuracy of home-based community rapid HIV testing in rural Malawi. J Acquir Immune Defic Syndr. 2010;55(5):625–30. 446. Bemelmans M, van den Akker T, Ford N, Philips M, Zachariah R, Harries A et al. Providing universal access to antiretroviral therapy in Thyolo, Malawi through task shifting and decentralization of HIV/AIDS care. Trop Med Int Health. 2010;15(12):1413–20. 447. Jackson D, Naik R, Tabana H, Pillay M, Madurai S, Zembe W et al. Quality of home-based rapid HIV testing by community lay counsellors in a rural district of South Africa. J Int AIDS Soc. 2013;16:18744. 448. Iwu EN, Holzemer WL. Task shifting of HIV management from doctors to nurses in Africa: clinical outcomes and evidence on nurse self-efficacy and job satisfaction. AIDS Care. 2014;26(1):42–52. 449. Leon N, Naidoo P, Mathews C, Lewin S, Lombard C. The impact of provider-initiated (opt-out) HIV testing and counseling of patients with sexually transmitted infection in Cape Town, South Africa: a controlled trial. Implement Sci. 2010;5:8. 450. Kanal K, Chou TL, Sovann L, Morikawa Y, Mukoyama Y, Kakimoto K. Evaluation of the proficiency of trained non-laboratory health staffs and laboratory technicians using a rapid and simple HIV antibody test. AIDS Res Ther. 2005;2(1):5. 451. Lloyd AR, Clegg J, Lange J, Stevenson A, Post JJ, Lloyd D et al. Safety and effectiveness of a nurse-led outreach program for assessment and treatment of chronic hepatitis C in the custodial setting. Clin Infect Dis. 2013;56(8):1078–84. 452. Hill WD, Butt G, Alvarez M, Krajden M. Capacity enhancement of hepatitis C virus treatment through integrated, community-based care. Can J Gastroenterol. 2008;22(1):27–32. 453. UNITAID. Hepatitis C medicines and diagnostics in the context of HIV/HCV co-infection: a scoping report. Geneva: World Health Organization; 2013 (http://www.unitaid.eu/images/marketdynamics/ publications/Hepatitis-C_October-2013.pdf, accessed 23 January 2017). 454. Pant Pai N, Sharma J, Shivkumar S, Pillay S, Vadnais C, Joseph L et al. Supervised and unsupervised self-testing for HIV in high- and low-risk populations: a systematic review. PLoS Med. 2013;10(4):e1001414. 455. Krause J, Subklew-Sehume F, Kenyon C, Colebunders R. Acceptability of HIV self-testing: a systematic literature review. BMC Public Health. 2013;13:735. 456. Figueroa C, Johnson C, Verster A, Baggaley R. Attitudes and acceptability on HIV aelf-testing among key populations: a literature review. AIDS Behav. 2015;19(11):1949–65. 457. Choko AT, MacPherson P, Webb EL, Willey BA, Feasy H, Sambakunsi R et al. Uptake, accuracy, safety, and linkage into care over two years of promoting annual self-testing for HIV in Blantyre, Malawi: a community-based prospective study. PLoS Med. 2015;12(9):e1001873. 458. HIV/AIDS prevention, care, treatment and support in prison settings: a framework for an effective national response. Vienna: United Nations Office on Drugs and Crime; 2006 (http://www.who.int/hiv/ pub/idu/framework_prisons.pdf?ua=1, accessed 6 February 2017). 459. Policy brief. HIV prevention, treatment and care in prisons and other closed settings: a comprehensive package of interventions. Vienna: United Nations Office on Drugs and Crime; 2013 (https://www. unodc.org/documents/hiv-aids/HIV_comprehensive_package_prison_2013_eBook.pdf, accessed 6 February 2017). 460. Grebely J, Bruggmann P, Treloar C, Byrne J, Rhodes T, Dore GJ et al. Expanding access to prevention, care and treatment for hepatitis C virus infection among people who inject drugs. Int J Drug Policy. 2015;26(10):893–8.

170

461. Integrating collaborative TB and HIV services within a comprehensive package of care for people who inject drugs: consolidated guidelines. Geneva: World Health Organization; 2016 (http://apps.who.int/ iris/bitstream/10665/204484/1/9789241510226_eng.pdf?ua=1, accessed 6 February 2017). 462. Benhamou Y, Bochet M, Di Martino V, Charlotte F, Azria F, Coutellier A et al. Liver fibrosis progression in human immunodeficiency virus and hepatitis C virus coinfected patients. The Multivirc Group. Hepatology. 1999;30(4):1054–8. 463. Di Martino V, Rufat P, Boyer N, Renard P, Degos F, Martinot-Peignoux M et al. The influence of human immunodeficiency virus coinfection on chronic hepatitis C in injection drug users: a long-term retrospective cohort study. Hepatology. 2001;34(6):1193–9. 464. Graham CS, Baden LR, Yu E, Mrus JM, Carnie J, Heeren T et al. Influence of human immunodeficiency virus infection on the course of hepatitis C virus infection: a meta-analysis. Clin Infect Dis. 2001;33(4):562–9. 465. Lo Re V 3rd, Kallan MJ, Tate JP, Localio AR, Lim JK, Goetz MB et al. Hepatic decompensation in antiretroviral-treated patients co-infected with HIV and hepatitis C virus compared with hepatitis C virus-monoinfected patients: a cohort study. Ann Intern Med. 2014;160(6):369–79. 466. Ni JD, Xiong YZ, Wang XJ, Xiu LC. Does increased hepatitis B vaccination dose lead to a better immune response in HIV-infected patients than standard dose vaccination: a meta-analysis? Int J STD AIDS. 2013;24(2):117–22. 467. Getahun H, Gunneberg C, Sculier D, Verster A, Raviglione M. Tuberculosis and HIV in people who inject drugs: evidence for action for tuberculosis, HIV, prison and harm reduction services. Curr Opin HIV AIDS. 2012;7(4):345–53. 468. Getahun H, Baddeley A, Raviglione M. Managing tuberculosis in people who use and inject illicit drugs. Bull World Health Organ. 2013;91(2):154–6. 469. Padmapriyadarsini C, Chandrabose J, Victor L, Hanna LE, Arunkumar N, Swaminathan S. Hepatitis B or hepatitis C co-infection in individuals infected with human immunodeficiency virus and effect of anti-tuberculosis drugs on liver function. J Postgrad Med. 2006;52(2):92–6. 470. Liu R, Li Y, Wangen KR, Maitland E, Nicholas S, Wang J. Analysis of hepatitis B vaccination behavior and vaccination willingness among migrant workers from rural China based on protection motivation theory. Hum Vaccin Immunother. 2016;12(5):1155–63. 471. Guidance on couples HIV testing and counselling including antiretroviral therapy for treatment and prevention in serodiscordant couples: recommendations for a public health approach. Geneva: World Health Organization; 2012 (http://apps.who.int/iris/bitstream/10665/44646/1/9789241501972_eng. pdf?ua=1, accessed 6 February 2017). 472. Patton H, Tran TT. Management of hepatitis B during pregnancy. Nat Rev Gastroenterol Hepatol. 2014;11(7):402–9. 473. Celen MK, Mert D, Ay M, Dal T, Kaya S, Yildirim N et al. Efficacy and safety of tenofovir disoproxil fumarate in pregnancy for the prevention of vertical transmission of HBV infection. World J Gastroenterol. 2013;19(48):9377–82. 474. HIV and adolescents: guidance for HIV testing and counseling and care for adolescents living with HIV. Recommendations for a public health approach and consideration for policymakers and managers. Geneva: World Health Organization; 2013 (http://apps.who.int/iris/ bitstream/10665/94334/1/9789241506168_eng.pdf?ua=1, accessed 6 February 2017).

Global Hepatitis Programme Department of HIV/AIDS 20, avenue Appia 1211 Geneva 27 Switzerland Email: hepatitis@who.int www.who.int/hepatitis/ 978 92 4 154998 1

9 789241 549981

GUIDELINES ON HEPATITIS B AND C TESTING February 2017 GUIDELINES

ANNEXES

Page | i

WHO guidelines on hepatitis B and C testing ISBN 978-92-4-154998-1 © World Health Organization 2017 Some rights reserved. This work is available under the Creative Commons AttributionNonCommercial-ShareAlike 3.0 IGO licence (CC BY-NC-SA 3.0 IGO; https://creativecommons. org/licenses/by-nc-sa/3.0/igo). Under the terms of this licence, you may copy, redistribute and adapt the work for noncommercial purposes, provided the work is appropriately cited, as indicated below. In any use of this work, there should be no suggestion that WHO endorses any specific organization, products or services. The use of the WHO logo is not permitted. If you adapt the work, then you must license your work under the same or equivalent Creative Commons licence. If you create a translation of this work, you should add the following disclaimer along with the suggested citation: “This translation was not created by the World Health Organization (WHO). WHO is not responsible for the content or accuracy of this translation. The original English edition shall be the binding and authentic edition”. Any mediation relating to disputes arising under the licence shall be conducted in accordance with the mediation rules of the World Intellectual Property Organization. Suggested citation. WHO guidelines on hepatitis B and C testing. Geneva: World Health Organization; 2017. Licence: CC BY-NC-SA 3.0 IGO. Cataloguing-in-Publication (CIP) data. CIP data are available at http://apps.who.int/iris. Sales, rights and licensing. To purchase WHO publications, see http://apps.who.int/ bookorders. To submit requests for commercial use and queries on rights and licensing, see http://www.who.int/about/licensing. Third-party materials. If you wish to reuse material from this work that is attributed to a third party, such as tables, figures or images, it is your responsibility to determine whether permission is needed for that reuse and to obtain permission from the copyright holder. The risk of claims resulting from infringement of any third-party-owned component in the work rests solely with the user. General disclaimers. The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of WHO concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted and dashed lines on maps represent approximate border lines for which there may not yet be full agreement. The mention of specific companies or of certain manufacturers ’ products does not imply that they are endorsed or recommended by WHO in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters. All reasonable precautions have been taken by WHO to verify the information contained in this publication. However, the published material is being distributed without warranty of any kind, either expressed or implied. The responsibility for the interpretation and use of the material lies with the reader. In no event shall WHO be liable for damages arising from its use.

Page | ii

CONTENTS Annex 1: The Global Hepatitis Health Sector Strategy – Global targets ……………………………… 1 Annex 2: Guidelines for the prevention, care and treatment of persons with chronic hepatitis B infection – Summary of recommendations ……………………………………………………………………… 3 Annex 3: Guidelines for the screening, care and treatment of persons with chronic hepatitis C infection – Summary of recommendations…………………………………………………………………………..6 Annex 4: PICO questions and decision making tables …………………………………………………………. 9 Annex 4.1: Who to test HBV- What is the impact, cost and cost-effectiveness of different HBV testing approaches and scenarios? ……………………………………………………………………….. 10 Annex 4.2: Who to test HCV- What is the impact, cost and cost-effectiveness of different HCV testing approached and scenarios? ………………………………………………………………………. 28 Annex 4.3: How to test HBV - PICO 1- HBsAg testing: Among persons identified for hepatitis B testing, what is the diagnostic accuracy of available assays for detecting HBsAg (RDT, EIA)? ......................................................................................................................... 52 Annex 4.4: How to test HCV - PICO 2- To ascertain exposure to HCV through anti-HCV testing: Among individuals identified for hepatitis C testing, what is the diagnostic accuracy of available assays for detecting anti-HCV (RDT, EIA)? ……………………………………………….66 Annex 4.5: How to test HBV - PICO 3- Testing strategy to diagnose chronic HBV infection through detection of HBsAg: Among individuals identified for hepatitis B testing, what is the best strategy (diagnostic accuracy and other outcomes) for detection of HBsAg? (One assay versus two-assay strategy) …………………………………………………………………. 76 Annex 4.6: How to test HCV - PICO 4- Testing strategy to ascertain exposure to HCV: Among persons identified for hepatitis C testing, what is the best testing strategy (diagnostic accuracy and other outcomes) for detection of HCV? (One assay versus two-assay strategy) ……………………………………………………………………………………………………………….. 86 Annex 4.7: How to test – HCV - PICO 5a and 6- Testing strategy for diagnosis of HCV active infection: Among individuals with confirmed exposure to HCV (HCV Ab positive), what is the best testing strategy (diagnostic accuracy and other outcomes); comparing HCV core antigen versus NAT for HCV RNA to diagnose active HCV infection? PICO 6Testing strategy for diagnosis of HCV active infection (quantitative or qualitative NAT): Among individuals with confirmed exposure to HCV (HCV Ab positive), what is the diagnostic test accuracy of qualitative NAT methods versus quantitative NAT methods to diagnose active HCV infection? ……………………………………………………………………….. 98 Annex 4.8: Dried blood spots - PICO 7- Dried blood spots as sample collection method for serology/NAT for HBV/HCV: Among persons identified for (1) hepatitis B, or (2)

Page | iii

hepatitis C testing, what is the diagnostic accuracy and impact of detecting HBsAg/HCV Ab or NAT from DBS samples versus venous samples? ……………………. 117 Annex 4.9: Treatment monitoring HCV - PICO 9- Monitoring for treatment response using HCV Ag testing in individuals with confirmed active HCV infection: Among individuals receiving antiviral treatment for HCV, what is the diagnostic accuracy of HCV core antigen versus NAT for HCV RNA qualitative detection (and/or) quantification to confirm successful treatment response with viral clearance? ……………………………. 129 Annex 4.10: Interventions to linkage to care - to optimize uptake of hepatitis testing and linkage to care across the viral hepatitis treatment cascade …………………………….. 138 Annex 5: Systematic reviews and evidence summaries …………………………………………………. 156 Annex 5.1: SR who to test HBV: Literature review on cost-effectiveness of HBV screening, treatment strategies and applicability to LMICs…………………………………………………………….. 157 Annex 5.2: SR who to test HCV: Literature review on cost-effectiveness of HCV screening, treatment strategies and applicability to LMICs…………………………………………………………….. 174 Annex 5.3: SR1 How to test HBV: Diagnostic accuracy of tests to detect hepatitis B surface antigen: a meta-analysis and review of the literature……………………………………………………. 188 Annex 5.4: SR2 How to test HCV: Diagnostic accuracy of tests to detect hepatitis C antibody: a meta-analysis and review of the literature……………………………………………………………………. 266 Annex 5.5: SR3 How to test HBV: Diagnostic strategies for hepatitis B surface antigen detection: a meta-analysis and review of the literature………………………………………………… 304 Annex 5.6: SR4 How to test HCV: Diagnostic strategies for hepatitis C antibody detection: a meta-analysis and review of the literature……………………………………………………………………. 324 Annex 5.7: SR6 How to test: Diagnostic accuracy of HCV RNA tests to detect active HCV infection: a meta-analysis and review of the literature…………………………………………………. 345 Annex 5.8: SR5a and 9 treatment monitoring HCV: HCV core antigen testing for presence of active HCV infection and monitoring for treatment response and cure: a systematic review… ……………………………………………………………………………………………………………………………………… 361 Annex 5.9.1: SR7a: Dried blood spots as a sample collection method for hepatitis B surface antigen serological testing…………………………………………………………………………………………….. 448 Annex 5.9.2: SR7b: Dried blood spots as a sample collection method for HBV DNA……… 471 Annex 5.9.3: SR7c: Dried blood spots as a sample collection method for hepatitis C virological testing: a systematic review and meta-analysis…………………………………………………………….. 495 Annex 5.9.4: SR7d: Dried blood spots as a sample collection method for HCV antibody: a systematic review and meta-analysis……………………………………………………………………………. 518

Page | iv

Annex 5.10: SR 8: Diagnostic accuracy of HBsAg/HBeAg test versus NAT to confirm successful treatment response: a meta-analysis and review of the literature………………………………… 538 Annex 5.11: SR Interventions to linkage to care - to optimize the chronic viral hepatitis care continuum: a systematic review and meta-analysis of interventions to improve hepatitis B and C screening, linkage to care, treatment uptake, treatment adherence, and viral suppression……………………………………………………………………………………………………………………. 553 Annex 6: Predictive modelling analysis ..……………………………………………………………………….. 592 Annex 6.1: Testing strategies for hepatitis B and C infection _Predictive modelling ……… 596 Annex 6.2: Decision analysis of strategies to identify hepatitis C_Predictive modelling … 610 Annex 7: Summary of declared interest ………………………………………………………………………… 630 Annex 8: Guideline Development group, Guideline Steering Group, Systematic review teams, External Review Group ……………..………………………………………………………………………………….. 632

Page | v

Page | vi

ANNEX 1: The Global Hepatitis Health Sector Strategy – Global targets TARGET AREA Impact targets Incidence: New cases of chronic viral hepatitis B and C infections Between 6 and 10 million infections are reduced to 0.9 million infections by 2030 (95% decline in hepatitis B virus infections, 80% decline in hepatitis C virus infections) 1.4 million deaths reduced to less than 500 000 by 2030 (65% for both viral hepatitis B and C ) 30% reduction (equivalent to 1% prevalence of HBsAg1 among children) 90% reduction (equivalent to 0.1% prevalence of HBsAg among children)

BASELINE 2015

2020 TARGETS

2030 TARGETS

Mortality: Viral hepatitis B and C deaths

10% reduction

65% reduction

Service coverage targets Hepatitis B virus vaccination: childhood vaccine coverage (third dose coverage) Prevention of hepatitis B virus mother-to-child transmission: hepatitis B virus birth-dose vaccination coverage or other approach to prevent mother-to-child transmission Blood safety 82%2 in infants 90% 90%

38%

50%

90%

39 countries do not routinely test all blood donations for transfusiontransmissible infections 89% of donations screened in a qualityassured manner3 5%

All countries have haemovigilance systems in place to identify and quantify viral hepatitis transfusion transmission rates

Reduce rates of transmission by 99% compared with 2020.

Safe injections: percentage of injections administered with safetyengineered devices in and out of health facilities Harm reduction: number of sterile needles and syringes provided per person who injects drugs per year Viral hepatitis B and C diagnosis

50%

90%

200

300

<5% of chronic hepatitis infections diagnosed

50%

90%

Page | 1

Viral hepatitis B and C treatment

<1% receiving treatment

5 million people receiving hepatitis B virus treatment 3 million people received hepatitis C virus treatment

80% of eligible persons with chronic hepatitis B virus infection treated 80% of eligible persons with chronic hepatitis C virus infection treated

1The 2

abbreviation “HBsAg” refers to hepatitis B virus surface antigen. WHO/UNICEF coverage estimates 2013 revision, July 2014, see: http://apps.who.int/immunization_monitoring/globalsummary/timeseries/tswucoveragebcg.html (accessed 1 November 2015). 3 Global Database on Blood Safety, Summary Report 2011, see: http://www.who.int/bloodsafety/global_database/GDBS_Summary_Report_2011.pdf?ua=1(accessed 1 November 2015).

Page | 2

ANNEX 2: Guidelines for the prevention, care and treatment of persons with chronic hepatitis B infection – Summary of recommendations WHO recommendations for the prevention, care and treatment of persons with chronic hepatitis B infection Non-invasive assessment of liver disease stage at baseline and during follow up: APRI (aspartate aminotransferase [AST]-to-platelet ratio index) is recommended as the preferred non-invasive test (NIT) to assess for the presence of cirrhosis (APRI score >2 in adults) in resourcelimited settings. Transient elastography (e.g. FibroScan) or FibroTest may be the preferred NITs in settings where they are available and cost is not a major constraint. (Conditional recommendation, low quality of evidence) Who to treat: As a priority, all adults, adolescents and children with CHB and clinical evidence of compensated or decompensated cirrhosis (or cirrhosis based on APRI score >2 in adults) should be treated, regardless of ALT levels, HBeAg status or HBV DNA levels. (Strong recommendation, moderate quality of evidence) Who to treat: Treatment is recommended for adults with CHB who do not have clinical evidence of cirrhosis (or based on APRI score ≤2 in adults), but are aged more than 30 years (in particular), and have persistently abnormal ALT levels and evidence of high- level HBV replication (HBV DNA >20 000 IU/mL), regardless of HBeAg status. (Strong recommendation, moderate quality of evidence) Who to treat: Where HBV DNA testing is not available: Treatment may be considered based on persistently abnormal ALT levels alone, regardless of HBeAg status. (Conditional recommendation, low quality of evidence) HBV/HIV coinfected persons: In HBV/HIV-coinfected individuals, a Tenofovir based ART regimen should be initiated in all HIV-infected persons, regardless of stage of liver disease or CD4 count. (Strong recommendation, low quality of evidence) UPDATED RECOMMENDATION 2015 ARV GUIDELINES Who not to treat but continue to monitor: Antiviral therapy is not recommended and can be deferred in persons without clinical evidence of cirrhosis (or based on APRI score ≤2 in adults), and with persistently normal ALT levels and low levels of HBV DNA replication (HBV DNA <2000 IU/mL), regardless of HBeAg status or age. (Strong recommendation, low quality of evidence) Who not to treat but continue to monitor: Where HBV DNA testing is not available: Treatment can be deferred in HBeAg-positive persons aged

Page | 3

30 years or less and persistently normal ALT levels. (Conditional recommendation, low quality of evidence) Who not to treat but continue to monitor: Continued monitoring is necessary in all persons with CHB, but in particular those who do not currently meet the above-recommended criteria for who to treat or not treat, to determine if antiviral therapy may be indicated in the future to prevent progressive liver disease. These include: persons without cirrhosis aged 30 years or less, with HBV DNA levels >20 000 IU/mL but persistently normal ALT levels; HBeAg-negative persons without cirrhosis aged 30 years or less, with HBV DNA levels that fluctuate between 2000 and 20 000 IU/mL, or who have intermittently abnormal ALT levels; Where HBV DNA testing is not available: Persons without cirrhosis aged 30 years or less, with persistently normal ALT levels, regardless of HBeAg status. Who not to treat but continue to monitor: It is recommended that the following be monitored at least annually: (Strong recommendation, moderate quality of evidence) ALT level (and AST for APRI), HBsAg, HBeAg, and HBV DNA levels (where HBV DNA testing is available) Non-invasive tests (APRI score or FibroScan) to assess for the presence of cirrhosis, in those without cirrhosis at baseline; If on treatment, adherence should be monitored regularly and at each visit. More frequent monitoring: In persons who do not yet meet the criteria for antiviral therapy: More frequent monitoring for disease progression may be indicated in: persons who have intermittently abnormal ALT levels or HBV DNA levels that fluctuate between 2000 IU/mL and 20 000 IU/mL (where HBV DNA testing is available), and in HIV-coinfected persons. (Conditional recommendation, low quality of evidence) More frequent monitoring: In persons on treatment or following treatment discontinuation: More frequent on-treatment monitoring (at least every 3 months for the first year) is indicated in: persons with more advanced disease (compensated or decompensated cirrhosis); during the first year of treatment to assess treatment response and adherence; where treatment adherence is a concern; in HIV-coinfected persons; and in persons after discontinuation of treatment. (Conditional recommendation, very low quality of evidence)

Page | 4

Monitoring for tenofovir and entecavir toxicity: Measurement of baseline renal function and assessment of baseline risk for renal dysfunction should be considered in all persons prior to initiation of antiviral therapy. Monitoring for tenofovir and entecavir toxicity: Renal function should be monitored annually in persons on long-term tenofovir or entecavir therapy, and growth monitored carefully in children. (Conditional recommendation, very low quality of evidence) Monitoring for hepatocellular carcinoma: Routine surveillance for HCC with abdominal ultrasound and alpha-fetoprotein testing every six months is recommended for: persons with cirrhosis, regardless of age or other risk factors (Strong recommendation, low quality of evidence) persons with a family history of HCC (Strong recommendation, low quality of evidence) persons aged over 40 years (lower age may apply according to regional incidence of HCC), without clinical evidence of cirrhosis (or based on APRI score ≤2), and with HBV DNA level >2000 IU/mL (where HBV DNA testing is available). (Conditional recommendation, low quality of evidence) Source: WHO, Guidelines for the screening, care and treatment of persons with chronic hepatitis B infection. 2015

Page | 5

ANNEX 3: Guidelines for the screening, care and treatment of persons with chronic hepatitis C infection – Summary of recommendations Existing and updated WHO recommendations (from 2016 guidelines) Screening to identify persons with HCV infection: It is recommended that HCV serology testing be offered to individuals who are part of a population with high HCV prevalence or who have a history of HCV risk exposure/behaviour. (Strong recommendation, moderate quality of evidence) When to confirm the diagnosis of chronic HCV infection: It is suggested that nucleic acid testing (NAT) for the detection of HCV ribonucleic acid (RNA) be performed directly following a positive HCV serological test to establish the diagnosis of chronic HCV infection, in addition to NAT for HCV RNA as part of the assessment for starting treatment for HCV infection. (Conditional recommendation, very low quality of evidence) Screening for alcohol use and counselling to reduce moderate and high levels of alcohol intake: An alcohol intake assessment is recommended for all persons with HCV infection followed by the offer of a behavioural alcohol reduction intervention for persons with moderate-to-high alcohol intake. (Strong recommendation, moderate quality of evidence) Assessing degree of liver fibrosis and cirrhosis: In resource-limited settings, it is suggested that the aminotransferase/platelet ratio index (APRI) or FIB4 tests be used for the assessment of hepatic fibrosis rather than other noninvasive tests that require more resources such as elastography or Fibrotest. (Conditional recommendation, low quality of evidence) Assessing for HCV treatment: All adults and children with chronic HCV infection, including people who inject drugs, should be assessed for antiviral treatment. (Strong recommendation, moderate quality of evidence) Treatment with direct-acting antiviral agents: it is recommended that DAA regimens be used for the treatment of persons with hepatitis C infection rather than regimens with pegylated interferon and ribavirin. (Strong recommendation, moderate quality of evidence) UPDATED in 2016 Subgroup considerations: for patients with HCV genotype 3 infection with cirrhosis and patients with genotypes 5 and 6 infection with and without cirrhosis, an interferon-based regimen: sofosbuvir/pegylated interferon and ribavirin is still recommended as an alternative treatment option. UPDATED in 2016

Page | 6

Removal of recommendation for treatment with telaprevir or boceprevir: the use of boceprevir- or telaprevir-containing regimens is no longer recommended for the treatment of persons with hepatitis C infection. (Strong recommendation, moderate quality of evidence) UPDATED in 2016 Source: WHO, Guidelines for the screening, care and treatment of persons with chronic hepatitis C infection. Updated version April 2016.

Page | 7

Summary of recommended preferred regimens for hepatitis C with treatment durations* Persons without cirrhosis Daclatasvir/s ofosbuvir Genotype 1 Genotype 2 Genotype 3 Genotype 4 Genotype 5 Genotype 6 12 weeks 12 weeks 12 weeks 12 weeks 12 weeks 12 weeks Ledipasvir/ sofosbuvir 12 weeks a

Persons with cirrhosis Sofosbuvir/ri bavirin Genotype 1 12 weeks 24 weeks Genotype 2 Genotype 3 Genotype 4 Genotype 5 Genotype 6 24 weeks 24 weeks 12 weeks 24 weeks 24 weeks 24 weeks 12 weeks 12 weeks 12 weeks b

Daclatasvir/ sofosbuvir 24 weeks

Daclatasvir/s ofosbuvir/ri bavirin 12 weeks

Ledipasvir/s ofosbuvir 24 weeks

Ledipasvir/s ofosbuvir /ribavirin 12 weeks b

Sofosbuvir/ri bavirin

16 weeks

* Treatment durations are adapted from the 2015 guidelines of the American Association for the Study of Liver Diseases (AASLD) and European Association for the Study of the Liver (EASL). a Treatment may be shortened to 8 weeks in treatment-naive persons without cirrhosis if their baseline HCV RNA level is below 6 million (6.8 log) IU/mL. The duration of treatment should be shortened with caution. b If platelet count <75 x 103/μL, then 24 weeks’ treatment with ribavirin should be given

Page | 8

World Health Organization Global Hepatitis Programme

ANNEX 4. PICO questions and decision-making tables

4.1

Who to test HBV- What is the impact, cost and cost-effectiveness of different HBV testing approaches and scenarios? Who to test HCV- What is the impact, cost and cost-effectiveness of different HCV testing approached and scenarios? How to test HBV - PICO 1- HBsAg testing: Among persons identified for hepatitis B testing, what is the diagnostic accuracy of available assays for detecting HBsAg (RDT, EIA)? How to test HCV - PICO 2- To ascertain exposure to HCV through anti-HCV testing: Among individuals identified for hepatitis C testing, what is the diagnostic accuracy of available assays for detecting anti-HCV (RDT, EIA)? How to test HBV - PICO 3- Testing strategy to diagnose chronic HBV infection through detection of HBsAg: Among individuals identified for hepatitis B testing, what is the best strategy (diagnostic accuracy and other outcomes) for detection of HBsAg? (One assay versus two-assay strategy) How to test HCV - PICO 4- Testing strategy to ascertain exposure to HCV: Among persons identified for hepatitis C testing, what is the best testing strategy (diagnostic accuracy and other outcomes) for detection of HCV? (One assay versus two-assay strategy) How to test – HCV - PICO 5a and 6- Testing strategy for diagnosis of HCV active infection: Among individuals with confirmed exposure to HCV (HCV Ab positive), what is the best testing strategy (diagnostic accuracy and other outcomes); comparing HCV core antigen versus NAT for HCV RNA to diagnose active HCV infection? PICO 6- Testing strategy for diagnosis of HCV active infection (quantitative or qualitative NAT): Among individuals with confirmed exposure to HCV (HCV Ab positive), what is the diagnostic test accuracy of qualitative NAT methods versus quantitative NAT methods to diagnose active HCV infection? Dried blood spots - PICO 7- Dried blood spots as sample collection method for serology/NAT for HBV/HCV: Among persons identified for (1) hepatitis B, or (2) hepatitis C testing, what is the diagnostic accuracy and impact of detecting HBsAg/HCV Ab or NAT from DBS samples versus venous samples? Treatment monitoring HCV - PICO 9- Monitoring for treatment response using HCV Ag testing in individuals with confirmed active HCV infection: Among individuals receiving antiviral treatment for HCV, what is the diagnostic accuracy of HCV core antigen versus NAT for HCV RNA qualitative detection (and/or) quantification to confirm successful treatment response with viral clearance? Interventions to linkage to care - to optimize uptake of hepatitis testing and linkage to care across the viral hepatitis treatment cascade

4.2

4.3

4.4

4.5

4.6

4.7

4.8

4.9

4.10

Page | 9

4.1. WHO to test HBV? Decision-making tables – PICO 1 What is the impact, cost, and cost–effectiveness of different HBV testing approaches and scenarios?

Population: 1. Risk-based screening in different high-risk populations: Injecting drug users (IDUs), men who have sex with men (MSM), immigrants, recipients of blood transfusion and blood products, sex workers, and health-care workers (HCW), HIV-infected persons 2. General population (excluding blood donors) or selected subpopulations of general population (women during pregnancy, those with raised alanine aminotransferase [ALT], Infants, schoolchildren and adolescents) 3. Other approaches: Birth cohort screening (based on different age cut-offs, born between 1945– 1960 or 1965 or 1970). 4. One off screening vs repeat screening every five years Intervention: Testing strategies for HBV in different populations (risk based and general population and birth cohort); and at different prevalence thresholds Comparator: No testing or current practice or comparison of different testing strategies Outcomes: Benefits, harms and costs, and cost–effectiveness with different screening strategies for different target populations Individual patient outcomes: No. of cases detected, overall mortality, liver-related mortality, cirrhosis, end-stage liver disease, rate of hospitalizations, serious adverse events, quality of life Prevention: New infections (mother to child, horizontal [IDUs needle sharing and sexual; and sexual, esp MSM]) Cost–effectiveness: Cost and incremental cost per case diagnosed; cost and incremental cost per case screened and treated; cost and incremental cost per life saved; cost and incremental cost per infections averted; quality-adjusted life years (QALYs) gained

Page | 10

Background: Epidemiology: Chronic hepatitis B (CHB) – defined as persistence of hepatitis B surface antigen (HBsAg) for six months or more – is a major public health problem. Worldwide, there are an estimated 250 million chronically infected persons, particularly in low- and middle-income countries (LMICs). Universal hepatitis B immunization programmes that target infants, with the first dose at birth, have been highly effective in reducing the incidence and prevalence of hepatitis B in many endemic countries. However, these programmes will not have an impact on HBV-related deaths until several decades after their introduction. The major complications of CHB are cirrhosis and hepatocellular carcinoma (HCC). Between 20% and 30% of those who become chronically infected will develop these complications, and an estimated 650 000 people will die annually due to CHB. The risk of developing chronic HBV infection decreases with age at infection, from about 90% when infected perinatally up to 6 months of age to 20–60% between the ages of 6 months and 5 years. Of those who acquire HBV as children 25% will develop primary liver cancer or cirrhosis as adults. Routes of transmission worldwide: In sub-Saharan Africa and east Asia, transmission predominantly occurs in infants and children by th eperinatal and horizontal routes (i.e. resulting from close contact that is not parenteral, perinatal, or sexual in nature) whereas in more industrialized countries, rates of new infection and acute disease are highest among young adults and transmission predominantly occurs via Injecting drug use and other high-risk behaviours. Worldwide, the majority of infections are acquired at birth or in early childhood. Low rates of diagnosis: The majority of people are unaware of their HBV infection, and therefore often present with advanced disease. At present, there is a massive burden of undiagnosed and untreated hepatitis B and C, with 40–85% of infected persons undiagnosed, but varies greatly by setting. By contrast, the estimated awareness of status among people living with HIV (PLHIV) is within 40%–60% range for two thirds of countries, but varies significantly (CHAI, UNAIDS Info).

 Based on still limited studies, overall <15% of the estimated 180 million who are chronically infected with HCV are aware of their diagnosis, based on data from higher-income setting – United States, Europe and China.

 And from a survey in the US, a similar proportion of those with chronic HBV infection are aware of their diagnosis.  The proportion in low-income settings is even higher, with only a tiny fraction diagnosed and aware.

Reasons for low uptake of testing are multifactorial, and include lack of awareness at all levels, lack of clear guidelines, competing health-care priorities, limited health-care budgets and political will. Page | 11

This leads to many people remaining undiagnosed until the later stages of the disease, when prognosis is poor. In addition to the very low access to and uptake of testing, there is also further attrition in the care cascade with very poor linkage to care and therefore treatment, among those who test positive. Hepatitis B and C testing and diagnosis are at the core of entry to both the prevention and treatment cascade.

 Testing is required to identify those with are positive, linking them with care, counselling them on measures to reduce transmission to others then assessing who needs treatment, initiating treatment, achieving treatment response (sustained virological response [SVR] for hepatitis C) or long-term viral suppression for HBV and retaining in care for HBV.

 Hepatitis testing is also needed to identify those who are negative, to provide hepatitis B vaccination, and the opportunity to implement individual or facility-level prevention measures, counsel to reduce risk behaviours, or institute facility-level prevention measures on measures to acquisition. There are three key approaches to screening: 1. Population- or community-based screening (including antenatal). This means that all members of the population have access to the screening programme under consideration. It may also include home-based testing (house to house); campaigns (e.g. HTC plus – malaria, safe water, noncommunicable diseases e.g. diabetes and hypertension); outreach (mobile) in general and key populations; workplaces and schools; and health-care facility0based screening. 2. Health-care facilities. Testing could also be offered in special dedicated clinics, e.g. HIV, STI clinics. Screening at health-care facilities may include primary care settings, inpatient and outpatient settings, and may involve screening on the basis of clinical presentation or focus on only those with abnormal liver function tests, abnormal ultrasound scan, family history of liver disease or other clinical suspicion of liver function test. 3. Targeted risk factor-based screening. This refers to screening of specific groups including key populations, who are generally at higher risk of being infected than the general population. This includes people who inject drugs (PWID), people in prisons and other closed settings, migrant populations, some indigenous populations, MSM and sex workers, but may also include healthcare workers. People attending services providing care and treatment for viral hepatitis or HIV can be encouraged to bring their partners to be tested. 4. Birth cohort screening for HBV and HCV. Existing guidelines: what are countries doing? 1. Most countries have based their list of high-risk groups as defined by the Centers for Disease Control and Prevention (CDC), and are largely based on known modes of transmission. Generally they include recommendations for three main screening approaches: (Apata, MMWR Morb Mortal Wkly Rep. 2014;63:613–19; Weinbaum, Hepatology. 2009;49:S35–S44; Han, Vaccine. Page | 12

2013;31 Suppl 9:J36–J42)

 Population-based screening that includes antenatal clinic screening  Screen those with high-risk behaviours, exposures and other conditions  Family members and household contacts of hepatitis B patients  MSM  PWID  HIV-positive patients  Patients on immunosuppression or chemotherapy  Persons with liver disease of unclear etiology  Health-care workers.  Birth cohort for HCV screening in US and Japan. 2. At present, there is no universally accepted recommended screening programme. There is widespread testing of blood donors (but not necessarily universal), and widespread antenatal screening and infant vaccination in Asia. In addition, there is a risk factor-based testing in highrisk groups in Asia (PWID, liver disease, renal dialysis) and use of a birth cohort approach in the US and Japan. Survey of guidelines (Surjo De) Evidence: systematic reviews of prevalence of HbsAg 1. General population: systematic review (Ott, Lancet 2015) 161 countries included. High endemicity (>5%): Most countries in Africa were of higher–intermediate endemicity (HBsAg prevalence 5–7·99%), or highly endemic for HBV (HBsAg prevalence ≥8%. The Western Pacific Region was also a high–intermediate endemicity region (5–7.99%), especially in the Pacific Island States such as the Solomon Islands. Intermediate endemicity (2–5%): The Eastern Mediterranean Region was of lower–intermediate endemicity (2·00–4·99%), but Djibouti, Somalia and Sudan showed a higher prevalence of HBsAg than other countries in the region such as Iran. Low endemicity: Countries in the Americas, such as Mexico, Guatemala, and the USA had mostly low endemicity levels (HBsAg prevalence <2%), ranging from 0.01% (95% CI 0.01–0.01) in the UK to 10.32% (8.56–12.38) in Kyrgyzstan. Overall, the South-East Asia Region had low endemicity levels but on country level, an HBsAg prevalence below 2% was only noted in India, Indonesia and Nepal.

Page | 13

Summary of prevalence across risk groups General Population (Schweitzer et al. 2015) PWID (Nelson et al. 2011) MSM (Hope, et al. 2014 - data is for European countries outside of the EU) Albania 18% Sex Workers (Hope, et al. 2014 - data is for (Rossi et al. European 2012; Hanhe et countries al. 2013) outside of the EU) Migrants & Refugees Prisoners Pregnant Women (Hanhe et al, 2013; Nilgun, 2011)) HIV-infected Healthcare persons (general Workers population) (Easterbrook et al. 2015)

East Asia and Pacific 57.99% South Asia Region 2-4% Central and Eastern Europe and Central Asia Region 2-4% North Africa and Middle East Region 24% Sub-Saharan Africa Region 57.99% Latin America

Southeast Asia and East Asia 2.9 - 19.5% South Asia 5.8 - 17.3% Central Asia 7.9% Eastern Europe 0.521.3% North Africa and Middle East region 0.0 - 18.5% Central Africa 3.8 - 9.0% Andean Latin America 2.3 -

Azerbaijan 3.3% East Asia and Pacific 11.3% Azerbaijan 4% Bosnia 1.4% South Asia Croatia 0.90% Serbia 18.3% Region 4.6% Georgia 10% Serbia (incl. Kosovoa) 8.70% Turkey 3.60% Ukraine 9.80% Turkey 2.4%5 Ukraine 9.1% Central and Eastern Europe and Central Asia Region 5.8% North Africa and Middle East Region 2% Sub-Saharan Africa Region 10.3% Latin America and Caribbean Region 1.7% Europe 1.0-

Europe East Africa 60.1(Spain)-4.4% 11% (Slovakia) West, Central Middle East 1% Africa 6-15% (Qatar) -2.9% Latin America (Lebanon) 0.6-2%

Tanzania tertiary hospital 5.6-7% (Mueller et al. 2015)

Uganda tertiary hospital 8.1% (Ziraba et al. South East Asia 2010). 1-2% Eastern Mediterranean 10%

2. PWID: Systematic review ((Nelson et al. 2011)) PWID are a key population who are at particularly high risk of HBV, HCV and HIV infection. In many high-income countries and some developing countries, ongoing HCV transmission is driven mainly by PWID populations. A review of global prevalence data from 77 countries estimated that exposure to HCV (anti-HCV positive) among PWID is estimated to be between 60% and 80% in 25 countries, and over 80% in 12 countries. Similarly, of 59 countries where data were available, prevalence of HBsAg Page | 14

among PWID ranged from 5% to 10% in 21 countries and over 10% in 10 countries. PWID data are global by 20 Global Burden of Disease regions Southeast Asia and East Asia: 2.9–19.5% South Asia: 5.8–17.3% Central Asia: 7.9% Eastern Europe: 0.5–21.3% Central Africa: 3.8–9.0% Andean Latin America: 2.3–8.6% 3. MSMs and sex workers: systematic review ((Hope et al. 2014))) MSM can acquire HBV and HCV sexually. In many populations, there are higher rates of HBV infection among MSM, requiring targeted HBV screening and vaccination. MSM who are HIV positive are at significantly higher risk of acquiring HCV infection than HIV-negative MSM. Sex workers are a key population who are at high risk of acquiring HBV and HCV infection. Multiple factors may contribute to this vulnerability, including unsafe working conditions, barriers to negotiating consistent condom use, and difficulties accessing health-care services. MSM (12 countries), sex workers (5 countries) Country Albania Azerbaijan Croatia Georgia Serbia (incl. Kosova) Turkey Ukraine MSM 18% 4% 0.9% 10% 8.7% 3.6% 9.8% 3.3% 1.4% (Bosnia) 11.1% 18.3% 2.4% 9.1% Sex workers

4. Migrants and refugees: systematic review (Rossi et al. 2012) sub-Saharan Africa Region: 10.3% East Asia and Pacific: 11.3% Central and Eastern Europe and Central Asia Region: 5.8% 5. HIV-infected persons: systematic review (Easterbrook et al. 2015) HIV/HBV (483 estimates from 75/193 (39%) countries

Page | 15

HBsAg prevalence based on a total of 170 estimates in HIV-infected persons, based on population type (general population, PWID, MSM, heterosexual, and pregnant women) and by eleven geographical regions. 1. First, reflecting the epidemiology of HBV in Africa whereby the majority of HBV infections are acquired perinatally or in childhood, the prevalence among key populations of HIV-infected PWID and MSM is not substantially higher than the background rate in the general population or among heterosexuals, especially in Africa, 2. Only in the South-East Asia Region is there a higher prevalence in among PWID and MSM. 6. Prisoners The prevalence of HBV in prisons is often significantly higher than in the general population. Globally, the prevalence of HIV, STIs, hepatitis B and C and tuberculosis in prison populations is estimated to be two to ten times higher than in the general population, and in some settings, 50 times higher. People in prisons and closed settings may be at particular risk for HBV, HCV and HIV infection for a number of reasons. Most commonly, this is due to sharing of needles and syringes and other injecting equipment; often because prevention hardware such as clean needles and syringes are not accessible to prisoners. 7. Indigenous populations: In some settings, indigenous populations are also disproportionately affected by viral hepatitis infection, along with a number of other health problems. Contributing factors to these disparities may include higher rates of injecting risk behaviours among indigenous people who inject drugs and higher rates of incarceration. Epidemic scenarios The broad categories of “generalized” and “concentrated” epidemics are not necessarily helpful in determining how best to prioritize hepatitis testing services. But some general principles apply. Page | 16

Recommendations Guiding principles for hepatitis B and C testing: 1. Promotion of health equity and human rights in national hepatitis B and C testing so that: expanded testing and access is fair and equitable; priority for testing is on diagnosing the undiagnosed; identifying those in greatest need of treatment and those with ongoing risk of infection; and that testing is voluntary and care is provided in a supportive environment free of stigma and discrimination. This is critical as many of the affected population are those who are systematically excluded from access to testing, treatment and care, such as sex workers, injection drug users, men who have sex with men, and prisoners. 2. All persons who test positive for hepatitis B and C (in addition to HIV) should have access to and be linked to hepatitis care and treatment services. 3. Testing of key populations should be undertaken where possible in conjunction with other risk or harm-reduction services. DRAFT recommendation(s): Existing recommendations on prisons, for sex workers and PWID on HBV vaccination 1. Prisons should have a comprehensive hepatitis programme, including the provision of free hepatitis B vaccination for all prisoners, free hepatitis A vaccination to those at risk, and other interventions to prevent, diagnose and treat hepatitis B and C equivalent to those available in the community (including condom, needle and syringe programmes and drug dependence treatment as needed). 2. Include sex workers as targets of catch-up hepatitis B immunization strategies in settings where infant immunization has not reached full coverage (Source: WHO, 2012).36 Page | 17

3. It is suggested to offer people who inject drugs the rapid hepatitis B vaccination regimen.

1.1.1 1. Summary and quality of evidence (see SR_Who to screen_HBV modelling report for references)

□ High □ Moderate □ Low □ Very low

Summary of evidence base for different screening approaches The evidence base for these different screening approaches remains very limited, and largely relies on observational data and modelling. • There are descriptive data showing that targeted testing and community-based screening programme approaches can increase uptake of testing and detection of cases, but very limited data to show impact on patient important outcomes (Pollack, Health Aff (Millwood) 2011;30:1974–83; Bryce BD, Yartel AK. Am J Prev Med 2014;47:23341). Community-based: BFreeNYC screening program (~9000 people screened, 6 cases HCC + 22 end-stage liver failure diagnosed and managed) Lack of evidence and uncertainty as to whether risk-based targeted screening is reaching targeted populations.

•

Cost–effectiveness evidence summary: overview of report – summary of existing studies on cost–effectiveness of screening and treatment for HBV, with an analytic summary of key considerations.  32 studies all from high-income countries in settings with low HBV prevalence. No data on cost or cost–effectiveness of screening for HBV in LMICs was identified. Eight published studies, and one unpublished study (PROLIFICA screening study in Gambia) met inclusion criteria. Two studies evaluated HBV screening in the general population and seven studies in “high-risk” groups (all but one concerned screening in migrant or refugee populations). There was one previously published study in the USA and one forthcoming study in the Gambia, looking at the cost–effectiveness of offering screening and treatment to the general population. The studies used different methods of screening the “high-risk groups” Page | 18



including, in the clinical setting (Wong, Rein), community outreach methods (Rein) and overseas screening (Jezwa). Various outcome measures were used, including cost per quality-adjusted life year (QALY) gained, cost per LY saved and cost per case screened. Many of the models were simulated using hypothetical cohorts.  Overall, data show that offering screening to the general population with subsequent antiviral treatment strategy is cost–effective in HICs (Eckman), as well as LICs (Nayagam), even down to a population prevalence as low as 0.3% and 2%, respectively, in these studies. PROLIFICA study of HBV community-based screening in Gambia: the feasibility of large-scale screening and treatment in sub-Saharan Africa (SSA) has been demonstrated by the ongoing PROLIFICA (Prevention of Liver Fibrosis and Liver Cancer in Africa) study in West Africa (Lemoine et al., forthcoming). This implementation study has screened nearly 10 000 adults for HBsAg at the community level in the Gambia and Senegal using an active outreach method. This is followed by full clinical assessment of those found to be HBsAg positive and antiviral treatment if meeting eligibility criteria. A cost–effectiveness analysis of this community-based screen and treat strategy in the Gambia (Nayagam et al., forthcoming), compared to status quo, revealed an incremental cost–effectiveness ratio (ICER) of $705/LY gained (other outcome measures also calculated: $476/QALY gained or $575/DALY averted). The authors acknowledge that willingness to pay (WTP) thresholds levels, and their use, are highly debated in LMICs. However, it can be regarded as cost-effective if using the WHO WTP threshold of three times the country’s GDP per capita to define a cost–effective intervention (3 times GDP per capita = $1460 in the Gambia). This is the only cost–effectiveness study of screening and treatment we have found in LMIC settings. Furthermore, it is furnished with real-life cost and effectiveness data from a large-scale screening and treatment intervention programme. Furthermore, screening also has benefits that extend beyond the person screened to also others, for example, prevention-of-mother to child transmission.



Conclusions: The data on the cost–effectiveness of screening for HBV is lacking, especially in LMICs. Difficult to draw conclusions regarding the best screening strategy, in terms of who to screen and where to screen, based on cost–effectiveness alone. Currently, there is not enough literature to make strong recommendations for screening based on cost–effectiveness arguments alone. Relatively low screening costs, highly effective and relatively lowPage | 19

cost antiviral therapy at generic price and a fraction of HBsAg-positive persons requiring antiviral therapy should help drive the cost–effectiveness of a test-and-treat strategy. However, this has to be balanced against longterm treatment and the fact that a high proportion with CHB will survive without treatment. Limitations of comparing models/generalizability of results: WHO recommendations are primarily aimed for use in LMICs. All models were from HICs (except PROLIFICA); making generalizations of results from cost– effectiveness analyses between countries or regions with such differing health-care structures, costs, patient behaviours, disease prevalence profiles and willingness-to-pay thresholds can be misleading. Key determinants of testing approach for countries (from cost– effectiveness review): HBsAg prevalence: HBsAg prevalence had a relatively small influence on cost–effectiveness in most of the studies. General population screening was found to remain cost–effective (i.e. ICER below the respective WTH threshold) down to HBsAg prevalence of 0.3% in the USA (Eckman) and 2% in the Gambia (PROLIFICA). Costs: Cost components that need to be considered in economic evaluations of screening and treatment for HBV include costs of screening, diagnostics, monitoring and drugs. This should involve both the cost of consumables, as well as other costs, including human resource costs (which are included to various extents between different studies). A key driver of cost– effectiveness of a screen-and-treat strategy reported in some studies is the cost of antiviral drug (Rossi, Hutton, PROLIFICA). Screening costs varied between the studies, and were only found to be drivers of cost– effectiveness in the Wong and PROLIFICA studies. Linkage to care and adherence: Adherence to treatment and linkage to care were reported as key drivers of cost–effectiveness in several studies (Rossi, Veld). In the PROLIFICA study, variation in treatment adherence was also a key driver of cost-effectiveness. Uptake of screening is not reported to be a key driver of ICER in the studies; however, this does not imply that high participation levels in screening is not important, as when considering health impact alone, increasing uptake Page | 20

is key. The implication of this result is that it is likely to be worthwhile performing screening and treatment, even if participation in screening is assumed to be low. This could be because screening costs are low, relative to the costs and health benefits of treatment for those who are infected. Distribution of patients between different disease states The proportion of people who would benefit from treatment in a population will guide cost–effectiveness, but by how much is difficult to quantify based on current evidence, and needs further research.

2. Risks/benefits Community-based testing (outreach, mobile or venue-based) Benefits   Leads to earlier diagnosis and access to treatment before development of cirrhosis Worldwide, the majority of infections are acquired at birth or in early childhood, and there is therefore generalized high prevalence throughout population, which requires population-based testing approaches. Highly acceptable with index partner testing, home-based and mobile outreach for HIV Generally good uptake Way of accessing missing populations, such as men, key populations and young women who are not pregnant Community-based testing is a critical approach for reaching people from key populations and vulnerable populations who are unlikely to go to a facility, particularly those who are asymptomatic.

□ Benefits clearly outweigh harms □ Benefits and harms are balanced □ Potential harms clearly outweigh potential benefits Are the desirable anticipated effects large? □ No □ Probably □ Uncertain □ Yes □ Varies

   

Risks  May lead to lower-than-expected positivity rates with home-based testing, testing within campaigns, key population outreach and testing of index partners. Suboptimal linkage to care is highly variable and may be problematic. Unit costs may be higher, but may be cost–effective.

 

Provider-initiated testing and counselling (PITC) in health-care facilities Benefits  89/117 low- or middle-income countries recommend HIV PITC in all patient encounters Page | 21

  

High HIV PITC acceptance in antenatal care (ANC) and TB settings Introduction of PITC increased paediatric HIV testing Many clinical settings in generalized epidemic settings not offering hepatitis testing – e.g. STI clinics, primary care, and so many missed opportunities for HBV diagnosis in health-care facilities.

Key and other populations targeted testing Benefits   Key populations are disproportionately affected by hepatitis in all regions. Key populations are less likely to have received HBV vaccination and offer of HBV testing will facilitate higher rates of completion of vaccination.

Partner testing Benefits  Participating in couples and partner HBV testing has a number of benefits. These include adoption of prevention strategies by the couple (for example, condom use, safe injecting practices) and promotion of linkage to and retention in appropriate health-care services. Also applies to opportunistically offering HBV testing and vaccination to family members and other close household contacts of people diagnosed with CHB re access to vaccination and care. Couples and partner testing helps more people know their HBV and/or HCV status, particularly men, who in generalized epidemic settings may be less likely to test than women. Partners: <5% of people currently HIV test with their partners and similar low rates for HBV. Note: HIV serodiscordance is common (half to two thirds of HIV-positive adults with a co-habiting relationship have an HIV-negative partner Offering partner testing for persons with HBV and HCV – highest possible yield. Although risk of infection may be low, a negative test in the partner provides reassurance and the opportunity to provide counselling on reducing future risk including vaccination.



Risks

 People may be reluctant to admit risk behaviours, or may be unaware they are at risk, and so a screening approach that relies on history may miss a substantial proportion of cases. Page | 22

Acceptability, values and preferences PITC  High HIV PITC acceptance in ANC and TB settings

□ No major variability □ Major variability

Is the option acceptable to key stakeholders?

Partner testing   Offering partner testing for persons with HBV and HCV – highest possible yield Need to overcome reluctance to provide partner testing/index partner testing

□ No □ Probably □ Uncertain □ Yes □ Varies

Community-based testing  Community-based testing services would need to be made available in settings acceptable and convenient to people from key populations and vulnerable populations. Services need to be convenient and available, through flexible opening hours and/or walk-in or same-day appointments. Involving affected populations, including adolescents in design, delivery and evaluation of testing services is necessary to ensure that these programmes address their need. Need to address concerns that older relatives, neighbours or family friends will see them attending viral hepatitis/HIV services, including testing services.

 



Equity, ethics and human right implications Will recommendation raise questions around equity?  As for all testing services, programmes for key populations need to emphasize WHO’s “5 Cs” – particularly consent, confidentiality and connection to comprehensive prevention, care and treatment. The use of community-based and hepatitis B and C rapid testing can increase the likelihood of some key populations, such as prisoners, receiving their results. Testing in certain populations, such as in prisons may increase the chances of stigmatization.

□ Less equitable □ More equitable



Are there ethical implications to this recommendation?  No major concerns.

Resource use and financial implications Page | 23

Resource use (see parameter matrix for sample HIV testing programme costs) Estimating the costs associated with a given hepatitis testing approach can be challenging. Costs for similar hepatitis testing may differ significantly between countries and by programme type within a country. Differences in programme costs may be due to general cost differences between countries, in what specific services are provided (referral to clinic for those testing hepatitis-positive vs enhanced linkage support), cadre of staff employed (nurses vs community health workers), the ease of reaching different populations, the capacity of the health system, and the level of HIV testing coverage. Standardized approach to costing of hepatitis testing: A common approach to estimating costs involves identifying and estimating costs incurred by the health-care provider within the following broad categories:    personnel (for example, health-care providers at facilities, counsellors, other paid programme staff, volunteers); recurrent costs (for example, HIV test kits and commodities, printed materials, office supplies); capital expenses, often amortized over their useful life and discounted annually at 3% (for example, office space, transportation, equipment);

Are the resources required small? □ No □ Probably □ Uncertain □ Yes □ Varies

Materials: • • • • Cost of testing kits, buffer/reagents Cost of sterile lancets, pipettes, gloves, sharps-bins or other method of disposal of used-kits Cost of automated reading machine, if applicable Quality-control reagents, if applicable (some kits are supplied with positive and negative controls)

Training and supervision: • • Cost of training testing providers and appropriate assessment, validation and revalidation of their skills From included studies, excellent robust specificity of all tests is reassuring in terms of ensuring cost–effective initiation of algorithms for further investigation and treatment. If being utilized at the point of care, it will be the responsibility of the testing provider to record and report the result appropriately.

•

Other: Page | 24

• •

Creation of a database into which results obtained by POC can be recorded Linkage to care, e.g. antenatal clinics.

Possible test procurement cost: Test RDT EIA Cost (US$) per test 0.3–0.95 (procurement cost) 0.4–2.8 (procurement cost) Source WHO database WHO database

Costs In the PROLIFICA study, despite an active community-based screening campaign, screening costs were low ($7.43 per person offered screening) and the intervention remained cost–effective even if there was a 3-fold increase in screening costs. The Rein study in USA reported costs per person screened between $40 and $280, with the higher costs representing the more active outreach strategies.

3. Feasibility and constraints to implementation Are any major barriers expected for the implementation of this recommendation? The feasibility of large-scale screening and treatment in sub-Saharan Africa (SSA) has been demonstrated by the ongoing PROLIFICA (Prevention of liver fibrosis and liver cancer in Africa) study in West Africa (Lemoine et al., forthcoming). This implementation study has screened nearly 10 000 adults for HBsAg at the community level in the Gambia and Senegal using an active outreach method. This is followed by full clinical assessment of those found to HBsAg positive and antiviral treatment if meeting eligibility criteria. A cost–effectiveness analysis of this community-based screen and treat strategy in the Gambia (Nayagam et al., forthcoming), compared to status quo, revealed an ICER of $705/LY gained (other outcome measures also calculated: $476/QALY gained or $575/disability-adjusted life year [DALY] averted). They authors acknowledge that WTP thresholds levels, and their use, are highly debated in LMICs. However, it can be regarded as cost– effective if using the WHO WTP threshold of three times the country’s GDP per ca–ita to define a cost-effective intervention (3 times GDP per capita = $1460 in the Gambia). This is the only cost–effectiveness study of screening and treatment we have found in LMIC settings. Couples and partners Page | 25 Is the option feasible to implement? □ No □ Probably □ Uncertain □ Yes □ Varies

HIV testing for couples and partners has been conducted in various settings, including ANC and community-based TB services, through ART services and during premarital health visits. Couples and partner HIV testing for the partners of women attending ANC, in particular, is a focus in the 21 priority eMTCT countries. These countries are all highly endemic for HBV, and this provides a unique opportunity to integrate concurrent HBV testing for partners of women with CHB, or chronic HCV infection if risk factors are present.

4. Relevance to different settings/populations Will this recommendation be most relevant for particular settings (e.g. endemicity)? Adolescents In high HBV-prevalence settings there are two groups of adolescents (that is, people 10 –19 years of age) who may need access to HBV testing: (1) undiagnosed adolescents who were exposed perinatally or in early childhood and; (2) adolescents who acquire HBV sexually (through early sex, sex with multiple partners or sex with a person with CHB), or through injecting drug use. Perinatally infected adolescents urgently need to be diagnosed so that they can be linked to HBV monitoring and care and start antiviral treatment if and when this is clinically indicated. In many highly endemic HBV settings, there are a significant number of undiagnosed perinatally infected adolescents. Perinatally exposed adolescents who do not have evidence of CHB need to be vaccinated if this has not yet been done. In many countries, adolescents and young adults may have missed out on HBV vaccination depending on the timing of introduction of universal infant vaccination. Children Universal HBV immunization, including a vaccine birth dose within 24 hours after birth, is key to preventing MTCT of HBV, but many countries have not been able to implement this crucial intervention, due to economic and logistic constraints. Most infants whose mothers have been diagnosed with HBV or HCV should be followed-up and routinely offered EID, and those diagnosed with either with should be regularly monitored for signs of liver disease so that treatment can be offered when necessary. However, some infants are lost to follow-up, so additional pediatric case finding is important. This can be achieved through the routine offer of PITC in health facilities, particularly in high prevalence settings, and also through testing the family members of index cases where appropriate. HBV testing services for infants should be implemented with the aim of identifying as many HBVinfected infants as early as possible. Although a conservative approach to treatment is usually indicated, children born to HBV-infected mothers should be screened early so that monitoring for progression of liver disease can be organized and so that testing and vaccination of household contacts can be carried out. In high-prevalence settings: HBV and HCV testing of mothers and infants should be routinely Page | 26

available through a variety of services – child health services, immunization clinics, under-5 clinics, malnutrition services, well-child services and services for hospitalized and all sick children, TB clinics, and services for orphans and vulnerable children. Testing the family members of index cases Gaps in HBV testing and in documenting the HBV status of children of HBV-positive parents constitute significant missed opportunities. These gaps can be closed by following up the families of cases identified in ANC or facilities offering HBV testing. In all settings all children with an HBVpositive parent or close household contact should be tested for HBV as a priority. 5. Rationale for recommendation:

6. Strength of recommendation

Implementation considerations • As for all testing services, programmes for key populations need to emphasize WHO’s “5 Cs” – particularly consent, confidentiality and connection to comprehensive prevention, care and treatment. • Need to overcome reluctance to provide partner testing/index partner testing • Make use of lay providers/peer testing for outreach especially among key populations • Viral hepatitis testing for key populations needs to be delivered alongside other key primary prevention interventions. • Accessibility and coverage of testing would need to be high to have an impact on the prevalence of HBV among PWID and other key populations. Offering DBS testing for HCV to PWID attending drug treatment programmes increased uptake of testing services.

1. Research gaps    Further research and large scale-implementation studies should be performed to evaluate this further in other high-endemic, low-income settings. What proportion of HBV- or HCV-positive cases will be missed by a testing policy based on screening for at risk behaviours and exposures? Evaluation of different testing approaches in terms of cost, impact and cost–effectiveness and evaluation of key drivers in a range of different settings.

Page | 27

4.2. Who to test HCV Decision-making tables – PICO 2 What is the impact, cost and cost–effectiveness of different HCV testing approaches and scenarios? 7. Topic for analysis: who to screen? Population: 1. Risk-based screening in different high-risk populations Injecting drug users (IDUs), men who have sex with men (MSM), immigrants, recipients of blood transfusion and blood products, sex workers, and health-care workers (HCW), HIV-infected persons 2. General population (excluding blood donors) or selected subpopulations of general population (women during pregnancy, those with raised alanine aminotransferase (ALT), infants, schoolchildren and adolescents) 3. Other approaches: Birth cohort screening (based on different age cut-offs; born between 1945– 1960 or 1965 or 1970) 4. One-off screening vs repeat screening every five years. Intervention: Testing strategies for HBV in different populations (risk-based and general population and birth cohort); and at different prevalence thresholds Comparator: No testing or current practice or comparison of different testing strategies Outcomes: Benefits, harms and costs, and cost–effectiveness with different screening strategies for different target populations Individual patient outcomes: Number of cases detected, overall mortality, liver-related mortality, cirrhosis, end-stage liver disease, rate of hospitalizations, serious adverse events, quality of life Prevention: New infections (mother to child, horizontal (IDUs needle sharing and sexual; and sexual, especially MSM) Cost–effectiveness: Cost and incremental cost per case diagnosed; cost and incremental cost per case screened and treated; cost and incremental cost per life saved; cost and incremental cost per infections averted; quality-adjusted life-years (QALYs) gained

Page | 28

Background: Hepatitis C virus (HCV) is a global public health burden and major cause of morbidity and mortality including liver failure and hepatocellular carcinoma. Current global HCV seroprevalence is estimated to be 2.8%, or >185 million infected individuals worldwide.

Routes of transmission: In many countries, HBV, HCV and HIV transmission occurs predominantly in high-risk key populations, often via common routes of transmission. Key populations include people who inject drugs (PWID), people in prisons and other closed settings, some mobile populations, some indigenous populations, MSM and sex workers. PWID are a key population who are at particularly high risk of HCV infection. In many high-income countries and some developing countries, ongoing HCV transmission is driven mainly by PWID populations. The advent of high-efficacy, low duration therapy, however, generates prioritization for testing for HCV infection, linking infected patients to care, and curing HCV before patients begin to experience the consequences of cirrhosis and end-stage liver disease.

Low rates of diagnosis and linkage to care: The majority of people are unaware of their HCV infection, and therefore often present with advanced disease. Based on still limited studies, overall <15% of the estimated 180 million who are chronically infected with HCV are aware of their diagnosis, based on data from higher-income settings – United States, Europe and China. In addition to the very low access to and uptake of testing, there is also further attrition on the care cascade with very poor linkage to care and therefore treatment, among those who test positive.

Hepatitis B and C testing and diagnosis are at the core of entry to both the prevention and treatment cascade. Testing is required to identify those with are positive, linking them with care, counselling them on measures to reduce transmission to others then assessing who needs treatment, initiating treatment, achieving treatment response (sustained virological response [SVR] for hepatitis C). Hepatitis testing is also needed to identify those who are negative, to provide hepatitis B vaccination, and the opportunity to implement individual or facility level prevention measures counsel to reduce risk behaviours.

There are three key approaches to HCV screening

1. Targeted risk factor-based screening. This refers to screening of specific groups including key populations, who are generally at higher risk of being infected than the general population. This includes PWID, people in prisons and other closed settings, migrant populations, some indigenous populations, MSM and sex workers, but may also include health-care workers. People attending services providing care and treatment for viral hepatitis or HIV can be encouraged to bring their partners to be tested. This involves screening those with high-risk behaviours, exposures and other conditions. Page | 29

Most countries have based their list of high-risk groups as defined by the Centers for Disease Control and Prevention (CDC), and are largely based on known modes of transmission.

i. ii. iii. iv. v. vi. vii.

Family members and household contacts of hepatitis B patients MSM PWID HIV-positive patients Patients on immunosuppression or chemotherapy Persons with liver disease of unclear etiology Health-care workers.

2. Birth cohort screening for HBV and HCV 3. Population or community-based screening (including antenatal). Routine general population screening: i.e. testing among the general population without attempt to identify high-risk behaviours or characteristics (“routine testing”). This means that all members of the population have access to the screening programme under consideration. It may also include home-based testing (house to house); campaigns (e.g. HTC plus – malaria, safe water, non-communicable diseases (diabetes and hypertension); outreach (mobile) in general and key populations; workplaces and schools; and health-care facility-based screening.

4. Health-care facilities. Testing could also be offered in special dedicated clinics, e.g. HIV, STI clinics. Screening at health-care facilities may include primary care settings, inpatient and outpatient settings, and may involve screening on the basis of clinical presentation or focus on only those with abnormal liver function tests, abnormal ultrasound scan, family history of liver disease or other clinical suspicion of liver function test. Persons in whom there is clinical suspicion of viral hepatitis: even when risk factors for HBV and/or HCV are not present, screening is indicated wherever there is clinical suspicion of viral hepatitis infection. This may occur, for example, where there is existing liver disease, including liver cirrhosis or hepatocellular carcinoma, or where there is unexplained liver disease including abnormal liver function tests. Existing guidelines: what are countries doing? The main approach to HCV testing is a targeted risk factor-based testing for those with high-risk behaviours, exposures and other conditions, e.g. (PWID, liver disease, renal dialysis) and use of a birth cohort approach in the US and Japan. At present, no country guidelines recommend routine testing for all individuals regardless of demographics or specific behavioural risk. 1. Screen those with high-risk behaviours, exposures and other conditions. Most countries have based their list of high-risk groups as defined by CDC, and are largely based on known modes of transmission.

i. Family members and household contacts of hepatitis B patients ii. MSM iii. PWID Page | 30

iv. v. vi. vii.

HIV-positive patients Patients on immunosuppression or chemotherapy Persons with liver disease of unclear etiology Health-care workers.

Survey of guidelines (Surjo De) Global prevalence of hepatitis C virus General population PWID MSM (Hope, et al. 2014 – data is for European countries outside of the EU) (Nelson et al. (Hanafiah et al. 2011) 2013) Migrants and Prisoners (Hope, et al. refugees (Larney et al. 2014 – data is (Hanhe et al. 2013) for European 2013) countries outside of the EU) Sex workers Azerbaijan 9.30% Bosnia and Herzegovina 4.30% Croatia 4% Kazakhstan 11% Kyrgyzstan 3.9–28% Republic of Moldova 13% Russian Federation 14–40% Serbia 3.30% Tajikistan 6.30% Turkey 2. 40% Ukraine 32% Uzbekistan 11–12.8% Europe 0– 23.4% (Hungary) SSA 7–26% Western Europe 26–34% Eastern Europe 14–31% Latin America 8– 19% Australasia 28– 43% North America 24–34% South Asia 4–11% Middle East and North Africa 1– 5% East and SE Asia 13–38% Central Asia 32– 43% Extrapolated global 23–29% Pregnant women (Hanhe et al., 2013) HIVinfected persons Health-care workers

Southeast Asia Southeast Asia Azerbaijan 14% 2% and East Asia Bosnia 12% East Asia 3.7% 41–89.8% Croatia 3% South Asia Oceana 2.6% Georgia 16% South Asia 3.4% 36.0– 87.3% Kazakhstan Central Asia 3.8% Central Asia 51.7–61.3 % 4.20% Kyrgyzstan 1.20% Republic of Moldova 11% Russian Federation 2.30% Ukraine 20%

Europe 0 (Slovakia) –1.7% (Italy)

United States Hospital workers 1% (Alter, 1997) Egypt National Liver Institute HCWs 16.8% (Adelwahab et al. 2013)

Central Europe Eastern Europe 22.6–90.5% 2.4% Eastern Europe North Africa and Middle 2.9% East region North Africa and Middle East 28.–67.6% region 3.6% Central Africa 2.3% East Africa 2% West Africa 22.2 - 97.3% Andean Latin America 9.8 – 97.4%

Southern Africa Australasia 2.1% 51.9-54.6 % West Africa Western 2.8% Europe 20.7 – Andean Latin 86.2% America 2% Central Latin America 1.6% Southern Latin America 1.6% Tropical Latin America 1.2% Caribbean 2.1% Asia Pacific High income 1.4% Australasia 2.7% Western Europe 2.4%

Page | 31

North America High Income 1.3%

Global epidemiology of hepatitis C virus infection: new estimates of age‐specific antibody to HCV seroprevalence (Hanafiah, 2013)

Prevalence of hepatitis C antibodies in injecting drug users A review of global prevalence data from 77 countries estimated that exposure to HCV (anti-HCV positive) among PWID is estimated to be between 60% and 80% in 25 countries and over 80% in 12 countries.

MSM MSM can acquire HBV and HCV sexually. MSM who are HIV positive are at significantly higher risk of Page | 32

acquiring HCV infection than HIV negative MSM. Incidence rates of sexually acquired HCV among MSM have been rising in several industrialized countries since 2000, and outbreaks have been described in some less industrialized nations.

Systematic review of HCV prevalence in HIV-infected persons (Platt et al. 2015). Coinfection estimates were identified for 78 of the 194 countries (40%). There were 760 HIV/HCV coinfection prevalence estimates. Findings suggest that globally HCV/HIV coinfection is 1.9% (IQR = 0·4–6·6%) among general population samples, 7% (IQR = 2·6–11·1%) among people living with HIV (PLHIV) who are pregnant or where heterosexual transmission is reported, 6.2% (IQR = 3·3–13·5%) among men who have sex with men (MSM) and 83% (IQR = 55·2–94·1%) among people who inject drugs (PWID). Odds of HCV infection are 11 times higher in the presence of HIV infection, but varied by risk group. The global estimate of HCV coinfections among PLHIV is 3.2 million (IQR = 1.4–4.3 million) of whom 1.2 million (IQR = 0.9–1.4 million) are PWID Mid-point co-infection prevalence (Interquartile range) Number of studies Gen pop PWID MSM 20%(1-38) 2 8% 1 2% 1 4% (0-16) 6 13 (8-15) 16 6% (5-8) 5 Hetero 4% (3-9) 10 Pregnant 0.6% (0.1-5) 3

East Africa 1.3% (0-4.9) 5 71 % (42-99) 2 Cental and West 5 % (2-12) 9 Africa South Africa Latin America North America South East Asia 5% (3-29) 7 Eastern Europe and Central Asia Europe East Med East Asia Western Pacific 7% (0.8-16.1) 3 82% (52-88) 4 84 (41-89) 25 90 (86-97) 18 82% (68-95) 8

8% (4-12.4) 19 10.1 (5-16) 4 0.5% (0-1) 3 11% (8-15) 2 12 (9-25) 9 5% (1.5-7) 5 10% (5-18) 4 4% 1

6% (0.3-30) 3 82% (53-91) 41 8% (4-17) 40 1% 1 81% (74-89) 7 96% (80-98) 15 4% (2-9) 3 9% (7-10) 4

11% (4-23) 11 3% 1

51% (6-89) 7

We compared the prevalence of HCV among 105 samples of HIV-positive and -negative population groups (general population, PWID, MSM, sex workers, prison inmates and high-risk populations). This is summarized in Figure 4 and in the online table. Overall, there was a 12-fold (95% CI 11·2– 11·8) increased odds for HCV positivity across all population groups among HIV-positive compared to HIV-negative persons. Odds of HCV were highest among HIV-positive prison inmates (OR=16·5 95% CI 15·9–17·1) and other high-risk populations (OR = 11·7 95% CI 11·0–12·4), PWID (OR = 5·1 95% CI 4·7–5·5), followed by MSM (OR = 3·8 95% CI 3·1–4·5) and general population samples (OR = 3·7 95% CI 3·3–4·3) and sex workers (OR = 2·5 95% CI 2·0–3·2). Epidemic scenarios hepatitis C virus Generally, HCV epidemics around the world are heterogeneous and represent mixtures of three core Page | 33

epidemic components: 1. Infection related to high-risk behaviours: In essentially every geographical region, the highest prevalence of HCV infection is among persons who use injection drugs (PWID). The prevalence of injection drug use differs between countries and regions, but within those who do inject drugs, HCV prevalence is nearly universally high. Commercial sex workers and prisoners also have increased prevalence (presumably related to both drug use and perhaps sexual transmission) as do men who have sex with men, especially those who are HIV infected. In many cohorts of PWID in North America, Europe and Asia, HCV prevalence ranges from 30% to 75%. 2. Infection related to past generalized exposures that have since been identified and removed: this epidemic pattern, in which there is a high prevalence of HCV within a given age group, is commonly referred to as a “birth cohort epidemic.” While typically identified as being the infection pattern in North America and Europe, many nations have some element of birth cohort epidemics with their unique HCV epidemiology (Table 1). Birth cohort epidemics reflect an HCV exposure source that was once present and to which a large portion of the population was exposed, but that has since been identified and removed. For example, before it was identified and sequenced, HCV infected the blood supply of many countries in all regions of the world. When the blood supply began to be screened for the presence of HCV, the exposure was removed. As a result, the incidence of HCV fell dramatically among the general population, but there remains a burden of prevalent, chronic HCV among patients who were alive and likely to get a blood transfusion during the time that HCV existed in the blood supply.

3. Generalized population epidemic: This pattern is related to a widespread exposure, often iatrogenic, that results in high prevalence (8–10%) across essentially all age groups. Note that the primary difference between a “birth cohort” pattern and a generalized pattern of infection is the duration of time that the generalized exposure existed and whether it has been removed or mitigated. An example of a generalized exposure is the common use of reusable hypodermic syringes and needles in medical settings without adequate sterilization between uses. Few epidemics fall into one of the above three categories. Rather, most are mixed, and represent some combination of all components (Table). Epidemic scenarios Definition Generalized High (>5%) Disaggregation With birth cohort Without birth cohort High intermediate (3–5%) With birth cohort Without birth cohort Low intermediate (2–3%) With birth cohort Without birth cohort 1.1.2 Country example Egypt, Pakistan 1.1.3 Congo, Ukraine 1.1.4 Cote d’Ivoire, Thailand 1.1.5

Page | 34

Mixed

Generalized population prevalence, low, moderate or high with a sizeable risk population (PWID)

High generalized High intermediate generalized

Uzbekistan Taiwan

Low intermediate generalized Low (1–2%) with PWID Low (1–2 %) without PWID To check (UNDOC % of population) Concentrated Generalized population prevalence <1% with high-risk groups Country example Brazil, Portugal

Syria, Hong Kong Mexico, Switzerland The Gambia, Singapore 1.1.6 United Kingdom, Turkey

Extra risk classification Unsafe blood transfusions prior to 1990

8. Draft recommendation(s):

Existing recommendations (Source: WHO, 20149) 1. Risk-based: It is recommended that HCV serology testing be offered to individuals who are part of a population with high HCV seroprevalence or who have a history of HCV risk exposure/behaviour. These include:           Persons who inject drugs (PWID) Persons with HIV infection (HIV-positive men who have sex with men) Prisoners and persons previously incarcerated Persons who have had tattoos, body piercing or scarification procedures where infection control is not guaranteed Children born to mothers infected with HCV Close contacts of persons infected with HCV Persons who have used intranasal drugs Persons from a country with intermediate or high prevalence (2% or greater) of hepatitis C Persons who have received medical or dental interventions in health-care settings where infection control is not guaranteed Persons who have received blood transfusions prior to the time when HCV serologic testing of blood donors was initiated or in countries where HCV serologic testing of blood donations is not routinely performed.

2. General population and birth cohort: It is recommended that in settings with a high HCV seroprevalence (>8%) in the general population, testing be offered, especially at least once, to persons born between 1945 and 1955. Page | 35

Existing WHO recommendations HIV testing services should be routinely offered to all key populations in the community, closed settings such as prisons, and clinical settings. PWID: WHO recommends delivery of a comprehensive package of nine evidence-based interventions for HIV prevention, treatment and care for PWID, all of which are also directly relevant to prevention, treatment and care for HBV and HCV, and one of which is specific to viral hepatitis testing (vaccination, diagnosis and treatment of viral hepatitis, see Box). The nine interventions in the comprehensive package for HIV prevention, treatment and care for people who inject drugs 1. 2. 3. 4. 5. 6. 7. 8. 9. Needle and syringe programmes Opioid substitution therapy and other drug dependence treatment HIV testing and counselling Antiretroviral therapy Prevention and treatment of sexually transmitted infections Condom programmes for people who inject drugs and their sexual partners Targeted information, education and communication for people who inject drugs and their sexual partners Vaccination, diagnosis and treatment of viral hepatitis Prevention, diagnosis and treatment of tuberculosis. Source: WHO, UNODC, UNAIDS,

Sex workers WHO outlines a comprehensive set of interventions and approaches, both to promote enabling environments and to provide prevention, testing, care and treatment, in relation to HIV and STI programming for sex workers (see Box). These recommendations are directly relevant to the response to viral hepatitis among sex worker populations. Essential interventions include enabling sex workers to access and consistently use condoms, access prevention and care, treatment and support services, diagnosis and treatment of important comorbid conditions such as for TB (particularly in HIV endemic settings, incarcerated persons, PWID and sex workers living in exposed to poor cramped working and living conditions), other STIs, and access to harm reduction services for sex workers who inject drugs. Importantly, as sex workers are a key population highly affected by the HBV epidemic, particularly in settings where vaccine coverage is suboptimal, and at high risk for HCV if they inject drugs, they must have access to hepatitis testing services (HepTS), repeat viral hepatitis testing, and partner and family testing wherever appropriate. The offer of HIV testing services (HTS) should be offered alongside HepTS, and HepTS can be integrated into HTS wherever possible. Delivery of HepTS should be informed by recommendations for the delivery of HTS. A variety of settings may be Page | 36

appropriate in which to implement testing services, including health-care settings as well as community settings, and via multiple different approaches. Outreach peer-led testing is likely to be particularly effective and acceptable in many sex worker populations. Good practice recommendations 1. All countries should work toward decriminalization of sex work and elimination of the unjust application of non-criminal laws and regulations against sex workers. 2. Governments should establish antidiscrimination and other rights-respecting laws to protect against discrimination and violence and other violations of rights faced by sex workers in order to realize their human rights and reduce their vulnerability to HIV infection and the impact of AIDS. Antidiscrimination laws and regulations should guarantee sex workers’ right to health and financial services. 3. Health services should be made available, accessible and acceptable to sex workers based on the principles of avoidance of stigma, non-discrimination and the right to health. 4. Violence against sex workers is a risk factor for HIV and must be prevented and addressed in partnership with sex workers and sex worker-led organizations. Evidence-based recommendations 1. 2. 3. 4. 5. 6. 7. 8. Offer a package of interventions to enhance community empowerment among sex workers. Promote correct and consistent condom use among sex workers and their clients. Offer periodic screening for asymptomatic STIs to female sex workers. Offer female sex workers, in settings with high prevalence and limited clinical services, periodic presumptive treatment for asymptomatic STIs. Offer voluntary HIV testing and counselling to sex workers. Use the current WHO recommendations on the use of antiretroviral therapy for HIV positive general populations for sex workers. Use the current WHO recommendations on harm reduction for sex workers who inject drugs (in particular needle and syringe programme and opioid substitution therapy). Include sex workers as targets of catch-up hepatitis B immunization strategies in settings where infant immunization has not reached full coverage. Source: WHO, 2012

Prisons Key WHO recommendations around testing services for people in prisons and closed settings have, to date, mostly focused on HIV prevention, testing and treatment. In 2013, UNODC and partners developed a comprehensive package of 15 key interventions for HIV prevention and treatment in prisons and other closed settings. Due to common transmission routes, these recommendations are equally applicable to viral hepatitis, and HBV vaccination and diagnosis and treatment of viral hepatitis is one of the specific recommendations (see box). This recommendation stipulates that Page | 37

“prisons should have a comprehensive hepatitis programme, including the provision of free hepatitis B vaccination for all prisoners, free hepatitis A vaccination to those at risk, and other interventions to prevent, diagnose and treat hepatitis B and C equivalent to those available in the community (including condom, needle and syringe programmes and drug dependence treatment as needed).” Existing recommendations 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. Information, education and communication Condom programmes Prevention of sexual violence Drug dependence treatment, including opioid substitution therapy Needle and syringe programmes Prevention of transmission through medical or dental services Prevention of transmission through tattooing, piercing and other forms of skin penetration Post-exposure prophylaxis HIV testing and counselling HIV treatment, care and support Prevention, diagnosis and treatment of tuberculosis Prevention of mother-to-child transmission of HIV Prevention and treatment of sexually transmitted infections Vaccination, diagnosis and treatment of viral hepatitis Protecting staff from occupational hazards Source: WHO, 2013

Recommendations and principles surrounding HIV testing in prisons apply equally to HBV and HCV testing in prisons (see box). Existing recommendations      It is important to guard against negative consequences of testing in prisons – for example, segregation of prisoners – and to respect confidentiality. It is also important that people who test positive have access and are linked to HIV care and treatment services. HIV testing and counselling should be voluntary. The use of HIV rapid testing can increase the likelihood of prisoners receiving their results. Testing in conjunction with other risk-reduction services such as the provision of condoms with lubricants and STI screening can increase the benefits of testing and counselling.

Source: WHO, 2014.

Page | 38

9. Summary and quality of evidence Summary of evidence base for different screening approaches Cost–effectiveness evidence summary: overview of report

1.1.7 □ High □ Moderate □ Low □ Very low

Cost–effectiveness analyses of screening for HCV 1. Gleue systematic search identified 19 studies. Majority of studies evaluated the cost–effectiveness of screening for HCV in Europe or the US; one study was carried out in Japan and another one in Italy. Ten studies evaluated screening in the general population; eleven studies screening IDUs; three studies looked at recipients of blood transfusions; one study evaluated screening in women during pregnancy and a further study looked at HCWs. Studies evaluated a one-off screening intervention, with the exception of one study that analysed screening every five years. Comparators were either: no screening or the status quo or different strategies were compared with each other. Recent studies concentrated on birth cohort screening in the US, evaluating the cost–effectiveness of oneoff screening for a cohort born between 1946 and 1970 and a cohort born between 1945 and 1965. HCV prevalence in this population was comparatively high. The initiation of a one-off screening intervention was assessed and compared with current risk-based screening interventions in these US studies. 2. Targeted review of the literature to determine the state of evidence about the cost–effectiveness of testing for HCV in different types of epidemics and among different risk groups. We provide a qualitative assessment of conclusions.

 Testing in high-risk groups such as PWID, MSM, prisoners, HIVinfected persons, and commercial sex workers is likely cost– effective. Testing in settings with a high prevalence of high-risk patients is almost certainly cost–effective in all locations. It is important, however, to ensure adequate follow up after diagnosis.

 The best approach to testing outside of high-risk risk groups depends a great deal on a country’s unique HCV epidemiology. Most countries have at least some component of “birth cohort” epidemic, and “birth cohort” testing is likely cost–effective in most settings.

 In most epidemics, routine screening in the entire population is likely not to be cost–effective. The specific threshold at which a country should alter its approach to routine testing, however, is a Page | 39

function of multiple factors and cannot be identified more generally. This report does not represent the results of a full systematic review. It is meant to serve as a summary of existing studies on cost–effectiveness of screening and treatment for HCV, with an analytic summary of key considerations. Due to the lack of relevant literature from low- and middleincome countries (LMICs), existing studies from high-income countries (HICs) are described and their potential uses and limitations, when drawing conclusions are discussed. Summary of findings and conclusions

 Testing in high-risk groups such as PWID, MSM, prisoners, HIV-infected persons, and commercial sex workers is likely cost–effective. Testing in settings with a high prevalence of high-risk patients is almost certainly cost–effective in all locations. It is important, however, to ensure adequate follow up after diagnosis.  Persons who inject drugs: Multiple analyses in many geographical regions concur that routine testing for HCV in venues with a high prevalence of PWID is cost–effective, even when the studies assume very poor follow-up rates and limited access to therapy. Further, dynamic HCV transmission models suggest that aggressive diagnosis and treatment among current drug users could reduce the incidence of HCV – “cure as prevention”. With typical prevalence estimates of 40%, but ranging as high as 75% in some cohorts, routine screening for HCV is almost certainly cost–effective.

1. Men who have sex with men: Men who have sex with men (MSM) are also at an increased risk of HCV incidence, particularly if they are also HIV-positive. Cost–effectiveness modelling has found testing using liver function tests in combination with HCV Ab testing to be cost–effective in the HIV-positive MSM population. The results of these studies are dependent on appropriate linkage to effective therapy and retention in care.

2. Prisoners: Prisons are likely to have a high HCV prevalence as the result of a high prevalence of PWID in prisons. A UK-based study, however, found that HCV case detection, using dried blood spot testing, was cost–effective, even when the model assumed low rates of HCV treatment initiation. A second study concurs that screening in prisons can be cost–effective, but this study concluded that targeting to screening to those prisoners with a history of injection drug use improves cost–effectiveness. Page | 40

3. HIV-infected persons: Although nearly every guideline for HCV care recommends HCV screening at enrolment in care, there are no cost–effectiveness analyses that address the specific question of the cost–effectiveness of HCV testing at enrollment in HCV care. Because the prevalence of HCV is known to be high in HIV-infected persons, the pace of fibrosis progression in HIV/HCV-coinfected patients is high, and new therapies to treat HCV are effective in HIV/HCV coinfection, testing for HCV at enrolment in HCV is almost certainly cost–effective.

4. Sex workers: Because many sex workers are also PWID or noninjection drug users, the prevalence of HCV in this group is likely high. No studies were identified that address cost–effectiveness of HCV testing in sex workers, and therefore uncertain whether cost– effective to routinely screen all sex workers, compared to an approach that targets testing to sex workers who report a history of injection drug use.  Birth-cohort: i.e. testing among easily identified age or demographic groups known to have high HCV prevalence (“birth-cohort testing”). Most countries have at least some component of “birth cohort” epidemic, and “birth cohort” testing is likely cost–effective in most settings. Whenever there is an easily identified demographic group that has high HCV prevalence (for example, all individuals born in a certain time period) it is likely cost–effective to routinely test for HCV within that cohort. Several cost–effectiveness studies in the US and in Portugal show that birth cohort testing was cost–effective. Routine testing of the entire population: Routine general population screening: testing among the general population without attempt to identify high-risk behaviours or characteristics (“routine testing”). The data about population screening typically come from HICs such as the US and UK, and such studies find that routine testing in the general population is not cost–effective. When compared to “birth cohort testing”, however, universal testing resulted in worse outcomes and higher costs than the birth cohort approach. This analysis raises the spectre that in countries whose HCV epidemic is largely concentrated to a specific birth cohort or demographics group, attempting to identify cases by routine testing of the entire population can dilute the testing effort and result in fewer cases of HCV being identified. An older study, conducted in the UK, also found that although screening high-risk groups in primary care settings was cost–effective, extending screening beyond high-risk individuals was not.



Page | 41

Importantly, all of the above studies reflect the epidemiology of HCV in HICs. One recent paper explicitly studied the cost–effectiveness in Egypt of one-time, routine screening for HCV followed by treatment with either pegylated interferon and ribavirin (PEG-RBV) or PEG-RBV plus an HCV protease inhibitor. Given the very high prevalence of disease, screening was always cost–effective, and often cost-saving. Drivers of cost–effectiveness HCV prevalence: Screening provides increasing value as prevalence rises. In one U.S.-based study, screening was cost–effective (compared to no screening) at a US willingness-to-pay threshold down to a prevalence of 0.53%. In addition to the underlying prevalence of HCV infection, studies identified the rate of progression from chronic HCV to cirrhosis as an important factor together with levels of linkage to care and treatment that substantially influenced cost–effectiveness results.

Risks/benefits Targeted risk-based testing Benefits   Key populations are disproportionately affected by hepatitis in all regions. Key populations are less likely to have received HBV vaccination, and offer of HBV testing will facilitate higher rates of completion of 1.1.8 vaccination. In HIV-infected persons: Benefits of testing and treatment greatest in HIV-coinfected persons as they have more rapid progression of liver disease (HCV-associated liver disease in coinfected patients is emerging as a major cause of morbidity and mortality in HIVinfected persons) and risk of onward transmission than those without HIV infection. Leads to earlier diagnosis and access to treatment before development of cirrhosis. Some studies of PWID have shown that HCV testing and knowledge of serostatus results in reduction in injecting risk behaviours, including frequency of injecting and number of people injected with, in addition to other important risk behaviours, such as heavy alcohol consumption, and increases treatment uptake. Modelling studies suggest that scaling up treatment with DAAs, across different prevalence settings, would have a major impact on the prevalence of HCV.



 



Page | 42

 

HCV treatment outcomes among people who actively inject drugs have been found to be acceptable. Community-based testing is a critical approach for reaching people from key populations and vulnerable populations who are unlikely to go to a facility, particularly those who are asymptomatic.

Risks   Challenging to identify and engage high-risk groups People may be reluctant to admit risk behaviours, or may be unaware they are at risk, and so a screening approach that relies on history may miss a substantial proportion of cases. Much high-risk behaviour is stigmatized and underreported. Health-care workers are not always skilled at identifying high-risk behaviours. Because HCV risk behaviours are stigmatized and underreported, trying to identify high-risk individuals is difficult and prone to under testing high-risk patients. For key populations, especially those whose behaviour is criminalized, testing services are sometimes misused in punitive or coercive ways. As a result, people from key populations avoid the health services that they need. In addition, stigma, discrimination, lack of confidentiality, coercion and fear of repercussions, as well as lack of appropriate health services, resources and supplies, prevent people from testing and, if positive, linking to care. Provider-initiated testing and counselling (PITC) among key populations and vulnerable populations is recommended, so long as it is not compulsory or coercive and it is linked to treatment and care. often reluctant to offer antiviral treatment to PWID. PITC in health-care facilities Benefits

  



 In many settings, and in spite of guidelines, service providers are

   

89/117 LMICs recommend HIV PITC in all patient encounters. High HIV PITC acceptance in antenatal care (ANC) and TB settings Introduction of PITC increased paediatric HIV testing Many clinical settings in generalized epidemic settings not offering hepatitis testing – e.g. STI clinics, primary care, and so many missed opportunities for HBV diagnosis in health care facilities. Page | 43

Partner testing Benefits  Participating in couples and partner HBV testing has a number of benefits. These include adoption of prevention strategies by the couple (for example, condom use, safe injecting practices) and promotion of linkage to and retention in appropriate health-care services. Also applies to opportunistically offering HBV testing and vaccination to family members and other close household contacts of people diagnosed with CHB re access to vaccination and care. Couples and partner testing helps more people know their HBV and/or HCV status, particularly men, who in generalized epidemic settings may be less likely to test than women. Partners: <5% of people currently HIV test with their partners and similar low rates for HBV. Note: HIV serodiscordancy is common (half to two thirds of HIV-positive adults with cohabiting relationship have HIV-negative partner. Offering partner testing for persons with HBV and HCV – highest possible yield. Although risk of infection may be low, a negative test in the partner provides reassurance and the opportunity to provide counselling on reducing future risk, including vaccination.



Birth cohort testing Benefits     Being a member of a birth cohort is easily determined. Screening by age group is less stigmatizing. Targeting testing to birth cohorts is feasible and often cost– effective. In countries with a strong birth cohort dynamic, birth cohort screening is likely preferred.

Risks  Feasibility and successful implementation of birth cohort screening not well established

Routine generalized testing Benefits  Leads to earlier diagnosis and access to treatment before development of cirrhosis Page | 44

 

Way of accessing missing populations, such as men, key populations and young women who are not pregnant Community-based testing is a critical approach for reaching people from key populations and vulnerable populations who are unlikely to go to a facility, particularly those who are asymptomatic.

Risks  When the HCV epidemic is concentrated to a specific age or risk group, routine generalized testing and screening may dilute the screening effort in the cohort with the highest prevalence of HCV and result in fewer cases of HCV identified and higher cost than “birth cohort testing.” If an epidemic is highly concentrated with a specific risk or demographic group, screening outside of that group can be inefficient and increase costs. In most epidemics, routine screening in the entire population may not be cost–effective. May lead to lower-than-expected positivity rates with home-based testing, testing within campaigns, key population outreach and testing of index partners. Countries with high HCV prevalence across the entire population should implement routine screening, but The specific threshold at which a country should alter its approach to routine testing, however, is a function of multiple factors and cannot be identified more generally. problematic



 

 Suboptimal linkage to care is highly variable and may be

10. Acceptability, values and preferences A values and preferences survey of implementers and users of hepatitis B and C testing services was carried out by FIND in September 2015. A total of 104 respondents from 43 (20 high-income, 23 low- and middle-income) countries. Relating to this PICO,  Respondents from LMICs identified following target populations as priority for hepatitis B and C testing: blood donors (>85% for B and C), children born to HCV-infected mothers (55% vs 75% for HBV), persons living with HIV (50% vs 65%), and pregnant women (40% vs 78%), MSM (25% vs 45% for HBV), sex workers (<10% for HCV and 45% for HBV), and prisoners (25% for HCV and HBV) and those chronically ill (around 25% for HCV and HBV%).

□ No major variability □ Major variability Is the option acceptable to key stakeholders? □ No □ Probably □ Uncertain □ Yes □ Varies

Page | 45



Population – wide screening was suggested by less than 10% for HCV and around 15% for HBV (30% of respondents), and in blood donors (22%) was less supported.

PITC  There is generally high acceptance of testing in ANC settings

Partner testing   Offering partner testing for persons with HBV and HCV – highest possible yield Need to overcome reluctance to provide partner testing/index partner testing.

Targeted testing in drug treatment programmes, STI clinics, HIV clinics

 Routine HIV, hepatitis B and C testing in these clinic settings less stigmatizing

11. Equity, ethics and human right implications Will recommendation raise questions around equity?  As for all testing services, programmes for key populations need to emphasize WHO’s “5 Cs” – particularly consent, confidentiality and connection to comprehensive prevention, care and treatment. The use of community-based and hepatitis B and C rapid testing can increase the likelihood of some key populations, such as prisoners receiving their results. Testing in certain populations, such as in prisons may increase chance for stigmatization.

□ Less equitable □ More equitable



Are there ethical implications to this recommendation?  No major concerns.

Resource use and financial implications Resource use (see parameter matrix for sample HIV testing programme costs) Estimating the costs associated with a given hepatitis testing approach can be challenging. Costs for similar hepatitis testing may differ significantly between countries and by programme type within a country. Differences in

1.1.9 Are the resources required small? □ No □ Probably □ Uncertain □ Yes □ Varies

Page | 46

programme costs may be due to general cost differences between countries, in what specific services are provided (referral to clinic for those testing hepatitis-positive vs enhanced linkage support), cadre of staff employed (nurses vs community health workers), the ease of reaching different populations, the capacity of the health system, and the level of HIV testing coverage. Standardized approach to costing of hepatitis testing: A common approach to estimating costs involves identifying and estimating costs incurred by the health-care provider within the following broad categories:    personnel (for example, health-care providers at facilities, counsellors, other paid programme staff, volunteers); recurrent costs (for example, HIV test kits and commodities, printed materials, office supplies); capital expenses, often amortized over their useful life and discounted annually at 3% (for example, office space, transportation, equipment);

Materials: • • • • Cost of testing kits, buffer/reagents Cost of sterile lancets, pipettes, gloves, sharps-bins or other method of disposal of used kits Cost of automated reading machine, if applicable Quality-control reagents, if applicable (some kits are supplied with positive and negative controls)

Training and supervision: • • Cost of training testing providers and appropriate assessment, validation and revalidation of their skills From included studies, excellent robust specificity of all tests is reassuring in terms of ensuring cost effective initiation of algorithms for further investigation and treatment. If being utilized at the point of care, it will be the responsibility of the testing provider to record and report the result appropriately.

•

Page | 47



Feasibility and constraints to implementation Is the option feasible to implement? □ No □ Probably □ Uncertain □ Yes □ Varies

Are any major barriers expected for the implementation of this recommendation?  As for all testing services, programmes for key populations need to emphasize WHO’s “5 Cs” – particularly consent, confidentiality and connection to comprehensive prevention, care and treatment. Need to overcome reluctance to provide partner testing/index partner testing Make use of lay providers/peer testing for outreach especially among key populations Viral hepatitis testing for key populations needs to be delivered alongside other key primary prevention interventions. Accessibility and coverage of testing would need to be high to have an impact on the prevalence of HBV among PWID and other key populations. Offering dried blood spot (DBS) testing for HCV to PWID attending drug treatment programmes increased uptake of testing services.

   

12. Relevance to different settings/populations Will this recommendation be most relevant for particular settings (e.g. endemicity)? Key populations 1. In all settings, a number of social and structural barriers exist to PWID being able to access testing, health-care and harm reduction services, including inadequate coverage and delivery of interventions, stigma and discrimination, high incarceration rates, unstable living conditions, comorbid health problems, poor health literacy and social difficulties. This results in inadequate uptake of prevention, testing, treatment and HBV vaccination,

2. Viral hepatitis testing and treatment for PWID must always be delivered alongside other evidence-based essential primary prevention interventions also. Studies have shown that the uptake of opioid substitution therapy (OST) and adequate needle–syringe programme (NSP) coverage, in isolation, have been shown to reduce the odds of acquiring HCV, but the combination of both interventions together had a much larger impact. Additionally, the high prevalence of comorbidities in populations of PWID, including viral hepatitis/HIV coinfection, TB, mental health problems and poly-drug use alongside social and economic predictors of poor health means that it is particularly important that comprehensive prevention, treatment, care and social services are integrated and accessible to this population. 3. In many settings, responses to viral hepatitis, with particular regard for the need to reach key populations, can be integrated in order to be most effective. When this is not possible, strong links among health services working with priority populations should be established and maintained. Additionally, integration of viral hepatitis testing and treatment with existing services for HIV diagnosis and care is likely to be effective and less resource intensive. Page | 48

4. Community-based testing is a critical approach for reaching people from key populations and vulnerable populations who are unlikely to go to a facility, particularly those who are asymptomatic. To improve access to and uptake of HBV, HCV and HIV testing, community-based testing services should be made available in settings acceptable and convenient to people from key populations and vulnerable populations 5. Accessibility and coverage of viral hepatitis testing will need to be high to have an impact on the prevalence of HBV and HCV among PWID. In order to achieve this, approaches must be acceptable to PWID populations. For example, studies from the UK suggest that offering DBS testing for HCV to PWID attending substance misuse services may increase uptake of testing services. Prisoners Many prisons around the world have implemented viral hepatitis and HIV prevention programmes, including HepTS and HTS as part of a comprehensive package of interventions; however, they are often small in scale and lack the necessary combination of essential interventions, greatly reducing their effectiveness. Access to testing and counselling for viral hepatitis and HIV which is voluntary in nature and easily accessible alongside access to prevention, care and treatment interventions must be urgently scaled up in prisons. Particular attention must go to providing accurate information, obtaining informed consent and maintaining confidentiality. Additionally, there are often major challenges to continuity of care within closed settings and between prisons and the community that need to be addressed. Pregnant women Although the risk of mother-to-child transmission (MTCT) of HCV is much lower than that of HBV, perinatal transmission of HCV does occur in 4–6% of births, and the risk is two to three times higher if the mother is coinfected with HIV. MTCT is the most common cause of HCV infection in young children. In some settings, such as in west sub-Saharan Africa, there is a relatively high prevalence of HCV among young children, which may be a result of the high prevalence of HIV/HCV coinfection in that region. HCV risk factor information should be elicited from pregnant women, and if present, testing of pregnant women for HCV should also be considered alongside testing initiatives for HIV and HBV. There is currently no effective public health intervention to decrease the risk of MTCT of HCV. However, as DAAs become more widely available, they may have a potential role to play in preventing MTCT of HCV if found to be safe and effective for use in pregnancy. Diagnosis of women with HCV before pregnancy should be prioritized so that appropriate treatment and potential HCV clearance can be achieved. Children Consideration should be given to testing children born to HCV-infected mothers, particularly if the Page | 49

mother is coinfected with HIV or has other risk factors for HCV infection, such as injecting drug use. Similar to HBV, the progression of HCV liver disease is usually slow in infected children, but early diagnosis is important to enable monitoring for liver disease and for linkage to appropriate care and treatment when necessary. As DAAs become more widely available, HCV testing in infants and children will become important to be able to offer curative treatment at an early stage before progression of liver disease. Most infants whose mothers have been diagnosed with HBV or HCV should be followed up and routinely offered EID, and those diagnosed with either with should be regularly monitored for signs of liver disease so that treatment can be offered when necessary. However, some infants are lost to follow up, so additional paediatric case-finding is important. This can be achieved through the routine offer of PITC in health facilities, particularly in high-prevalence settings, and also through testing the family members of index cases where appropriate. Integration of HBV and HCV testing into child health programmes In high-prevalence settings: HBV and HCV testing of mothers and infants should be routinely available through a variety of services – child health services, immunization clinics, under-5 clinics, malnutrition services, well-child services and services for hospitalized and all sick children, TB clinics, and services for orphans and vulnerable children. For example, in Malawi it was reported that integrated testing for HIV-exposed infants at six weeks of age into routine postnatal, under-5 and immunization clinics was acceptable and feasible. Potential viral hepatitis testing approaches to improve case-finding among infants and children In all settings   Offer early infant diagnosis for HBV- and HCV-exposed infants. Offer testing to all children and adolescents presenting with indicator conditions/signs and symptoms that suggest acute HBV or HCV, including anorexia, nausea, jaundice, right upper quadrant discomfort and abnormal liver function tests. Consider offering viral hepatitis testing to all children and adolescents attending HIV services, STI clinics and TB clinics.



In high-prevalence settings and for high-risk individuals  Offer viral hepatitis testing or retesting to mothers or infants in immunization clinics or under-5 clinics. If mothers are not available for testing or refuse testing, infant testing is an acceptable alternative. Offer viral hepatitis testing to all children with parents or siblings receiving any HIV service (for example, PMTCT, ART) through home-based or facility-based HTS. Consider viral hepatitis testing in all children and adolescents attending paediatric inpatient health services.

 

13. Rationale for recommendation Page | 50

14. Strength of recommendation

15. Implementation considerations

16. Research gaps  Large-scale implementation studies in range of different LMICs should be performed to evaluate different testing approaches and the extent to which providers can accurately identify and test high risk patients when employing a targeted approach or birth cohort approach, as well as estimate of linkage to HCV care and the HCV cascade. Terms of cost, impact and cost–effectiveness and evaluation of key drivers in a range of different highendemic, low-income settings. A formal cost–effectiveness analysis that compares “targeted” vs “birth cohort” vs “routine” testing requires estimates of the prevalence of high-risk behaviours, stratified by age, the prevalence of HCV among those with high- and low-risk behaviours, stratified by age, and the age structure of the population. This will need to be informed by cost of both HCV therapy in a country, as well as the costs associated with untreated HCV and end-stage liver disease. What proportion of HCV-positive cases will be missed by a testing policy based on screening for at-risk behaviours and exposures?



Page | 51

4.3. How to test HBV Decision-making tables – PICO 1 HBsAg testing: Among persons identified for hepatitis B testing, what is the diagnostic accuracy of available assays for detecting HBsAg (RDT, EIA)? Topic for analysis: How to test Population: Persons identified for HBV testing Intervention: Rapid diagnostic test for HBsAg detection Comparison: Enzyme immunoassays for HBsAg detection Outcomes: Diagnostic accuracy (Sensitivity, Specificity, Positive predictive value, Negative predictive value, TN, TP, FN, and FP). Background:  The most important marker for the diagnosis of hepatitis B infection that may require treatment remains the detection of hepatitis B surface antigen (HBsAg).  Chronic hepatitis B infection is defined by the detection of HBsAg on two occasions six months apart.  However, after initial testing, further characterization of the individual’s HBV infection is based on a sequential testing strategy of for other markers of HBV infection (supplementary testing) triggered by the detection of HBsAg in the first instance.

Immunoassays (laboratory-based)  The most widely used HBsAg assays are laboratory-based immunoassays.  This can be in the form of an enzyme immunoassay (EIA), chemiluminescence immunoassay (CLIA) or electrochemiluminescence immunoassay (ECL).  These are best suited to settings with high throughput of specimens and where infrastructure (electricity, cold storage, climate-controlled rooms) and skilled staff are consistently available.  Other simple assays such as agglutination assays are also available for detection of HBsAg but these generally require serum/plasma specimens and cold storage. The results of simple assays may be read visually.

Rapid diagnostic tests (RDTs) – performed in-laboratory or at the point-of-care  Many laboratories in resource-limited settings may not have access to specialized equipment and Page | 52

process few specimens, per day. Hence, individual tests, including rapid diagnostic tests (RDTs), may be more appropriate.  RDTs for detection of HBsAg come in immunochromatographic (lateral flow) formats. immunofiltration (flow through) and

 In general, RDTs do not require cold storage and may be tested using capillary (finger-stick) whole blood.  The manufacturer’s instructions for use should always be followed. The results of RDTs are read visually.  RDTs may be deliverable at the point of care (POC).  The expansion of their use depends on their performance and operational characteristics in the setting of intended use, ultimately with the aim being to reach resource-limited settings and offer cost-efficient testing services as an alternative to assays that require specific laboratory infrastructure and staff skills to perform.

The selection of EIA or RDTs should not be mutually exclusive. Choice of appropriate technology can be complex but can usually be distilled down to three main factors: performance, cost and accessibility. There are inevitably trade-offs, based not only on disease prevalence and the healthcare infrastructure, but also on technical, socioeconomic, cultural, behavioural considerations.

DRAFT recommendation(s):

Summary pooled diagnostic accuracy of rapid HBsAg assays stratified by □ High study, patient, index and reference test Subanalysis n Study Pre 2005 19 Pooled clinical accuracy Sen (95% CI) 96.9 (96.0– 97.7) 86.4 (85.2– 87.5) 96.7 (96.0– Spec (95% CI) 99.7 (99.6– 99.8) 99.4 (99.2– 99.5) 99.3 (99.0– Likelihood ratios (REM) LR+ (95%CI) 266 (106– 665) 84.6 (43.6– 165) 105 (48.0– LR– (95% CI) 0.056 (0.033– 0.095) 0.126 (0.087– 0.183) 0.028 (0.010– Heterogeneity (Tau-squared) PLR 2.72 NLR 0.91

□ Moderate □ Low □ Very low

Post 2005

4.10

1.27

Case– control

2.23

4.86

Page | 53

97.3) Patient Blood donors HIV+ 19 91.6 (90.1– 92.9) 72.3 (67.9 – 76.4) 92.6 (89.8– 94.8) 91.7 (89.1– 93.9) 90.8 (88.9– 92.4) 97.6 (96.2– 98.6) 82.5 (77.5– 86.7) 82.5 (77.5– 86.7) 80.4 (77.9– 82.6)

99.5) 99.5 (99.3– 99.7) 99.8 (99.5– 99.9) 99.6 (99.0– 99.9) 99.9 (99.8– 99.9) 99.1 (98.9– 99.4) 100 (99.7– 100) 99.9 (99.8– 100) 99.9 (99.8– 100) 99.0 (99.6– 99.3)

230) 89.2 (32.8– 243) 193 (77.4– 497) 79.5 (11.6– 545) 347 (158– 762) 239 (17.1– 33300) 221 (36.1– 1350) 1070 (376– 3060) 285 (71.4– 1140) 58.5 (31.3– 109)

0.076) 0.106 (0.055– 0.204) 0.29 (0.22– 0.38) 0.08 (0.05– 0.13) 0.089 (0.058– 0.136) 0.077 (0.035– 0.168) 0.045 (0.016– 0.128) 0.108 (0.026– 0.458) 0.045 (0.029– 0.069) 0.141 (0.074– 0.268) 3.82 1.86

0.384

0.0059

HIV–

2.97

0.080

Index Test

Whole blood Determine

0.81

0.24

20.2

1.56

BinaxNOW

3.53

1.20

VIKIA

<0.005

1.472

Serodia

<0.005

Reference Test

CMIA

0.44

0.73

EIA: enzyme immunoassay; RDT: rapid diagnostic test; CI: confidence interval; *with EIA reference

Quality of evidence *Refer GRADE table in footnote

Conclusions:  Rapid diagnostic tests, including those performed on whole blood specimens, have good clinical sensitivity and excellent clinical specificity compared to the reference standard (laboratory-based EIA for HBsAg detection). Improvement in both clinical and analytical sensitivity could potentially enhance their impact globally.  Caution in HIV-positive individuals is important with significantly reduced clinical sensitivity compared to HIV-negative individuals Reassuring accuracy of whole blood specimens compared to plasma or serum specimens further facilitates use in the field. □ Benefits clearly outweigh harms

Risks/benefits

Page | 54

Benefits Advantages of testing by RDT compared to laboratory-exclusive EIAs  Does not require capital investment in laboratory infrastructure, e.g. EIA plate washers, readers, incubators, analysers, cartridge or random-access analysers  Concurrent reduction in maintenance costs and reagents

□ Benefits and harms are balanced □ Potential harms clearly outweigh potential benefits

 May be deliverable at the point of care (POC). This may allow greater access to testing and eliminate need for mechanisms for transportation of □ No specimens to the laboratory

Are the desirable anticipated effects large?

 If testing at POC, may reduce number of individuals “lost to follow up”, i.e. □ Uncertain never receive their test results  May be carried out by trained lay providers and health-care workers, in □ Varies addition to trained laboratory scientists  Dedicated venepuncture may not be required. Risks Disadvantages of testing by RDT compared to laboratory-exclusive EIAs  Possible reduction in clinical sensitivity/specificity compared to laboratorybased methods.  RDTs appear to be less sensitive in HIV-positive individuals.  Higher cost per test after expense of laboratory infrastructure has been met.  User variability and subjectivity in reading of a visual assay, second reader suggested.  Performance characteristics may vary with environmental factors, e.g. heat, humidity, storage conditions.  Internal quality control measures may be inferior to standardised laboratory assays, e.g. lack of test kit controls, no specimen addition controls.  Although RDTs using capillary whole blood negate the need for venipuncture and maintenance of laboratory equipment, significant heterogeneity and sub-optimal clinical and analytical sensitivity must be considered. □ Yes

□ Probably

 Recording of results in a database which can be subsequently interrogated and audited as is the case with centralised laboratory testing may be compromised with testing at POC. This may impact on reporting and epidemiological surveillance of the burden of disease. □ No major Acceptability, values and preferences A values and preferences survey of implementers and users of hepatitis B and variability

Page | 55

C testing services was carried out by FIND in September 2015. A total of 104 □ Major variability respondents from 43 (20 high-income, 23 low- and middle-income) countries. Relating to this PICO,  47% of respondents from low- and middle-income countries would prefer Is the option an RDT method of testing using capillary whole blood compared to acceptable to key dedicated venepuncture, even at the cost of reduced clinical sensitivity). stakeholders?  50% of respondent would accept an assay with a minimal sensitivity of 95%, 43.5% would accept 98% and 4.3% would accept 90%. However, when the □ No notion of cost was introduced, only 7% responded that 95% sensitivity □ Probably would not be acceptable. □ Uncertain  77.3% of respondents preferred results of be available on the same day or □ Yes sooner. Respondents commented that delay in individuals receiving results □ Varies was likely to result in a loss to follow-up. Community:  Support for the most effective testing approach in order to impact on availability of testing, especially in resource-limited settings and remote areas and optimize access to at-risk groups. Patients/caretakers:  In the setting of HIV, use of RDTs has facilitated scaling up of testing services in terms of widening access to testing services. Health-care workers:  If RDTs are utilized at the POC, this will allow HCWs to carry out testing and organize follow up potentially in one consultation. There is a need for appropriate training of testing-providers and laboratory staff.  The intervention was considered likely to be acceptable to key stakeholders as the sensitivity and the specificity of RDT for screening of chronic HBV infections are comparable with EIAs. □ Less equitable □ More equitable

Equity, ethics and human right implications Will the recommendation raise questions around equity?  No. The recommendation of the possibility of testing using RDT at POC offers new opportunities for enhancing screening, referral, and treatment for the individuals with chronic HBV infection especially in the resourcelimited settings, thus will reduce transmission, morbidity and mortality associated with undetected and untreated HBV infection.

Page | 56

Are there ethical implications to this recommendation?  No major concerns.

Resource use and financial implications Materials:  Cost of test kits  Cost of sterile lancets, alcohol swabs, gloves, sharps bins or other method of disposal of used kits  Cost of automated RDT readers, if applicable  Quality control reagents, if applicable (some kits are supplied with positive and negative test kit controls). Are the resources required small? □ No □ Probably □ Uncertain □ Yes □ Varies

Training and supervision:  Cost of training testing providers and appropriate competency assessment, certification and re-certification of their skills  From included studies, excellent clinical specificity of all assay formats is reassuring in terms of ensuring cost effective initiation of testing strategies for further investigation and treatment.  If being utilized at the point-of-care, it will be the responsibility of the testing provider to record and report the result appropriately.

Other:  Creation of a database into which results obtained by POC can be recorded  Linkage to care, e.g. antenatal clinics

Possible test procurement cost: Test RDT EIA Cost (US$) per test 0.3–0.95 (procurement cost) 0.4–2.8 (procurement cost) Source WHO database WHO database

Page | 57

Feasibility and constraints to implementation

1.1.10

Are any major barriers expected for the implementation of this Is the option feasible recommendation? to implement?  High-throughput EIAs require certain laboratory infrastructure and □ No equipment with precision and expertise required in its operation.  Delivery of RDTs requires appropriate training of test providers in performing and reading of the test result, storage of materials and recording and reporting of the status. Decentralisation of testing puts tremendous stress on already fragile health systems in terms of training needs, supply chain management, quality assurance, and monitoring and evaluation of effectiveness and impact. External quality assessment of quality of tests and testing possible but challenging when the need for proficiency panels is increased from a few laboratories to hundreds and possibly thousands of POC sites.

□ Probably □ Uncertain □ Yes □ Varies

Feasibility survey report to be presented at meeting.

Relevance to different settings/populations Will this recommendation be most relevant for particular settings (e.g. endemicity)?  The introduction of HBsAg testing using RDTs will be most relevant in settings where there is poor access to existing laboratory testing-services, either access to centralised laboratory testing or lack of testing-infrastructure in existing laboratories.  Delivery of RDTs at the point-of-care in remote or resource-limited settings, e.g. HBsAg testing in antenatal clinics may significantly affect the future burden of disease.  Useful for testing of both symptomatic and asymptomatic individuals.  It will be most relevant to key affected populations who may be at risk of infection but who may be reluctant to or have poor access to health-care services, such as individuals who attend drugrehabilitation clinics or prisoners. These individuals require screening, may require treatment if infected or vaccination if not currently infected.  It will be less relevant in individuals who have good access to health care and in settings where laboratory testing for hepatitis B is already well established.

Rationale for recommendation:

Page | 58

Strength of recommendation

Implementation considerations  Symptomatic vs asymptomatic individuals; in a symptomatic individual, you may not need such good analytical sensitivity than when screening an asymptomatic individual.

Research gaps  Impact of using RDTs for HBsAg at the point-of-care on delivery and implementation of testing services.

Page | 59

GRADE Summary of findings

Page | 60

Page | 61

Question: Should RDTs be used to diagnose HBsAg in HIV-positive individuals? Sensitivity Specificity Outcome

0.72 (95% CI: 0.68–0.76) 1.00 (95% CI: 0.99–1.00) No. of studies (No of patients) Study design

Prevalence

20%

Factors that may decrease quality of evidence

Effect per 1000 patients/year

Test accuracy QoE

Risk of bias

Indirectness

Inconsistency

Imprecision

Publication bias

pre-test probability of 5% 36 (34–38)

pre-test probability of 20% 145 (136–153)

True positives (patients with HBsAg) False negatives (patients incorrectly classified as not having HBsAg) True negatives (patients without HBsAg) False positives (patients incorrectly classified as having HBsAg) 1.

5 studies 2566 patients

cross-sectional (cohort type accuracy study)

serious1

not serious2

serious3

serious4

none

⨁◯◯◯ Very low1–4

14 (12–16)

55 (47–64)

5 studies 2566 patients

cross-sectional (cohort type accuracy study)

serious1

not serious2

not serious5

not serious6

none

948 (945–949)

798 (796–799)

⨁ ⨁ ⨁ ◯ Moderate1–2, 5–6

2 (1– 5)

2 (1–4)

Downgraded for one for risk of bias: all studies were prospective cohort studies of consecutive patients. Studies used different specimens (serum, 2; capillary whole blood, 1; venous whole blood, 1), reference standards (CMIA, EIA confirmed by neutralization), and had patients with different ART status (four studies ART naive). Not downgraded for indirectness: all studies performed in cohorts of consecutive patients in Tanzania (2), Ghana (3), Malawi (4), South Africa (5) and Bissau (6).

Page | 62

Downgraded by one for inconsistency with sensitivities ranging from 62% to 100%: unexplained heterogeneity may arise from differences between studies in specimen type, specimen processing and study population. Two studies had very high sensitivities (100%, 96%) while the remainder (3,5,6) had low sensitivities (range 62–70%). Tau-squared <1 for studies

Downgraded by one for imprecision: confidence intervals 67.9–76.4%. Two studies had very high sensitivities (100%, 96%) while the remainder (3,5,6) had low sensitivities (range 62–70%). Not downgraded for inconsistency: specificities ranged from 99% to 100%, with tau-squared <1 Not downgraded for imprecision: narrow confidence interval

5. 6.

Question: Should Determine HBsAg be used to diagnose HBsAg in a global setting? Sensitivity Specificity Outcome

0.91 (95% CI: 0.89–0.92) 0.99 (95% CI: 0.99–0.99) No. of studies (No. of patients) Study design Risk of bias serious

Prevalences

20%

Factors that may decrease quality of evidence Indirectness Inconsistency Imprecision Publication bias none

Effect per 1000 patients/year pre-test probability of 5% 45 (44–46) pre-test probability of 20% 182 (178–185)

Test accuracy QoE

True positives (patients with HBsAg) False negatives (patients incorrectly classified as not having HBsAg) True negatives (patients without HBsAg) False positives

12 studies cohort and 7552 patients case–control 1 type studies

not serious

very serious

not serious

⨁ ◯ ◯ ◯ 2 Very low

5 (4–6)

18 (15–22)

12 studies cohort and 7552 patients case–control type studies

serious

not serious

serious

not serious

none

941 (940–944)

793 (791–795)

⨁ ⨁ ◯◯ 3 Low

9 (6–10)

7 (5–9)

Page | 63

Outcome

No. of studies (No. of patients)

Study design Risk of bias

Factors that may decrease quality of evidence Indirectness Inconsistency Imprecision Publication bias

Effect per 1000 patients/year pre-test probability of 5% pre-test probability of 20%

Test accuracy QoE

(patients incorrectly classified as having HBsAg) 1. 2. 3.

Lin (7), Lien (8) and Randrianna (9) used a case–control design. Significant heterogeneity across studies for sensitivity; tau-squared 20.2. Heterogeneity exists, but with lower clinical impact; tau-squared 1.56.

Page | 64

World Health Organization Global Hepatitis Programme

References 1. Njai HF, Shimakawa Y, Sanneh B, Ferguson L, Ndow G, Mendy M et al. Validation of rapid point-of-care (POC) tests for detection of hepatitis B surface antigen in field and laboratory settings in the Gambia, Western Africa. J Clin Microbiol. 2015;53(4):1156‒63. 2. Franzeck FC, Ngwale R, Msongole B, Hamisi M, Abdul O, Henning L et al. Viral hepatitis and rapid diagnostic test based screening for HBsAg in HIV-infected patients in rural Tanzania. PLoS ONE. 2013;8(3):e58468. 3. Geretti AM, Patel M, Sarfo FS, Chadwick D, Verheyen J, Fraune M et al. Detection of highly prevalent hepatitis B virus coinfection among HIV-seropositive persons in Ghana. J Clin Microbiol. 2010;48(9):3223‒30. 4. Davies J, van Oosterhout JJ, Nyirenda M, Bowden J, Moore E, Hart IJ et al. Reliability of rapid testing for hepatitis B in a region of high HIV endemicity. Trans R Soc Trop Med Hyg. 2010;104(2):162‒4. 5. Hoffmann CJ, Dayal D, Cheyip M, McIntyre JA, Gray GE, Conway S et al. Prevalence and associations with hepatitis B and hepatitis C infection among HIV-infected adults in South Africa. Int J STD AIDS. 2012;23(10):e10‒3. 6. Honge BL, Jespersen S, Te DS, da Silva ZJ, Laursen AL, Krarup H et al. Hepatitis B virus surface antigen and anti-hepatitis C virus rapid tests underestimate hepatitis prevalence among HIV-infected patients. HIV Med. 2014;15(9):571‒6. 7. Lin YH, Wang Y, Loua A, Day GJ, Qiu Y, Nadala EC, Jr et al. Evaluation of a new hepatitis B virus surface antigen rapid test with improved sensitivity. J Clin Microbiol. 2008;46(10):3319‒24. 8. Lien TX, Tien NTK, Chanpong GF, Cuc CT, Yen VT, Soderquist R et al. Evaluation of rapid diagnostic tests for the detection of human immunodeficiency virus types 1 and 2, hepatitis B surface antigen, and syphilis in Ho Chi Minh City, Vietnam. Am J Trop Med Hyg. 2000;62(2):301‒9. 9. Randrianirina F, Carod JF, Ratsima E, Chretien JB, Richard V, Talarmin A. Evaluation of the performance of four rapid tests for detection of hepatitis B surface antigen in Antananarivo, Madagascar. J Virol Methods. 2008;151(2):294‒7.

Page | 65

4.4. How to test HCV Decision-making tables – PICO 2 To ascertain exposure to HCV through anti-HCV testing: Among individuals identified for hepatitis C testing, what is the diagnostic accuracy of available assays for detecting anti-HCV (RDT, EIA)?

1. Topic for analysis: How to test Population: Individuals identified for HCV testing to ascertain exposure to HCV Intervention: Rapid diagnostic tests and enzyme immunoassays for detection of antibodies to HCV

Comparison: 1. Nucleic acid testing (NAT) 2. EIA and immunoblot 3. EIA only

Outcomes: Diagnostic accuracy (Sensitivity, Specificity, Positive predictive value, Negative predictive value, TN, TP, FN and FP).

2.  

Background: Screening for exposure to hepatitis C virus (HCV) is dependent on assays that detect antibodies to HCV (anti-HCV) in the first instance. Once antibody status is confirmed, the individual should undergo supplementary testing for active HCV infection using an assay designed to detect viral replication, such as HCV RNA or core antigen (HCV cAg). Assays designed solely to detect antibodies to viral antigens will inevitably have a “window period” of infectivity in early infection in an individual who has been recently infected whose infection will not be detected by a given serological assay. This diagnostic window period can be shortened by direct detection of viral antigen or nucleic acid. The improvements in assay performance over time, in particular the EIAs, have been termed as “generations” of the assays. *(see footnote) It is important to note that the latest generation of assays designed to detect anti-HCV also are combined with cAg in order in increase seroconversion sensitivity of the assay, but these “4th generation” or “combination” assays should not be used to differentiate exposure from chronic infection. Page | 66



 



Note that sensitivity of antibody-only assays may be reduced if the patient is immunocompromised, e.g. HIV, immunosuppressive therapy, renal dialysis.

Immunoassays (laboratory-based)   The most widely used anti-HCV assays are laboratory-based immunoassays. They detect antibodies to core and non-structural antigens. In the case of 4th generation assays, the assay combines detection of antibodies to HCV along with detection of hepatitis C core (P22 Ag) antigen directly. These can be in the form of an enzyme immunoassay (EIA), chemiluminescence immunoassay (CLIA), electrochemiluminescence immunoassay (ECL) or recombinant immunoblot assay. These are best suited to settings with high throughput of specimens and where infrastructure (electricity, cold storage, climate-controlled rooms) and skilled staff are consistently available.

 

Rapid diagnostic tests (RDTs)  Many laboratories in resource-limited settings may not have access to this specialized equipment and process fewer specimens, per day. Hence, individual tests, including rapid diagnostic tests (RDTs), may be more appropriate. RDTs for detection of anti-HCV are simple to perform and do not require instrumentation, they come in immunofiltration (flow through) and immunochromatographic (lateral flow) formats and may be read visually. In general, RDTs do not require cold storage and may be tested using capillary (fingerstick) whole blood or oral fluid. However, the manufacturer’s instructions for use should always be followed. RDTs may be deliverable at the point of care (POC). The expansion of their use depends on their performance and operational characteristics in the setting of intended use, ultimately with the aim being to reach resource-limited settings and offer cost-efficient testing services as an alternative to assays that require specific laboratory infrastructure and staff skills to perform.



 

* Summary of assay “generations” 1st generation assays:    Detection of antibodies to NS4 antigen only Becomes detectable 12–26 weeks after exposure High false-positive rate, i.e. poor positive-predictive value in low-prevalence populations.

2nd generation assays:   Detects antibodies to NS3, NS4 and core antigen Decreased window period of infectivity to 10–24 weeks. Page | 67

3rd generation assays:   Detects antibodies to NS3, NS4, NS5 and core antigen Further decreased window period of infectivity.

4th generation assays A.K.A. combination assays:   Combination of detection of circulating antibodies to viral antigens as above, but also addition of monoclonal antibodies to detect hepatitis C antigens (P22 Ag) directly. Immunoassays solely for the detection of HCV core antigen were developed initially to close the diagnostic window in seronegative infection and subsequently for the detection of antigenaemia in the presence of antibody. Draft recommendation(s):

4. Summary and quality of evidence Systematic review report

1.1.12 □ High □ Moderate □ Low □ Very low

A systematic review was commissioned in order to assess this PICO question (see SR PICO 2). The purpose of this review was to determine the sensitivity and specificity of assays used to detect hepatitis C antibody using multiple specimen types, including serum, whole blood and oral fluid.

Summary of the evidence Method: A literature search was conducted focused on hepatitis C, diagnostic tests, and diagnostic accuracy. Studies were included if they evaluated an assay to determine the sensitivity and specificity of a single hepatitis C antibody (HCVAb) test among humans. Two reviewers performed a quality assessment of the studies and extracted data for estimating test accuracy.

Results: A total of 52 studies were included, evaluating 30 RDT devices of varying generation of assay. Page | 68

    

RDTs vs EIA only: Pooled clinical sensitivity and specificity were 0.98 and 1.00, respectively. RDTs vs EIA, immunoblot and NAT: Pooled clinical sensitivity and specificity were 0.96 and 1.00, respectively. RDTs vs NAT or immunoblot: Pooled clinical sensitivity and specificity were 0.93 and 0.98, respectively. RDTs vs Ag/Ab combination assay: Pooled clinical sensitivity and specificity were 0.86 and 0.99, respectively. RDTs on oral fluid compared to a serological reference standard using serum/plasma: Pooled clinical sensitivity and clinical specificity were 0.94 and 1.00, respectively. Results were comparable across general populations, hospital patients and key populations.



Pooled diagnostic accuracy for HCV antibody tests Pooled test accuracy for different tests (52 research studies). Comparison (number of studies) Pooled SE 95%CI Tau-square P-value for heterogeneity 1.00 0.95 0.98 <0.001 <0.001 <0.001 Pooled SP 95% CI Tau-square P-value for heterogeneity 1.00 0.99 1.00 <0.001 <0.001 <0.001

RDT versus EIA only (n = 5) RDT versus NAT or Immunoblot (n = 13) RDT versus EIA, NAT or Immunoblot (n = 14) Oral RDT versus blood reference (n = 12) Sample type Blood samples (n = 45) Oral samples (n = 12) Source population General screening (n = 17) Key population (n = 19) Hospital patients (n = 16) Antibody and Antigen Combo testing (n=6) Oral kits brand OraQuick (n = 8)

0.99 0.93 0.97

0.98 0.91 0.96

1.00 0.98 1.00

1.00 0.97 1.00

0.94

0.93

0.96

<0.001

1.00

<0.001

0.98

0.97

0.98

<0.001

0.98

0.99 1.1.11

0.94

0.93

0.96

<0.001

1.00

<0.001

0.95 0.97 0.97 0.86

0.94 0.96 0.96 0.79

0.96 0.98 0.98 0.94

<0.001 <0.001 <0.001 <0.001

0.99 0.94 1.00 0.99

0.98 0.94 1.00 0.98

0.99 0.95 1.00 1.00

<0.001 <0.001 <0.001 <0.001

0.98

0.97

0.99

<0.001

1.00

<0.001

Page | 69

Other brands (n = 6)

0.88

0.84

0.92

<0.001

0.99

1.00

<0.001

SE: sensitivity; SP: specificity; CI: confidential interval; RDT: rapid diagnostic test; EIA: enzyme immunoassay; NAT: nucleic acid testing Note:* #

Studies conducted across these regions were not included here.

Conclusions:  Rapid diagnostic tests, including RDTs for oral fluid, have excellent sensitivity and specificity compared to laboratory-based methods, across different populations for detection of antibodies to HCV. This suggests that RDTs can be used to test for HCV antibody. Sensitivity/specificity was less for RDTs for anti-HCV compared to newer combination antibody/antigen assays.



Issues raised from the review: • • The comparison of RDT versus immunoblot would include HCV-cleared person: HCV Ab+ but HCV RNA–. Publication bias, as studies with poor test performance were less likely to be published, lead to exaggerated estimates of the accuracy.

Quality of evidence *Refer GRADE table in footnote □ Benefits 5. Risks/benefits Benefits Advantages of testing by RDT compared to laboratory-exclusive EIAs    Does not require capital investment in laboratory infrastructure, e.g. EIA washers, readers, incubators, analysers, cartridge or random-access analysers Concurrent reduction in maintenance costs and reagents May be deliverable at the point of care (POC). This may allow greater access to testing and eliminate need for mechanisms for transportation of specimens to the laboratory clearly outweigh harms □ Benefits and harms are balanced □ Potential harms clearly outweigh potential benefits

  

If testing at POC, may reduce number of individuals “lost to follow-up”, i.e. Are the never receive their test results

May be carried out by trained lay-providers and health-care workers, in anticipated addition to trained laboratory scientists Dedicated venepuncture may not be required as some assays are validated

desirable

effects large?

Page | 70

for capillary whole blood or oral fluid.  Rapid diagnostic tests (RDTs) have potential for scaling up access to hepatitis B testing, particularly for key populations.

Risks Disadvantages of testing by RDTs compared to laboratory-exclusive EIAs       Increased cost per test after expense of laboratory infrastructure has been met User variability and subjectivity in reading the visual assay, suggest second reader Performance characteristics may vary with environmental factors, e.g. heat, humidity, storage conditions. Internal quality control measures may be inferior to standardized laboratory assays, e.g. lack of test kit controls, no specimen addition controls. Performance characteristics may vary in certain individuals, e.g. HIV infection, immunosuppressed – lower sensitivity of anti-HCV compared to NAT testing. Recording of results in a database which can be subsequently interrogated and audited as is the case with centralized laboratory testing may be compromised with testing at POC. This may impact on reporting and epidemiological surveillance of the burden of disease.

□ No □ Probably □ Uncertain □ Yes □ Varies

□ No major

6. Acceptability, values and preferences variability A values and preferences survey of implementers and users of hepatitis B and C □ Major testing services was carried out by FIND in September 2015. A total of 104 variability respondents from 43 (20 high-income, 23 low- and middle-income) countries participated. Relating to this PICO,  47% of respondents from low- and middle-income countries would prefer an Is the option RDT method of testing, even at the cost of reduced sensitivity). acceptable to key stakeholders?

Community: 

Support for the most effective testing approach in order to impact on availability of testing, especially in resource-limited and remote areas and □ No □ Probably optimize access to at-risk groups.

Patients/caretakers:  In the setting of HIV, use of RDTs has facilitated scaling up of testing services in terms of widening access to testing services.

□ Uncertain □ Yes □ Varies

Health-care workers:  If RDTs are utilized at POC, this will allow HCWs to carry out testing and Page | 71

organize follow up potentially in one consultation. There is a need for appropriate training of testing providers and laboratory staff.  Appropriate pre- and post-test counselling was mentioned in the values and preferences survey, which suggested that in low- and middle-income countries, not all individuals were being offered this service. 47% of respondents (n = 23) preferred POC testing using capillary whole blood even at the expense of clinical sensitivity. The intervention was considered likely to be acceptable to key stakeholders as clinical sensitivity and clinical specificity of RDTs for ascertaining HCV exposure are comparable with EIAs. □ Less equitable □ More equitable

 

Equity, ethics and human right implications

Will the recommendation raise questions around equity?  No. The recommendation of the possibility of testing using RDTs offers new opportunities for enhancing screening, referral, and treatment for the individuals with chronic HCV infection especially in the resource-limited settings, thus will reduce transmission, morbidity and mortality associated with undetected and untreated HCV infection.

Are there ethical implications to this recommendation?  No major concerns.

Resource use and financial implications Materials:     Cost of test kits Are the resources required small?

Cost of sterile lancets, alcohol swabs, gloves, sharps-bins or other method of □ No disposal of used-kits □ Probably □ Uncertain Cost of automated RDT readers, if applicable Quality control reagents, if applicable (some kits are supplied with positive □ Yes □ Varies and negative test kit controls)

Training and supervision:   Cost of training testing providers and appropriate competency assessment, certification and re-certification of their skills From included studies, excellent specificity of all assays is reassuring in Page | 72

terms of ensuring cost–effective initiation of algorithms for further investigation and treatment.

Other:   Creation of a database into which results obtained by POC can be recorded Linkage to care, e.g. antenatal clinics

Possible test-procurement costs: RDTs Cost (US$) per test 0.50–1.70 (>10 for oral fluid) Laboratory-based assays HCV EIA HCV RNA 2.60–4.30 (procurement costs) ~20 WHO database Source MSF, WHO database

Feasibility and constraints to implementation Is the option feasible to implement?

Are any major barriers expected for the implementation of this recommendation?  High-throughput EIAs require extensive laboratory infrastructure and equipment with staff expertise in its operation. Batching can lead to long 1.1.13 delays before results are available.



 Decentralization of testing puts tremendous stress on already fragile health systems in terms of training needs, supply chain management, quality assurance, and monitoring and evaluation of effectiveness and impact. External quality assessment of quality of tests and testing possible but challenging when the need for proficiency panels is increased from a few laboratories to hundreds and possibly thousands of POC sites. Feasibility survey report to be presented at meeting.

Delivery of RDTs requires appropriate training of test providers in performing □ Probably and reading of the test result, storage of materials and recording and □ Uncertain reporting of status. □ Yes □ Varies

□ No

9. Relevance to different settings/populations Will this recommendation be most relevant for particular settings (e.g. endemicity)?  This recommendation of the introduction of RDTs will be most relevant in settings where there is poor provision of laboratory testing-services, either access to centralized laboratory testing or lack of testing infrastructure in existing laboratories. Delivery of RDTs at the POC will have relevance more in remote or resource-limited settings, Page | 73



compared to settings where there is good access to health care and established screening programmes.  It will be most relevant to groups of patients at risk of infection but who may be reluctant to or have poor access to health-care services, such as individuals who attend drug-rehabilitation clinics or closed settings such as prisoners. These individuals require screening and may require assessment and treatment if found to be infected. It will be less relevant in individuals who have good access to health care and in settings where laboratory testing for hepatitis C is already well established.



10. Rationale for recommendation:

11. Strength of recommendation

12. Implementation considerations

13. Research Gaps  Impact of testing at POC using RDTs to rule out/rule in HCV exposure on HCV-associated morbidity and mortality and onward transmission testing and treatment programmes.

Page | 74

World Health Organization Global Hepatitis Programme

GRADE Summary of findings Table: Strength of evidence for diagnostic accuracy Unit of analysis Type of samples Studies, n Risk of bias Consistency Directness Precision # of samples Strength of evidence Sen (95% CI) Sp (95% CI) Pretest probability (%) Positive LR (95% CI) PPV Negative NPV Strength LR evidence (95% CI)

RDT versus General population, Oral fluid, EIA, hospital patients, serum or or immunoblot blood donors, plasma injection drug users and other high-risk populations Oral RDT versus blood reference General population, Oral fluid, hospital patients, serum or blood donors, plasma injection drug users and other high-risk populations

Mod

Se: Inconsistent

Indirect

Precise

42,239

Se: Mod

0.97 (0.0.96 –0.98)

1.00 (1.00 –1.00)

0.05

372.92 (267.56 –574.12)

0.95

0.03 (0.02 –0.04)

1.00

Sp: Consistent 12 Mod Se: Inconsistent Indirect Precise 14,547

Sp: Mod

Se: Mod

0.94 (0.93 –0.96)

1.00 (1.00 –1.00)

0.05

314.5 (202.02 –684.07)

0.94

0.06 (0.04 –0.07)

1.00

Sp: Consistent

Sp: Mod

Mod: Moderate; Sen: sensitivity; Sp: specific

Page | 75

World Health Organization Global Hepatitis Programme

4.5. How to test - testing strategy HBV Decision-making tables – PICO 3 Testing strategy to diagnose chronic HBV infection through detection of HBsAg: Among individuals identified for hepatitis B testing, what is the best testing strategy (diagnostic accuracy and other outcomes) for detection of HBsAg? (One-assay versus two-assay strategy) 1. Topic for analysis: How to test Population: Persons identified for HBV testing Intervention: One-assay testing strategy; one HBsAg assay Comparison: Two-assay testing strategy; two different HBsAg assays Outcomes: Diagnostic accuracy (Sensitivity, Specificity, Positive predictive value, Negative predictive value, TN, TP, FN, and FP). Background:

 

The most important marker for the diagnosis of chronic hepatitis B infection that may require treatment remains the detection of hepatitis B surface antigen (HBsAg). Although the case definition of chronic hepatitis B is the detection of HBsAg twice six months apart, if facilities exist, after an initial positive result, supplementary testing can be undertaken in order to facilitate entry into a care pathway. HBsAg will be detectable in the blood if there is current hepatitis B infection. Confirmation of the specificity of a reactive HBsAg first-line test result is usually carried out by either; i) repeating the HBsAg testing in a different assay of similar sensitivity, or ii) performing a neutralization test using a specific anti-HBs-containing reagent in the same first-line assay after appropriate dilution of the specimen under test. Specificity is confirmed when this reagent abolishes reactivity in the assay. This PICO question addresses the issue of whether a positive result from a single HBsAg assay has sufficient specificity in order to proceed to supplementary testing and/or entry into a care pathway, or whether confirmatory testing on the same specimen with a different HBsAg assay Page | 76



(or neutralization), performed sequentially after the first assay is required.

Testing strategies:   WHO recommends standardized testing strategies to maximize the accuracy of hepatitis B and C testing while minimizing cost and increasing simplicity. A testing strategy describes a testing sequence for a specific testing objective, taking into consideration the presumed prevalence of the analyte to be tested in the population. In both high and low prevalence settings, more than one assay may be required. See footnote for further detailed explanation of the one- and two-assay strategies.

Figs 1–3: Possible testing strategies for detecting HBsAg

HBsAg one-assay strategy HBsAg (RDT/EIA) (A1)

NonReac ve

Reac ve

Interpreta on: No evidence of HBV infec on Advise retes ng +\- immunisa on if ongoing risk or known exposure

Interpreta on: Compa ble with HBV infec on Proceed to supplementary tes ng

Page | 77

HBsAg (RDT/EIA) (A1)

HBSsAg Two-assay strategy

NonReac ve

Reac ve

Interpreta on: No evidence of HBV infec on Advise retes ng +\- immunisa on if ongoing risk or known exposure

HBsAg (RDT/EIA) (A2) Nonreac ve Interpreta on: Inconclusive result Further tes ng as appropriate

Reac ve

Interpreta on: Compa ble with HBV infec on Proceed to supplementary tes ng

HBsAg (RDT/EIA) (A1)

Two-assay strategy or neutralisa on

Reac on not neutralised

NonReac ve

Reac ve

HBsAg neutralisa on (EIA) (A1+)

Interpreta on: Inconclusive result Further tes ng as appropriate

Interpreta on: No evidence of HBV infec on Advise retes ng +\- immunisa on if ongoing risk or known exposure

HBsAg (RDT/EIA) (A2) Nonreac ve Interpreta on: Inconclusive result Further tes ng as appropriate

Reac on neutralised

Reac ve

Interpreta on: Compa ble with HBV infec on Proceed to supplementary tes ng

Since the decision for PICO 3 critically depends on whether currently available tests for detecting HBsAg show acceptable sensitivity and specificity, the following conclusions from PICO 1 should be considered:   Overall compared to an EIA reference, RDTs had pooled clinical sensitivity 90% (95% CI: 89, 91) and clinical specificity 100% (95% CI: 99, 100). RDTs were more accurate among HIV-negative patients, with pooled clinical sensitivity of 93% (95% CI: 90, 95), and pooled clinical specificity of 100% (95% CI: 99, 100) compared to results in HIV-positive patients in whom RDTs are 72% sensitive (95% CI: 68, 76) and 100% specific (95% CI: 100, 100) compared to an EIA reference. Results for capillary whole blood specimens were comparable to serum and less heterogeneous.



Page | 78

The two-test strategy involves the use of a more sensitive test as the screening test and a second test that is more specific to reduce false-positive results. Since the review in PICO 1 showed that HBsAg serological assays have excellent specificity, then there is no need for the use of another test if a simple testing strategy is desired.

2. DRAFT recommendation(s):

Summary and quality of evidence Systematic review  A systematic review by the London School of Hygiene and Tropical Medicine (LSHTM) was commissioned to assess this PICO question.

□ High □ Moderate □ Low □ Very low

Summary of results • No studies directly compared diagnostic accuracy, cost, costeffectiveness of one- vs two-assay HBsAg testing strategies.   The reviewers found four documents that demonstrated a strategy for hepatitis B testing. The Public Health England UK standard for microbiology investigation (SMI) (under review) suggested that either neutralization or repeat testing of the same specimen with a different HBsAg assay would be acceptable, but did not directly quote any evidence for this. The Australian “National Testing Policy for hepatitis B” suggested that HBsAg be confirmed with neutralization but also anti-HBc and anti-HBs should be assessed in order to form a profile and suggested management, e.g. immunization. Fan et al. (2014) used a decision-tree model of sequential HBeAg and HBV-DNA in a cohort of pregnant women in order to calculate costeffectiveness.



Page | 79

Conclusions  Diagnosis of HBV is very complex and there may not be 1–2 simple algorithms that will cover all settings.

Quality of evidence   Study quality was not evaluated using the QUADAS-2 tool and the STARD checklist, as it was not applicable. None of the studies met inclusion criteria.

GRADE Summary of findings Not applicable as no studies met the inclusion criteria for the systematic review. • All PICOs related to HBV will need to be looked at together in order to try and answer PICO 3.

Modelling studies of one-assay vs two-assay strategies To be added at meeting.

3. Risks/benefits Benefits   The intervention of testing using a single HBsAg assay would simplify the process of testing. Sensitivity, i.e. detecting those individuals with HBV infection should not be compromised with a one-assay strategy. This will identify those who require further assessment and possible treatment. Cost of overall testing may be reduced.

□ Benefits clearly outweigh harms □ Benefits and harms are balanced □ Potential harms clearly outweigh potential benefits

 

May allow more rapid reporting of the result, so that the patient can be appropriately followed up for further assessment, vaccination or given health protection advice, e.g. measures to reduce onward transmission. □ No

Are the desirable anticipated effects large?

Risks  Possible reduced positive predictive value of a single test result on one

□ Probably □ Uncertain □ Yes □ Varies

Page | 80

assay may lead to more false-positives and associated anxiety and cost associated with follow up, before the correct diagnosis is ultimately made. This is especially relevant in low-prevalence settings at which even a test of 99% specificity may yield more false-positive than truepositive results.  If analytical or clinical sensitivity of the first-line assay is lower than the combination of two different assays, this may result in more falsenegatives, resulting in individuals not referred for the appropriate further assessment and potential ongoing transmission The recommendation of confirmation with neutralization is reliant on adequate facilities being available. If performed incorrectly, may result in indeterminate results, leading to increased cost of further testing. □ No major variability □ Major variability



4. Acceptability, values and preferences Community/patients/caretakers 

Patients require a simple and rapid testing strategy, yet one that is Is the option acceptable acceptably accurate for the purpose of testing to key stakeholders? □ No □ Probably □ Uncertain □ Yes □ Varies

Health-care workers    HCWs will need to understand the strengths and limitations of any given testing strategy and appropriately counsel patients who are screened. HCWs will need to appropriately act on results, either positive and negative. If performing testing at POC, HCWs need to be adequately trained in the use of RDTs and record/report results appropriately.

Other  Laboratory staff need to be appropriately trained in performance of the various assays according to the manufacturers’ instructions for use, especially if HBsAg neutralization is being utilized □ Less equitable □ More equitable

Equity, ethics and human right implications Will recommendation raise questions around equity?  Yes, availability of HBsAg neutralization assays or other supplementary testing, such as nucleic acid testing and/or further hepatitis B markers with be more available in settings with established hepatitis B testing

Page | 81

facilities.  However, note that where an RDT for HBsAg is used and no neutralization reagents are available, confirmation of an acute or chronic infection for diagnostic purposes may be concluded by repeat testing for HBsAg after a period of time. Consecutive reports of HBsAgpositive status twice at least 6 months apart will confirm a diagnosis of chronic HBV infection.

Are there ethical implications to this recommendation?  The ethical implication of recommending a strategy with possible suboptimal PPV.

5. Resource use and financial implications Diagnostics/laboratory monitoring  Are the resources required small?

Cost of a one-test vs two-test strategy will be less due to fewer tests □ No being used. □ Probably □ Uncertain □ Yes □ Varies

Assay format RDT EIA

Indicative cost (US$) per test 0.30–0.95 (procurement cost) 0.40–2.80 (procurement cost)

Source WHO database WHO database

Training and supervision  Appropriate training for laboratory staff and health-care workers delivering testing services at the point of care.

6. Feasibility and constraints to implementation Are any major barriers expected for the implementation of this implement? recommendation?  Procurement of appropriate test sensitivity/specificity/analytical sensitivity kits meeting required □ No □ Probably 1.1.14 Is the option feasible to

Page | 82



Delivery of testing services to remote/resource-limited settings

□ Uncertain □ Yes □ Varies

Feasibility survey report to be presented at meeting.

7. Relevance to different settings/populations Will this recommendation be most relevant for particular settings (e.g. endemicity)?  This recommendation will be most relevant in settings that do not have established hepatitis B testing programmes and of less relevance where laboratory testing for hepatitis B is already well established.

8. Rationale for recommendation:

9. Strength of recommendation

10. Implementation considerations

11. Research gaps   Impact of a one- vs two-step testing strategy on hepatitis B testing programmes Cost-effectiveness studies of one- versus two-assay or neutralization testing strategies

Footnote: Explanation of one- and two-assay strategies and PPV and NPV calculation method based on prevalence

One-assay strategy  A single test is performed. If the test result is reactive, an “HBsAg positive” status is reported, with need for complementary testing and follow-up HBsAg testing in 6 months to diagnose chronic infection recommended. If the initial test result is non-reactive, an “HBsAg negative” status is reported. This strategy efficiently identifies most uninfected individuals and rules out chronic HBV infection; it identifies those likely to be infected and in need of additional testing.

 

Page | 83



It is a suitable testing strategy for resource-limited settings because only a single test or one assay is performed as long as an assay that meets high standards for analytical and clinical sensitivity and clinical specificity is used. Limitations of this approach are that a small percentage of test results may be false positive, so appropriate procedures to follow up individuals need to be in place.



Two-assay strategy   The test results of two different assays are used sequentially (i.e. not in parallel), to increase the positive predictive value of the overall test strategy. If the test result on the first-line assay is non-reactive, an “HBsAg negative” result is reported. However, if the test result on the first-line assay is reactive, a second test with an assay from a different manufacturer is performed. If both test results are reactive, the status is reported as: “HBsAg positive” with need for supplementary testing and follow-up HBsAg testing in 6 months to diagnose chronic infection recommended.” If the test result on the second-line assay is non-reactive, the result is reported as “HBsAg inconclusive; requires additional testing.” This strategy efficiently identifies most uninfected individuals and more definitively rules out chronic HBV infection than a one-assay testing strategy. It improves the positive predictive value when the test results of two different assays are both reactive. It can be used by non-laboratory staff, provided that adequate quality assurance standards are in place. The two-assay strategy may produce a small number of false-positive results (particularly in lowprevalence settings); some persons with recent HBV infection may receive false-negative results, which will depend on the analytical sensitivity of the assays used.



Worked example to illustrate the effect of prevalence on predictive values for the two different testing strategies Assuming the following assay performance characteristics:   If Assay 1 has sensitivity of 99% and specificity of 98%. If Assay 2 has sensitivity of 99.4 and specificity of 99.5%. Prevalence of analyte 0.1% Positive predictive values Assay 1 Assay 1 + Assay 2 (serial) Negative predictive values Assay 1 99.9% 99.99% 99.99% 4.7% 90.7% 33.3% 99% 84.6% 99.9% 1% 10%

Using the following equation for PPV and NPV that incorporates prevalence more correctly, Page | 84

Reference: Altman DG, Bland JM. Diagnostic tests 2: predictive values. BMJ. 1994 Jul 9; 309(6947): 102. Available at: https://www.ncbi.nlm.nih.gov/pmc/ articles/PMC2540558 /pdf/bmj004480038a.pdf

Page | 85

4.6. How to test - testing strategy HCV Decision-making tables – PICO 4 Testing strategy to ascertain exposure to HCV: Among persons identified for hepatitis C testing, what is the best testing strategy (diagnostic accuracy and other outcomes) for detection of antibodies to HCV? (One-assay versus two-assay strategy) 1. Topic for analysis: How to treat Population: Persons identified for hepatitis C virus (HCV) testing Intervention: One-assay testing strategy; one HCV serological assay Comparison: Two-assay testing; two different HCV serological assays Outcomes: Diagnostic accuracy (Sensitivity, specificity, positive predictive value, Negative predictive value, TN, TP, FN, and FP).

Page | 86

2. Background:

 

WHO recommends standardized testing strategies to maximize accuracy while minimizing cost and increasing simplicity. A testing strategy describes a testing sequence for a specific testing objective, taking into consideration the presumed prevalence of the analyte to be tested in the population. In both high- and low-prevalence settings, more than one serological assay may be required to establish exposure to HCV. Screening for exposure to HCV is dependent on assays that detect antibodies to HCV (anti-HCV) in the first instance. Once antibody status is confirmed, the patient will undergo supplementary testing for active HCV infection using an assay designed to detect viral replication, such as HCV RNA or core antigen (HCV cAg). It is important to note that the latest generation of assays designed to detect anti-HCV also are combined with cAg in order in increase sensitivity of the assay, but these “combo” assays should not be used to differentiate HCV exposure from active HCV infection. The question this PICO aims to address is whether one or two serological assays (anti-HCV or HCV Ag/Ab combo assays) performed sequentially are required in terms of specificity and positive predictive value in order to proceed to supplementary testing.



See footnotes for explanation of one-test and two-test strategies.

Page | 87

Figs 1–4: Possible testing strategies for detection of anti-HCV and diagnosis of active HCV infection

An -HCV one-assay strategy (HCV-exposure) An -HCV (RDT/EIA) (A1)

NonReac ve

Reac ve

Interpreta on: No serological evidence of exposure to HCV

Interpreta on: Compa ble with exposure to HCV Proceed to supplementary tes ng

Fig. 1. One-assay testing strategy for exposure to HCV (detection of anti-HCV)

An -HCV (RDT/EIA) (A1)

NonReac ve

Reac ve

An -HCV one-assay strategy for HCV exposure, with addi onal step for diagnosis of ac ve HCV infec on

HCV RNA or HCV cAg (A2)

Detected

Not detected

Interpreta on: No exposure to HCV

Interpreta on: Ac ve HCV infec on

Interpreta on: No ac ve HCV infec on

Link to care

Addi onal tes ng as appropriate

Fig. 2. One-assay testing strategy for diagnosis of HCV (detection of anti-HCV, followed by HCV RNA/core Ag)

Page | 88

An -HCV (RDT/EIA) (A1) NonReac ve

Reac ve

An -HCV two-assay strategy (HCV-exposure)

An -HCV (RDT/EIA)(A2)

Nonreac ve

Inconclusive status

Reac ve

Interpreta on: Compa ble with exposure to HCV Interpreta on: No exposure to HCV Proceed to supplementary tes ng

Addi onal tes ng as appropriate, e.g. request second specimen, consider HCV NAT

Fig. 3. Two-assay testing strategy for exposure to HCV (detection of anti-HCV) An -HCV (RDT/EIA) (A1) NonReac ve

Reac ve

An -HCV two-assay strategy for HCV exposure, with addi onal step for diagnosis of ac ve HCV infec on Nonreac ve Inconclusive status

An -HCV (RDT/EIA)(A2)

Reac ve

Addi onal tes ng as appropriate

HCV RNA or HCV cAg Not detected Likely past infec on with viral clearance OR if liver tests abnormal possible low VL. Repeat HCV RNA in 6 months if infec on s ll a concern

Detected Interpreta on: No exposure to HCV Ac ve HCV infec on

Fig. 4. Two-assay testing strategy for diagnosis of HCV (detection of anti-HCV, followed by HCV RNA/core Ag)

Page | 89

3. Draft recommendation(s):

Page | 90

4. Summary and quality of evidence Systematic review: A systematic review was commissioned in order to assess this PICO question. The purpose of this review was to determine from the published literature the evidence for a one-assay strategy for one-assay testing compared to two assays (sequential), before testing to diagnose active HCV infection is undertaken.

□ High □ Moderate □ Low □ Very low

Summary of results No study compared the diagnostic accuracy, cost, or cost-effectiveness of one- versus two-assay testing strategies for HCV exposure. The result of the PICO 2 review showed that the overall pooled clinical sensitivity and clinical specificity for HCVAb RDT versus EIA were 0.98 (95% CI 0.98–1.00) and 1.00 (95% CI 1.00–1.00). The overall pooled clinical sensitivity and clinical specificity for RDTs that use oral fluid compared to a serological reference standard using serum/plasma were 0.94 (95% CI 0.93–0.96) and 1.00 (95% CI 1.00–1.00).

See footnote for tables illustrating the impact of prevalence and one- versus twoassay strategies on PPV.

Quality of evidence   Study quality was not evaluated using the QUADAS-2 tool and the STARD checklist, as it was not applicable. None of the studies met inclusion criteria.

GRADE Summary of findings Not applicable as no studies met the inclusion criteria for the systematic review.

Predictive modelling analysis: (Parry, Public Health England)      There is a strong influence of prevalence and assay specificity on positive predictive value (PPV). At high prevalence, the use of a highly specific single screening test yields a PPV in excess of 0.99, and the ratio of true-positive (TP):false-positive (FP) results is high (164). At the lowest prevalence (e.g. 0.4%), PPV might be expected to fall below 0.5, yielding more false-positive diagnoses than true ones, particularly if the test specificity falls below 0.995. Even in middling prevalence, the PPV will struggle to exceed 0.9 should the specificity performance of the test employed fall below 0.99. The negative predictive value (NPV) is generally high except in high-prevalence populations (e.g. 45%) when a test with relatively poor sensitivity (<99%) is employed.

Page | 91



In all but the highest population prevalence, the use of less sensitive test kits in the two-test strategy had only a modest impact on NPVs. In all but the highest prevalence, the calculations demonstrate the potential for substantial gains in diagnostic accuracy in return for a modest investment in a second, independent, test kit to be applied to initially reactive individuals. For example, taking a model population of 100 000 with an anti-HCV prevalence of 2%, the PPV of an algorithm employing two independent test kits with modest sensitivities and specificities of 0.98 is improved from 0.500 to 0.980 in return for the supplementary use of <4000 of Test Kit B.

Predictive modelling analysis: (Linas, Boston University) To be added at meeting 5. Risks/benefits □ Benefits clearly outweigh harms □ Benefits and harms are balanced □ Potential harms clearly outweigh potential benefits

Benefits    The intervention of testing using a single test on one anti-HCV/core Ag assay will simplify the process of testing. Cost of overall testing may be reduced. May allow more rapid reporting of the result, so that the patient can be appropriately followed up for further assessment or given health protection advice.

Risks  Possible reduced PPV of a single test on one assay may lead to more falsepositives, especially in low-prevalence settings, with associated anxiety and cost associated with follow-up testing or treatment. Are the desirable anticipated effects large? □ No □ Probably □ Uncertain □ Yes □ Varies

Page | 92

□ No major

6. Acceptability, values and preferences variability A values and preferences survey of implementers and users of hepatitis B and C □ Major testing services was carried out by FIND in September 2015. A total of 104 variability respondents from 43 (20 high-income, 23 low- and middle-income) countries participated. Relating to this PICO,  50% of respondents to the values and preference survey from low- and middle- Is the option income settings (n = 23) preferred a one-assay screening strategy before acceptable confirmatory testing (compared to a one-assay strategy for diagnosis of HCV to key active infection). □ No □ Probably □ Uncertain □ Yes □ Varies

stakeholders?

Patients/caretakers  Patients require a simple and rapid testing strategy, yet one that is acceptably accurate for the purpose of testing.

Health-care workers   HCWs will need to understand the strengths and limitations of any given testing strategy and appropriately counsel patients who are screened. HCWs will need to appropriately act on results, either positive or negative.

Other  Laboratory staff needs to be appropriately trained in performance of the various assays according to the manufacturers’ instructions for use. □ Less equitable □ More equitable

Equity, ethics and human right implications Will recommendation raise questions around equity?  Yes, availability of HCV RNA NAT assays or other supplemental testing, such as core antigen testing, will be more easily available in settings with established hepatitis C testing facilities.

Are there ethical implications to this recommendation?

Page | 93

7. Resource use and financial implications Diagnostics/Laboratory monitoring  Are the resources required small?

Cost of a one-test vs two-assay strategy will be less due to fewer tests being used but the positive predictive value is higher and therefore the status will be more □ No accurately assigned.

Assay format RDTs EIA HCV Ag Quantitative NAT for HCV RNA Qualitative NAT for HCV RNA)

Indicative cost (US$) per test 0.50–2.00 (10 for oral fluid RDTs) 0.50–1.70 25–50 10–45 43–51

Source MSF, WHO WHO MSF MSF, UNITAID UNITAID

□ Probably □ Uncertain □ Yes □ Varies

8. Feasibility and constraints to implementation Are any major barriers expected for the implementation of this recommendation? • • • None, provided there is internal and country commitment to HCV testing Regional and country variability in access to treatment and procurement of testing equipment and services With regard to any diagnostic assay, availability of a local laboratory, which is able to procure the testing platform and reagents required for testing. Is the option feasible to implement?

□ No □ Probably □ Uncertain □ Yes □ Varies

Feasibility survey report to be presented at meeting

9. Relevance to different settings/populations Will this recommendation be most relevant for particular settings (e.g. endemicity)?

Page | 94

10. Rationale for recommendation:

11. Strength of recommendation

12. Implementation considerations

13. Research gaps   Development/implementation projects evaluating use of HCV core antigen as a one-stop diagnostic test Impact of one-test vs two-test screening on hepatitis C testing services

GRADE Summary of findings Footnote: Explanation of one- and two-assay strategies One-assay testing strategy  A single test is performed. If the test result is reactive, an “anti-HCV positive” status is reported. A specimen should be collected for supplemental laboratory testing to pursue the testing algorithm. If the initial test is non-reactive, an “anti-HCV negative” status is reported. This testing strategy efficiently identifies most uninfected individuals, it identifies those likely to be infected and in need of additional testing. It is suitable for resource-limited settings because one single test on one assay is performed as long as the assay used has high clinical sensitivity and clinical specificity. Limitations of this approach are that a small percentage of test results may be false positive, so appropriate procedures to follow up individuals need to be in place.

 

Two-assay testing strategy     The test results of two different assays are used sequentially (i.e. not in parallel), to improve the positive predictive value of overall testing strategy. If the test result for the first-line assay is non-reactive, an “anti-HCV negative” status is reported. If the test result for the first-line assay is reactive, a second test performed on an assay from a different manufacturer is performed. If both test results are reactive, the result is reported as: “anti-HCV positive for further diagnostic testing.” If the test result of the second-line assay is non-reactive, the result is reported as “anti-HCV inconclusive; requires additional testing”. This testing strategy efficiently identifies most uninfected individuals and more definitively rules out exposure to HCV than a one-assay testing strategy. It improves the positive predictive value when the test results for two different assays are both reactive. It can be used by non-laboratory staff, provided that adequate quality assurance measures are in place. Page | 95



The two-assay testing strategy may still produce a small number of false-positive results (particularly in low-prevalence settings); some persons with recent exposure may receive falsenegative results which will depend on the analytical sensitivity of the assays used

Modelling tables: Table 1: Outcomes for single-test strategy based on typical estimates of test accuracy

Page | 96

Table 2:

Outcomes for two-test strategy based on typical estimates of test accuracy (see Fig. 4)

Page | 97

4.7. How to test – confirmation of viraemia HCV Decision-making tables – PICO 5a Testing strategy for diagnosis of HCV active infection: Among individuals with confirmed exposure to HCV (HCV Ab positive), what is best testing strategy (diagnostic accuracy and other outcomes); comparing HCV core antigen versus NAT for HCV RNA to diagnose active HCV infection?

Decision-making tables – PICO 6 Testing strategy for diagnosis of HCV active infection (quantitative or qualitative NAT): Among individuals with confirmed exposure to HCV (Ab positive), what is the diagnostic test accuracy of qualitative NAT methods versus quantitative NAT methods to diagnose active HCV infection? 1. Topic for analysis: How to test PICO 5a PICO 6 Population: Persons with confirmed Population: Individuals with detectable HCV RNA with exposure to HCV (HCV seropositivity) or without confirmed exposure to HCV (HCV seropositivity) Intervention: Qualitative HCV RNA NAT Intervention: HCV core antigen Comparison: HCV RNA NAT Comparison: Quantitative HCV RNA NAT

Outcomes: Diagnostic accuracy (Sensitivity, Specificity, Positive predictive value, Outcomes: Diagnostic accuracy (Sensitivity, Negative predictive value, TN, TP, FN, and Specificity, Positive predictive value, Negative FP). predictive value, TN, TP, FN, and FP).

2. Background:  Determination of exposure to HCV through detection of antibodies to HCV (anti-HCV) is a commonly used first-line diagnostic tool to identify those who are infected with HCV who might benefit from antiviral treatment, yet diagnosis of active HCV infection requires evidence of active viral replication, traditionally ascertained by the detection of HCV ribonucleic acid (RNA) by nucleic acid testing (NAT). Detection of HCV RNA may also be used as the preferred first-line investigation in individuals who may be persistently seronegative due to underlying immunosuppression, e.g. poorlycontrolled HIV infection, renal dialysis. Qualitative NAT for HCV allows for detection of the virus as well as evidence of the level of HCV RNA circulating in the peripheral blood falling below a clinically relevant threshold, i.e. a qualitative measurement. Page | 98



 

Assays to detect HCV-RNA have been developed for the near point-of-care (POC) setting. NAT assays can be very sensitive and specific but more costly than serological methods such as HCV core antigen testing, and require sophisticated laboratory equipment and therefore skilled staff. HCV core antigen (HCVcAg) testing was developed as an alternative to NAT for the diagnosis of active HCV infection. HCV nucleocapsid peptides 22 (p22) are released into plasma during viral assembly and can be detected throughout the course of HCV infection. There are also several assays that have been commercialized for stand-alone detection of HCV cAg as a replacement to NAT. Furthermore, HCV cAg is detectable earlier than antibodies and therefore detection of HCV cAg has also been applied as an additional analyte for serological assays for use to ascertain exposure to HCV as a combination HCV Ag/Ab (4th generation) assay. The addition of cAg in combination was intended to increase sensitivity of the assay in early infection. Although the output may be interrogated in order to ascertain whether antigen, antibody or both were detected, the purpose of the assays is not to differentiate seropositivity from active infection. Both HCV cAg and HCV RNA (either qualitative or quantitative detection) have been shown to have clinical utility for the detection of active HCV infection though there is scarce research comparing the HCVcAg, and qualitative and quantitative NAT for this purpose.



3. Draft recommendation(s):



Summary and quality of evidence

1. Distribution of viraemia in persistent infection: A threshold of >10 000 IU/mL will capture 95% of persistent infections (except, temporarily, for a minority of those with partial viral control) between 5 and 12 months post infection. In 95% of cases of hepatitis C infections: • Those with persistent infection and viral plateau will have a viral load (VL) >100 000 IU/mL at month 5 and remain at least >10 000 IU/mL between months 5 and 12. Those with persistent infection but partial viral control will also have >100 000 IU/mL at month 5 and remain having a VL at least >1000 IU/mL temporarily, going back to a viral load >100 000 IU/mL between months 10 and 12. Between 2 and 3–4 months after infection, acute infection (giving VLs up to >100

□ High □ Moderate □ Low □ Very low

•

Page | 99

000 IU/mL) may be captured that would spontaneously resolve but would likely be antibody negative.

Publications are: Hajarizadeh B, Grady B, Page K, Kim AY, McGovern BH, Cox AL et al. Patterns of hepatitis C RNA levels during acute infection: the C3 study. PLOS One. 2015;10 (4):e 0122232.

Hajarizadeh B, Grebely J, Applegate T, Matthews GV, Amin J, Petoumenos K et al. J Med Virol. 2014;86 (10):1722–9. Glynn SA, Wright DJ, Kleinman SH, Hirschkorn D, Tu Y, Heldebrant C et al. Transfusion. 2005;45 (6):994–1002.

2. Systematic reviews of the evidence: Two systematic reviews were commissioned to address the above PICO questions. These aimed to summarize (1) the diagnostic accuracy of HCV cAg testing (PICO 5a), (2) the diagnostic accuracy of qualitative versus quantitative RNA for diagnosis of active HCV infection (PICO 6).

Summary of results of HCV cAg review Diagnostic accuracy of HCV cAg for diagnosis of active infection (Fig. 1)  7 assays utilizing HCV core antigen were assessed. Note that two of these were HCV antigen/antibody combination assays and not designed to differentiate active infection from seropositivity, but are included as the antigen/antibody components of the assays were reported separately. Effect Accuracy (95% CI) Effect LR

n Index test (samples)

Unit of analysis

Sensitivity

Specificity

Positive LR

Negative LR

Abbott ARCHITECT

20 (11,820)

Sample

93.4% (88.7, 96.2)

98.7% (96.9, 99.4)

71.8 (28.6, 160.3)

0.07 (0.04, 0.12)

Page | 100

HCV Ag Assay

Ortho ELISA-Ag

5 (1,177)

Sample

93.2% (81.6, 97.7)

99.2% (87.9, 99.9)

116.5 (6.7, 977)

0.06 (0.02, 0.07)

Bio-RAD Monolisa HCV Ag-Ab ULTRA

5 (525)

Sample

28.6–95%*

94.9% (89.9, 99.8)**

EIKEN Lumispot HCV Ag

2 (235)

Sample

97.5–98.1%*

Fujirebio Lumipulse Ortho HCV Ag

1 (80)

Sample

95% (90.2, 99.8)**

Hunan Jynda HCV Core Ag ELISA

4 (524)

Sample

59.5% (46, 71.7)

82.9% (58.6, 94.3)

3.5 (1.1, 12.6)

0.28 (0.2, 0.3)

DiaSorin S.A. Murex Ag/Ab EIA

4 (730)

Sample

50–100%*

83.8–100%*

n: study number, CI: confidence interval; LR: likelihood ratio; ND: no data, NA: not applicable — if sensitivity and specificity results were not available from meta-analysis, likelihood ratios were not calculated. * Meta-analysis not possible. Range of results seen across studies reported. **Result from one study only.

Limits of detection of HCV cAg assays  The limit of detection for the most evaluated and best performing assay (Abbott Architect) is 3 fmol/L HCV cAg or 0.06 pg/mL, which equals to a limit of detection of about ~1000–3000 IU/mL with a NAT. Impact of findings in different prevalence settings (for assays where a meta-analysis was possible) (Fig. 2)

Page | 101

Effect per 1000 patients with presumed HCV for varying Outcome prevalence settings comparing HCV core Ag against HCV RNA Prevalence 2%* Abbott ARCHITECT HCV Ag Assay TP TN FP FN Ortho ELISA-Ag TP TN FP FN 19 (16, 20) 970 (862,980) 10 (0, 118) 1 (0,4) 93 (82, 98) 891 (792, 900) 9 (0, 108) 7 (2, 18) 279 (246, 294) 693 (616, 700) 7 (0, 84) 21 (6, 54) 19 (18, 19) 967 (951, 974) 13 (6, 29) 1 (1, 2) 93 (89, 96) 888 (873, 895) 12 (5, 27) 7 (4, 11) 279 (267,288) 691 (697, 696) 9 (4, 21) 21 (12, 33) Prevalence 10%* Prevalence 30%*

Hunan Jynda HCV Core Ag ELISA TP TN FP FN 12 (9, 14) 813 (578, 921) 167 (59, 402) 8 (6, 11) 60 (46, 72) 747 (531, 846) 153 (54, 369) 41 (28, 54) 179 (138, 216) 581 (413, 658) 119 (42, 287) 122 (84, 162)

TR: true positives (individuals with active HCV); TN: true negatives (individuals without active HCV); FP: false positives (individuals incorrectly classified as having active HCV); FN: false negatives (individuals incorrectly classified as not having active HCV) *Numbers in parentheses consider 95% confidence interval of accuracy estimate.

Limits of detection of hepatitis C NAT assays This systematic review shows that for diagnosis of active HCV infection, the lower limit of detection of most commercial qualitative assays was in the 10–15 IU/mL range measured against a WHO standard, whereas the lower limit of detection for quantitative assays is at 600–1100 IU/mL.

Additional information: Newer quantitative viral load assays report limits of detection similar to qualitative viral load but might not quantitate results at that level. NAT are in the pipeline for detection on capillary whole blood. Furthermore, detection from dried blood spot is also promoted to increase access. For both those strategies, the limit of detection is likely to be substantially higher (e.g. in the range of 2000 IU/mL) than for existing assays (because of volume tested and technical feasibility).

Page | 102

Fig. 3: Limit of detection of qualitative vs quantitative NAT Limits of detection Study Lee 2000 Qualitative NAT (IU/mL) COBAS TM AMPLICOR HCV Test v2.0 assay. Roche 15 95% 25 95% >50 100% Quantitative NAT (IU/mL) AMPLICOR HCV test, v 2.0. Roche Sarrazin 2008 CAP/CTM Roche 100% – – 300 57% 600 95% 1100 100%

RealTime HCV assay Abbott

87%

Lee SC, Antony A, Lee N, Leibow J, Yang JQ, Soveiro S et al. Improved version 2.0 qualitative and quantitative AMPLICOR reverse transcription-PCR tests for hepatitis C virus RNA: calibration to international units, enhanced genotype reactivity, and performance characteristics. J Clin Microbiol. 2000;38 (11):4171–9. Sarrazin C, Dragan A, Gartner BC, Forman MS, Traver S, Zeuzem S et al. Evaluation of an automated, highly sensitive, real-time PCR-based assay (COBAS Ampliprep/COBAS TaqMan) for quantification of HCV RNA. J Clin Virol. 2008;43 (2):162–8.

Conclusions:  HCV core Ag assays can have high sensitivity (up to 93.4% for certain commercialized assays), high specificity, and good correlation with HCV RNA to a detection limit of roughly 1000–3000 IU/mL. NAT on plasma and serum are able to achieve higher sensitivity than cAg. Qualitative assays from published data are more sensitive than quantitative assays. This might not be the case for newer assays.



Quality of evidence *Refer GRADE table in footnote Predictive modelling (Linas, Boston University) To be added at meeting Page | 103

4. Risks/benefits cAG vs NAT

Benefits    HCV cAg testing by immunoassay format has the potential to be less costly given that the cost of goods is lower. Also HCV cAg is more stable and thus does not require a cold chain. Further assessment and treatment may be administered more promptly to individuals when diagnosis of active HCV infection is undertaken in a more decentralized manner. There are both HCV cAG and NAT tests in the pipeline that might be possible on capillary blood at the point of care. However, current cAg immunoassays still require sophisticated laboratory equipment, electricity and therefore skilled staff to operate, while there are already some NAT assays available that can be done near the patient (however, still require plasma).

□ Benefits clearly outweigh harms □ Benefits and harms are balanced □ Potential harms clearly outweigh potential benefits



HCV cAg assay can conceivably be used in a one-step testing strategy (i.e. without an antibody test) particularly in high-prevalence settings. With such a strategy, patients can be identified earlier in their infection (than with antibody testing), □ No □ Probably results may be available faster to the patient and provider, resulting in less loss to □ Uncertain follow up and faster treatment initiation. □ Yes □ Varies

Are the desirable anticipated effects large?

Risks  The patient with low-level viraemia (<3000 IU/mL) could be missed with HCV cAg assays or HCV RNA NAT with lower sensitivity. The clinical implications for the individual person and on a population level are not well understood. However, if the test does reach more patients (e.g. because it can be done on capillary blood or is less costly), then this risk might be outweighed by the benefits. Although data exist on the utility of HCV cAg in seronegative individuals, no studies examined the use of HCV cAg and HCV RNA NAT in diagnosing HCV infection in key affected populations, in a one-step testing strategy rather than using an antibody assay. Such a strategy would only be cost–effective in certain high-prevalence settings.



Qual versus Quant NAT Benefits  Qualitative NATs generally are more or at least as sensitive than quantitative assays.

Page | 104



The cost of a qualitative NAT assay may be lower than that of a quantitative assay.

Risk  Recent publications suggest that monitoring for HCV will not be necessary with direct-acting antivirals (DAA) (as viral load at EVR is not predictive of cure). Therefore, quantitation at baseline will not be necessary. This is currently not yet widely confirmed or reflected in guidelines.

□ No major

5. Acceptability, values and preferences variability A values and preferences survey of implementers and users of hepatitis B and C □ Major testing services was carried out by FIND in September 2015. A total of 104 variability respondents from 43 (20 high-income, 23 low- and middle-income) countries. Relating to this PICO,  As assay (platform) for detection of HCV cAg is currently available in India, Is the option Indonesia, Macedonia, Viet Nam, Turkey (although the representativeness of acceptable these data are limited). The platform is available in South Africa but not currently to key being used for HCV cAg detection. 50% found a diagnostic sensitivity of >95% acceptable, particularly if the test cost is lower, therefore increasing access to testing (respondents felt that then the test cost should be less than US$ 10 (83%) with a sensitivity of 95%. For a test with a sensitivity of 98%, forty-one per cent of respondents would accept a cost of US$ 11–20). Free text comments suggested that HCV cAg was easier to do. A larger number of respondents felt that the cost of testing for HCV RNA by NAT was considered more of a barrier than that of HCV cAg. 47% of respondents in low- and middle-income countries favoured a decentralized test and a test on capillary blood even at the cost of sensitivity. Also, 50% of patients preferred a test result in <2 hours (which could only be achieved at the point of care) while 27% found a result on the same day acceptable.



stakeholder? □ No □ Probably □ Uncertain □ Yes □ Varies

   

Patient: • Patients at risk of progressive liver disease will benefit from reduced disease progression and related mortality and morbidity, if treatment is provided as a result of wider access to testing programmes.

Page | 105

Community: • • To identify the individuals who require assessment and treatment would be an effective use of resources. As testing and treatment programmes are scaled up, the numbers developing progressive disease and serious outcomes (HCC and complications of advanced liver disease), premature morbidity and mortality within the community will be reduced, and so also the burden of disease to societies where the disease is most prevalent.

Health-care workers: • • • Appropriate use of resources to channel treatment to patients with higher risk of complications in the medium- and short term Will require training in the use of testing equipment if being used in the near-POC setting Appropriate reporting and recording of results.

Laboratory: • Will require training for HCV cAg test: careful sample processing is necessary for HCV cAg assay to lyse viral particles, expose antigen and dissociate antibody from antigen and optimize the detection for HCV cAg.

6. Equity, ethics and human right implications Will the recommendation raise questions around equity? • • Equity will improve as a result of decreased cost and increased decentralization of testing; however, still improvement of access to testing facilities is necessary. Regional and country variability in access to treatment.

□ Less equitable □ More equitable

Are there ethical implications to this recommendation? • Ethical consideration for the possibility that WHO could recommend a testing strategy.

7. Resource use and financial implications Input from modelling team

Are the resources required

Page | 106

Materials/equipment:   Cost of testing platform and reagents Other laboratory consumables

small? □ No □ Probably □ Uncertain □ Yes □ Varies

Training and supervision:    Appropriate training of laboratory staff Quality control programmes If using near-POC assays, appropriate training of testing providers.

Other:  Cost of transportation of specimens to the laboratory

Possible procurement costs (Fig. 4):

Cepheid: <20 NAT. With volume-based pricing down to~15 Note that new POC or near-POC assay platforms for HCV RNA are in development. See figures at the end of the table.

Page | 107

Feasibility and constraints to implementation Are any major barriers expected for the implementation of this recommendation? • • • Is the option feasible to implement?

None, provided there is internal and country commitment to HCV testing. Regional and country variability in access to treatment and procurement of testing equipment and services □ No With regard to any diagnostic assay, availability of a local laboratory, which is able □ Probably □ Uncertain to procure the testing platform and reagents required for testing. □ Yes □ Varies

Feasibility survey report to be presented at meeting.

8. Relevance to different settings/populations Will this recommendation be most relevant for particular settings (e.g. endemicity)?    Currently available HCV cAg testing will be more relevant to populations that presently rely on centralized laboratory testing for HCV RNA for confirmation of status. One-step strategies for testing with HCV cAg or NAT may be cost-effective only in highprevalence settings. The recommendations are less likely to be relevant in high-income settings where there is already access to established hepatitis C testing and treatment programmes.

9. Rationale for recommendation:

10. Strength of recommendation

11. Implementation considerations • Optimize test for asymptomatic patients in primary-care settings or in the community where the HCV endemic is high. 12. Research gaps • Development/implementation projects evaluating use of HCV core antigen or HCV RNA as a onestep diagnostic strategy. • Surveillance data: how many patients are missed by assays that have limits of detection of 2000 IU. • Outcomes of patients with low viral loads • More information on patients with high viral loads and negative HCV cAg to inform the

Page | 108

• • •

optimization of antigen detection. The kinetics of HCVcAg with treatment needs to be evaluated further, particularly in the context of new DAA regimens. More rigorous assessment of covariates in accuracy studies is required, such as HIV or HBV coinfection or genotype (particularly genotypes 5 and 6 where there are limited data). Development of muliplex instrument with other disease diagnosis with HIV, HBV, and TB at health centre.

Page | 109

World Health Organization Global Hepatitis Programme

GRADE Summary of findings Table: Strength of evidence for diagnostic accuracy SR outcome: diagnostic accuracy Quality Strength of evidence

Index Test

Outcome Measure

# Studies (# samples)

Design

Risk of Bias

Inconsistency

Indirectness

Imprecision

Abbott ARCHITECT HCV Ag Assay

Sensitivity

30 (12 788)

Cohort and cross-sectional

Low

Moderate (–1)

Low

Moderate ⨁⨁⨁◯

Specificity

20 (11 820)

Cohort and cross-sectional

Low

Moderate (–1)

Low

Moderate ⨁⨁⨁◯

Ortho ELISA-Ag

Sensitivity

6 (1 423)

Cohort and cross-sectional

High (–2)

Moderate (–1)

Low

Very low ⨁◯◯◯

Specificity

5 (1 177)

Cohort and cross-sectional

High (–2)

Moderate (–1)

Low

Very low ⨁◯◯◯

Bio-RAD Monolisa HCV Ag-Ab ULTRA:

Sensitivity

5 (525)

Cohort and cross-sectional

Low

High (–2)

Moderate (–1)

Low

Very low ⨁◯◯◯

Page | 110

Sensitivity

Specificity

1 (337)

Cross-sectional

Moderate (–1)

Moderate (–1) 2

Very low ⨁◯◯◯

(–1) 1

EIKEN Lumispot HCV Ag: Sensitivity

Sensitivity

2 (235)

Cross-sectional

Moderate (–1)

Low

Moderate (–1)

Low

Low ⨁⨁◯◯

Specificity

NA 1

NA 2

NA 3

NA 4

Fujirebio Lumipulse Ortho HCV Ag: Sensitivity

Sensitivity

1 (80)

Cross-sectional

Moderate (–1)

Very low ⨁◯◯◯

(–1)

Specificity

NA 1

NA 2

NA 3

NA 4

Hunan Jynda HCV Core Ag ELISA

Sensitivity

4 (524)

Cohort and cross-sectional

Moderate (–1)

High (–2)

Moderate (–1)

Low

Very low ⨁◯◯◯

Specificity

4 (524)

Cohort and cross-sectional

Moderate (–1) 1

High (–2)

Moderate (–1)

Low

Very low ⨁◯◯◯

DiaSorin S.A. Murex Ag/Ab EIA

Sensitivity

4 (770)

Cohort and cross-sectional

Low

High (–2)

Moderate (–1) 2

None

Very low ⨁◯◯◯

Specificity

Cohort

Low

Moderate

Low

Page | 111

(658)

(–1)

⨁⨁◯◯

NA: not applicable Footnotes: For each index test, quality of evidence started high when there were several high-quality observational studies (prospective cohort studies, cross-sectional studies with direct comparison of index test results with a reference standard). We then downgraded one point when there was moderate concern identified and two points when a there was a high concern identified in any of the four factors that may decrease the quality of evidence: risk of bias, inconsistency, indirectness, and imprecision. 1

We used QUADAS-2 to assess risk of bias.  For ARCHITECT, in half of the studies it was unclear how participants were selected and one study used only healthy blood donors, however, the data from all studies are consistent and unclear selection does not appear to cause bias thus we did not downgrade.  For the Ortho ELISA, two studies of five used convenience enrolment for participant selection, and one enrolled only healthy blood donors thus we downgraded 2 points.  For the Monolisa, four of five studies had unclear patient selection. For one it was unclear if the index and reference tests were performed within 30 days. Given that there were no high-risk concerns for bias we did not downgrade. For specificity, there was only one study with data that had unclear participant selection, thus we downgraded one point as there were no data from studies with random or consecutive selection to compare to and identify possible selection bias (as was possible with the ARCHITECT).  For the Lumispot, both studies had unclear patient selection. As there were no data from studies with random or consecutive selection to compare, we downgraded one point.  The Lumipulse only included one study with unclear participant selection and was downgraded one point.  The Hunan Jynda had one of four studies with unclear participant selection, one in only healthy blood donors, and one for which it was unclear whether the index and reference were performed within 30 days. As the use of only healthy blood donors was considered a high-risk category, in combination with the other unclear factors, we downgraded one point.  For the Murex test, three of four studies had unclear participant selection but no other high-risk concerns for bias and thus we did not downgrade.

Page | 112

Unexplained heterogeneity in the remaining studies may be related to covariates that could not be adjusted for in meta-regression due to limited data (HIV and HBV coinfections, HCV genotype). Additionally, not all studies identified HCV antibody status or stratified by acute and chronic infection thus variability of HCV replication could contribute to higher false-negative HCV cAg.  There was little heterogeneity noted in the ARCHITECT studies thus we did not downgrade.  For the Ortho ELISA, there was moderate heterogeneity with largely one outlier study, thus we downgraded 1 point.  For the Monolisa sensitivity outcome, heterogeneity between studies precluded meta-analysis and thus we downgraded 2 points. For specificity, there is only 1 study and we cannot assess heterogeneity and downgraded 1 point.  For the Murex sensitivity outcome there was too much heterogeneity to pool the data, and thus we downgraded 2 points. For specificity, there were not enough studies to perform meta-analysis and heterogeneity could not be formally assessed; however, there is a broad range among results and thus we downgraded one point.  The EIKEN Lumispot was only used in 2 studies. Sensitivity was similar in both studies suggesting little heterogeneity, thus we did not downgrade.  For the Fujirebio Lumipulse, there is only 1 study and we cannot assess heterogeneity and downgrade 1 point.

All studies were performed in reference laboratories, and the majorities were in high- and middle-income countries. Thus the patient population, the viral population tested (e.g. genotype distribution), and the test users are not representative of the limited-resource settings for which these guidelines are envisioned. All were downgraded 1 point. We considered imprecision as present when the pooled confidence intervals were >10% and when there were fewer than 250 samples in the analysis. As such, we downgraded the Ortho ELISA, and Hunan Jynda one point for wide confidence intervals, and downgraded the Lumispot one point for small sample size. Additionally, imprecision could not be graded for the Monolisa specificity outcome, and the Lumipulse test as these only included one study.

Page | 113

World Health Organization Global Hepatitis Programme

Fig. 1: POC HCV platforms available within the next 2 years

Page | 114

World Health Organization Global Hepatitis Programme

Fig. 2 and Table 2: Roche Molecular Systems (1) Qualitative Roche COBAS AmpliPrep/COBAS TaqMan HCV Qualitative Test v.2 Roche COBAS Abbott RealTime HCV AmpliPrep/COBAS TaqMan Assay HCV Quantitative Test v.2 VERSANT HCV RNA 1.0 Assay (kPCR) HCV Real-TM Quant Dx Assay Artus HCV QS-RCQ Kit VERIS MDx RT-TMA Technology for the Panther® System Abbott Diagnostics (2) Siemens Healthcare Diagnostics Inc. (3) Sacace Biotechnologies (4) QIAGEN (5) Beckman Coulter (6) Hologic Inc (7)

Quantitative

Lower limit of Detection

15 IU/mL

12 IU/mL –0.5 mL sample; 30 IU/mL –0.2 mL sample Plasma or serum

15 IU/mL

13 IU/mL with 1 mL sample

35 IU/mL with automated extraction

Sample type

Plasma or serum

Plasma

Plasma, serum and culture

Plasma

Cost/test (US$)

36–38/43–51

13–35

72–100

>20

16–45 Pricing for the assay and instrument is available from Qiagen. Continuous loading in batches of up to 24 samples plus internal controls

10–15 Pricing for the assay and Pricing for the assay and instrument is available instrument is available from Qiagen. from Qiagen. 48 samples can be lined up on 12 racks. DNA tests is approximately 70 minutes and for RNA tests is 110 minutes First reportable results available within three hours after loading samples and five results after every five minutes thereafter. Samples can be continuously loaded, with up to 120 samples on the Panther® System

Price of instrument 80 000–100 000

248 000 Pricing for the assay and 113 000 (45 000 + 162 000 + 80 instrument is available (95 000+18 645) 000) from Siemens.

# Specimen/run

24 specimens in two hours, 96 samples at a time in 89 samples per run with 5–6 hours for a run of 24 but it can process up to 72 about three hours of a total time to result of samples samples at one time cycling time less than six hours

Page | 115

World Health Organization Global Hepatitis Programme

Qualitative/Quantitative HCV RNA platforms currently available (1–5) and soon to be available (6– 7)

Page | 116

4.8. Dried blot spots Decision-making tables – PICO 7 Dried blood spots as sample collection method for serology/NAT for HBV/HCV: Among persons identified for (1) hepatitis B, or (2) hepatitis C testing, what is the diagnostic accuracy and impact of detecting HBsAg/HCV Ab or NAT from DBS samples versus venous samples? 1. Topic for analysis: How to test Population: • • • • Samples for serology for HBV (HBsAg) Samples for serology for HCV (HCV Ab) Samples for HBV DNA Samples for HCV RNA

Intervention: Using dried blood spot (DBS) samples Comparison: Using venous samples Outcomes: Diagnostic accuracy (Sensitivity, specificity, positive predictive value, negative predictive value, TN, TP, FN, and FP)

2. Background: In high-prevalence of hepatitis B virus (HBV) among low- and middle-income countries (LMICs) there is a need for improved HBV screening, especially in decentralized settings. And the entry of new alloral direct acting antiviral therapy for hepatitis C provides an opportunity to scale up HCV care in LMICs and dramatically simplify diagnosis and monitoring. In HIV, use of DBS has facilitated the diagnosis of HIV in children under 18 months and is a promising tool for HIV management in resource-limited settings (viral load monitoring is strongly recommended by WHO as the preferred ART monitoring test). While commercial rapid diagnostic tests (RDTs) exist for HBsAg and HCV antibody for HBV diagnosis and HCV screening, respectively, the use of DBS sent to centralized lab facilities for diagnosis or screening purposes may be useful in certain contexts where RDTs are not available or not feasible due to human resource, procurement, quality, regulatory or other constraints, particularly as they can be prepared from capillary blood, thus obviating the need for phlebotomy services. Molecular tests for HBV DNA and HCV RNA must currently be performed in centralized facilities, where barriers to sample collection and transport, such as phlebotomy services, plasma separation Page | 117

and cold chain, may make testing less feasible. One decentralized molecular platform for HCV RNA does exist but it too relies on plasma as a sample type. Thus DBS may be used in a similar way to HIV to facilitate diagnosis and monitoring in certain contexts. Two main advantages of using DBS for HCV compared to HIV are that (1) there is no proviral DNA to overestimate the quantity of virus in the blood compared to plasma (although cell-associated RNA may still contribute) and (2) qualitative testing should suffice, both for diagnosis and measurement of SVR, as (i) viral load monitoring of DAA-based treatment is not useful and, (ii) since DAA therapy is non-toxic, a log drop calculation for treatment continuation at week 12 (EVR), as was the case for IFN-based therapy, is not needed. In prioritizing the validation of serological testing on DBS versus NAT testing on DBS, depending on the context, they may be equally useful. If affordable, good-quality RDTs are available that can be performed off capillary blood then the impetus may be to prioritize that validation of NAT testing for HBV DNA and HCV RNA. However, if RDTs are not available then DBS testing may be equally important to increase access to serological testing, for example, for hard-to-reach populations and those with poor venous access. Equally, DBS may be useful where polyvalent screening for multiple diseases, such as HIV/HBV/HCV, is useful and where multiplex RDTs for this purpose are not available or more costly. Thus the choice and combination of test may be context specific whereby different programmes may opt for difference combinations of: (1) DBS serology + DBS NAT (remote settings), (2) RDT serology + DBS NAT (clinics, e.g. antental), or (3) EIA serology + plasma-based NAT (urban settings or more central hospitals).

See Feasibility section for current use of DBS.

3. Draft recommendation(s):

4. Summary and quality of evidence Summary of results 7a: Meta-analysis for HBsAg No. of studies included 10 (SR) 9 (meta-anal)

□ High □ Moderate □ Low □ Very low

Total sample size Overall sensitivity Overall specificity

2481 92.9% [86.2–96.5] 99.0% [96.2–99.7]

Page | 118

Positive LR Negative LR Impact of cut-off

92.9 0.072 DBS LOD cut-off may be higher due to smaller sample volume (50 μL) but OD cut-off lower to maximize sensitivity (ROC curve needed to set cut-off) Std: SN 88% [74–95] SP 98.6% [89–100] High: SN 95.6% [91–98] SP 99.1% [97–100]

Impact of storage

Cold chain: SN 78.7% [70–85] SP 98.6% [68–100] ≥RT: SN 96.1% [92–98] SP 99.7% [98–100]

Impact of duration of storage Impact of assay

Accuracy not affected if RT –33°C for ≤15 d (1 study) or 63 d (another study) DBS provides good rule in test for diagnosing HBV but may not be able to rule out HBV in a minority of cases.

LR: likelihood ratio

7b: Meta-analysis for HCV Ab No. of studies included 18 (SR) 14 (meta-anal)

Total sample size Overall sensitivity Overall specificity Positive LR Negative LR Impact of cut-off

6120 98% [94–99] 99% [97–100] 171 0.02 None identified (varied widely, some used ROCs to set own cut-offs) –20°C = less variation compared to RT

Impact of storage

Page | 119

Impact of duration of storage Impact of assay

Accuracy not affected if RT for ≤3 d (1 study) or ≤6 d (another study) or ≤60 d (another study) DBS provides good rule in and rule out test for screening for HCV.

7c: Meta-analysis for HBV DNA No. of studies included 9 (SR) 7 (meta-anal)

Total sample size Overall sensitivity Overall specificity Positive LR Negative LR Impact of threshold

905 96% [91–98] 100 [54–100] 287 0.04 LOD: 914 IU/mL (one study), 3000–4000 IU/mL (other studies) Clinical Tx threshold: 2000 IU/mL

Impact of storage

Not possible to calculate because all accuracy studies at –20°C. No affect of accuracy if 4–37°C for ≤7 d (2 studies).

Impact of duration of storage Impact of assay

DBS good to rule in HBV but may not rule out HBV in a minority of cases, particularly if viral loads are <3000 IU/mL.

7d: Meta-analysis for HCV RNA

Page | 120

No. of studies

included

9 (SR & meta-anal)

Total sample size Overall sensitivity Overall specificity Positive LR Negative LR Impact of threshold Impact of storage

1250 96.0% [93.4–97.6] 97.7% [94.7–99.0] 41.74 0.041 LOD: ≥150–250 IU/mL Better result at –20°C compared to RT; conflicting results re deterioration of sample at RT Conflicting results re deterioration over time

Impact of duration of storage Impact of assay

DBS good to rule in active HCV infection but may not rule out infection in a minority of cases with lower viral loads.

Quality of evidence     Overall HBsAg: moderate (no significant indirectness, imprecision or inconsistency but significant risk of bias). Overall HCV Ag: moderate (no significant indirectness, imprecision or inconsistency but moderate risk of bias). Overall HBV DNA: low (no significant indirectness or imprecision but significant inconsistency and high risk of bias). Overall HCV RNA: moderate (no significant indirectness, imprecision or inconsistency but significant risk of bias).

*Refer GRADE table in footnote □ Benefits clearly outweigh harms □ Benefits and harms are balanced

5. Risks/Benefits Benefits of DBS

Page | 121

  

DBS is likely stable over time and maintains good accuracy in conditions with □ Potential harms clearly outweigh higher temperatures and with higher humidity. Lower biohazard, easier transport, no venepuncture needed potential benefits

Greater access to testing, especially for remote settings or specific programmes servicing key populations (people who inject drugs, prisoners, Are the desirable etc). anticipated effects large?

Risks of DBS  Accuracy may be negatively affected when storing for prolonged durations (greater than 14 days) at higher temperatures (room temperature and above) and humidity. Use of DBS may require higher cut-offs to determine test positivity as DBS uses a small volume of blood, in order to maintain sensitivity, a higher cut-off may be required as compared to when using plasma samples. This may decrease the ability to rule out HBV/HCV in a minority of cases. DBS not regulatory approved as a sample type for HBsAg, HCV Ab, HBV DNA or HCV RNA testing. No guidance from manufacturers on use of commercial assays with DBS. The best type of filter paper to use is not known (but has been established for HIV DNA and RNA testing). Differences between capillary versus venous blood are not known (but has been established for HIV DNA and RNA testing with insignificant difference where training and proficiency testing are provided for capillary blood sampling). Most appropriate volume of capillary blood to be used (e.g. one versus more than one spot of 50 μL) is not known (but has been established for HIV DNA and RNA testing). Best types of commercial test kits to use with DBS are not known.

□ No □ Probably □ Uncertain □ Yes □ Varies



  



6. Acceptability, values and preferences 1. Acceptable accuracy of DBS for testing HBsAg and HCV Ab as compared to use of plasma samples (as compared to WHO performance acceptance criteria): HBsAg WHO EIA Sensitivity Specificity 100 ≥98 WHO RDT 100 ≥98 SR 92 99 Acceptable No Yes

□ No major variability □ Major variability

Is the option acceptable to key stakeholders? □ No □ Probably □ Uncertain

Page | 122

EIA: enzyme immunoassays; SR: systematic review

□ Yes □ Varies

HCV Ab WHO EIA Sensitivity Specificity 100 ≥98 WHO RDT ≥98 ≥97 SR 98 99 Acceptable Yes Yes

Accuracy is acceptable except for sensitivity of HBsAg testing, which may lead to a minority of cases being missed.

2. Acceptable accuracy of DBS for testing HBV DNA and HCV RNA as compared to use of plasma samples is not known because there are no WHO performance acceptance criteria. However accuracy should be measured in the context of clinical relevance, i.e.: For HBV DNA: 1. Treatment thresholds of 2000 and 20 000 IU/mL 2. Ability to measure suppression in treatment monitoring (threshold not known but most people fail at viral loads >20 000 IU/mL).

For HCV RNA: 1. Confirming all those with chronic and active HCV infection (threshold not known but viral loads are generally high, with 95% of people having viral loads >1000 IU/mL if chronically infected and those with lower early viral loads more likely to clear infection). Following new DAA therapy, confirming those with sustained virological response (SVR), whether at SVR 12/24 (threshold is not known but preliminary evidence suggests those failing therapy have higher viral loads).

7. Equity, ethics and human right implications Will the recommendation raise questions around equity? DBS may allow more equitable access to testing.

□ Less equitable □ More equitable

Page | 123

Are there ethical implications to this recommendation? No. 8. Resource use and financial implications   May reduce costs associated with sample collection, storage and transport. May facilitate task shifting to lay workers to decrease human resource limitations. □ No Are the resources required small?

□ Probably □ Uncertain □ Yes □ Varies

9. Feasibility and constraints to implementation 10. Relevance to different settings/populations Are any major barriers expected for the implementation of this recommendation? Will this recommendation be most relevant for particular settings (e.g. endemicity)? Is the option feasible to implement? □ No □ Probably □ Uncertain □ Yes □ Varies

There is limited programmatic experience with the use of DBS for viral hepatitis testing, although this is well established for HIV DNA and RNA testing. 1. Studies are limited. 2. There are no current standards for test OD cut-offs for HBsAg or HCV Ab using DBS. 3. There is a dearth of studies that systematically examine the effects of storage and transport conditions on the accuracy of DBS.

Hepatitis B and C testing using DBS have been used in several screening programmes in non-clinical settings and research studies. There is encouraging evidence from pilot schemes and where dried blood spot collection is used for hepatitis testing (detection of HBsAg and HCV Ab). DBS collection is view as an interesting alternative testing technology for peoples at increased risk of infection. For example, hepatitis testing using DBS is used by associations such as “Hepatitis C Trust” in the UK1 or le “Réseau Hépatites LR” in France,2 and by community pharmacists in the UK.3 The “CheckPoint-Paris” from the “Kiosque”, a Page | 124

voluntary counselling and testing service in France, offers rapid tests for screening and DBS for confirmation since 2010.4 The National Institute for Health and Care Excellence (NICE) points out that DBS tests for hepatitis B and C can be useful in certain settings for people with poor venous access and where there may be no facilities or expertise to take venous blood samples.3 The NICE recommends access to hepatitis B and C testing using DBS for prisoners and drug users. In France, guidelines from the “Haute Autorité Sanitaire”5 and AFEF-ANRS6 underlined that DBS tests have good performance and are an alternative to venous blood tests. However, the absence of standardization limits the usefulness of DBS and thus there is no clear recommendation to expanded DBS testing. Critically, as part of Scotland’s Action Plan on HCV, “the introduction of DBS testing in the specialist drug services setting has had the greatest impact” in terms of increasing access to diagnosis.

References 1. 2. Health Protection Agency. Hepatitis C in the UK: 2011 report. London: HPA; 2011. Accueillir, accompagner, dépister les personnes à risque d'hépatites B ou C en Languedoc-Roussillon (ABCD-LR) – Projet DGS. (http://www.chumontpellier.fr/fileadmin/user_upload/Pole_Digestif/ ReseauHepatitesLR/Depistage/20141014_EVALUATION_PROJET_BUVARD_DGS_VF.p df, accessed 03 June 2016). National Institute for Health and Clinical Excellence. Hepatitis B and C: ways to promote and offer testing to people at increased risk of infection. NICE public health guidance 43. London: NICE; 2012. (https://www.nice.org.uk/Guidance/PH43, accessed 06 June 2016). Le kiosque [webpage] (http://www.lekiosque.org/checkpoint, accessed 06 June 2016). Haute Autorité Sanitaire, 2014. Place des tests rapides d’orientation diagnostique (TROD) dans la stratégie de dépistage de l’hépatite C (http://www.hassante.fr/portail/upload/docs/application/pdf/201405/place_des_trod_dans_la_strategie_de_depistage_de_vhc-_rapport.pdf) Prise en charge des personnes infectées par le virus de l'hépatite B ou de l'hépatite C (http://socialsante.gouv.fr/IMG/pdf/Rapport_Prise_en_charge_Hepatites_2014.pdf, accessed 06 June 2016). Scotland’s Action Plan on HCV. (http://www.inhsu.com/international -symposia.html.

4. 5.

Page | 125

11. Rationale for recommendation:

12. Strength of recommendation

13. Implementation considerations • Regulatory approval for DBS as a sample type • Where HR and phlebotomy services are lacking, training of lay workers to perform fingerprick DBS • QA/QC; e-health/m-health to improve turnaround time of results

• Training clinicians to act on the result. 14. Research gaps • Operational research to evaluate the accuracy of DBS using field specimens prepared and stored in real-life conditions will be needed. • Validation of the use of DBS with commercially available HBsAg and HCV Ab tests and determining cut-off values with the use of ROC curves will be needed. • The use of DBS for HCV RNA viral load measurement and of the rate of degradation of HCV RNA when stored in DBS at ambient temperatures and high humidity for different time periods will be needed as well as further study in HIV-coinfected patients and for use in treatment monitoring, specifically for measuring SVR 12/24 post DAA therapy. • A systematic review on the LOD for HCV RNA testing on DBS that will serve to capture (1) everyone with chronic and active HCV infection, and (2) everyone failing DAA therapy at SVR12/24 (which will then inform the best time-point for SVR). • A systematic review on the LOD for HBV DNA testing on DBS that will serve to capture (1) everyone who needs HBV treatment, and (2) everyone failing therapy.

GRADE Summary of findings for PICO7a Number of studies Type of study Directness Precision Consistency Risk of bias Overall quality

Sensitivity 92.9% (95% CI 86.2–96.5) 10 studies (370 HBsAg positive among 1516 samples) Crosssectional or cohort No significant indirectness No significant imprecision Significant inconsistency Significant risk of bias (patient enrolment not consecutive or (One paper random in some reported lower studies; pre-specified sensitivity) cut-off not used in Moderate

Page | 126

some studies) Specificity 99.9% (95% CI 97.6–100%) 10 studies (370 HBsAg positive among 1516 samples) Crosssectional or cohort No significant indirectness No significant imprecision No significant inconsistency Significant risk of bias (patient enrolment not consecutive or random in some studies; pre-specified cut-off not used in some studies) Moderate

DBS: dried blood spot GRADE Summary of findings for PICO7b Number of studies Type of study Directness Precision Consistency Risk of bias Overall quality

Sensitivity 98% (95% CI 93.0%–99.0%) 14 studies (1549 HCV positive among 4304 samples) Diagnostic accuracy No significant indirectness No significant imprecision Significant inconsistency Moderate risk of bias Moderate (patient enrolment only partly consecutive or random; several case–control studies)

Specificity 99% (95% CI 97–100%) 13 studies (2756 HCV positive among 4304 samples) Diagnostic accuracy No significant indirectness No significant imprecision Significant inconsistency Moderate risk of bias Moderate (patient enrolment only partly consecutive or random; several case–control studies)

GRADE Summary of Findings for PICO7c Number of studies Type of study Directness Precision Consistency Risk of bias Overall quality

Sensitivity 98% (95% CI 92.0%–98.0%) 7 studies (154 HBsAg positive among Diagnostic accuracy No significant indirectness No significant Significant imprecision inconsistency High risk of bias Low (patient enrolment not consecutive or random in all studies;

Page | 127

552 samples)

several case–control studies)

Specificity 100% (95% CI 39–100%) 4 studies (125 HBV DNA pos positive among 1648 samples) Diagnostic accuracy No significant indirectness Significant imprecision with small sample size Significant inconsistency High risk of bias Low (patient enrolment not consecutive or random in all studies; several case control studies)

GRADE Summary of findings for PICO7d Number of studies Type of study Directness Precision Consistency Risk of bias Overall quality

Sensitivity of DBS for HCV VL: 96.0% (upper-lower bounds 93.4–97.6) 9 studies 1335 samples Cross-sectional, case–control or cohort No significant indirectness No significant imprecision No significant inconsistency Significant risk of bias (nonrandomized or consecutive patient recruitment or case–control design) Moderate

Specificity of DBS for HCV VL: 97.7% (upper-lower bounds 94.7–99.0) 9 studies 1335 samples Cross-sectional, case–control or cohort No significant indirectness No significant imprecision No significant inconsistency Significant risk of bias (nonrandomized or consecutive patient recruitment or case–control design) Moderate

Page | 128

4.9. Monitoring treatment response HCV Decision-making tables – PICO 9 Monitoring for treatment response using HCV Ag testing in individuals with confirmed active HCV infection: Among individuals receiving antiviral treatment for HCV, what is the diagnostic accuracy of HCV core antigen versus NAT for HCV RNA qualitative detection (and/or) quantification to confirm successful treatment response with viral clearance? Population: Patients receiving treatment for HCV Intervention: HCV core antigen assay Comparison: NAT for HCV RNA detection (and/or) quantification Outcomes: Diagnostic accuracy (Sensitivity and specificity, TN, TP, FN, and FP)

1. Background:  HCV core antigen (HCV cAg) testing was developed as an alternative to NAT for diagnosis of active HCV infection. HCV nucleocapsid peptides 22 (p22) are released into plasma during viral assembly and can be detected throughout the course of HCV infection. Detection of HCV viraemia is also important during treatment of chronic HCV infection. Current guidelines recommend verification of virological activity pre-treatment with the measurement of a baseline HCV RNA quantitative measurement (viral load) by NAT. For interferon-based treatments, HCV RNA viral load is assessed at week 4 of therapy for the “rapid viral response” (RVR) to help predict efficacy of therapy, and repeated at week 6 if elevated at week 4 to see further viral response and guide whether treatment should be continued. NAT for HCV RNA is performed again at week 12 (early viral response, EVR), at the end of treatment, and 12 and 24 weeks after therapy is completed to test for cure, “sustained viral response” (SVR). New, direct-acting antivirals (DAA) have made treatment for HCV much easier with oral rather than parenteral administration and shorter, more effective regimens that are likely to be easier to adhere to making access to affordable diagnostic and monitoring assays even more important. However, it is important to note that, ultimately, treatment monitoring may not be required with the routine use of DAAs. This PICO addresses the question of whether HCV cAg can be used as a tool for assessing response to treatment for HCV infection.

 



Page | 129

2. Draft recommendation(s):

3. Summary and quality of evidence

1.1.15 □ High □ Moderate □ Low □ Very low

A systematic review (see SR_PICO 9) was commissioned to address the above PICO question. This aimed to examine the utility of HCV cAg monitoring for those on HCV treatment (PICO 9).

Summary of results Sensitivity and specificity of Abbott ARCHITECT HCV cAg assay compared to HCV RNA assessed at baseline, at week 4 of interferon-based therapy (early viral response), and at week 24 after completion of treatment (sustained viral response) Baseline Early viral response (EVR) Sustained viral response (SVR)

Author, Year

Sensitivity (95% CI)

Specificity (95% CI)

Sensitivity (95% CI)

Specificity (95% CI)

Sensitivity (95% CI)

Specificity (95% CI)

Feng, 2014

100%

N/A

100%

88.9% (68.4%, 100%)

100%

Loggi, 2013

100%

N/A

73.5% (58.7%, 88.4%)

100%

94.1% (82.9%, 100%)

Moscato, 2010

N/A

100%

70% (41.6%, 98.4%)

N/A

N: number of subjects; HCV: hepatitis C virus, Ag: antigen, Se: sensitivity, Sp: specificity, CI: confidence interval, N/A: not applicable as cannot be calculated from study data

Sensitivity and specificity of HCV core antigen in prediction of sustained viral response (SVR) after initiation of interferon-based treatment

Page | 130

Author, Year

N (N to achieve SVR)

Index test

Timing of test after treatment start

Change in HCVcAg

Sensitivity

Specificity

Feng, 2014

32 (21)

Abbott ARCHITECT

6 days

Log 10

95.2%

70%

Loggi, 2013

90 (57)

Fujirebio Lumipulse

7 days

Absolute

79.4%

88.5%

Moscato , 2010

44 (10)

Fujirebio Lumipulse

7 days

Absolute

57.1%

93.3%

HCV: hepatitis C virus; N: number

Conclusions: • HCV core Ag assays can have high sensitivity (up to 93.4% for certain commercialized assays), high specificity, and good correlation with HCV RNA to a detection limit of roughly 3000 IU/mL. • The data on HCV core Ag for treatment monitoring and as a test of cure is too limited to reach reliable conclusions.

Quality of evidence *Refer GRADE table in footnote 4. Risks/benefits Benefits • HCV cAg testing by immunoassay format has the potential to be less costly and less complicated to perform than HCV RNA by NAT. However, these immunoassays still require sophisticated laboratory equipment and therefore skilled staff to operate. Access to cold storage and constant electricity is required for the current types of assays available for HCV cAg testing. Results for patients on antiviral treatment being monitored for sustained viral response may be available more rapidly as a result of decentralized testing. □ Benefits clearly outweigh harms □ Benefits and harms are balanced □ Potential harms clearly outweigh potential benefits

•

Are the desirable anticipated

Page | 131

Risks  Due to reduced analytical sensitivity and limited understanding of kinetics of HCV cAg compared to HCV RNA by NAT, individuals on antiviral treatment may be misclassified as responding to treatment, but may have persisting viraemia below the limits of detection of the assay.

effects large? □ No □ Probably □ Uncertain □ Yes □ Varies □ No major variability □ Major variability

5. Acceptability, values and preferences A values and preferences survey of implementers and users of hepatitis B and C testing services was carried out by FIND in September 2015. A total of 104 respondents from 43 (20 high-income, 23 low- and middle-income) countries participated. Relating to this PICO,  As assay (platform) for detection of HCV cAg is available in India, Indonesia, Former Yugoslav Republic of Macedonia, Viet Nam, Turkey. The platform is available in South Africa but not currently being used for HCV cAg detection. 44% of respondents preferred a 12-week follow up for testing after completion of therapy, while 19% and 15% preferred a 4- or 8-week follow up. 47% of patients preferred to have the same test for monitoring and detection, ideally in decentralized settings. Free text comments from respondents included concerns regarding the sensitivity and specificity of cAg, but that it was potentially easier to do. One comment stated that it could be acceptable if it increased access to treatment. A larger number of respondents felt that the cost of testing for HCV RNA by NAT was considered more of a barrier than that of HCV cAg. As stated previously, 47% of respondents in low- and middle-income countries would prefer testing at POC, even at the cost of sensitivity.

Is the option acceptable to key stakeholders? □ No □ Probably □ Uncertain □ Yes □ Varies

 Currently only 11% of respondents are using HCV cAg as a test of cure. 

 

Patients: • Patients at risk of progressive liver disease will benefit from reduced disease progression and related mortality and morbidity, if treatment is provided as a result of wider access to testing programmes.

progressive disease and serious outcomes (HCC and complications of advanced liver disease), premature morbidity and mortality within the community will be reduced, and so also the burden of disease to societies where the disease is most prevalent.

Health-care workers: • • • Appropriate use of resources to channel treatment to patients with higher risk of complication in the medium- and short term Will require training in the use of testing equipment if being used in the nearPOC setting Appropriate reporting and recording of results.

Laboratory:  Will require purchasing of the appropriate platform and reagents for the HBsAg detection assay Will require training for HCV cAg test: careful sample processing is necessary for HCV cAg assay to lyse viral particles, expose antigen and dissociate antibody from antigen and optimize the detection for HCVcAg. □ Less equitable □ More equitable

6. Equity, ethics and human right implications Will the recommendation raise questions around equity? • • Equity will improve as a result of decentralization of testing, however, still improvement of access to the testing facilities is necessary. Regional and country variability in access to treatment.

Are there ethical implications to this recommendation? • Ethical consideration for the possibility that WHO could recommend a suboptimal testing strategy.

7. Resource use and financial implications Materials/equipment:   Cost of testing platform and reagents Other laboratory consumables Are the resources required small? □ No □ Probably □ Uncertain Page | 133

Training and supervision:    Appropriate training of laboratory staff Quality control programmes If using near-POC assays, appropriate training of testing providers.

□ Yes □ Varies

Other:  Cost of transportation of specimens to the laboratory

Possible procurement costs:

Assay format

Indicative cost (US$) per test 0.50–2.00 (10 for oral fluid RDTs) 0.50–1.70 25–50 10–45 43–51

Source

RDTs

MSF, WHO

EIA HCV Ag Quantitative NAT for HCV RNA Qualitative NAT for HCV RNA)

WHO MSF MSF, UNITAID UNITAID

•

Feasibility and constraints to implementation Is the option feasible to implement? □ No □ Probably □ Uncertain □ Yes □ Varies

Are any major barriers expected for the implementation of this recommendation? • • Regional and country variability in access to treatment and procurement of testing equipment and services With regard to HCV cAg, availability of a local laboratory, which is able to procure the testing platform and reagents required for testing.

Page | 134

8. Relevance to different settings/populations Will this recommendation be most relevant for particular settings (e.g. endemicity)?   HCV cAg testing will be more relevant to populations that presently rely on centralized laboratory testing for HCV RNA for confirmation of status. The recommendations are less likely to be relevant in high-income settings where there is already access to established hepatitis C testing and treatment programmes.

9. Rationale for recommendation:

10. Strength of recommendation

11. Implementation considerations Optimize test for asymptomatic patients in primary-care settings or in the community where the HCV endemic is high.

12. Research gaps • The kinetics of HCV cAg with treatment needs to be evaluated further, particularly in the context of new DAA regimens. • More rigorous assessment of covariates is required in studies assessing HCV cAg or NATs for treatment monitoring or as a test of cure, such as HIV or HBV coinfection or genotype. • Is treatment monitoring and/or confirmation of cure necessary with DAA regimens? If so, what would be the optimal timing of testing? • When is the best time-point to test for cure with HCV core Ag? • Development of muliplex instrument with other disease diagnosis such as HIV, HBV, and TB at health centre. • HCVcAg assay to detect the variants of HCV.

Page | 135

World Health Organization Global Hepatitis Programme

GRADE Summary of findings SR outcome 1: Diagnostic accuracy at SVR 1.1.16 1.1.17 1.1.18 Quality Effect* Strength of evidenc e

Index test

Outcom e measure

# Studies (# samples)

Design

Risk of bias

Inconsiste ncy

Indirectn ess

Imprecisi on

1.1.19

1.1.20

Abbott ARCHITECT HCVAg Assay

Sensitivi ty

2 (67)

RCT, cohort

Low1

Low2 1.1.21

Moderat e3 (–1)

Low4 1.1.22

100%*

Moderat e 

Specificit y

2 (67)

RCT, cohort

Low1

Moderate2 (–1)

Moderat e3 (–1)

Low4 1.1.23

94– 100%*

Low 

SR outcome 2: Predictive accuracy of SVR

Quality

Effect*

Strength of evidence

Outcome measure Index test

# Studies (# individuals)

Design

Risk of bias

Inconsistency

Indirectness

Imprecision

Abbott ARCHITECT HCV Ag Assay

Sensitivity

1 (23)

Cohort

Low1

NA2 (–1)

Moderate3 (–1)

NA4

95.2%**

Low 

Specificity

1 (23)

Cohort

Low1

NA2 (–1)

Moderate3 (–1)

NA4

70%**

Low 

Fujirebio Lumipulse Ortho HCV Ag

Sensitivity

2 (134)

Cohort

Moderate1 (–1)

Moderate2 (–1)

Moderate3 (–1)

Moderate4 (–1)

57.1– 79.4%*

Very low 

Specificity

2 (134)

Cohort

Moderate1 (–1)

Moderate2 (–1)

Moderate3 (–1)

Moderate4 (–1)

88.5– 99.3%*

Very low 

* Results reported are a range across studies or **individual result, NA: not applicable

Page | 136

References 1. Feng B, Yang RF, Xie Q, Shang J, Kong FY, Zhang HY et al. Hepatitis C virus core antigen, an earlier and stronger predictor on sustained virological response in patients with genotype 1 HCV infection. BMC Gastroenterol. 2014;14:47. 2. Loggi E, Cursaro C, Scuteri A, Grandini E, Panno AM, Galli s et al. Patterns of HCV-RNA and HCV core antigen in the early monitoring of standard treatment for chronic hepatitis C. J Clin Virol. 2013;56(3):207–11. 3. Moscato GA, Gianelli G, Grandi B, Pieri D, Marsi O, Guarducci I, Batini I et al. Quantitative determination of hepatitis C core antigen in therapy monitoring for chronic hepatitis C. Intervirology. 2011;54(2):61–5.

Page | 137

4.10. Intervention to promote linkage to care Decision-making tables – PICO 10 Interventions to optimize uptake of hepatitis testing and linkage to care across the viral hepatitis treatment cascade 17. Topic for analysis Population: Individuals living with chronic hepatitis B or C (diagnosed or undiagnosed) or providers caring for these patients Intervention: Psychosocial or structural interventions delivered in conjunction with screening, care, or treatment of hepatitis Comparison: Standard of care or no intervention Outcomes: Retention and progression along the continuum of care

18. Background: Globally, 250 million people are chronically infected with hepatitis B virus (HBV), and 80–140 million are infected with chronic hepatitis C virus (HCV). Viral hepatitis is now responsible for 1.45 million deaths every year and is the seventh leading cause of mortality worldwide. Chronic HBV and HCV are responsible for over 90% of these deaths. But therapeutic advances are rapidly changing clinical management of both HBV and HCV infection, especially HCV infection is increasingly curable. Reaping the clinical benefits of novel HBV and HCV therapies will require a continuum of care that start with screening and ultimately reaches and sustains viral suppression (Fig. 1). Similar to the HIV continuum of care, each step of the HBV/HCV continuum of care necessarily requires all prior steps and must be maintained over time. Screening is the critical entry way into the continuum and allows positive individuals to proceed and negative individuals to receive vaccination in the case of HBV. However, several barriers to screening have been reported, such as lack of knowledge, awareness and/or denial, lack of access to the infrastructure where screening is available, cultural beliefs, financial conditions including health insurance status, fear of blood taking and stigmatization as a patient aspects, and lack and gaps of awareness of risk factors, knowledge, experience of patient management, lack of infrastructure for testing and lack of access to treatment centres as a health-care aspect. Interventions can enhance chronic viral hepatitis management as part of each step along the continuum of care, including screening, linkage to care, treatment uptake, adherence, and viral suppression. Page | 138

People living with undiagnosed chronic viral hepatitis

Screening

Fig. 1. Overview of the Patients withstages a positive comprising serological resultthe for viral hepatitis treatment continuum, including HBV or HCV exposure, but without testing, linkage confirmatory to care, testing enrolment in care, treatment uptake, treatment adherence, and or liver disease staging treatment outcome Linkage to care

Patients with confirmed chronic viral hepatitis infection and qualified for treatment, but have not initiated treatment

Treatment uptake

Patients who have initiated treatment for chronic viral hepatitis

Treatment adherence Patients who have completed HCV treatment or Patients who are maintained on HBV treatment

Treatment outcome

Patients who have achieved HCV SVR or Patients who have achieved HBV virologic suppression

Page | 139

19. DRAFT recommendation(s):

20. Summary and quality of evidence Summary of results Evidence from the systematic review for linkage to care for HIV ARV GL 2015: among people living with HIV, what interventions facilitate linkage to care compared to standard of care? (PICO F.2.1) Category 1. Counselling and support (n = 14) Summary Most interventions improved linkage to and engagement in care but not ART initiation Mixed: one intervention worked, two did not Most interventions increased outcomes targeted; Quality 3 moderate, 6 low, 5 very low

□ High □ Moderate □ Low □ Very low

2. Incentives (n = 4)

1 moderate, 2 low, 1 very low 2 moderate, 5 very low

3. Quality improvement (n = 7)

Page | 140

interventions in category were multifaceted and incorporated interventions from other categories

Evidence from the systematic review for linkage to care for hepatitis B and C testing GL: We conducted a systematic review to identify interventions that improve the continuum of care in adults with chronic hepatitis B and C infection, quantify the effect size of these interventions, and recognize gaps in knowledge in interventional studies that target the chronic viral hepatitis continuum of care.

I. Meta-analyses for interventions to improve HBV screening Single culturally tailored lay health worker educational session to improve HBV knowledge and promote testing vs no or unrelated educational session for selfreported HBV screening.

II. Meta-analyses for interventions to improve HCV screening Clinician reminder to use HCV screening algorithm during clinical visit with or without supplementary provider education vs no clinician reminder for HCV screening.

Page | 141

III. Meta-analyses for interventions to improve HCV linkage to care Facilitated referral and scheduling to specialist visit by staff at site of established care with or without supplementary HCV education and post-test counselling vs no facilitated referral for attendance at HCV specialist visit.

Unadjusted results

Adjusted results

Results of educational interventions data from a systematic review: For HBV • Targeted population was HBV non-infected but at-risk individuals. • Educational interventions increased: knowledge about the disease, HBsAg testing and HBV vaccination rates.

For HCV • 50% of studies targeted HCV non-infected and the rest were HCV-infected individuals.

• Educational interventions increased: knowledge about the disease, the number of anti-HCVAb testing, willingness to undergo therapy, and treatment adherence.

Page | 142

Ref: Shah HA1, Abu-Amara M. Education provides significant benefits to patients with hepatitis B virus or hepatitis C virus infection: a systematic review. Clin Gastroenterol Hepatol. 2013;11(8):922–33.

Quality of evidence *Refer GRADE table in footnote

21. Risks/Benefits Benefits of peer support, clinician reminder, and quality improvement initiatives/integration  Increasing the rate for hepatitis B and C screening will increase the opportunity to link infected individuals to further hepatitis care and treatment.

□ Benefits clearly outweigh harms □ Benefits and harms are balanced □ Potential harms clearly outweigh potential benefits

 •

The infected individuals will be offered health-protection advice to prevent Are the desirable disease progression and transmission. Increase HCV treatment initiation, improve treatment completion, and anticipated effects large? increase SVR. □ No □ Probably □ Uncertain □ Yes □ Varies

Risks • • Stigmatization by identifying HBV and/or HCV infection Might be a challenge to differentially incentivize for receiving hepatitis care where poverty is prevalent and the rest of the population have limited access to health services in general. There are few studies on quality improvement initiatives/integration dealing with HBV.

22. Acceptability, values and preferences Counselling and peer support  Counselling is recognized as standard of HIV-testing services and is also an essential intervention to support adherence. Although counselling for hepatitis testing services has not been established, simple educational interventions for lay health-care workers (HCWs) require little training, generate minimal costs, are highly feasible, and may substantially scale up HBV screening.

□ No major variability □ Major variability

Is the option acceptable to key stakeholders?

Page | 143

Clinician reminder  HCWs will need to understand the strengths and limitations of appropriately counsel patients who are screened.  HCWs will need to aware the importance of reminder to enhance hepatitis screening during medical consultations. Quality improvement initiatives/integration  Coordinated mental health/substance abuse services within HCV service delivery will require new training and systems; however, linkage to initiation for HCV treatment, improved HCV treatment completion, and increased SVR will increase the rate of cure and prevent further HCV transmission.

□ No □ Probably □ Uncertain □ Yes □ Varies

23.

Equity, ethics and human right implications

□ Less equitable □ More equitable

Counselling and peer support Will recommendation raise questions around equity?  No. Increased rate for hepatitis B and C screening will increase equity to access for further hepatitis care and treatment, especially in resourcelimited settings. Are there ethical implications to this recommendation?  No major concerns.

Clinician reminder Will recommendation raise questions around equity?  No; increase in rate for hepatitis B and C screening will increase equity to access for further hepatitis care and treatment, especially in resourcelimited settings.  Might be a challenge to differentially send clinician reminder to patients for receiving hepatitis care where poverty is prevalent and the rest of the population has limited access to health services in general.  Linkage to care of at-risk populations, e.g. antenatal clinics, tohose who attend drug-rehabilitation clinics, prisoners might be a challenge.

Page | 144

Are there ethical implications to this recommendation?  No major concerns.

Quality improvement initiatives/integration Will recommendation raise questions around equity?  Some quality improvement initiatives may have a broader benefit for the health system. Are there ethical implications to this recommendation?  No major concerns.

24. Resource use and financial implications Counselling and peer support  As testing and treatment programmes are scaled up, the numbers developing progressive disease and serious outcomes (HCC and complications of advanced liver disease), premature morbidity and mortality within the community will be reduced, and so also the burden of disease to societies where the disease is most prevalent. Are the resources required small? □ No □ Probably □ Uncertain □ Yes □ Varies

Clinician reminder  Creation of a new proper patient database in which the clinician reminder to prompt HCV testing require attention could be recorded might be costly; however, no extra cost is required to expand facility-based HCV screening in settings that have electronic records or analogous reminder systems.

Quality improvement initiatives/integration  Variable and context specific, detailed costing of each intervention is not done

25.

Feasibility and constraints to implementation

1.1.24 option to implement?

Are any major barriers expected for the implementation of this Is the feasible recommendation? Counselling and peer support and clinician reminder

Page | 145

Feasible.  Simple educational intervention and clinician reminder, either electronic records or analogous reminder are highly feasible. Quality improvement initiatives/integration Feasible.

□ No □ Probably □ Uncertain □ Yes □ Varies

 Coordinated mental health/substance abuse services within HCV service delivery will require involvement of stakeholders. 26. Relevance to different settings/populations Will this recommendation be most relevant for particular settings (e.g. endemicity)?  These recommendations will increase the opportunity of hepatitis screening and will be relevant to any circumstances.

27. Rationale for recommendation:

28. Strength of recommendation

29. Implementation considerations  Establish linkages and referral pathways for the individuals infected with hepatitis viruses and suffering from mental health problems.  Ensure support from management and from service providers. 30. Research gaps  Scientific implementation research, especially in low- and middle-income countries, is urgently needed to inform optimize chronic viral hepatitis service delivery systems.  Costing studies need to be performed.

GRADE Summary of Findings

Page | 146

Evidence from the systematic review for linkage to 2015: interventions providing counselling and support (GRADE) # (type) studies Risk of: Bias Inconsistency Indirectness Imprecision

care

for

HIV Effect

ARV

N intervention Risk (control) intervention (control)

Quality

Linkage to care: trials 1 (individual) Not serious Not serious Not serious Serious 200 (200) 0.67 (0.38) RR 1.8 (1.4–2.1) Moderate1

Linkage to care: observational 7 (6 cohort, 1 Serious other) Engagement in care: trials 1 (individual) Not serious Not serious Not serious Not serious 188 (191) 0.92 (0.83) pRR 1.1 Low3 (1.03–1.20) Not serious Serious Not serious 5271 (7195)* 0.81 (0.64)* RR 1.25 Very low2 (1.22–1.28)*

Engagement in care: observational 1 (Pre/post) Serious Serious Serious Not serious 1147 (1210) 0. 57 (0.45) RR 1.28 (p<0.0001) Low

ART initiation: trials 2 (individual) Not serious Not serious Not serious Serious 806 (719)* 0.41 (0.43)* RR 0.95 Moderate1 (0.84–1.07)*

PMTCT: trials 1 (individual) Not serious Not serious Not serious Serious 197 (181) 0.64 (0.53) aHR 1.39 Moderate1 (1.01–1.91)

PMTCT: observational 1 (cohort) Not serious Not serious Not serious Serious 63 (332) 0.54 (0.25) aOR 3.18 Low (1.76–5.73)

Interventions offering incentives (GRADE) # (type) studies Risk of: N intervention (control) Risk intervention (control) Effect Quality

Bias

Inconsistency Indirectness Imprecision

Linkage to care: trials Moderate1

2 (1 individual, 1 cluster)

Not serious

Serious

60 (33)*

0.82 (0.48)*

RR 1.68 (1.16–2.44)*

Engagement in care: observational Very low1

1 (cohort)

Serious

Not serious

Serious

100 (80)

0.94 (0.98)

2.00 (0.41–9.64)

Page | 147

ART initiation: trials Low2

1 (individual)

Serious

Not serious

Serious

60 (60)

0.45 (0.26)

aHR 2.93 (1.39–6.20)

Interventions introducing quality improvement (GRADE) # (type) studies Bias Inconsistency Risk of: N intervention (control) (control) Risk intervention Effect Quality

Indirectness

Imprecision

ART initiation: trials Moderate1

1 (1 cluster)

Not serious

5390 (3862)

0.68 (0.62)

RR 1.24 (0.88–1.73)

Engagement in care: trials

1 (1 cluster)

Not serious

5390 (3862)

0.62 (0.58)

RR 1.1 (1.04–1.16)

High

PMTCT (ART initiation): observational Very low3

3 (3 prepost)

Serious

Not serious

Serious

619 (1296)*

0.36 (0.1)*

RR 3.48 (2.87– 4.22)*

PMTCT (EID access): observational Very low2

1 (1 prepost)

Serious

Not serious

Serious

63 (332)

0.54 (0.25)

aOR 3.18 (1.76– 5.73)

PMTCT (Receipt of AZT): trials Very low2

1 (1 prepost)

Serious

Not serious

Serious

Not serious

1258 (776)

0.87 (0.71)

RR 1.22 (1.16–1.28)

Page | 148

World Health Organization Global Hepatitis Programme

Evidence from the systematic review for linkage to care for hepatitis B and C testing GL: I. Meta-analyses for interventions to improve HBV screening Quality assessment No. of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations No of patients Single LHW educational session No or unrelated educational session Relative (95% CI) Effect Absolute (95% CI) Quality

HBV screening 6 Randomized trials Serious 1

Not serious

None

255/1344 (19.0%)

92/1413 (6.5%)

RR 2.68 (1.82–3.93)

109 more per 1000 (from 53 more to 191 more) 110 more per 1000 (from 54 more to 192 more)

⨁ ⨁ ⨁ ◯ Moderate

6.6%

6/6 studies are at high risk of detection bias because the outcome was self-reported HBV screening 6 months post intervention. 5/6 studies are at high risk of attrition bias because the ratio of participants with missing data to participants with HBV screening outcome was high (>1.0). Although all included studies involved Asian immigrants in North America, this was not judged to be a significant enough difference in populations to downgrade because the intervention strategies are not exclusive to Asian immigrant populations. The confidence interval is not wide. The OIS was calculated to be 222, and the pooled sample size exceeded the OIS. 3/6 included studies were cluster RCTs, none of which performed analyses that accounted for clustering. Consequently, this meta-analysis commits a unit-of-analysis error and produces over-precise results. Additionally, no ICC were reported in the included studies, so statistical methods could not be used to reduce the effective sample size of the cluster RCTs. Despite this limitation, it is unlikely proper adjustment for cluster design would significantly impact the precision of the pooled results.

Page | 149

II. Meta-analyses for interventions to improve HCV screening Quality assessment № of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations HCV screening 3 Other 1 design Serious 2

No. of patients Clinical testing reminder No reminder Relative (95% CI)

Effect Absolute (95% CI)

Quality

Serious

Not serious

Serious

None

5185/33253 (15.6%)

976/19694 (5.0%)

RR 3.70 (1.81–7.57)

134 more per 1000 (from 40 more to 326 more) 161 more per 1000 (from 48 more to 393 more)

⨁ ◯ ◯ ◯ Very low

6.0%

1. 2.

This meta-analysis includes 1 cluster RCT and 2 NRS. Drainoni (2012) is at high risk of performance bias and did not employ methods to adjust for confounding potentially introduced by its non-randomized study design. Krauskopf (2014) did not report comparability of randomized clusters and therefore was at high risk of bias. All included studies report a risk ratio >1.0. However, I² = 99%. The high degree of heterogeneity may be due to differences between HCV screening algorithms used in each intervention. Although the pooled sample size exceeds the calculated OIS, the confidence interval is wide. Additionally, Krauskopf (2014) was a cluster RCT that did not account for clustering in its analysis. Consequently, this meta-analysis commits a unit-of-analysis error and produces over-precise results. No ICC was reported, so statistical methods could not be used to reduce the effective sample size of the cluster RCT. All included studies report a risk ratio >2.0. However, the pooled results have not been upgraded for large effect because the non-randomized design of 2/3 studies introduces a significant possibility of confounding.

Page | 150

III. Meta-analyses for interventions to improve HCV linkage to care Quality assessment No. of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations No. of patients Facilitated referral to specialist visit at site of established care No facilitated referral Relative (95% CI) Effect Absolute (95% CI) Quality

Attended HCV specialist visit 3 Randomized trials Not 1 serious Serious 2

Not serious

None

151/243 (62.1%)

72/194 (37.1%)

RR 1.57 (1.03–2.41)

212 more per 1000 (from 11 more to 523 more) 212 more per 1000 (from 11 more to 525 more)

⨁ ⨁ ⨁ ◯ Moderate

37.2%

2. 3.

Rosenberg (2010) relied on self-reported HCV status and self-reported attendance to an HCV specialist visit, putting the study at high risk of detection bias. However, because this study had a relatively small sample size it was not judged to put the entire meta-analysis at high risk of bias. I² = 85%. This high degree of heterogeneity may be due to differences between the intensity of interventions in the included studies. The confidence interval is not wide. The OIS was calculated to be 124, and the pooled sample size exceeded the OIS.

Page | 151

References 1. Drainoni ML, Litwin AH, Smith BD, Koppelman EA, McKee MD, Christiansen CL et al. Effectiveness of a risk screener in identifying hepatitis C virus in a primary care setting.Am J Public Health. 2012;102(11):e115–21. 2. Krauskopf K, Kil N, Sofianou A, Toribio W, Lyons J, Singer M et al. Evaluation of an electronic health record prompt for hepatitis c antibody screening of baby boomers in primary care – a cluster randomized control trial. J Gen Intern Med2014;29:S88–S89. 3. Rosenberg SD, Goldberg RW, Dixon LB, Wolford GL, Slade EP, HImelhoch S et al. Assessing the STIRR model of best practices for blood-borne infections of clients with severe mental illness. Psychiatr Serv. 2010;61(9):885–91.

Page | 152

Individually tailored mental health counselling and motivational therapy for HCV+ patients with mental health and/or substance use comorbidities vs usual care for physician referral to initiate treatment Quality assessment No. of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations No. of patients Individually tailored mental health counselling and motivational therapy Usual care Relative (95% CI) Effect Absolute (95% CI) Quality

Physician referral to initiate treatment 2 Other 1 design Serious 2

Not serious

None

66/120 (55.0%)

35/130 (26.9%)

RR 2.04 (1.48–2.80)

280 more per 1000 (from 129 more to 485 more) 263 more per 1000 (from 121 more to 455 more)

⨁ ◯ ◯ ◯ Very low

(25.3%)

Adjusted physician referral to initiate treatment 2 Other design Serious 1 5

Not serious

Serious

None

–/120

–/165

OR 3.43 (1.81–6.49)

0 fewer per 1000 (from 0 fewer to 0 fewer)

⨁ ◯ ◯ ◯ Very low

1. 2.

Evon (2011) is a RCT, while Knott (2006) is an NRS. Knott (2006) is at high risk of detection bias because the outcome was subjective and determined by the physician overseeing treatment who was not blinded. Unadjusted results from Knott (2006) were used in this meta-analysis that did not employ methods to adjust for confounding potentially introduced by its non-randomized study design. The decision to not downgrade for indirectness assumes guidelines are applied to other contexts where mental health or substance use comorbidities are also contraindications to recommending HCV+ patients for treatment. The confidence interval is not wide. The OIS was calculated to be 94, and the pooled sample size exceeded the OIS. Knott (2006) is at high risk of detection bias because the outcome was subjective and determined by the physician overseeing treatment who was not blinded. The confidence interval for the pooled adjusted outcomes is wide. Page | 153

Page | 154

References 4. Evon DM, Simpson K, Kixmiller S, Galanko J, Dougherty K, Golin C et al. A randomized controlled trial of an integrated care intervention to increase eligibility for chronic hepatitis C treatment. Am J Gastroenterol. 2011;106(10):1777–86. Knott A, Dieperink E, Willenbring ML, Heit S, Durfee JM, WIngert M et al. Integrated psychiatric/medical care in a chronic hepatitis C clinic: effect on antiviral treatment evaluation and outcomes. Am J Gastroenterol. 2006;101(10):2254–62.

Page | 155

ANNEX 5: Systematic reviews and evidence summaries 5.1 SR who to test HBV: Literature review on cost-effectiveness of HBV screening, treatment strategies and applicability to LMICs 5.2 SR who to test HCV: Literature review on cost-effectiveness of HCV screening, treatment strategies and applicability to LMICs 5.3 SR1 How to test HBV: Diagnostic accuracy of tests to detect hepatitis B surface antigen: a meta-analysis and review of the literature 5.4 SR2 How to test HCV: Diagnostic accuracy of tests to detect hepatitis C antibody: a metaanalysis and review of the literature 5.5 SR3 How to test HBV: Diagnostic strategies for hepatitis B surface antigen detection: a meta-analysis and review of the literature 5.6 SR4 How to test HCV: Diagnostic strategies for hepatitis C antibody detection: a metaanalysis and review of the literature 5.7 SR6 How to test: Diagnostic accuracy of HCV RNA tests to detect active HCV infection: a meta-analysis and review of the literature 5.8 SR5a and 9 treatment monitoring HCV: HCV core antigen testing for presence of active HCV infection and monitoring for treatment response and cure: a systematic review 5.9.1 SR7a: Dried blood spots as a sample collection method for hepatitis B surface antigen serological testing 5.9.2 SR7b: Dried blood spots as a sample collection method for HBV DNA 5.9.3 SR7c: Dried blood spots as a sample collection method for hepatitis C virological testing: a systematic review and meta-analysis 5.9.4 SR7d: Dried blood spots as a sample collection method for HCV antibody: a systematic review and meta-analysis 5.10 SR 8: Diagnostic accuracy of HBsAg/HBeAg test versus NAT to confirm successful treatment response: a meta-analysis and review of the literature 5.11 SR Interventions to linkage to care - to optimize the chronic viral hepatitis care continuum: a systematic review and meta-analysis of interventions to improve hepatitis B and C screening, linkage to care, treatment uptake, treatment adherence, and viral suppression

Page | 156

Annex 5.1

PICO 1 - Who to test (HBV)

Literature review on cost-effectiveness of HBV screening, treatment strategies and applicability to LMICs

Nayagam S, Sicuri E, Lemoine M, Conteh L, Hess S, Thursz M, Hallett TB (Team Lead) For correspondence: s.nayagam01@imperial.ac.uk Imperial College, London, United Kingdom

September 2015

Page | 157

1. Background Hepatitis B virus (HBV) infection is highly prevalent worldwide, with a disproportionately high burden in low- and middle-income countries (LMICs).1 There is mounting evidence regarding the efficacy of antiviral therapy in the reduction of disease progression to cirrhosis and hepatocellular carcinoma (HCC). However, this impact is not fully translated into practice as many people still remain unaware of their infection status, even in high-income countries (HICs),2–4 and this value is likely to be even lower in LMICs. For example, in the Gambia, only 0.4% of screening participants in PROLIFICA had been tested in the past. Wilson and Jungner criteria have been used to assess whether a disease should screened.5 However, despite fulfilling most of these criteria, screening for HBV is not performed systematically. The reasons surrounding this are likely multifactorial, including lack of awareness at all levels, lack of clear guidelines, competing health-care priorities, limited health-care budgets and political will. This leads to many people remaining undiagnosed until later stages of the disease, when prognosis is poor. Furthermore, even if diagnosed, access to appropriate antiviral therapy and ongoing clinical management is lacking.

Clarifications of terminology used in this report 1.1.1. High-risk group For the purposes of HBV screening in LMIC, the categorisation of populations into “high-risk” groups is not helpful or informative in guiding policy. In most LMICs, the adult population prevalence (unvaccinated) falls into the intermediate- to high-endemicity categories.1 Furthermore, within countries HBsAg prevalence is more homogenous within the population, than for example, with hepatitis C virus (HCV) infection. In this report, we therefore only refer to “high-risk” groups when referring to literature from HIC settings.

1.1.2. General population screening This is used to refer to the fact that all members of the population have access to the screening programme under consideration. This can include community outreach screening, health-care facility-based screening, etc.

Page | 158

1.1.3. Targeted screening This refers to screening of specific groups, e.g. pregnant women. Targeted groups are not necessary at higher risk of being infected than the general population.

2. Overview of the report The purpose of this report is not to represent the results of a full systematic review. It is meant to serve as a summary of existing studies on cost-effectiveness of screening and treatment for HBV, with an analytic summary of key considerations. It was envisaged that there was a lack of relevant literature in LMICs, so existing studies from HICs are described and their potential uses and limitations when drawing conclusions are discussed.

3. Search strategy We searched the bibliography of two previous systematic reviews on the cost-effectiveness of HBV screening by Hahné et al.6 and Gueue et al. (unpublished, shared by WHO team) and included these in the discussion, where appropriate. Hahne and Gueue searches were performed upto 2011 and 2012, respectively. We therefore performed an updated search using PubMed to retrieve any further relevant articles to be included in this report. We searched PubMed for articles published between January 2000 and September 2015, with terms incorporating “hepatitis B”, “HBV”, or “CHB” and “cost” or “economic” and “screen”, “test” or “Diagn”. We excluded studies prior to 2000, as older studies were mainly studying cost-effectiveness of pre-vaccination screening, rather than screening for consideration of antiviral therapy. Furthermore, Geue et al. reports the low methodological standards of costeffectiveness analyses in older studies. We selected articles published in English only. We did not search any databases other than PubMed, nor did we search the grey literature. However, we attempted to include any known ongoing HBV screening programmes in LMIC by consulting colleagues at WHO, in order to include any unpublished studies in this report. We excluded studies that considered screening in the following groups, unless the study reported further linkage into care and treatment – blood banks and health-care workers. We excluded evaluations that included screening prior to vaccination, unless the analysis also considered antiviral therapy for the person found to be HBsAg-positive. We also excluded studies around screening for HBV prior to chemotherapy, as this was only likely to be relevant to higher income settings and would only concern a small subset of the populations in LMIC. We also excluded studies looking at coinfection with HIV and comparing diagnostic methods. The PubMed search retrieved 32 studies, many of which overlapped with the bibliographies of the existing reviews. All studies were performed in HICs. We were unable to find any previous studies describing cost or cost-effectiveness of screening for HBV in LMICs. Due to the lack of published literature in LMICs, to better inform the report, we also included

Page | 159

data from the PROLIFICA study (forthcoming).7 Finally, eight published studies and one unpublished study met inclusion criteria and are discussed in further detail below.

4. Summary of main literature The existing published studies on the cost-effectiveness of screening and treatment for HBV have been performed in HICs where the prevalence in the general population is low.1 We have also included discussion of unpublished PROLIFICA data, which is the only study in a LIC setting. Two studies evaluated HBV screening in the general population8,9 and seven studies in “high-risk” groups (all but one concerned screening in migrant or refugee populations).10–16 We excluded studies of ANC screening as they did not consider antiviral therapy to the mother and only looked at the benefit of screening in order to guide vaccination strategies to reduce mother-to-child transmission. However, a brief summary is given below. The studies used different methods of screening the “high-risk” groups including, in the clinical setting,10,14 community outreach methods14 and overseas screening.16 Various outcome measures were used including cost per quality-adjusted life-year (QALY) gained, cost per life-year (LY) saved and cost per case screened. Many of the models were simulated using hypothetical cohorts.

4.1. General population level screening There was one previously published study in the USA and one forthcoming study in The Gambia, looking at the cost-effectiveness of offering screening and treatment to the general population. Eckman et al.8 looked at the cost-effectiveness of HBsAg testing of asymptomatic outpatients in primary care settings in the USA, using a hypothetical cohort (35-year-old male) with a general population prevalence of 2%. Screening was then followed by treatment with one of four regimens and compared to a no screening strategy. Screening and treatment were found to be cost-effective with an incremental cost-effectiveness ratio (ICER) of US$ 29 230/QALY. The ICER remained below their willingness to pay (WTP) threshold of US$ 50 000/QALY gained, even down to a population prevalence of 0.3%. The feasibility of large-scale screening and treatment in sub-Saharan Africa (SSA) has been demonstrated by the ongoing “Prevention of liver fibrosis and liver cancer in Africa” (PROLIFICA) study in West Africa (Lemoine et al., forthcoming). This implementation study has screened nearly 10 000 adults for HBsAg at the community level in The Gambia and Senegal using an active outreach method. This is followed by full clinical assessment of those found to HBsAg positive and antiviral treatment if meeting eligibility criteria. A cost-effectiveness analysis of this community-based screen and treat strategy in The Gambia (Nayagam et al., forthcoming), compared to status quo, revealed an ICER of US$ 705/LY gained (other outcome measures also calculated: US$ 476/QALY gained or US$ 575/DALY averted). The authors

Page | 160

acknowledge that WTP thresholds levels, and their use, are highly debated in LMICs. However, it can be regarded as cost-effective if using the WHO WTP threshold of three times the country’s GDP per capita to define a cost-effective intervention (3 times GDP per capita = US$ 1460 in The Gambia).17 This is the only cost-effectiveness study of screening and treatment we have found in LMIC settings. Furthermore, it is furnished with real-life cost and effectiveness data from a large-scale screening and treatment intervention programme.

4.2. Screening of “high-risk” groups in HIC There were six studies looking at the cost-effectiveness of screening and treatment in migrant or refugee populations in HICs,10-14,16 and one looking at screening all groups classified as “high–risk” in Italy.15 The study by Wong and colleagues in 2011 looked at the cost-effectiveness of screening and treatment of immigrants for chronic hepatitis B (CHB) in Canada.10 They considered a screen and treat strategy and a screen, treat or vaccinate strategy, with status quo (no screening). Screening was offered by the primary-care physician at a visit scheduled for another reason, described by the authors as a “case-finding” strategy. They used a hypothetical cohort (35-year-old male) with a baseline HBsAg prevalence among the immigrant population of 4.81%. The screen and treat strategy had an ICER of US$ 69 000/QALY gained. The authors acknowledge the uncertainty around WTP thresholds, but quotes range from US$ 50 000 to US$ 120 000 for Canada, implying a cost-effective intervention. This model is more clinically representative than many of the other models; however, it uses high and probably unrealistic uptake and adherence rates. Another Canadian study by Rossi et al. (2013)11 looked at combinations of scenarios involving screening, treatment and vaccination among newly arrived immigrants and refugees. The screen and treat scenario was found to be the most cost-effective with an ICER of US$ 40 880/QALY gained. This strategy exceeds the Canadian WTP threshold adopted in this study of US$ 50 000/QALY, when HBsAg prevalence is less than 3%. A societal perspective for the analysis was used. A hypothetical cohort of 250 000 immigrants was used, with baseline assumptions of 70% acceptance of screening, 60% linkage to care, 75% of those eligible will have treatment and annual cost of antiviral drugs at US$ 8089. An earlier study by Hutton et al.12 looked at the cost-effectiveness of screening and vaccination of Asian Pacific Islander adults for HBV by using a hypothetical cohort of 20–60 years old with a HBsAg prevalence of 10%. They compared four strategies of combinations of screening, treatment and vaccination, similar to the study described above. The screen and treat strategy was the most cost-effective with an ICER of US$ 36 000/QALY gained (compared to no screening), even down to an HBsAg prevalence of 1%. This study used a societal perspective. Another, more recent, US study by Jezwa and colleagues16 compared the cost-benefits of two overseas programmes for reducing HBV infection among refugees. They compared two strategies: (i) vaccination only, and (ii) screening and vaccination; and suggested onward

Page | 161

treatment on arrival in USA if HBsAg was positive. The strength of this study was the use of original data sets of refugee populations in two US states. Their baseline assumptions included a prevalence of 6.8%, 100% adherence with screening, 60% of those tested positive for HBsAg will link to specialist care and that 90% will adhere to treatment. The study by Veldhuijzen et al.13 was the only European study which looked at the cost-effectiveness of HBV screening and early treatment of migrants. An active screening method was used, where the target population is identified using the municipal population registry and receives a postal invitation to attend screening. Compared to status quo, screening and treatment had an ICER of €8966/QALY saved and was therefore reported as cost-effective compared to the authors’ reported WTP threshold of €20 000/QALY. Their baseline HBsAg prevalence was 3.35%, 58% linkage to specialist care and 75% adherence. A study by Rein et al.14 looked at different methods of screening for HBV among the Asian migrant population in the USA. This was a descriptive rather than a formal costeffectiveness analysis, with outcome measures given as cost per person screened. The screening methods analysed included testing at a community clinic and other more active community outreach models, where screening was performed at various events in the Asian community. The costs per person screened ranged from US$ 40 to US$ 280 depending on the method used. Integrating screening into clinical services was found to be the least costly method, but reached the least people, whereas extending screening outside the clinical setting was more costly as it included costs of organizing events and volunteer time, but reached more people. This study provides useful insights into the relative costs of various screening methods and, unlike some of the other studies, it includes full costs including those associated with recruiting patients. However, it does not provide long-term outcomes following on from a positive screening test and is therefore limited in its generalizability. Ruggeri et al.15 looked at screening of all groups defined as “high–risk” (according to local Italian guidelines), and compared the cost-effectiveness of screening followed by treatment for CHB using one of five alternative antiviral drugs. This was compared to the status quo strategy of no screening, but treatment for cirrhosis and HCC stages only. A hypothetic cohort of 100 000 individuals was considered and screening and treatment had an ICER of €17000/QALY.

4.3. Pregnant women The screening of pregnant women for HBsAg (with or without HBeAg testing) in antenatal care (ANC) settings has also been considered in previous cost-effectiveness analyses. However, all these studies consider only the reduction in mother-to-child transmission and benefits to the child (using various outcome measures—cost per case detected, cost per infant carrier prevented or cost per LY gained). None of these ANC studies include onward linkage into care or treatment for the mother, to reduce her risk of progression of liver disease. A full discussion of these studies is therefore not included in this report. Furthermore, many of the studies are older studies published before 2000 (see Hahne review for summary of these

Page | 162

studies18) and performed in HICs (or one in upper-middle income category). They are also heterogeneous in terms of their research question and the baseline strategy under consideration, e.g. Barbosa study is comparing a comprehensive programme to a status quo which already includes screening and birth dose (BD), hepatitis B immune globulin (HBIG) and infant vaccination.19 Fan compares whether to screen for HBeAg or HBV VL in order to guide the use of PPT antiviral therapy in USA.20 Vimloket compared universal neonatal vaccination to screening for HBsAg and HBeAg to stratify whether HBIG is needed in Thailand, using cost per infection averted.21 A full discussion of these studies is therefore not included in this report, as they were unlikely to be useful in helping guide these current recommendations for HBV screening and treatment in ANC settings in LMICs, but would be relevant to consider for reduction of HBV mother-to-child transmission strategies.

5. Drivers of cost-effectiveness From the studies reviewed, some of the main drivers of whether a HBV screening and treatment strategy will be cost-effective are discussed below. This is not meant to provide an exhaustive list of drivers of cost-effectiveness but a descriptive analysis of key considerations, which will hopefully be useful in informing discussions. The main factors influencing the costeffectiveness result are usually presented as the results of one-way sensitivity analyses, meant to be performed over plausible parameter ranges. However, it should be noted that the contribution of each parameter depends on the underlying type of model used and its baseline parameters.

5.1. HBsAg prevalence Although the studies varied in the baseline HBsAg prevalence used in the model, they reported how the cost-effectiveness of the intervention would change over wide HBsAg prevalence ranges. HBsAg prevalence was found to have a relatively small influence on costeffectiveness over the wide ranges tested in most of the studies. General population screening was found to remain cost-effective, i.e. ICER below the respective WTP threshold down to a HBsAg prevalence of 0.3% in the USA8 and 2% in The Gambia (PROLIFICA). Screening of migrants in North America remained cost-effective down to a prevalence of 1– 3%.12,22 Other studies did not explicitly state a prevalence cut-off when the intervention is no longer cost-effective.10,15 It is important to note that “cost-effectiveness” is assessed using differing scales of cost and WTP thresholds between these studies. Therefore, extrapolation of the HIC results to LMICs is difficult, and absolute threshold cut-off for HBsAg prevalence should not be decided on the basis of this literature from HICs. However, the fact that all analyses revealed that a

Page | 163

screen and treat strategy remained cost-effective down to low HBsAg prevalence in the groups analysed increases the confidence of this finding. This has important implications for strategy choice when considering screening in other countries with different prevalence profiles to the study in question. Also, importantly, as prevalence begins to fall as vaccination coverage increases, will it still remain cost-effective to continue screening once prevalence is low, and down to what HBsAg prevalence level does it still remain cost-effective to continue?

5.2. Costs Cost components that need to be considered in economic evaluations of screening and treatment for HBV include costs of screening, diagnostics, monitoring and drugs. This should involve both the cost of consumables as well as other costs including human resource costs (which are included to various extents in different studies). A key driver of cost-effectiveness of a screen and treat strategy reported in some studies is the cost of antiviral drugs.9,11,12 The Rossi study used a drug cost of US$ 8089/year to represent the average cost of tenofovir and entecavir and varied this between US$ 7000 and US$ 9100, changing ICER by US$ 10 000, while still remaining cost-effective. Other costs were less important drivers of cost-effectiveness in their study. In the PROLIFICA study, the generic price of tenofovir (US$ 48) available for use in HIV programmes in SSA23 was used as the base case. It should be noted that this price is not currently available for most countries to treat HBV mono-infection. Using the current pharmaceutical drug price of US$ 20724 was reported to increase the ICER to US$ 1042/LY saved, whilst still remaining below the WTP threshold. Screening costs varied between the studies, and were only found to be drivers of costeffectiveness in the Wong10 and PROLIFICA studies. In the PROLIFICA study, despite an active community-based screening campaign, screening costs were low (US$ 7.43 per person offered screening) and the intervention remained cost-effective even if there was a 3-fold increase in screening costs. The Rein14 study in USA reported costs per person screened between US$ 40 and US$ 280, with the higher costs representing the more active outreach strategies. It should also be noted that in HICs there are different cost components incurred (and included in these studies) for the management of end-stage liver disease, e.g. liver transplant. The cost-effectiveness of screen and treat strategies in HIC settings is partly due to the fact that early management reduces the risk of long-term sequelae, which can incur significant costs, e.g. estimated costs of managing cirrhosis is US$ 9000 per patient per year (pppy) and HCC is US$ 15 000 pppy in the Canadian study by Rossi et al.11 However, in LICs, where there are currently limited options for management of end-stage liver disease (no transplant, limited endoscopy facilities, limited palliative care) and where patients often die at home, with the family as the primary care-giver, the costs of the intervention might not offset the cost avoided of end-stage liver disease. Furthermore, the annual costs of managing liver disease are variable and largely unknown.25 The addition of a societal perspective analysis might be more appropriate in these settings.

Page | 164

5.3. Patient behaviour Adherence to treatment and linkage to care were reported as key drivers of cost-effectiveness in some of the studies.11,13 Veldhuijzen et al. reported that variation in rates of linkage to care and treatment adherence had the largest influence on ICER (ICER varied by about €3000 over the ranges tested—39–75%, 50–100%, for linkage and adherence, respectively). In the PROLIFICA study, variation in treatment adherence was also a key driver of cost-effectiveness. However, rates of linkage into care were reported to be less influential on ICER in this study. The baseline value of linkage into care was high at 81%, likely aided by re-imbursement of transportation fees, clinics held in rural sites to facilitate access to treatment, active reminders about appointments, as well as good sensitization and counselling of screened participants during the study. Linkage into care and adherence rates being drivers of cost-effectiveness should be unsurprising if one considers that in order to gain the health benefits of a screening programme, the infected person needs to start antiviral therapy to reduce their chance of progression to end-stage liver disease. Furthermore, when people drop out at later stages of the care cascade, the impact is reduced, but the initial costs have already been incurred. This highlights the importance of educating patients on the need for continued treatment that has potential implications for successful programmatic implementation. Many barriers exist to successful linkage to care including both health service and patient factors – poor health infrastructure, distance from screening site to health facility, lack of education and patient fear. Uptake of screening is not reported to be a key driver of ICER in the studies; however, this does not imply that high participation levels in screening is not important, as when considering health impact alone, increasing uptake is the key. The implication of this result is that it is likely to be worthwhile performing screening and treatment even if participation screening is assumed to be low. This could be because screening costs are low relative to the costs and health benefits of treatment for those who are infected.

5.4. Age of cohort Age of the cohort screened was reported as a significant driver of cost-effectiveness in the Hutton and Wong studies. The former varied aged of screened cohort from 20 to 60 years, showing variation of ICER of US$ 23 000–US$ 58 000; the latter showed ICER between US$ 60 000 and US$ 136 000 over similar ranges, and Rossi found that the screen and treat intervention is no longer cost-effective if the cohort is over 55 years, with a non-linear relationship between ICER and age. However, despite the finding in HICs that it is more costeffective to screen and treat younger, rather than older people, there are ethical

Page | 165

considerations around using age cut-offs and whether this should be used to guide these type of decisions.

5.5. Disease progression rates Although the HBV models used slightly differing natural history structures and parameter assumptions, most of them showed that the cost-effectiveness was relatively sensitive to variations in disease progression rates used. The Dutch study13 showed that varying parameters between a range representing fast to slower disease progression showed significant variation in ICER between €5000 and €60 000/QALY gained, respectively, a trend which was also seen in other studies.10,12 The Eckman study showed that the ICER was most sensitive to the rate of spontaneous HBeAg seroconversion assumed to be 5% at baseline, but exceeded the WTP threshold if increased to 10%. PROLIFICA study also showed that many of the transition rates were influential on ICER. However, given the complex and heterogeneous natural history of HBV both within and between populations, and lack of natural history progression rate data specific to all populations, this is likely to remain an inherent limitation of all CE models for HBV. However, the ICER did remain below the WTP threshold used in the respective studies for most of the ranges used.

5.6. Effectiveness of antiviral therapy Effectiveness of antiviral therapy was found to be influential on ICER in some studies.10,12 However, different antivirals and different efficacy assumptions (which have often been superceded with more current data) were used by different authors (the older studies often included low-barrier to resistance drugs like lamivudine or interferon, whilst the newer studies mainly used tenofovir or entecavir). Therefore, conclusions as to the influence of these parameters on the result, as well as comparisons between studies have to be interpreted with caution. With the recommendation of the use of newer drugs like tenofovir and entecavir, with similar high efficacy rates and better data on efficacy, model inconsistencies regarding efficacy assumptions should be less of a problem with economic analyses in the future. It needs to be noted that this will be dependent on the assumption that efficacy of antiviral therapy will be the same in HBV infected populations in LMICs as in HICs where most of the efficacy literature originates from.

Page | 166

5.7. Distribution of patients between different disease states The proportion of HBsAg-positive patients with “stable infection”, i.e. CHB not requiring treatment was seen as one of the drivers of cost-effectiveness in some studies. Rossi estimated that 50% of migrants diagnosed with CHB would be eligible for treatment, i.e. they had active chronic infection. They found that the ICER was sensitive to the proportion with stable infection, which when decreased from 70% to 30% increased ICER from US$ 37 000 to US$ 48 000/QALY saved. Veldhuijzen et al. assumed that 26% of HBeAg positive patients and 19% of HBeAg negative patients would be eligible for treatment according to Dutch HBV treatment guidelines, but did not comment on its influence on ICER. In contrast, within the PROLIFICA study, less than 10% of patients were considered eligible for treatment (in states of chronic active hepatitis, compensated cirrhosis or decompensated cirrhosis), and when a lower proportion of HBsAg positive people had stable CHB infection, the ICER decreased, i.e. the intervention became more cost-effective. The explanation for the differences in eligibility criteria is beyond the scope of this current report, but might be partially explained by population characteristics (especially between HBV in Asian and African populations)26,27 and the use of different local guidelines to classify treatment eligibility. The natural history structures are different between models, therefore direct conclusions cannot be drawn from these studies. The proportion of people who would benefit from treatment in a population, is likely to guide cost-effectiveness, but by how much is difficult to quantify based on current evidence and needs further research.

5.8. Others Other drivers of cost-effectiveness included factors that are inherent to some of the techniques used in economic analysis, e.g. health utility values used for QALY assumptions8–10 and discount rate used.10,16 However, these are not discussed further in this report.

6. Limitations of comparing models/generalizability of results WHO recommendations are primarily aimed for use in LMICs. Therefore, most of the studies summarized in this report have to be interpreted with extreme caution as they have mostly been conducted in HICs. The application of results from one setting cannot be translated into another setting. Conclusions drawn by making generalizations of results from costeffectiveness analyses between countries or regions with such differing health-care structures, costs, patient behaviour, disease prevalence profiles and WTP thresholds can be misleading.

Page | 167

Comparison of model results are also hindered by differences in model structures, base-line scenarios used, populations under consideration, costs components included and varying assumptions around models parameters. The most useful health outcome measures to be used for cost-effectiveness analyses are also debated, and vary between studies, as do WTP thresholds. In order to fully answer the question of what the most cost-effective approach is, ideally, a cost-effectiveness analysis is needed which is as specific as possible to the setting being considered as well as the strategies under consideration. However, this is obviously time and labour intensive.

7. Other considerations regarding place of screening 7.1. Community-level Community-level screening could be considered the most active type of case-finding strategy with outreach components and therefore likely the most labour and resource intensive. However, within PROLIFICA, it has been found to be cost-effective, with low screening costs of US$ 7.43 per person offered screening. Various examples of community outreach programmes exist in the field of HIV,28 and comparable strategies could be considered for HBV, with the caveat that “high-risk” groups will not be as applicable to HBV infection.

7.2. Health-care facilities Screening at health-care facilities could include primary-care settings, inpatient and outpatient settings. It could include testing everyone, regardless of the reason for presentation or focus on only those with abnormal liver function tests, abnormal ultrasound scan, family history of liver disease or other clinical suspicion of liver function test. Testing could also be offered in special dedicated clinics, e.g. HIV, STD clinics. A clinically guided testing approach is likely to reveal a higher proportion of people with HBV in highly endemic settings and therefore a lower cost per positive person found. Preliminary data from Mboup et al. (Senegal – verbal communication) where HBsAg screening is performed in the hospital guided by clinical reasons in the health facility (inpatient and outpatient settings), shows that out of 1000 people screened, 567 have been found to be HBsAg-positive (56.7% of those tested). However, when considering performing a cost-effectiveness analysis of health facilitybased screening, the difficulty arises in adjusting for background mortality among those seeking health care. It will depend on many factors, including underlying comorbidities and

Page | 168

age distribution and is likely to be highly heterogeneous between settings. Research into this is ongoing (Hess et al.).

7.2.1. ANC clinics Cost-effectiveness of ANC screening, linkage into care and antiviral treatment for the mother (for the health benefit of the mother, rather than just the child), could be affected by the fact that women have been shown to have slower rates of progression to HCC29 and have lower prevalence of HBsAg than men.30 However, women attending ANC screening are likely to be of a younger age group than those reached by community-based screening, with a longer life expectancy, and therefore can potentially have more impact. The prevalence of HBsAg in women of childbearing age will also depend on the historical vaccination coverage in the country and the percentage of HBeAg-positive mothers will partly depend on the average childbearing age of the country and the rate of HBeAg loss in the region under consideration.31 However, most importantly, since screening of mothers for HBV has benefits to both the mother and child, this is likely to be cost-effective. Since there is variable percentage of attendance to antenatal care depending on the world region (ref), with this being the lowest in sub-Saharan Africa (SSA) (77% of women have at least one ANC visit, only 48% have four ANC visits),32 this approach should also take into factors which will help strengthen ANC coverage in general and awareness campaigns.

7.2.2. Blood banks Blood donor screening for HBV already forms part of WHO recommendations in order to prevent transmission of blood-borne viruses to the recipient.33 However, this is rarely accompanied by the HBsAg positive donor being informed of this positive result, counselled and linked into care for clinical evaluation and treatment.34 As part of the PROLIFICA study, linkage into specialist care for blood donors who had tested HBsAg-positive at the blood bank was performed (Lemoine et al., forthcoming). The main difference found between the cohort of blood donors and those screened in the community were a higher proportion who were tested HBsAg-positive, a majority of whom were males, of younger ages, with a higher proportion requiring treatment and a lower proportion who linked to care. A formal cost-effectiveness analysis has not yet been performed, but these factors are likely to make it even more cost-effective for this cohort, compared to the cohorts who were screened in the community. However, as blood donors form only a small fraction of the population, this strategy is likely to be limited in its reach and population level effectiveness and probably should be seen as a complementary, rather than as an alternative to a wider screening strategy.

Page | 169

7.3. Workplace Other ongoing research in West Africa as part of the PROLIFICA programme includes HBsAg screening in workplaces in Senegal (Mboup et al., unpublished data). Epidemiological and cost-effectiveness studies are underway. Provisional data shows that compared to community screening, there is a higher HBsAg prevalence, higher proportion of males uptaking screening and a higher proportion requiring treatment.

7.4. Others Although other methods of screening are used, to varying levels, worldwide, including screening of health workers, couples pre-marriage, military recruits or pre-employment screening, etc., implementation and guidance of these methods are highly heterogeneous between countries;35 and apart from the study in Iran (below), no data was found regarding their cost-effectiveness. Therefore, they will not be considered here in further detail. The study in Iran36 looked at premarital HBsAg testing, but this was in order to determine whether to offer the partner of someone who is tested HBsAg positive vaccination. This does not include linkage for treatment. Mandatory premarital testing is not policy in many countries and would therefore have limited reach and applicability.

8. Further research needed to fill this information gap More implementational research in LMICs needs to be done to assess feasibility, impact and cost-effectiveness of different screening methods. Further research into the simplification of care, as well as health systems research into integration of hepatitis programmes with other health services (e.g. HIV services), could also help guide how impact can be maximised and cost-effectiveness improved. Ongoing HBV cost-effectiveness screening analyses that are being conducted are as follows:     Screening in OPD settings – Sarah Hess, WHO Screening in ANC – benefits to the mother, Sarah Hess, WHO Screening in ANC – benefits to the child, Jess Howell, Imperial College Screening in work places, Senegal – Shevanthi Nayagam, Imperial College.

9. Conclusions The data on the cost-effectiveness of screening for HBV is lacking, especially in LMICs. Therefore, it is hard to draw conclusions regarding the best screening strategy in terms of who to screen and where to screen, based on cost-effectiveness alone. However, the data that is

Page | 170

available shows that offering screening to the general population with subsequent antiviral treatment strategy is cost-effective in HICs8 as well as LICs,9 even down to a population prevalence as low as 0.3% and 2%, respectively in these studies. Furthermore, screening also has benefits that extend beyond the person screened but also others, e.g. prevention of mother-to-child transmission. Relatively low screening costs, highly effective and relatively low-cost antiviral therapy at generic price and a fraction of HBsAg-positive persons requiring antiviral therapy should help drive the cost-effectiveness of a test and treat strategy. However, this has to be balanced against long-term treatment and the fact that a high proportion with CHB will survive without treatment. Finite treatment courses in certain patient groups are showing promising results and this could help increase cost-effectiveness further.37 Improving country access to generic priced tenfovir for HBV mono-infection in all LMICs is vital to allowing adoption of wide scale HBV treatment programmes. Other strategies for reducing costs further include integration of HBV services into existing health-care structures, particularly in SSA where enormous progress has been made in the scale-up of HIV services, which may be expanded to also deliver HBV interventions using existing infrastructure, trained health-care professionals and field teams. Although general guidance cannot be given based on the evidence, a pragmatic approach is to encourage screening anywhere that it is feasible within the country context, e.g. it can include ANCs, health-care facilities and blood banks. PROLIFICA has shown that population-level screening is feasible and cost-effective in The Gambia, but further research and large-scale implementation studies should be performed to evaluate this further in other high-endemic, low-income settings. Furthermore, HBV screening costs could be shared across other disease programmes, as there are overlapping benefits and synergies with maternal and child health goals and HIV infrastructure and experience. This report aims to summarize key components of the existing literature which has highlighted that apart from the PROLIFICA study in West Africa, there is no data about the cost-effectiveness of screening and treatment in LMICs. Currently, there is not enough literature to make strong recommendations for screening based on cost-effectiveness arguments alone, and further research needs to be done to fill this gap, using similar real life screening data in LMICs like the PROLIFICA project. However, cost-effectiveness analyses form only a small part of guiding public health recommendations, and the overall health impact and key drivers should be considered.

References 1. Schweitzer A, Horn J, Mikolajczyk RT, Krause G, Ott JJ. Estimations of worldwide prevalence of chronic hepatitis B virus infection: a systematic review of data published between 1965 and 2013. Lancet. 2015;386(10003):1546‒55. 2. Papatheodoridis G, Sypsa V, Kantzanou M, Nikolakopoulos I, Hatzakis A. Estimating the treatment cascade of chronic hepatitis B and C in Greece using a telephone survey. J Viral Hepat. 2015;22(4):409‒15. Cohen C, Holmberg SD, McMahon BJ, Block JM, Brosgart CL, Gish RG, et al. Is chronic hepatitis B being undertreated in the United States? J Viral Hepat. 2011;18(6):377‒83.

Page | 171

Allard NL, MacLachlan JH, Cowie BC. The cascade of care for Australians living with chronic hepatitis B: measuring access to diagnosis, management and treatment. Aust N Z J Public Health. 2015;39(3):255‒9. Andermann A, Blancquaert I, Beauchamp S, Dery V. Revisiting Wilson and Jungner in the genomic age: a review of screening criteria over the past 40 years. Bull World Health Organ. 2008;86(4):317‒9. Hahne SJ, Veldhuijzen IK, Wiessing L, Lim TA, Salminen M, Laar M. Infection with hepatitis B and C virus in Europe: a systematic review of prevalence and cost-effectiveness of screening. BMC Infect Dis. 2013;13(1):181. Lemoine et al. Screen and treat as an intervention programme for hepatitis B virus infection in subSaharan Africa: the PROLIFICA experience in the Gambia (Manuscript in progress). Eckman MH, Kaiser TE, Sherman KE. The cost-effectiveness of screening for chronic hepatitis B infection in the United States. Clin Infect Dis. 2011;52(11):1294‒306. Nayagam et al. An economic evaluation of community-based screening and treatment for chronic hepatitis B in the Gambia (Manuscript in progress).

7. 8. 9.

10. Wong WW, Woo G, Jenny Heathcote E, Krahn M. Cost effectiveness of screening immigrants for hepatitis B. Liver Int 2011;31(8):1179‒90. 11. Rossi C, Schwartzman K, Oxlade O, Klein MB, Greenaway C. Hepatitis B screening and vaccination strategies for newly arrived adult Canadian immigrants and refugees: a cost-effectiveness analysis. PLoS One. 2013;8(10):e78548. 12. Hutton DW, Tan D, So SK, Brandeau ML. Cost-effectiveness of screening and vaccinating asian and pacific Islander adults for hepatitis B. Ann Intern Med. 2007;147(7):460‒9. 13. Veldhuijzen IK, Toy M, Hahné SJ, De Wit GA, Schalm SW, de Man RA, et al. Screening and early treatment of migrants for chronic hepatitis B virus infection is cost‒effective. Gastroenterology. 2010;138(2):522‒30. 14. Rein DB, Lesesne SB, Smith BD, Weinbaum CM. Models of community-based hepatitis B surface antigen screening programs in the US and their estimated outcomes and costs. Public Health Rep. 2011;126(4):560‒7. 15. Ruggeri M, Cicchetti A, Gasbarrini A. The cost-effectiveness of alternative strategies against HBV in Italy. Health Policy. 2011;102(1):72‒80. 16. Jazwa A, Coleman MS, Gazmararian J, Wingate LT, Maskery B, Mitchell T, et al. Cost –benefit comparison of two proposed overseas programs for reducing chronic hepatitis B infection among refugees: is screening essential? Vaccine. 2015;33(11):1393‒9. 17. United Nations Statistics Division. UN data – Gambia: Country profile. In: World Statistics Handbook. New York: United Nations; 2014 ( http://data.un.org/CountryProfile.aspx?crName=gambia, accessed 07 June 2016). 18. Hahne SJ, Veldhuijzen IK, Wiessing L, Lim TA, Salminen M, Laar M. Infection with hepatitis B and C virus in Europe: a systematic review of prevalence and cost-effectiveness of screening. BMC Infect Dis. 2013;13:181. 19. Barbosa C, Smith EA, Hoerger TJ, Fenlon N, Schillie SF, Bradley C, et al. Cost‒effectiveness analysis of the National Perinatal Hepatitis B Prevention Program. Pediatrics. 2014; 133(2):243‒53. 20. Fan L, Owusu-Edusei K, Jr, Schillie SF, Murphy TV. Cost-effectiveness of testing hepatitis B-positive pregnant women for hepatitis B e antigen or viral load. Obstet Gynecol. 2014;123(5):929‒37. 21. Vimolket T, Poovorawan Y. An economic evaluation of universal infant vaccination strategies against hepatitis B in Thailand: an analytic decision approach to cost‒effectiveness. Southeast Asian J Trop Med Public Health. 2005;36(3):693‒9.

Page | 172

22. Rossi C, Shrier I, Marshall L, Cnossen S, Schwartzman K, Klein M, et al. Seroprevalence of chronic hepatitis B virus infection and prior immunity in immigrants and refugees: a systematic review and meta-analysis. PLoS One. 2012;7:e44611. 23. HIV/AIDS: Global Price Reporting Mechanism. Geneva: World Health Organization; 2011 (http://apps.who.int/hiv/amds/price/hdd/, accessed 07 June 2016). 24. MSF Access Campaign. Untangling the Web of Antiretroviral Price Reductions. 17th Edn. Geneva: Médecins Sans Fronti res; 2014. 25. Aggarwal R, Ghoshal UC, Naik SR. Assessment of cost‒effectiveness of universal hepatitis B immunization in a low-income country with intermediate endemicity using a Markov model. J Hepatol. 2003;38(2):215‒22. 26. Hadziyannis SJ. Natural history of chronic hepatitis B in Euro-Mediterranean and African countries. J Hepatol. 2011;55(1):183‒91. 27. Lin X, Robinson NJ, Thursz M, Rosenberg DM, Weild A, Pimenta JM, et al. Chronic hepatitis B virus infection in the Asia–Pacific region and Africa: review of disease progression. J Gastroenterol Hepatol. 2005;20(6):833‒43. 28. Suthar AB, Ford N, Bachanas PJ, Wong VJ, Rajan JS, Saltzman AK, et al. Towards universal voluntary HIV testing and counselling: a systematic review and mta‒a nalysis of community-based approaches. PLoS Med. 2013;10(8):e1001496. 29. Kirk GD, Lesi OA, Mendy M, Akano AO, Sam O, Goedert JJ, et al. The Gambia Liver Cancer Study: Infection with hepatitis B and C and the risk of hepatocellular carcinoma in West Africa. Hepatology. 2004;39(1):211‒9. 30. Blumberg BS. Sex differences in response to hepatitis B virus. I. History. Arthritis Rheum. 1979;22(11):1261‒6. 31. Ott JJ, Stevens GA, Wiersma ST. The risk of perinatal hepatitis B virus transmission: hepatitis B e antigen (HBeAg) prevalence estimates for all world regions. BMC Infect Dis. 2012;12:131. 32. Global Health Observatory Data Repository. Geneva: World Health Organization; 2016 (http://apps.who.int/gho/data/node.main, accessed 07 June 2016). 33. WHO Global Database on Blood Safety Summary Report. Geneva: World Health Organization; 2011 (http://www.who.int/bloodsafety/global_database/GDBS_Summary_Report_ 2011.pdf, accessed 07 June 2016). 34. Allain JP, Opare-Sem O, Sarkodie F, Rahman R, Owusu-Ofori S. Deferred donor care in a regional hospital blood center in Ghana. Transfusion. 2009;49(4):669‒75. 35. Hatzakis A, Van Damme P, Alcorn K, Gore C, Benazzouz M, Berkane S, et al. The state of hepatitis B and C in the Mediterranean and Balkan countries: report from a summit conference. J Viral Hepat. 2013;20(Suppl2):1‒20. 36. Adibi P, Rezailashkajani M, Roshandel D, Behrouz N, Ansari S, Somi M H, et al. An economic analysis of premarriage prevention of hepatitis B transmission in Iran. BMC Infectious Diseases. 2004;4(31). 37. Berg T, Simon KG, Mauss S, Schott E, Heyne R, Klass D, et al. O119. Stopping tenofovir disoproxil fumarate (TDF) treatment after long term virologic suppression in HBeAg-negative CHB: week 48 interim results from an ongoing randomized, controlled trial (“finite CHB”). J Hepatol. 2015;62(2):S253.

Page | 173

Annex 5.2 PICO 2 - Who to test (HCV) Literature review on cost-effectiveness of HCV screening, treatment strategies and applicability to LMICs

Benjamin Linas (Team Lead), Jake Morgan Boston University School of Medicine, Boston, USA

Page | 174

1. Executive summary We conducted a targeted review of the literature to determine the state of evidence about the cost–effectiveness of testing for HCV in different types of epidemics and among different risk groups. We provide a qualitative assessment of conclusions. 5. Testing in high-risk groups such as persons who inject drugs (PWID), men who have sex with men (MSM), prisoners, HIV-infected persons, and commercial sex workers is likely to be cost–effective. Testing in settings with a high prevalence of high-risk patients is almost certainly cost–effective in all locations. It is important, however, to ensure adequate follow up after diagnosis. 6. The best approach to testing outside of high-risk risk groups depends a great deal on a country’s unique HCV epidemiology. Most countries have at least some component of “birth cohort” epidemic, and “birth cohort” testing is likely cost–effective in most settings. 7. Routine testing of the entire population carries two risks. First, when the HCV epidemic is concentrated to a specific age or risk group, generalized testing can dilute the testing effort and reduce the number of HCV cases identified. Second, if an epidemic is highly concentrated with a specific risk or demographic group, screening outside of that group can be inefficient and increase cost. Countries with high HCV prevalence across the entire population should implement routine screening, but in most epidemics, routine screening in the entire population is likely not be cost-effective. The specific threshold at which a country should alter its approach to routine testing, however, is a function of multiple factors and cannot be identified more generally.

2. Background Hepatitis C virus (HCV) is a global public health burden and major cause of morbidity and mortality including liver failure and hepatocellular carcinoma.1,2 Current global HCV seroprevalence is estimated to be 2.8%, or more than 185 million infected individuals worldwide.3 Historically, it has been very difficult to treat HCV and most cases of HCV have gone unidentified. The advent of high-efficacy, low-duration therapy, however,4 generates new enthusiasm for testing for HCV infection, linking infected patients to care, and curing HCV before patients begin to experience the consequences of cirrhosis and end-stage liver disease. It is not clear, however, exactly who should be targeted for HCV testing. Similar to the conversation around HIV testing, there are several approaches to screening for HCV that may provide high yield and improve outcomes including: 1) targeted testing of the highest-risk groups, 2) routine testing among specific demographic groups that are readily identified and who have a high prevalence of HCV infection, and 3) routine testing throughout the entire population. This review develops a rubric by which to measure and characterize the HCV epidemic within a country, surveys the literature about the cost–effectiveness of screening for HCV in various populations, and discusses how epidemiology within a country should inform decision-making about who to test for HCV.

Page | 175

3. Overview of report This report does not represent the results of a full systematic review. It is meant to serve as a summary of existing studies on cost–effectiveness of screening and treatment for HCV, with an analytic summary of key considerations. It was envisaged that there was a lack of relevant literature in low- and middle-income countries (LMICs), so existing studies from high-income countries (HICs) are described and their potential uses and limitations, when drawing conclusions are discussed.

4. Summary of global HCV epidemiology Generally, HCV epidemics around the world are heterogeneous and represent mixtures of three core epidemic components: 4. Infection related to high-risk behaviours: In essentially every geographical region, the highest prevalence of HCV infection is among persons who use injection drugs (PWID). 5,6 The prevalence of injection drug use differs between countries and regions, but within those who do inject drugs, HCV prevalence is nearly universally high. Commercial sex workers and prisoners also have increased prevalence (presumably related to both drug use and perhaps sexual transmission),7,8 as do men who have sex with men, especially those who are HIV infected.9 In many cohorts of PWID in North America, Europe, and Asia, HCV prevalence ranges from 30% to 75%. 5. Infection related to past generalized exposures that have since been identified and removed: This epidemic pattern, in which there is a high prevalence of HCV within a given age group, is commonly referred to as a “birth cohort epidemic”.10 While typically identified as being the infection pattern in North America and Europe, many nations have some element of birth cohort epidemics with their unique HCV epidemiology (Table 1).11 Birth cohort epidemics reflect an HCV exposure source that was once present and to which a large portion of the population was exposed, but that has since been identified and removed. For example, before it was identified and sequenced, HCV infected the blood supply of many countries in all regions of the world. When the blood supply began to be screened for the presence of HCV, the exposure was removed. As a result, the incidence of HCV fell dramatically among the general population, but there remains a burden of prevalent, chronic HCV among patients who were alive and likely to get a blood transfusion during the time that HCV existed in the blood supply. 6. Generalized population epidemic: This pattern is related to a widespread exposure, often iatrogenic, that results in high prevalence (8–10%) across essentially all age groups. Note that the primary difference between a “birth cohort” pattern and a generalized pattern of infection is the duration of time that the generalized exposure existed and whether it has been removed or mitigated. An example of a generalized exposure is the common use of reusable hypodermic syringes and needles in medical settings without adequate sterilization between uses. Few epidemics fall into one of the above three categories. Rather, most are mixed, and represent some combination of all components (Table 1). The nature of an epidemic

Page | 176

within a specific country determines a great deal about the appropriate approaches to who to screen.

Epidemic scenarios HCV Generalized

Definition

Disaggregation

Country example

High (>5%)

With birth cohort

(N=23) Cameroon, CAR, Armenia, Egypt, Liberia, Gabon, Guinea, Ghana, Guinea-Bissau, Mongolia, Sierre Leone, Uzbekistan, Cape Verde, Chad, Mali, Niger, Nigeria, Pakistan, San Tome et Principe, Senegal, Togo, Burkina Faso, Georgia

Without birth cohort High intermediate (3–5%) With birth cohort (N=14) Angola, Bahrain, Congo, Democratic Republic of the Congo, Equatorial Guinea, Estonia, Lebanon, Moldova, Russia, Taiwan, Turkmenistan, Ukraine, United Arab Emirates, Oman

Without birth cohort Low intermediate (2–3%) With birth cohort (N=44) American Samoa, Anguilla, Azerbaijan, Benin, Bermuda, British Virgin Isles, Cayman Islands, Cook Islands, Cote d'Ivoire, Falkland Isalnds, Faroe Islands, Gibraltar, Greenland, Holy See, Hong Kong, Iraq, Isle of Man, Jordan, Kazakstan, Kuwait, Kyrgzstan, Latvia, Lichtenstein, Lithuania, Macau, Monaco, Monserrat, Nauru, Niue, Northern Mariana, Islands, Palau, Palestine, Romania, St Helena, St Kitts and Nevis, St Pierre Miquelon, San Marino, St Martin, Tajikistan, Thailand, Tokelau, Turks and Kakos, Tuvalu, Wallis and Futuna

Without birth cohort Mixed Generalized population prevalence, low, moderate or high with a sizeable risk population (PWID) High generalized High intermediate generalized Low intermediate generalized Low (1–2%) with PWID (N=3) Pakistan, Egypt, Uzbekistan (N=5) Estonia, Kazakstan, Taiwan, Turkmenistan, Ukraine (N=8) Hong Kong, Latvia, Lithuania, Palestine, Romania, Thailand, Tajikistan, Syria (N=46) Albania, Algeria, Argentina, Australia, Belarus, Bhutan, Bosnia, Brazil, Cambodia, Chile, China, Colombia, Costa Rica, Croatia, El Salvador, Greece, Honduras, Israel, Italy, Japan, Kenya, Former Yugoslav Republic of Macedonia, Malaysia, Mauritius, Mexico, Montenegro, Morocco, Myanmar, Nepal, New Zealand, Nicaragua, Panama, Paraguay, Portugal, Puerto Rico, Serbia, Slovakia, Slovenia, South Sudan, Spain, Switzerland, United States, United States Virgin Islands, Uruguay, Viet Nam, Yemen

Page | 177

Low (1–2 %) without PWID

(N=8) Botswana, Ethiopia, The Gambia, Guadalope, Lesotho, Rwanda, Saudi Arabia, Singapore (N=21) Andora, Antigua Comorres, Curacao, Democratic People’s Republic of Korea, Dominica, French Guinea, French Polynesia, Guatemala, Martinique, Mauritania, Mayotte, Namibia, New Caledonia, Reunion, Seychelles, South Africa, Sri Lanka, Swaziland, Western Sahara, Zimbabwe

Concentrated

Generalized population prevalence <1% with high-risk groups

(N=68) Afghanistan, Aruba, Austria, Bahamas, Bangladesh, Barbados, Belgium, Belize, Bolivia, Brunei, Bulgaria, Burundi, Canada, Cuba, Cyprus, Czech Republic, Denmark, Djibouti, Dominican Republic, Ecuador, Eritrea, Fiji, Finland, France, Germany, Grenada, Guyana, Haiti, Hungary, Iceland, India, Indonesia, Iran, Ireland, Jamaica, Kiribati, Korea Republic of, Laos, Libya, Luxembourg, Madagascar, Maldives, Marshall Islands, Micronesia, Mozambique, Netherlands, Norway, Papua New Guinea, Peru, Philippines, Poland, Qatar, St. Lucia, Samoa, Solomon Islands, Somalia, Surinam, Sweden, Tanzania, TimorLeste, Tonga, Trinidad & Tobago, Tunisia, Turkey, Uganda, United Kingdom, Vanuatu, Zambia

5. Summary of the literature – Who to screen? What is the evidence base from modelling of the impact and cost – effectiveness of different screening approaches using different prevalence thresholds? Testing in high-risk groups, including persons who inject drugs, MSM, prisoners, HIV-infected persons, and commercial sex workers

5.1. Persons who inject drugs Multiple analyses in many geographical regions concur that routine testing for HCV in venues with a high prevalence of persons who inject drugs is cost–effective, even when the studies assume very poor follow-up rates and limited access to therapy.12,13 Further, dynamic HCV transmission models suggest that aggressive diagnosis and treatment among current drug users could reduce the incidence of HCV – “cure as prevention”.14,15 With typical prevalence estimates of 40%, but ranging as high as 75% in some cohorts, routine screening for HCV is almost certainly cost–effective. It is essential to consider the HCV cascade of care when screening recent or current drug users. Modelling studies demonstrate that even 100%

Page | 178

effective HCV therapy has almost no impact on population-level outcomes without efforts to significantly improve the number of HCV-infected patients who initiate therapy.16

5.2. Men who have sex with men Men who have sex with men (MSM) are also at an increased risk of HCV incidence, particularly if they are also HIV-positive. Since the mid-2000s, outbreaks have surfaced in the US, Europe and Australia among HIV-positive MSM.17,18 Cost–effectiveness modelling has found testing using liver function tests in combination with HCV Ab testing to be cost–effective in the HIVpositive MSM population.19 Preliminary studies have suggested that core-antigen testing has the potential to be cost–effective in this population as well.20 The results of these studies are dependent on appropriate linkage to effective therapy and retention in care. These studies do not fully account for either the reduction in secondary transmission or the possibility of reinfection in high-risk groups.

5.3. Prisoners Prisons likely have high HCV prevalence as the result of a high prevalence of persons who inject drugs in prisons. One challenge to HCV testing in prisons is that new treatments for HCV are costly and many prisoners do not have access to new therapies. A UK-based study, however, found that HCV case detection, using dried blood spot testing, was cost–effective, even when the model assumed low rates of HCV treatment initiation.21 A second study concurs that screening in prisons can be cost–effective, but this study concluded that targeting screening to those prisoners with a history of injection drug use improves cost– effectiveness22 A later study by several of the same authors found that routine screening of all prisoners is not cost–effective, although that study found significant uncertainty in the results. If the prison population had more advanced disease at the time of screening, or if the rate of HCV disease progression is faster than estimated in the base case, then routine screening in prisons can be cost–effective.23

5.4. HIV-infected persons Screening for incident and acute HCV in HIV-infected MSM is likely cost–effective.19 Although nearly every guideline for HCV care recommends HCV screening at enrolment in care, we were not able to find a cost-effectiveness analysis that answers the specific question of the cost-effectiveness of HCV testing at enrollment in HCV care. Because the prevalence of HCV is known to be high in HIV-infected persons, the pace of progression of fibrosis in HIV/HCV-

Page | 179

coinfected patients is high, and new therapies to treat HCV are effective in HIV/HCV coinfection, testing for HCV at enrolment in HCV is almost certainly cost–effective.

5.5. Sex workers We were not able to find a study that addresses the cost–effectiveness of HCV testing in sex workers. Because many sex workers are also PWID or non-injection drug users, the prevalence of HCV in this group is likely high. It is not clear at this time, however, if it is cost–effective to routinely screen all sex workers, compared to an approach that targets testing to sex workers who report a history of injection drug use.

5.6. Testing among easily identified age or demographic groups known to have high HCV prevalence (“birthcohort testing”) Whenever there is an easily identified demographic group that has high HCV prevalence (for example, all individuals born in a certain time period) it is likely cost–effective to routinely test for HCV within that cohort. Several cost–effectiveness studies in the US estimate incremental cost–effectiveness ratios of “birth cohort” screening that are below commonly cited willingness to pay thresholds for resource-rich countries.10, 24–5 Each of these studies compared “birth cohort testing” to the current standard of care, and shared the same qualitative conclusions. Similarly, one study from Portugal found that birth cohort testing was cost–effective in that country.26 Notably, based on Portugal’s local HCV epidemiology, the cohort to test is not identical to that in the US. Routine screening is preferred to targeted screening because providers are often not skilled at identifying high-risk behaviours, and because for many patients in such an epidemic, the “risk” to target is simply being a member of a high-prevalence age cohort (i.e. there are no specific behavioural risks to identify).

5.7. Testing among the general population without attempt to identify high-risk behaviours or characteristics (“routine testing”) At this time, no jurisdiction of which we are aware recommends routine testing for all individuals regardless of demographics or specific behavioural risk. The data about population screening typically come from HICs such as the US and UK, and such studies find that routine testing in the general population is not cost–effective. For example, one cost–effectiveness analysis, conducted in the US context, found that when the general population prevalence of

Page | 180

HCV infection exceeded 0.53%, the incremental cost–effectiveness ratio of routine universal screening compared to targeted screening was far below commonly cited US willingness to pay thresholds.27 When compared to “birth cohort testing,” however, universal testing resulted in worse outcomes and higher costs than the birth cohort approach. This analysis raises the spectre that in countries whose HCV epidemic is largely concentrated to a specific birth cohort or demographics group, attempting to identify cases by routine testing of the entire population can dilute the testing effort and result in fewer cases of HCV being identified. Similarly, a recent study in the US context investigated the cost–effectiveness of two approaches to testing for HCV in average-risk, asymptomatic adults accessing primary care: a) HCV EIA followed by quantitative RNA for those with positive EIA results, and b) quantitative RNA for all patients. Neither strategy was cost–effective.28 An older study, conducted in the UK, also found that although screening high-risk groups in primary care settings was cost–effective, extending screening beyond high-risk individuals was not.29 Importantly, however, that study pre-dates the existence of effective, antiviral therapy targeting HCV. Higher efficacy of therapy could improve cost–effectiveness conclusions. Similarly, a study conducted in Japan found that routine testing of the population was cost–effective compared to “no screening.”30 That paper, however, did not consider a birth cohort approach, and the conclusions therefore are not certain. Another analysis, conducted in Italy, used Markov modelling to compare “testing” to “no testing” among patients who had undergone surgery. They found that testing was not cost–effective in this group.31 Notably, individuals who have undergone surgery are more likely to have had exposure to blood products, and therefore likely have a higher HCV prevalence then the general population. If screening among these patients was not cost– effective, screening in even lower prevalence groups, such as the general population, will also not be cost–effective. There is one recent study, conducted in Canada, that found that one-time testing of patients outside the “birth cohort” of those aged 65 years or older would be cost–effective by Canadian standards.32 It is difficult from that manuscript to determine the epidemiological assumptions that led to this finding, which differs from most US-based studies. One assumption that could have influenced the results was that early-stage HCV had a low quality of life, which tends to make screening and treating HCV more cost–effective. Similarly, a modelling study based on ten years of retrospective data at a London antenatal clinic found that routine testing for HCV for pregnant women was cost–effective, even at baseline prevalence levels as low as 0.1%.33 This contradicts the findings of a 2005 paper based in the US, which found that screening of asymptomatic pregnant women in the US, even when coupled with elective caesarean delivery to minimize antenatal transmission risk, was not cost–effective.34 It also contradicts a paper based in the Netherlands that found that adding routine one-time testing for HCV in antenatal clinics would not be cost–effective.35 These disparate findings may be influenced by estimates of fibrosis progression, discounting rates, and health-care costs in each country.

Page | 181

Importantly, all of the above studies reflect the epidemiology of HCV in HICs. One recent paper explicitly studied the cost–effectiveness in Egypt of one-time, routine screening for HCV followed by treatment with either pegylated interferon and ribavirin (PEG-RBV) or PEG-RBV plus an HCV protease inhibitor.36 Given the very high prevalence of disease, screening was always cost–effective, and often cost-saving. It is important to consider, however, that assumptions about linkage to HCV care and availability of treatment after diagnosis impact cost–effectiveness conclusions. If general population screening will likely identify many cases of HCV, but those who are infected have limited options for treatment, screening may not be cost–effective.

Drivers of cost–effectiveness

The main benefit of testing is identifying cases of HCV before they lead to the sequelae of endstage liver disease; the resource implications of testing broadly are important. First, the cost of testing itself is not trivial. Second, if the testing strategy (i.e. the laboratory protocol one uses to identify HCV exposure and test for HCV viraemia) results in a large number of falsepositive tests, the cost of unnecessary HCV therapy could be very large. At the same time, trying to “over target” testing to only the highest-risk groups can be detrimental to public health. Many high-risk behaviours are stigmatized and underreported, and health-care workers are not always skilled at identifying high-risk behaviours. Balancing these considerations is a challenge, and requires country-level determination of best approaches. General themes that should inform the screening approach include the following: 1. HCV prevalence – screening provides increasing value as prevalence rises. In one US based study, screening was cost–effective (compared to no screening) at a US willingness to pay threshold down to prevalence of 0.53%.27 Importantly, however, choice of comparator impacts the incremental cost–effectiveness ratio of routine population testing. When routine testing was compared to “birth cohort testing” in that same paper, routine screening diluted the screening effort in the cohort with the highest prevalence of HCV and therefore resulted in fewer cases of HCV identified and higher cost than “birth cohort testing.” However, in any population subgroup that has HCV prevalence >1%, it is likely that some form of testing is cost–effective. The question in such scenarios is whether to routinely screen, or to attempt to identify risk and target screening to that group. 2. Degree of concentration of the epidemic – to the extent that an epidemic is concentrated to a specific risk or demographic group, targeting screening to that group becomes more cost–effective. This dynamic is most directly at play when considering “birth cohort testing.” Being a member of a birth cohort is easily determined and generally carries no stigma. Thus, targeting testing to birth cohorts is feasible and often cost–effective. In countries with a strong birth cohort dynamic, birth cohort screening is likely preferred. To the extent that epidemics are concentrated among high-risk groups such as PWID, however, targeted testing is more challenging. Because HCV risk behaviours are stigmatized and underreported, trying to identify high-risk individuals is difficult and prone to under-testing high-risk patients.

Page | 182

3. Treatment rates – screening clearly becomes less cost–effective when identified patients cannot link to effective therapy. US-based analyses typically assume availability of interferon-free regimens to cure HCV. If such treatments are not available, or only available to a limited proportion of identified cases, then the incremental cost– effectiveness ratio of screening increases. 4. Assumptions about the HCV cascade of care – similar to treatment rates, loss to follow up has an important impact on cost–effectiveness conclusions. As the proportion of patients with identified HCV infection who successfully link to HCV care decreases, the incremental cost–effectiveness ratio of screening also goes up. 5. Cost of testing – the cost of testing may impact the cost–effectiveness of one testing strategy compared to another (i.e. which tests to use and in what order), but it has little impact on the cost–effectiveness conclusions about who to screen. In one study conducted in the US, ranging the cost of testing by as much as 50% in either direction had no impact on cost–effectiveness conclusions.37 6. Efficacy of HCV therapy – the cost–effectiveness of HCV testing depends in part on the efficacy of HCV treatment. This dynamic is easily demonstrated by a hypothetical scenario, in which patients with identified HCV do not receive any therapy (efficacy = 0%). In such a case, the incremental benefit of screening would be zero, and the incremental cost– effectiveness ratio would approach infinity (no value). 7. Cost of HCV therapy – as the cost of treatment increases, the incremental cost– effectiveness ratio also increases. This is not surprising, as the cost of therapy has no impact on clinical outcomes (denominator of the cost–effectiveness ratio), but does increase cost (the numerator of the cost–effectiveness ratio). For example, in one US study, the incremental cost–effectiveness ratio of “birth cohort testing” compared to no testing increased more than 100% when one assumed treatment with pegylated interferon, ribavirin, and an HCV protease inhibitor compared to pegylated interferon and ribavirin alone. 8. Estimates of quality of life with early-stage HCV – if early-stage HCV has a large impact on quality of life, then testing (via any approach) becomes more cost–effective. If early-stage HCV has little impact on quality of life, then the benefits of testing accrue only to the minority of patients who become cirrhotic, and only in the distant future when those patients begin to experience complications of end-stage liver disease. In contrast, if earlystage HCV has an immediate impact on quality of life, then every patient with identified and cured HCV accrues lifetime benefits that greatly increase the benefits of testing. 9. HCV fibrosis progression rates – most of the sequelae of chronic HCV infection and essentially all HCV-attributable mortality, accrue only when a patient has reached cirrhosis. The time from HCV infection to development of cirrhosis is highly variable and can be as long as 25 years. Some patients never become cirrhotic. Faster rates of fibrosis progression tend to make testing for HCV more cost–effective, because faster fibrosis progression results in a larger proportion of the population experiencing sequelae of HCV.

Page | 183

7. What further research needs to be done to fill this information gap It is important to collect accurate epidemiological data to better inform decision-making around HCV testing. A formal cost–effectiveness analysis that compares “targeted” vs “birth cohort” vs “routine” testing requires estimates of the prevalence of high-risk behaviours, stratified by age, the prevalence of HCV among those with high- and low-risk behaviours, stratified by age, and the age structure of the population. Further, it is important to know the cost of both HCV therapy in a country, as well as the costs associated with untreated HCV and end-stage liver disease. In addition, more implementation research in LMICs is needed to determine the degree to which providers can accurately identify and test high-risk patients when employing a targeted approach, as well as estimate of linkage to HCV care and the HCV cascade.

8. Conclusions – who should be tested for HCV?  Testing in high-risk groups such as PWID, MSM, prisoners, HIV-infected persons, and commercial sex workers is likely cost–effective. Testing in settings with a high prevalence of high-risk patients is almost certainly cost–effective in all locations. It is important, however, to ensure adequate follow up after diagnosis. The best approach to testing outside of high-risk risk groups depends a great deal on a country’s unique HCV epidemiology. Most countries have at least some component of “birth cohort” epidemic, and “birth cohort” testing is likely cost–effective in most settings. Routine testing of the entire population carries two risks. First, when the HCV epidemic is concentrated to a specific age or risk group, generalized testing can dilute the testing effort and reduce the number of HCV cases identified. Second, if an epidemic is highly concentrated within a specific risk or demographic group, screening outside of that group can be inefficient and increase costs. Countries with high HCV prevalence across the entire population should implement routine screening, but in most epidemics, routine screening in the entire population may not be cost–effective. The specific threshold at which a country should alter its approach to routine testing, however, is a function of multiple factors and cannot be identified more generally.



References 1. de Oliveria Andrade LJ, D'Oliveira A, Melo RC, De Souza EC, Costa Silva CA, Paraná R. Association between hepatitis C and hepatocellular carcinoma. J Glob Infect Dis. 2009;1(1):33–7. doi: 10.4103/0974-777X.52979. [PMID: Accession Number]; PubMed Central. 2. Cooke GS, Lemoine M, Thursz M, Gore C, Swan T, Kamarulzaman A et al. Viral hepatitis and the Global Burden of Disease: a need to regroup. J Viral Hepat. 2013;20(9):600 –1. doi: 10.1111/jvh.12123. [PMID: Accession Number]; PubMed Central.

Page | 184

Hanafiah KM, Groeger J, Flaxman AD, Wiersma ST. Global epidemiology of hepatitis C virus infection: new estimates of age-specific antibody to HCV seroprevalence. Hepatology. 2013;57(4):1333–42. doi: 10.1002/hep.26141. [PMID: Accession Number]; PubMed Central. Younossi Z, Henry L. Systematic review: patient-reported outcomes in chronic hepatitis C – the impact of liver disease and new treatment regimens. Aliment Pharmacol Ther. 2015;41(6):497–520. Epub 2015/01/24. doi: 10.1111/apt.13090. [PMID: Accession Number]; PubMed Central. Johnston L, Saumtally A, Corceal S, Mahadoo I, Oodally F. High HIV and hepatitis C prevalence amongst injecting drug users in Mauritius: findings from a population size estimation and respondent driven sampling survey. Int J Drug Policy. 2011;22(4):252 –8. doi: 10.1016/j.drugpo.2011.05.007. [PMID: Accession Number]; PubMed Central. Nelson PK, Mathers BM, Cowie B, Hagan H, Des Jarlais D, Horyniak D et al. Global epidemiology of hepatitis B and hepatitis C in people who inject drugs: results of systematic reviews. Lancet. 2011;378(9791):571–83. doi: 10.1016/S0140-6736(11)61097-0. [PMID: Accession Number]; PubMed Central PMC3285467: PMC3285467. Alonso M, Gutzman A, Mazin R, Pinzon CE, Reveiz L, Ghidinelli M. Hepatitis C in key populations in Latin America and the Caribbean: systematic review and meta-analysis. Int J Public Health. 2015;60 (7):789–98. doi: 10.1007/s00038-015-0708-5. [PMID: Accession Number]; PubMed Central. Chen Y, Shen Z, Morano JP, Khoshnood K, Wu Z, Lan G et al. Bridging the epidemic: a comprehensive analysis of prevalence and correlates of HIV, hepatitis C, and syphilis, and infection among female sex workers in Guangxi Province, China. PloS One. 2015;10(2):e0115311. doi: 10.1371/journal.pone.0115311. [PMID: Accession Number]; PubMed Central PMC4344209: PMC4344209. Hagan H, Jordan AE, Neurer J, Cleland CM. Incidence of sexually transmitted hepatitis C virus infection in HIV-positive MSM: a systematic review and meta-analysis. AIDS. 2015;29 (17):2335–45. doi: 10.1097/QAD.0000000000000834. [PMID: Accession Number]; PubMed Central.

10. Rein DB, Smith BD, Wittenborn JS, Lesesne SB, Wagner LD, Roblin DW et al. The cost-effectiveness of birth-cohort screening for hepatitis C antibody in US primary care settings. Ann Intern Med. 2012;156(4):263–70. doi: 10.7326/0003-4819-156-4-201202210-00378. [PMID: Accession Number]; PubMed Central. 11. Mohd Hanafiah K, Groeger J, Flaxman AD, Wiersma ST. Global epidemiology of hepatitis C virus infection: new estimates of age-specific antibody to HCV seroprevalence. Hepatology. 2013;57(4):1333–42. doi: 10.1002/hep.26141. [PMID: Accession Number]; PubMed Central. 12. Schackman BR, Leff JA, Barter DM, DiLorenzo MA, Feaster DJ, Metsch LR et al. Cost-effectiveness of rapid hepatitis C virus (HCV) testing and simultaneous rapid HCV and HIV testing in substance abuse treatment programs. Addiction. 2015;110(1):129 –43. doi: 10.1111/add.12754. [PMID: Accession Number]; PubMed Central 4270906: 4270906. 13. Castelnuovo E, Thompson-Coon J, Pitt M, Cramp M, Siebert U, Price A et al. The cost-effectiveness of testing for hepatitis C in former injecting drug users. Health Technol Assess. 2006;10(32):iii –iv, ix-xii, 1–93. [PMID: Accession Number]; PubMed Central. 14. Grebely J, Matthews GV, Lloyd AR, Dore GJ. Elimination of hepatitis C virus infection among people who inject drugs through treatment as prevention: feasibility and future requirements. Clin Infect Dis. 2013;57(7):1014–20. doi: 10.1093/cid/cit377. [PMID: Accession Number]; PubMed Central. 15. Martin NK, Vickerman P, Grebely J, Hellard M, Hutchinson SJ, Lima VD et al. Hepatitis C virus treatment for prevention among people who inject drugs: modeling treatment scale-up in the age of direct-acting antivirals. Hepatology. 2013;58(5):1598–609. Epub 2013/04/05. doi: 10.1002/hep.26431. [PMID: Accession Number]; PubMed Central 3933734: 3933734.

Page | 185

16. Linas BP, Barter DM, Leff JA, Assoumou SA, Salomon JA, Weinstein MC et al. The hepatitis C cascade of care: identifying priorities to improve clinical outcomes. PloS One. 2014;9(5):e97317. doi: 10.1371/journal.pone.0097317. [PMID: Accession Number]; PubMed Central 4026319: 4026319. 17. Yaphe S, Bozinoff N, Kyle R, Shivkumar S, Pai NP, Klein M. Incidence of acute hepatitis C virus infection among men who have sex with men with and without HIV infection: a systematic review. Sex Transm Infect. 2012;88(7):558–64. doi: 10.1136/sextrans-2012-050566. [PMID: Accession Number]; PubMed Central. 18. van de Laar T, Pybus O, Bruisten S, Brown D, Nelson M, Bhagani S et al. Evidence of a large, international network of international hepatitis C virus transmission in HIV-positive men who have sex with men. Gastroenterology. 2009;136(5):1609–17. [PMID: Accession Number]; PubMed Central. 19. Linas BP, Wong AY, Schackman BR, Kim AY, Freedberg KA. Cost-effective screening for acute hepatitis C virus infection in HIV-infected men who have sex with men. Clin Infect Dis. 2012;55(2):279–90. doi: 10.1093/cid/cis382. [PMID: Accession Number]; PubMed Central PMC3403839: PMC3403839. 20. Cresswell FV, Fisher M, Hughes DJ, Shaw SG, Homer G, Hassan-Ibrahim MO. Hepatitis C core antigen testing: a reliable, quick and potentially cost-effective alternative to hepatitis C polymerase chain reaction in diagnosing acute hepatitis C virus infection. Clin Infect Dis. 2015;60 (2):263–6. doi: 10.1093/cid/ciu782. [PMID: Accession Number]; PubMed Central. 21. Martin NK, Hickman M, Miners A, Hutchinson SJ, Taylor A, Vickerman P. Cost-effectiveness of HCV case-finding for people who inject drugs via dried blood spot testing in specialist addiction services and prisons. BMJ Open. 2013;3(8). doi: 10.1136/bmjopen-2013-003153. [PMID: Accession Number]; PubMed Central PMC3752052: PMC3752052. 22. Sutton AJ, Edmunds WJ, Gill ON. Estimating the cost-effectiveness of detecting cases of chronic hepatitis C infection on reception into prison. BMC Public Health. 2006;6:170. doi: 10.1186/14712458-6-170. [PMID: Accession Number]; PubMed Central PMC1543636: PMC1543636. 23. Sutton AJ, Edmunds WJ, Sweeting MJ, Gill ON. The cost-effectiveness of screening and treatment for hepatitis C in prisons in England and Wales: a cost-utility analysis. J Viral Hepat. 2008;15(11):797–808. doi: 10.1111/j.1365-2893.2008.01008.x. [PMID: Accession Number]; PubMed Central. 24. McGarry LJ, Pawar VS, Panchmatia HR, Rubin JL, Davis GL, Younossi ZM et al. Economic model of a birth cohort screening program for hepatitis C virus. Hepatology. 2012;55(5):1344 –55. doi: 10.1002/hep.25510. [PMID: Accession Number]; PubMed Central. 25. Liu S, Cipriano LE, Holodniy M, Goldhaber-Fiebert JD. Cost-effectiveness analysis of risk-factor guided and birth-cohort screening for chronic hepatitis C infection in the United States. PloS One. 2013;8(3):e58975. doi: 10.1371/journal.pone.0058975. [PMID: Accession Number]; PubMed Central PMC3606430: PMC3606430. 26. Brogueira P, Costa A, Miranda A, Peres S, Baptista T, Aldir I et al. Improve screening of HCV infection by targeting high prevalence aged groups: analysis of a cohort of HCV and HIV co-infected patients. J Int AIDS Soc. 2014;17(4 Suppl 3):19601. doi: 10.7448/IAS.17.4.19601. [PMID: Accession Number]; PubMed Central PMC4224779: PMC4224779. 27. Coffin PO, Scott JD, Golden MR, Sullivan SD. Cost-effectiveness and population outcomes of general population screening for hepatitis C. Clin Infect Dis. 2012;54(9):1259 –71. doi: 10.1093/cid/cis011. [PMID: Accession Number]; PubMed Central PMC3404694: PMC3404694.

Page | 186

28. Singer ME, Younossi ZM. Cost effectiveness of screening for hepatitis C virus in asymptomatic, average-risk adults. Am J Med. 2001;111(8):614–21. doi: http://dx.doi.org/10.1016/S00029343(01)00951-2. [PMID: Accession Number]; PubMed Central. 29. Josset V, Torre JP, Tavolacci MP, Van Rossem-Magnani V, Anselme K, Merle V et al. Efficiency of hepatitis C virus screening strategies in general practice. Gastroenterol Clin Biol. 2004;28(4):351 –7. [PMID: Accession Number]; PubMed Central. 30. Nakamura J, Terajima K, Aoyagi Y, Akazawa K. Cost-effectiveness of the national screening program for hepatitis C virus in the general population and the high-risk groups. Tohoku J Exp Med. 2008;215(1):33–42. doi: 10.1620/tjem.215.33. [PMID: Accession Number]; PubMed Central. 31. Tramarin A, Gennaro N, Compostella FA, Gallo C, Wendelaar Bonga LJ, Postma MJ. HCV screening to enable early treatment of hepatitis C: a mathematical model to analyse costs and outcomes in two populations. Curr Pharm Des. 2008;14(17):1655–60. [PMID: Accession Number]; PubMed Central. 32. Wong WW, Tu HA, Feld JJ, Wong T, Krahn M. Cost-effectiveness of screening for hepatitis C in Canada. CMAJ. 2015;187(3):E110–E121. doi: 10.1503/cmaj.140711. [PMID: Accession Number]; PubMed Central PMC4330166: PMC4330166. 33. Selvapatt N, Ward T, Bailey H, Bennett H, Thorne C, See L-M et al. Is antenatal screening for hepatitis C virus cost-effective? A decade’s experience at a London centre. J Hepatol. 2015;63(4):797–804. doi: http://dx.doi.org/10.1016/j.jhep.2015.05.015. [PMID: Accession Number]; PubMed Central. 34. Plunkett BA, Grobman WA. Routine hepatitis C virus screening in pregnancy: a cost-effectiveness analysis. Am J Obstet Gynecol. 2005;192(4):1153–61. doi: http://dx.doi.org/10.1016/j.ajog.2004.10.600. [PMID: Accession Number]; PubMed Central. 35. Urbanus AT, van Keep M, Matser AA, Rozenbaum MH, Weegink CJ, van den Hoek A et al. Is adding HCV screening to the antenatal national screening program in Amsterdam, the Netherlands, costeffective? PloS One. 2013;8(8):e70319. doi: 10.1371/journal.pone.0070319. [PMID: Accession Number]; PubMed Central PMC3741285: PMC3741285. 36. Kim DD, Hutton DW, Raouf AA, Salama M, Hablas A, Seifeldin IA et al. Cost-effectiveness model for hepatitis C screening and treatment: implications for Egypt and other countries with high prevalence. Glob Public Health. 2015;10(3):296–317. doi: 10.1080/17441692.2014.984742. [PMID: Accession Number]; PubMed Central PMC4320005: PMC4320005. 37. Rein DB, Smith BD, Wittenborn JS, Lesesne SB, Wagner LD, Roblin DW et al. The cost-effectiveness of birth-cohort screening for hepatitis C antibody in US primary care settings. Ann Intern Med. 2012;156(4):263–70. doi: 10.7326/0003-4819-156-4-201202210-00378. [PMID: Accession Number]; PubMed Central.

Page | 187

Annex 5.3 PICO 1 - How to test (HBV) Diagnostic accuracy of tests to detect hepatitis B surface antigen: a meta-analysis and review of the literature

Ali Amini,* Helen Kelly,* Debi Boeras, Wen Chen, Jane Falconer, Helen Kelly, Weiming Tang, Joseph Tucker, Olivia Varsaneux, Rosanna Peeling (Team lead) London School of Hygiene and Tropical Medicine team *Co-leaders of this review

September 2015

Page | 188

1. Executive summary Background: Rapid diagnostic tests are potentially useful tools for the diagnosis of hepatitis B surface antigen (HBsAg) globally, particularly in low-resource areas. Expansion for global use depends on their performance characteristics clinically in the field, ultimately with the aim being to reach low-resource settings and offer cost–efficient screening as an alternative to laboratory tests. Objectives: The purpose of this review was to identify quantitative evidence on the clinical sensitivity1 and specificity of available in vitro diagnostics (hereafter referred to as assays) used to detect hepatitis B antibody, synthesize the evidence, and inform models. Methods: Two reviewers independently assessed the quality and extracted data for estimating accuracy. Meta-analysis was performed. We further performed stratified estimates based on individual products, HIV status, specimen type, study setting and design. Results: Thirty-three studies were included using an EIA reference standard. The overall pooled clinical sensitivity and specificity of rapid HBsAg tests were 90.0% (95% CI: 89.1, 90.8) and 99.5% (95% CI: 99.4, 99.5), respectively, compared to laboratory-based immunoassay reference standards. Pooled specificity was comparable and less heterogeneous. Pooled sensitivity in studies of HIV-positive was lower than in known HIV-negative patients; 72.3% (95% CI: 67.9, 76.4) compared to 92.6% (95% CI: 89.8, 94.8), respectively. Pooled sensitivity and specificity in blood donors were 91.6% (95% CI: 90.1, 92.9) and 99.5% (95% CI: 99.5, 99.9), respectively. Samples using whole blood specimens (venous or capillary) were 91.7% (95% CI: 89.1, 93.9) and 99.9% (95% CI: 99.8, 99.9) sensitive and specific compared to serum. Results were comparable for studies performed prior to the past ten years as those performed since 2005. Estimates of assay sensitivity demonstrated significant heterogeneity not entirely corrected by sub-analysis, although studies using whole blood specimens (venous or capillary), and the same reference standard (CMIA) were more robust (tau-squared <1 in all cases). The overall pooled clinical sensitivity and specificity of laboratory-based HBsAg tests were comparable, with 88.9% (95% CI: 87.0, 90.6) and 98.4% (95% CI: 97.8, 98.8) sensitivity and specificity, respectively, compared to immunoassay state-of-the-art chemiluminescent microparticle enzyme immunoassays. Conclusions: Assays for HBsAg detection such as rapid diagnostic tests (RDTs), including those performed on serum/plasma and capillary whole blood specimens, have good sensitivity and excellent specificity compared to a reference standard comprising laboratory-based methods of HBsAg detection. Improvement in sensitivity, or development of innovative testing strategies could potentially enhance their use as first-line screening globally. Caution in HIV-

Unless otherwise specified, sensitivity and specificity refer to clinical and not analytical sensitivity for tests.

Page | 189

positive individuals is important, while the reassuring accuracy of capillary whole blood specimens compared to plasma/serum further facilitates use in settings where phlebotomy is not available. 2. Background An estimated 240 million individuals worldwide1 are chronically infected with hepatitis B virus (HBV) and there are an estimated four million acute HBV infections each year. Twenty per cent to 30% of those with chronic hepatitis B infection will develop cirrhosis2 or hepatocellular carcinoma,3 leading to approximately 650 000 deaths each year.4 However, most individuals with chronic HBV infection are not aware of their serostatus, contributing to delayed diagnosis and complications from advanced disease.5 HBV testing is critically important in order to refer infected individuals to HBV treatment and care, to refer uninfected individuals for vaccination, and to mobilize prevention and control efforts. In March 2015, the World Health Organization (WHO) published the first guidelines for the prevention, care, and treatment of individuals with chronic HBV infection.5 These guidelines focused on assessment for treatment eligibility, initiation of first-line therapies, switching and monitoring. These initial guidelines did not include testing recommendations, and in particular which tests to use. Given the large burden of HBV in low- and middle-income settings where there are limited or no existing HBV testing guidelines, there is a substantial need for HBV testing guidelines. Chronic HBV infection is defined as persistence of hepatitis B surface antigen (HBSAg) for at least six months. However, interpretation of HBV serologies is complex. The serologies most frequently used for HBV testing include HBsAg, total anti-HBc, and anti-HBs. HBV screening includes both the one-test (e.g. HBsAg) and two-test strategies (e.g. HBsAg followed by hepatitia B core antibody [HBcAb] or nucleic acid testing [NAT]). Detection of HBsAg can include rapid diagnostic tests (RDTs) or immunoassays. Rapid tests developed for screening include solid-phase assays, flow-through, agglutination and lateral-flow. The majority, however, are immunochromatographic assays. Immunoassays use different methods for detection of HBsAg using polyclonal or monoclonal anti-HBs antibodies. Labelling to measure antigen–antibody complexes can include radioactive compounds, enzymes with a change in colour in solution, or substances emitting light. Advances in HBV detection technology create new opportunities for enhancing screening, referral, and treatment. Previous systematic reviews on hepatitis B infection have focused on immunological responses,6 surveillance of cirrhosis,7 and treatment.8 Existing systematic reviews9‒11 on hepatitis B testing focused on point-of-care (POC) tests and included tests with unclear reference standards or those not appropriate for assessing operational diagnostic accuracy in the field.

3. Objectives The purpose of this review was to identify quantitative evidence on the sensitivity and specificity of assays used to detect hepatitis B antibody, synthesize the evidence and inform models.

Page | 190

To our knowledge, this is the first study exclusively comparing the clinical performance of both RDTs and laboratory-based immunoassays, in addition to addressing the question of accuracy in the context of HIV specifically. PICO 1 Among persons identified for hepatitis B testing, what is the diagnostic accuracy of available assays for detecting HBsAg? Persons identified for HBV testing Rapid diagnostic tests or immunoassays for HBsAg detection Reference standard comprising a laboratory-based enzyme immunoassay (EIA)2 (one or more HBsAg enzyme immunoassays, with or without neutralization to confirm) Diagnostic accuracy (sensitivity, specificity, TN, TP, FN, and FP; positive predictive value, negative predictive value)

P I C

As a subanalysis, we also analysed data for studies comparing the accuracy of HBsAg assays against a nucleic-acid amplification test (NAT) reference standard. This is important given the importance of reducing transmission during the seroconversion period and in the diagnosis of occult hepatitis B, where HBsAg may not be detectable and which is more common in HIV coinfection. [Results in Annex 9.2]

4. Methods Search strategy and identification of studies Literature search strategies were developed by a medical librarian with expertise in systematic review searching. Our search algorithm consisted of the following components: hepatitis B, diagnostic tests and diagnostic accuracy (see Annex 1). We searched MEDLINE (OVID interface, 1946 onwards), EMBASE (OVID interface, 1947 onwards), the Cochrane Central Register of Controlled Trials (Wiley interface, current issue), Science Citation Index Expanded (Web of Science interface, 1970 onwards), Conference Proceedings Citation Index-Science (Web of Science interface, 1990 onwards), SCOPUS (1960 onwards), Literatura LatinoAmericana e do Caribe em Ciências da Saúde (LILACS) (BIREME interface) and WHO Global Index Medicus. The search was supplemented by searching for ongoing studies in WHO’s International Clinical Trials Registry. In addition to searching databases, we contacted individual researchers, experts working in the field and authors of major trials to address whether any relevant manuscripts are in preparation or in press. The references of published articles found in the above

For convenience we shall refer to all laboratory-based immunoassays for HBsAg detection (ELISA, MEIA, ECLIA, CMIA) as EIAs as most have some form of enzymatic amplification.

Page | 191

databases were searched for additional pertinent articles. The review was registered in PROSPERO and reported in accordance with PRISMA guidelines. Study selection proceeded in three stages. First, titles/abstracts were screened by a single reviewer (AA/HK) according to standard inclusion and exclusion criteria. Second, all articles identified for full manuscript review were obtained and assessed independently by the two reviewers (AA and HK) against inclusion criteria. Papers were accepted or rejected and reasons for rejection were specified. Third, discrepancies were resolved by discussion between review authors, with several studies resolved by a third independent reviewer (RP). Selection criteria Types of studies We included case–control, cross-sectional, cohort studies and randomized trials with a primary purpose of evaluating HBsAg tests published until May 2015. We excluded: conference abstracts, comments or review papers; studies with primary aims other than evaluation of both sensitivity and specificity of HBsAg detection; studies related to disease prevalence, drug resistance, genotyping, sequencing, or non-diagnostic purposes; studies that focus on detection of anti-HBsAg (antibody to hepatitis B antigen); articles in languages other than English. Participants We included studies using original data from patient specimens in defined populations. We included all age groups, settings, countries and specimen types, notably whole blood (venous and capillary), plasma or serum. Saliva specimens were considered, but no suitable studies were identified. We excluded studies using commercial reference panels or clinical panels not sourced by authors given applicability and bias concerns from unknown sampling in unclear populations. Index tests Studies utilizing commercially available HBsAg tests were eligible for inclusion. We excluded: in-house developed tests; laboratory-based immunoassays which are no longer commercially available. We did not, however, exclude rapid tests based on current commercial availability, in keeping with the methodology in recent systematic reviews. We did however subcategorize more recent studies between 2005 and 2015.

Reference standard The reference standard for definite diagnosis of hepatitis B is complicated, given the different viral kinetics of HBsAg and HBV DNA. We included studies using an established commercially available immunoassay as a reference standard for HBsAg detection. Studies using a NAT reference standard were included as a supplementary secondary analysis. For studies comparing immunoassays, we only included studies using chemiluminescent microparticle

Page | 192

immunoassays (CMIAs) as the reference standard, given the generally accepted higher analytical sensitivity of these assays. For studies of rapid tests, in keeping with previous systematic reviews, we did not limit based on type of immunoassays. Data extraction Information on the following variables were independently extracted by the two review authors (AA, KH): first author, total sample size, country (and city) of sampling, specimen type (oral fluid, capillary [finger-prick] whole blood, venous whole blood, etc.), eligibility criteria, reference standard, manufacturer, raw cell numbers (true positives, false negatives, false positives, true negatives), HIV coinfection, sources of funding, reported conflicts of interest. Assessment of methodological quality Study quality was evaluated using the QUADAS-2 tool12 and the STARD checklist.13 QUADAS includes domains to evaluate bias in the following categories: risk of bias (patient selection, index test, reference standard, flow and timing); applicability concerns (patient selection, index test, reference standard). GRADE summary of finding tables The GRADE system was used to rate the strength of evidence of each body evidence as high, moderate, low, or very low on the basis of aggregate quality, consistency, precision, directness, and reporting bias. We did not downgrade directness because the outcome evaluated was diagnostic accuracy (an intermediate outcome), as the PICO focuses on diagnostic accuracy. Summary of finding tables were prepared. We did not perform a formal assessment of publication bias. Data analysis and synthesis Standard methods for meta-analysis of diagnostic accuracy were used. For each study we calculated sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR) and 95% confidence intervals (CI). Likelihood ratios are metrics and are calculated using a combination of sensitivity and specificity. Positive likelihood ratio (PLR) is the ratio of sensitivity (1−specificity), whereas the negative likelihood ratio (NLR) is the ratio of specificity (1−sensitivity). When a diagnostic test has no discriminating capability, both likelihood ratios equal 1. We conducted meta-analysis pooling data using the DerSimonian–Laird bivariate random effect model (REM) to calculate pooled sensitivity, specificity and other related indices. The REM was more suitable than the fixed-effects model (FEM) given significant heterogeneity in studies found, as it takes into account variability both within studies (random error) and between studies (heterogeneity). Statistical heterogeneity was measured using the random-effect variance (tau-squared). Meta-regression was performed to investigate the source of heterogeneity within included studies (inverse variance weighted). We analysed sub-groups according to the following characteristics: study year (2005–2015); geographical area; individual tests; sample

Page | 193

type; patient type (blood donors); patient HIV status. We did not formally test for publication bias. Where meta-regression was not possible for covariates, we performed descriptive statistics. Meta-analysis of the collected data was conducted using the software: Meta-Disc© version 1.4.7. Statistical analysis was performed using Meta-Disc 1.4 for Windows (XI Cochrane Colloquium, Barcelona, Spain). QUADAS-2 analysis was performed using Microsoft Excel. 5. Results Study selection A total of 11 589 citations were identified and 6575 duplicates were removed. Each of the 5014 titles was examined according to pre-specified inclusion and exclusion criteria. A total of 33 research studies were included in the final primary analysis (Fig. 1), with studies comparing both rapid diagnostic tests (RDTs)14–42 and enzyme immunoassays43–46 against an immunoassay reference standard. Of these, 19 studies14–31, 47 were also included in a recent systematic review by Khuroo et al.;11 eight papers from that study were excluded as they were conference abstracts or letters to editors,48–50 foreign language articles,51, 52 or evaluated reference panels.53 Two reports by WHO and the International Consortium for Blood Safety (ICBS) were also excluded as they were not published in peer-reviewed journals and were case–control studies constructed using reference panels from populations of affected individuals. Our search identified 11 additional articles32–42 comparing RDTs against the EIA reference standard not found in the previous review. Six articles exclusively assessed accuracy in cohorts of HIV-positive individuals.20, 36–38, 54, 55 Studies evaluating laboratory-based immunoassays for HBsAg detection as the index test all used state of the art CMIAs as the reference standard. Seven studies were included in the supplementary analysis assessing diagnostic accuracy of HBsAg assays against a nucleicacid based reference standard.47, 54–59

Page | 194

Fig. 1. PRISMA flow diagram outlining study selection examining diagnostic accuracy of HBsAg assays in our systematic review

Page | 195

Study characteristics Immunoassay reference standard [Table 1a; 1b] Overall, the 33 included articles originated in 23 countries: Australia,35 Western Europe16, 18, Western Africa;15, 25, 28, 34, 36, 38‒40 Eastern Africa;19, 20, 26, 30, 41 South Africa;38 South-eastern Asia;24, 32 Eastern Asia;25, 27, 31, 43 Southern Asia14, 17, 21, 22, 29, 42 and the United States.23 Thirty-three RDTs were evaluated using serum (fresh or frozen), plasma, venous and capillary whole blood. Sample sizes varied from 25 to 3928 individuals (mean, 631). Sixty-three data points were generated from the articles. All had bivariate data for analysis; 38% of data points were generated from case–control studies, while the remainder were from cross-sectional cohort studies. Prevalence of hepatitis B in populations tested ranged from 1.9% to 84%. Populations studied included: general screening of healthy volunteers and blood donors; screening of at-risk populations such as pregnant women, incarcerated adults; and patients from hepatitis cohorts, including confirmed hepatitis B. The majority of studies were performed in laboratory settings, with some studies performed in the field. A mixture of serum, plasma, capillary and whole blood was used for RDTs, with serum or plasma used for EIAs. Seven studies assessed performance using capillary or venous whole blood specifically.15, 16, 20, 24, 28, 37, 40 In terms of tests used, only the Determine HBsAg and BinaxNOW had more than 3 data points. NAT reference standard [Table 8; Annex 9.2] Overall, the 7 articles originating in 4 countries (Iran, Nigeria, Uganda, South Africa), contributed 18 data points and included serum or plasma samples from 3304 individuals, with sample size ranging from 74 to 950. All studies were either cross-sectional, cohort or case– control studies. All had data for bivariate analysis.

Page | 196

Table 1a. Study characteristics – EIA vs EIA Study [Author, Year] Liu, 2014 Location [Country, City] China Sample size 250 CC Study design Hospital patients; outpatients (preselected based on CMIA quantitative results) Peng, 2011 China 498 CC Hospital patients (preselected based on S/CO from KHB screen) Geretti, 2010 Ghana, Kumasi 838 CS – CSQ HIV clinic (1/3 on lamivudine) Serum CMIA, Architect HBsAg CMIA, Liaison Ultra EIA, Murex v3 Ol, 2009 Viet, 2012 Cambodia Vietnam 120 119 CS – CSQ CS – CSQ Blood donors (rural community) Blood donors (rural community) Serum Serum ELISA, Monolisa EIA, Monolisa Ultra CMIA, Architect HBsAg CMIA, Architect HBsAg CMIA, Architect/ Liaison EIA, Murex v3 *

Setting

Sample

Assay under evaluation [Type, Brand] 3

Reference standard [Type, Brand] 3

Serum

ECLIA, Cobas ELISA, Wantai

CMIA, Architect HBsAg

Serum

ELISA, KHB

CMIA, Architect HBsAg

*Reactive all three assays OR reactive in one assay with neutralisation CMIA: chemiluminescent microparticle enzyme immunoassay; ECLIA: electrochemiluminescent immunoassay; EIA: enzyme immunoassay; ELISA: enzyme-linked immunosorbent assay; CC: case–control; CS: cross-sectional; CSQ: consecutive patients

Abbreviated names for table clarity – full product names in Annex

Page | 197

Table 1b. Study characteristics – RDT vs EIA Study [Author, Year] Location [Country, City] Sample size Study design Setting Sample RDT under evaluation [Type, Brand] Mvere, 1996 Zimbabwe 206 CS Blood bank S Dipstick (PATH) SimpliRed Sato, 1996 Japan 462 CC Hospital S Dainascreen Serodia Abraham, 1998 India, Vellore 50 400 Oh, 1999 Korea 250 CC – Panel CS –Screen CC – Panel Blood donor panel S Hospital patients (Multiply transfused; chronic liver disease; preop and antenatal patients) S QuickChaser Virucheck EIA, Auszyme or Hepanostika EIA, Auszyme EIA, Auszyme Reference test [Type, Brand]

Genedia Serodia

EIA, Cobas Core

Kaur, 2000 Lien, 2000

India Viet Nam

2754 328

CS – CSQ CC

Hospital surgery patients; blood donors; patients ruling out HBV High-risk volunteers; pregnant women; patients with other infectious diseases (including 10 with HIV); preselected HBsAg pos (101), HBsAg neg (99) Hospital laboratory samples (emergency preop screening; antenatal women in labour; haemodialysis; urgent donor screening) Hospital - patients with biopsy-proven HBV; healthy volunteers from a vaccine evaluation trial; blood donors Hepatology clinics Incarcerated offenders Chinese community health fair (random patients); known HBV-positive patients (liver clinic) Blood donors (male)

S SP

Hepacard Dainascreen Determine Serodia

EIA, Ortho 3rd generation EIA, Monolisa MEIA for discordant

Raj, 2001

India, Vellore

999

Hepacard

EIA, Auszyme MEIA, AxSYM v2

Clement, 2002

Belgium

942

WB, S

BinaxNOW

MEIA, AxSYM v2

Lau, 2003

USA

1011 827 625

CS – CSQ CS CS – CSQ

S fresh S frozen WB

Binax NOW

EIA, ETI-MAK2

Akanmu, 2006

Nigeria, Lagos

101

CS – CSQ

Binax NOW

ELISA, Monolisa

Page | 198

36 Nyirendra, 2008 Malawi 194 CS – CSQ

Chronic liver disease Hospital Hospital patients including 152 HIV+ P Determine EIA, Bioelisa Neutralisation (positives)

Lin, 2008

China Guinea

671 579 200

CC CC CC

Blood donors (500); Clinical specimens HBsAg+ (171) Blood donors (491); Stored positives (88) Not specified

S SP S

Determine DRW

EIA, Hepanostika Ultra

Randrianirina, 2008

Madagascar

Cypress Determine Hexagon Virucheck

EIA, AxSYM

Ola, 2009

Nigeria

25 55

CS - CSQ

Medical clinic Blood donors

WB S S

AMRAD GWHB Biotec Latex Accurate Onecheck

ELISA, Wellcozyme Kit

Khan, 2010

Pakistan

ELISA, 2nd generation

Davies, 2010

Malawi

CS – CSQ

HIV-positive adults (ART naive)

Determine

EIA, Biokit Neutralization (positives)

Bjoerkvoll, 2010

Cambodia Viet Nam

1200 1200 838

CS – CSQ

General screen – blood donors (rural)

ACON

EIA, Monolisa Ultra*

Geretti, 2010

Ghana, Kumasi

CS – CSQ

HIV clinic (1/3 on lamivudine)

Determine VIKIA

CMIA, Architect/ Liason EIA, Murex v3 ELISA, AxSYM ELISA, Monolisa Ultra Neutralization (positives)

Hoffman, 2012 Bottero, 2013

South Africa France, Paris

973 2472 3922 3928

CS – CSQ CS – CSQ

HIV-positive adults (ART naïve) – antenatal or primary care General screening (health-care centres) [general population prevention, screening, vaccination]

WB (cap) WB (ven)

Determine Determine QUICK PROFILE VIKIA

Chameera, 2013

Sri Lanka

Hospital (surgical, other)

Cortez

EIA, Surase B-96 (TMB)

Page | 199

Onsite Franzeck, 2013 Tanzania, Ifakara Unclear 272 CS – CSQ HIV clinic (ART naive) WB (ven) P 558 CC Chronic hep B (known mutants, blood donors); HBsAg negative (mix, including HIV, 34; HCV, 48) Acute hepatitis Pregnant - pregnant women at delivery Blood donors At risk health fairs, outreach; Vietnamese (72%) SP S ACON Nanosign ELISA, HBsAg Ultra EIA, Quest Diagnostics SP DRW v2.0 Determine EIA, Murex v3 Neutralization (positives) CMI, Architect

Chevaliez, 2014

408 802 Erhabor, 2014 Gish, 2014 Nigeria, Sokoto Australia, Melbourne Bissau China 130 297

CS CS CC CS – CSQ

Honge, 2014 Liu, 2014

438 250

CS – CSQ CC

HIV clinic - mixed ART/ naive Hospital patients; outpatients (preselected based on CMIA quantitative results)

S S

VEDA LAB Intec One Step

CLIA, Architect CMIA, Architect

Mutocheluh, 2014

Ghana

150

CS – CSQ

Blood donors

Abon Acull-Tell Core TM Rapid care Wondfo

ELISA, Human Gesellschaft

Upretti, 2014 Njai, 2015

Nepal Gambia

347 178 203 773 476

CS – CSQ CS

Children – pre- and post vaccination; mothers (8) Hepatitis patients CHB carriers (study 3), incl 3 coinfected HIV (treatment naive)

S S

SD Bioline Determine Espline

EIA, Surase B-96 (TMB) CMIA (quantitative), Architect

CS – CSQ

General community screen

Determine VIKIA

EIA (DBS), AxSym + Neutralisation

*Validated random sample with CMIA, Abbott CC: case–control study; CS: cross-sectional study; CSQ: consecutive patients; S: serum; P: plasma; WB (cap): capillary whole blood; WB (ven): venous whole blood

Page | 200

Assessment of the quality of the studies The methodological quality of included studies4 is summarized [Fig. 2, Table 2]. Patient selection We judged 15 studies to be at “high risk of bias”. Of these, 10 were case–control studies. Others with a high risk of bias included studies in blood donors and highly selected populations, such as patients with known hepatitis B. Applicability was judged to be “high risk” in 8 studies, notably those published over ten years ago or with tests which are no longer commercially available. Index test We judged 7 studies as high risk of bias, and 14 as unclear, with the most common reason being a lack of reported blinding while reading test results. Reference standard We judged 5 studies as high risk of bias, with 17 unclear; the most common reason was a lack of reported blinding interpreting reference tests, or utilisation. Flow and timing Bias was predominantly due to lack of reported flow and timing. Although the majority of studies did not specificity the exact time differences between performance of the index and reference assays, we can assume they were low risk as they were on the same sample.

Studies refers to either entire articles or individual sub-studies within a single publication using different patients, methods, index tests or reference standard.

Page | 201

Table 2. Risk of bias and applicability according to QUADAS-2 domains for individual studies Risk of bias Patient selection Abraham (CC) Abraham (CS) Akanmu (Blood) Akanmu (Liver) Bjoerkvoll Bottero Chameera Chevaliez (CC) Chevaliez (Hepatitis) Chevaliez (Pregnant) Clement Davies Erhabor Franzeck Geretti Gish Hoffman Honge Kaur Khan Lau (Hepatology clinic) Lau (Prison) Lau (Screen + Known) Lien Lin Liu Mutocheluh Mvere Njai (1) Njai (2) Njai (3) High Low Low Low Low Low Low High High High High Unclear High Low Low Low Low Low Low High Low Low Low Low High High High Low Low Low High Index test Reference standard Low Low Low Low Low Low Unclear Unclear Unclear Unclear High Unclear Unclear Low Unclear Unclear Low Unclear Unclear Low Low Low Low Low Unclear Low Low Low Low Low Low Flow and timing Low Low Low Low Low Low Low Low Low Low Low High Low Low Unclear Low Unclear Low Low Low Low Low Low Low Low High Low Low Unclear Unclear Low Patient selection Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Applicability Index test Reference standard Unclear Unclear Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Unclear Low Low Low Low Low Low Low Low Unclear Low Low Low

Low Low Low Low Low High High Unclear Unclear Unclear Unclear Low Low Low Unclear Unclear Unclear Low Unclear Low Low Low High High Unclear Unclear Low Low Low Low Low

High High Unclear Unclear Low Low Low Low Low Low Unclear Low Low Low Low Low Low Low High Low Unclear Unclear Unclear Low Low Low Low High Low Low Low

Page | 202

Risk of bias Patient selection Nyirendra Oh Ol Ola BD Ola Clinic Peng Raj Randrianirina Sato Upretti Viet Low High Low Low Low High Unclear High Unclear High Low Index test Reference standard Unclear High Unclear Unclear Unclear Unclear High High Low High Unclear Flow and timing High Low High Low Low High High Low Low High High Patient selection Low High Low Low Low Low Low Low Unclear High Low

Applicability Index test Reference standard Low Unclear Low Low Low Low Low Low Unclear Low Low

Low High Unclear High Low Low Unclear Unclear High Low Unclear

Low High Low High Low Low High Low High Low Low

Fig. 2. Risk of bias and applicability summary according to QUADAS-2 domains presented as percentages across included studies

FLOW AND TIMING QUADAS-2 Domain REFERENCE STANDARD

Low

High

INDEX TEST PATIENT SELECTION 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% Proportion of studies with low, high, or unclear Proportion of studies with low, high or unclear CONCERNS regarding APPLICABILITY RISK of BIAS

Page | 203

Diagnostic accuracy Overall clinical performance of assays against an immunoassay reference Pooled test accuracy for RDTs compared to EIAs A total of 21 studies14‒36, 38‒43 contributing 63 data points evaluated 25 brands of RDTs using 15 EIA reference assays, with 36 919 total samples, including serum, plasma, venous and capillary whole blood. Sample sizes ranged from 25 to 3983 (mean 586). Sensitivities ranged from 50% to 100% with overall pooled sensitivity of 90.0% (95% CI: 89.1, 90.8). Specificities ranged from 69% to 100%, with overall pooled specificity of 99.5% (95% CI: 99.4, 99.5). Pooled PLR and NLR were 117.5 (95% CI: 67.7, 204.1) and 0.095 (95% CI 0.067, 0.136), respectively, with tausquare 3.89, 1.72, respectively, suggestive of significant heterogeneity between studies [Fig. 3, Table3]. Pooled test accuracy for EIAs compared to other immunoassays Five studies36, 43‒46 contributed 8 data points evaluating 8 EIAs with reference to state-of-theart immunoassays alone, using a total 3751 serum samples, with sample sizes ranging from 119 to 838 (mean, 469). Studies were performed in China, Ghana, Cambodia and Viet Nam. Sensitivities ranged from 73% to 100%% with overall pooled sensitivity of 88.9% (95% CI: 87.0, 90.6). Specificities ranged from 88% to 100%, with overall pooled specificity of 98.4% (95% CI: 97.8, 98.8). Pooled PLR and NLR were 46.76 (95% CI: 12.86, 170.03) and 0.041 (95% CI: 0.013, 0.134), respectively, with tau-square 2.95, 2.46, respectively, suggestive of significant heterogeneity between studies [Fig. 4, Table 3]. Overall clinical performance of assays against a NAT reference Pooled test accuracy for RDTs compared to NAT Three articles47, 57, 59 contributed 9 data points evaluating 7 RDTs with a NAT reference, using a total 1710 serum or plasma samples, with sample sizes ranging from 74 to 950 (mean 190). Of note, 1440 were from the same 240 patients in one case–control study evaluating 6 tests. Only one study (Nna) 57 used plasma. Sensitivities ranged from 38% to 99% with overall pooled sensitivity of 93.3% (95% CI: 91.3, 94.9). Specificities ranged from 94% to 99%, with overall pooled specificity of 98.1% (95% CI: 97.0, 98.9). Pooled PLR and NLR were 39.42 (95% CI: 22.148, 70.185) and 0.051 (95% CI: 0.009, 0.275), respectively, with tau-square 0.2163, 6.264, respectively. [Fig. 5, Table 9; Fig. 8 are in the Annex] Pooled test accuracy for EIAs compared to NAT Five articles54‒56, 58, 59 contributed 9 data points evaluating EIAs with a NAT reference, using a total 1594 samples, with sample sizes ranging from 74 to 240 (mean, 177). Sensitivities ranged from 38% to 98%% with overall pooled sensitivity of 75.7% (95% CI: 72.1, 79.1). Specificities ranged from 70% to 98%, with overall pooled specificity of 86.1% (95% CI: 83.8, 88.2). Pooled PLR and NLR were 7.234 (95% CI: 4.441, 11.758) and 0.296 (95%CI: 0.192, 0.458), respectively,

Page | 204

with tau-square 0.3748, 0.3299, respectively suggesting acceptable interstudy heterogeneity. [Fig. 6, Table 9; and Fig. 9 are in the Annex] Clinical performance of assays in HIV-positive and negative individuals Pooled test accuracy for RDTs in HIV-positive individuals Five articles19, 20, 36‒38 contributed 6 data points evaluating 3 RDTs with an EIA reference, using a total 3434 samples from 2566 patients, with sample sizes ranging from 75 to 838. Sensitivities ranged from 62% to 100% with overall pooled sensitivity of 72.3% (95% CI: 67.9, 76.4). Specificities ranged from 99% to 100%, with overall pooled specificity of 99.8% (95% CI: 99.5, 99.9). Pooled PLR and NLR were 192.63 (95% CI: 77.4, 479.17) and 0.288 (95%CI: 0.217, 0.381), respectively, with tau-square 0.3838, 0.0585, respectively [Table 4 and 5; Fig. 11, Annex 9.3]. Pooled test accuracy for RDTs in HIV-negative individuals One article40 contributed 4 data points evaluating 3 RDTs with an EIA reference, using a total 1624 samples from 997 patients, with sample sizes ranging from 175 to 773. Sensitivities ranged from 88% to 95% with overall pooled sensitivity of 92.6% (95% CI: 89.8, 94.8). Specificities ranged from 93% to 100%, with overall pooled specificity of 99.6% (95% CI: 99.0, 99.9). Pooled PLR and NLR were 79.449 (95% CI: 11.575, 545.315) and 0.082 (95%CI: 0.053, 0.125), respectively, with tau-square 2.9668, 0.0803 respectively [Table 4 and 5; Fig. 12, Annex 9.3]. Pooled test accuracy for EIAs in HIV-positive individuals One article36 contributed 3 data points evaluating 3 EIAs with an EIA reference. 838 samples were tested with each index test. Sensitivities ranged from 97% to 99% with overall pooled sensitivity of 97.9% (95% CI: 96.0, 99.0). Specificities ranged from 99% to 100%, with overall pooled specificity of 99.4% (95% CI: 99.0, 99.7). Pooled PLR and NLR were 167.26 (95% CI: 95.135, 294.07) and 0.022 (95% CI: 0.012, 0.043), respectively, with tau-square <0.005, <0.005, respectively [Table 4; Fig. 14, Annex 9.3.] Clinical performance of RDTS compared to EIAs in other stratified subgroups Pooled test accuracy for RDTs using whole blood (capillary or venous) Seven studies15, 16, 20, 24, 28, 37, 40 contributed 11 data points evaluating 5 RDTs with an EIA reference, using a total 13731 samples, with sample sizes ranging from 25 to 3928 (mean 722). Sensitivities ranged from 75% to 100% with overall pooled sensitivity of 91.7% (95% CI: 89.1, 93.9). Specificities ranged from 99% to 100%, with overall pooled specificity of 99.9% (95% CI: 99.8, 99.9). Pooled PLR and NLR were 346.64 (95% CI: 157.598, 762.42) and 0.089 (95%CI: 0.058, 0.136), respectively, with tau-square 0.8124, 0.2367, respectively [Table 5; Fig. 18, Annex 9.3.2]. Pooled test accuracy for RDTs in studies using a case-control design Page | 205

Ten articles14, 18, 22, 24, 25, 27, 30, 31, 33, 34 contributed 21 data points evaluating 13 RDTs with an EIA reference, using a total 7258 samples, with sample sizes ranging from 50 to 698 (mean 345). Sensitivities ranged from 50% to 100% with overall pooled sensitivity of 96.7% (95% CI: 96.0, 97.3). Specificities ranged from 91% to 100%, with overall pooled specificity of 99.3% (95% CI: 99.0, 99.5). Pooled PLR and NLR were 105.16 (95% CI: 48.038, 230.212) and 0.028 (95%CI: 0.010, 0.076), respectively, with tau-square 2.2261, 4.8632, respectively. Of note, one study22 had significantly lower sensitivity for both index tests evaluated, with otherwise sensitivities ranging from 90% to 100% in remaining studies [Table 5; Fig. 19, Annex 9.3.3]. Pooled test accuracy for RDTs used in blood donors Seven articles25‒28, 32, 34, 39 contributed 19 data points evaluating 15 RDTs with an EIA reference, using a total 6881 samples, with sample sizes ranging from 25 to 1200 (mean, 362). Sensitivities ranged from 50% to 100% with overall pooled sensitivity of 91.6% (95% CI: 90.1, 92.9). Specificities ranged from 86% to 100%, with overall pooled specificity of 99.5% (95% CI: 99.3, 99.). Pooled PLR and NLR were 89.219 (95% CI: 32.782, 242.818) and 0.106 (95%CI: 0.055, 0.204), respectively, with tau-square 3.8171, 1.8505, respectively. [Table 5; Fig. 20, Annex 9.3.4] Pooled test accuracy for RDTs published before and after 2005 Twenty-one articles15‒17, 19, 20, 22, 25, 28, 30, 32‒43 contributed 44 data points evaluating 26 RDTs with an EIA reference, using a total 25 261 samples, with sample sizes ranging from 25 to 3928 (mean 574). Sensitivities ranged from 50% to 100 % with overall pooled sensitivity of 86.4% (95% CI: 85.2, 87.5). Specificities ranged from 69% to 100%, with overall pooled specificity of 99.4% (95% CI: 99.2, 99.5). Pooled PLR and NLR were 84.657 (95% CI: 43.553, 164.553) and 0.126 (95%CI: 0.087, 0.183), respectively, with tau-square 4.0986, 1.2712, respectively. Nine articles published before 200514, 18, 21, 23, 24, 26, 27, 29, 31 contributed 19 data points evaluating 10 RDTs with an EIA reference, using a total 25 253 samples, with sample sizes ranging from 25 to 3928 (mean 1122). Sensitivities ranged from 77% to 100% with overall pooled sensitivity of 96.9% (95% CI: 96.0, 97.7). Specificities ranged from 97% to 100%, with overall pooled specificity of 99.7% (95% CI: 99.6, 99.8). Pooled PLR and NLR were 265.5 (95% CI: 106.1, 664.5) and 0.056 (95%CI: 0.033, 0.095), respectively, with tau-square 2.72, 0.91, respectively. [Table 5; Figs 21 and 22, Annex 9.3.4] Pooled test accuracy for RDTs by brand Stratifying by test brand did not eliminate statistical heterogeneity. Data for all 50 brands used [Table 6] demonstrate heterogeneous results for sensitivity, with more robust specificity as previously noted. Determine HBsAg was evaluated in the most studies, with only one published before 2008. Ten articles16, 19, 20, 24, 25, 30, 36, 37, 40, 41 contributing 12 data points evaluated against an EIA reference, using a total 7553 samples, with sample sizes ranging from 75 to2472. Sensitivities ranged from 56% to 100% with overall pooled sensitivity of 90.8% (95% CI: 88.9, 92.4). Page | 206

Specificities ranged from 69% to 100%, with overall pooled specificity of 99.1% (95% CI: 98.9, 99.4). Excluding one particularly anomalous study,41 the lowest sensitivities and specificities would be 69% and 93%, respectively. Pooled PLR and NLR were 239.24 (95% CI: 17.139, 33339.4) and 0.077 (95%CI: 0.035, 0.168), respectively, with tau-square 20.17, 1.556, respectively. [Table 5, 6] BinaxNOW HBsAg was evaluated in three articles, (15, 18, 23) all published before 2007, contributing 6 data points evaluating against an EIA reference, using a total 3550 samples, with sample sizes ranging from 36 to 1011. Sensitivities ranged from 94% to 100% with overall pooled sensitivity of 97.6% (95% CI: 96.2, 98.6). Specificity was 100% in all studies, with overall pooled specificity of 100% (95% CI: 99.7, 100). Pooled PLR and NLR were 221.21 (95% CI: 36.160, 1354.1) and 0.045 (95%CI: 0.016, 0.128), respectively, with tau-square 3.53, 1.20, respectively. [Tables 5, 6] VIKIA HBsAg was also evaluated in three articles, 16, 36, 40 all published after 2010, contributing 3 data points evaluating against an EIA reference, using a total 5242 samples, with sample sizes ranging from 476 to 3928. Sensitivities ranged from 71% to 97% with overall pooled sensitivity of 82.5% (95% CI: 77.5, 86.7). Specificities ranged from 99.8% to 100%, with overall pooled specificity of 99.9% (95% CI: 99.8, 100). Pooled PLR and NLR were 1072.3 (95% CI: 376.082, 3057.2) and 0.108 (95%CI: 0.026, 0.458), respectively, with tau-square <0.005, 1.472, respectively. [Tables 5, 6] Serodia HBsAg was also evaluated in three articles24, 27, 31 all published before 2000, contributing 3 data points evaluating against an EIA reference, using a total 1040 samples. Sensitivities ranged from 71% to 97% with overall pooled sensitivity of 82.5% (95% CI: 77.5, 86.7). Specificities ranged from 99.8% to 100%, with overall pooled specificity of 99.9% (95% CI: 99.8, 100). Pooled PLR and NLR were 284.91 (95% CI: 71.42, 1136.6) and 0.045 (95%CI: 0.029, 0.069), respectively, with tau-square <0.005, <0.005, respectively. [Tables 5, 6] Pooled accuracy of RDTs evaluated against CMIA reference Five articles33, 36, 38, 40, 43 contributed 9 data points evaluating 6 RDTs against specifically a CMIA reference, using a total 4513 samples, with sample sizes ranging from 178 to 838 (mean 501). Sensitivities ranged from 62% to 100% with overall pooled sensitivity of 80.4% (95% CI: 77.9, 82.6). Specificities ranged from 93% to 100%, with overall pooled specificity of 99.0% (95% CI: 98.6, 99.3). Pooled PLR and NLR were 58.5 (95% CI: 31.3, 109.3) and 0.141 (95%CI: 0.074, 0.268), respectively, with tau-square 0.4375, 0.7337, respectively. [Table 5]

Page | 207

Fig. 3. Forest plots, RDT vs EIA, ordered by [Test, Author]* Specificity (95% CI) Abon - Mutocheluh Sensitivity 0.99 (0.96 -(95% 1.00) CI) Specificity (95% CI) 0.99 0.95 1.00 1.00 0.91 0.99 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.86 0.98 1.00 0.96 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.93 1.00 0.69 1.00 1.00 0.99 1.00 0.99 0.99 0.98 0.95 1.00 1.00 0.99 0.96 1.00 0.98 1.00 1.00 1.00 1.00 0.99 0.99 1.00 1.00 1.00 1.00 1.00 0.99 1.00 1.00 1.00 1.00 0.97 0.98 0.99 (0.96 (0.74 (0.99 (0.99 (0.84 (0.96 (0.40 (0.03 (0.89 (0.99 (1.00 (1.00 (0.99 (0.64 (0.94 (0.92 (0.91 (0.98 (0.99 (1.00 (0.93 (0.99 (0.99 (0.99 (0.98 (0.99 (0.99 (0.78 (0.99 (0.62 (0.97 (0.98 (0.98 (0.99 (0.97 (0.97 (0.97 (0.82 (0.96 (1.00 (0.98 (0.91 (0.95 (0.95 (0.82 (0.92 (1.00 (0.88 (0.98 (0.96 (0.99 (0.98 (0.96 (0.98 (0.98 (0.98 (1.00 (0.99 (0.99 (0.88 (0.95 (0.94 (0.96 1.00) 1.00) 1.00) 1.00) 0.96) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 0.97) 1.00) 1.00) 0.99) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 0.99) 1.00) 0.76) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 0.99) 0.99) 1.00) 1.00) 0.99) 0.99) 1.00) 0.99) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 0.98) 1.00) 1.00)

Sensitivity (95% CI) Abon - Mutocheluh Accurate - Khan ACON - Bjoerkvoll (Camb) ACON - Bjoerkvoll (Viet) ACON - Erhabor Acull-Tell - Mutocheluh AMRAD - Ola Binax - Akanmu (BD) Binax - Akanmu (CLD) Binax - Clement Binax - Lau (Fresh S) Binax - Lau (Frozen S) Binax - Lau (WB) Biotec - Ola Core TM - Mutocheluh Cortez - Chameera Cypress - Randrianirina Dainascreen - Lien Dainascreen - Sato Determine - Bottero Determine - Davies (H+) Determine - Franzeck (H+) Determine - Geretti (H+) Determine - Hoffman (H+) Determine - Lien Determine - Lin (China) Determine - Lin (Guinea) Determine - Njai (CHB) Determine - Njai (Screen) Determine - Nyirendra Determine - Randrianirina Dipstick (PATH) - Mvere DRW - Lin (China) DRW - Lin (Guinea) DRW v2.0 - Chevaliez (Hep) DRW v2.0 - Chevaliez (Preg) DRW v2.0 - Chevaliez CC Espline - Njai (CHB) Genedia - Oh Hepacard - Kaur Hepacard - Raj Hexagon - Randrianirina Intec - Liu Nanosign - Gish (H+) Onecheck - Khan Onsite - Chameera QUICK PROFILE - Bottero QuickChaser - Abraham CC QuickChaser - Abraham CS Rapid care - Mutocheluh SD Bioline - Upretti Serodia - Lien Serodia - Oh Serodia - Sato SimpliRed - Mvere VEDA LAB - Honge (H+) VIKIA - Bottero VIKIA - Geretti (H+) VIKIA - Njai (Screen) Virucheck - Abraham CC Virucheck - Abraham CS Virucheck - Randrianirina Wondfo - Mutocheluh 0.50 0.50 0.93 0.82 1.00 0.55 0.95 0.94 1.00 1.00 0.94 0.95 0.96 0.59 0.50 0.60 0.97 1.00 1.00 0.94 1.00 0.96 0.69 0.75 1.00 0.99 0.94 0.95 0.88 0.56 0.98 0.93 0.99 0.97 1.00 0.96 0.95 0.94 0.98 0.93 0.83 0.96 0.51 0.74 0.53 0.80 0.90 0.90 0.77 0.55 1.00 0.96 0.96 0.96 0.93 0.62 0.96 0.71 0.90 0.90 0.79 0.96 0.59 (0.28 - 0.72) (0.33 - 0.67) (0.86 - 0.98) (0.74 - 0.88) (0.88 - 1.00) (0.32 - 0.76) (0.76 - 1.00) (0.87 - 0.97) (0.40 - 1.00) (0.99 - 1.00) (0.79 - 0.99) (0.86 - 0.99) (0.89 - 0.99) (0.41 - 0.75) (0.28 - 0.72) (0.15 - 0.95) (0.91 - 0.99) (0.97 - 1.00) (0.98 - 1.00) (0.82 - 0.99) (0.86 - 1.00) (0.80 - 1.00) (0.61 - 0.77) (0.59 - 0.87) (0.97 - 1.00) (0.96 - 1.00) (0.90 - 0.97) (0.90 - 0.98) (0.81 - 0.94) (0.38 - 0.73) (0.92 - 1.00) (0.68 - 1.00) (0.97 - 1.00) (0.93 - 0.99) (0.98 - 1.00) (0.81 - 1.00) (0.76 - 1.00) (0.89 - 0.97) (0.94 - 1.00) (0.84 - 0.98) (0.61 - 0.95) (0.89 - 0.99) (0.43 - 0.58) (0.49 - 0.91) (0.36 - 0.69) (0.28 - 0.99) (0.82 - 0.96) (0.68 - 0.99) (0.55 - 0.92) (0.32 - 0.76) (0.63 - 1.00) (0.90 - 0.99) (0.91 - 0.99) (0.91 - 0.98) (0.68 - 1.00) (0.51 - 0.73) (0.90 - 0.99) (0.62 - 0.78) (0.79 - 0.96) (0.68 - 0.99) (0.54 - 0.94) (0.89 - 0.99) (0.36 - 0.79)

Accurate - Khan 0.95 (0.74 - 1.00) Abon - Mutocheluh Abon - Mutocheluh 0.50 (0.28 - 0.72) ACON - Bjoerkvoll 1.00 (0.99 - 1.00) Accurate - 0.50 Khan Accurate - Khan (Camb) (0.33 - 0.67) ACON --Bjoerkvoll (Viet) 1.00 (0.99 - 1.00) ACON (Camb) ACON Bjoerkvoll (Camb) - Bjoerkvoll 0.93 (0.86 - 0.98) (Viet) ACON Bjoerkvoll ACON (Viet) - Bjoerkvoll 0.82 (0.74 - 0.88) ACON --Erhabor 0.91 (0.84 - 0.96) ACON - Erhabor ACON - Erhabor 1.00 (0.88 - 1.00) Acull-Tell - Mutocheluh 0.99 (0.96 - 1.00) Acull-Tell - 0.55 Mutocheluh Acull-Tell - Mutocheluh (0.32 - 0.76) AMRAD - Ola 1.00 (0.40 - 1.00) AMRAD 1.00 - Ola AMRAD - Ola (BD) 0.95 (0.76 - 1.00) Binax - Akanmu (0.03 - 1.00) Binax - Akanmu (BD) Binax Akanmu (BD) 0.94 (0.87 - 0.97) Binax --Akanmu (CLD) 1.00 (0.89 - 1.00) Binax - Akanmu Binax - Akanmu (CLD) 1.00 (CLD) (0.40 - 1.00) Binax - Clement 1.00 (0.99 - 1.00) Binax - Clement Binax - Clement 1.00 (0.99 - 1.00) Binax - Lau (Fresh S) 1.00 (1.00 - 1.00) - Lau (Fresh S) - 0.99) Binax - Lau (FreshBinax S) 0.94 (0.79 Binax --Lau S)S) - Lau 1.00 (1.00 - 1.00) Binax (Frozen S)- 0.99) Binax Lau(Frozen (Frozen 0.95 (0.86 Binax --Lau 1.00 (0.99 - 1.00) (WB) Binax Lau(WB) (WB) Binax - Lau 0.96 (0.89 - 0.99) Biotec --Ola 0.86 (0.64 - 0.97) Biotec - Ola Biotec Ola 0.59 (0.41 - 0.75) Core TM (0.94 - 1.00) Core TM0.98 - Mutocheluh Core TM--Mutocheluh Mutocheluh 0.50 (0.28 - 0.72) Cortez --Chameera (0.92 - 1.00) Cortez - 1.00 Chameera Cortez Chameera 0.60 (0.15 - 0.95) Cypress --Randrianirina (0.91 - 0.99) Cypress 0.96 - Randrianirina Cypress Randrianirina 0.97 (0.91 - 0.99) - (0.98 Lien Dainascreen LienDainascreen 1.00 (0.97 - 1.00) Dainascreen - Lien 1.00 - 1.00) Dainascreen - (0.99 Sato Dainascreen Sato 1.00 (0.98 - 1.00) Dainascreen - Sato 1.00 - 1.00) Determine Bottero Determine Bottero 0.94 (0.82 - 0.99) Determine - -Bottero 1.00 (1.00 - 1.00) Determine Davies Determine Davies (H+) 1.00 (0.86 - 1.00) Determine - -Davies (H+) 1.00 (0.93 -(H+) 1.00) Determine Franzeck (H+) Determine Franzeck (H+) 1.00 0.96 (0.80 - 1.00) Determine - -Franzeck (H+) (0.99 - 1.00) Determine Geretti (H+) Determine - Geretti (H+) 0.69 (0.61 - 0.77) Determine - Geretti (H+) 1.00 (0.99 - 1.00) Determine Hoffman (H+) Determine - Hoffman (H+) 0.75 (0.59 - 0.87) Determine - Hoffman (H+) 1.00 (0.99 - 1.00) Lien (0.97 - 1.00) Determine - Lien Determine 1.00 Determine - -Lien 1.00 (0.98 - 1.00) Determine Lin (China) Determine Lin (China) 0.99 (0.96 - 1.00) Determine - -Lin (China) 1.00 (0.99 - 1.00) Determine Lin (Guinea) Determine Lin (Guinea) 0.94 (0.90 - 0.97) Determine Lin (Guinea) 1.00 (0.99 1.00) Determine Njai (CHB) Determine - Njai (CHB) 0.95 (0.90 - 0.98) Determine - -Njai (CHB) 0.93 (0.78 - 0.99) Determine Njai (Screen) Determine Njai (Screen) 0.88 (0.81 - 0.94) Determine - -Njai (Screen) 1.00 (0.99 - 1.00) Determine Nyirendra Determine Nyirendra 0.56 (0.38 - 0.73) Determine Randrianirina Determine Randrianirina 0.98 (0.92 - 1.00) Determine - -Nyirendra 0.69 (0.62 - 0.76) Dipstick 1.00 (PATH) -(0.68 Mvere Dipstick (PATH) - Mvere 0.93 - 1.00) Determine - Randrianirina (0.97 - 1.00) DRW - Lin (China) DRW - Lin (China) 0.99 (0.97 - 1.00) Dipstick (PATH) - Mvere 1.00 (0.98 - 1.00) DRW - Lin (Guinea) DRW Lin(China) (Guinea) 0.97 (0.93 - 0.99) DRW --Lin 0.99 (0.98 - 1.00) DRW v2.0 Chevaliez (Hep) DRW - Chevaliez (Hep) 1.00 (0.98 - 1.00) DRW -v2.0 Lin (Guinea) 1.00 (0.99 - 1.00) DRW v2.0 Chevaliez DRW v2.0 - Chevaliez (Preg) 0.96 (0.81 (Preg) - 1.00) DRW v2.0 - Chevaliez (Hep) 0.99 (0.97 - 1.00) DRW v2.0 Chevaliez DRW v2.0 - Chevaliez CC 0.95 (0.76 CC - 1.00) DRW v2.0 - Chevaliez (Preg) 0.99 (0.97 - 1.00) Espline - Njai (CHB) Espline - Njai (CHB) 0.94 (0.89 - 0.97) DRW v2.0 Chevaliez CC 0.98 (0.97 0.99) Genedia - Oh Genedia - Oh 0.98 (0.94 - 1.00) Espline - Njai (CHB) Hepacard 0.95 (0.82 - 0.99) -0.93 Kaur Hepacard - Kaur (0.84 - 0.98) Genedia Oh 1.00 (0.96 - 1.00) Hepacard -0.83 Raj Hepacard - Raj (0.61 - 0.95) Hepacard 1.00 (1.00 - 1.00) Hexagon - Randrianirina Hexagon - Kaur Randrianirina 0.96 (0.89 - 0.99) Hepacard - Raj 0.99 (0.98 - 0.99) Intec - Liu Intec - Liu 0.51 (0.43 - 0.58) Nanosign - 0.74 Gish (H+) Nanosign Gish (H+) (0.49 - 0.91) Hexagon - Randrianirina 0.96 (0.91 - 0.99) Khan Onecheck 0.53 (0.36 - 0.69) Intec - Liu - Khan Onecheck 1.00 (0.95 - 1.00) Onsite - 0.98 Chameera Onsite - Chameera 0.80 (0.28 - 0.99) Nanosign - Gish (H+) (0.95 - 0.99) -Bottero QUICK PROFILE Bottero PROFILE 0.90 (0.82 - 0.96) Onecheck - Khan -QUICK 1.00 (0.82 1.00) QuickChaser -(0.92 Abraham CC QuickChaser - Abraham CC1.00 0.90 (0.68 - 0.99) Onsite - Chameera - 1.00) QuickChaser - Abraham CS QuickChaser - Abraham CS 0.77 (0.55 - 0.92) QUICK PROFILE - Bottero 1.00 (1.00 - 1.00) Rapid care 0.55 - Mutocheluh Rapid care - Mutocheluh (0.32 - 0.76) QuickChaser - Abraham CC 1.00 (0.88 - 1.00) SD Bioline 1.00 - Upretti SD Bioline - Upretti (0.63 - 1.00) QuickChaser - Abraham CS 0.99 (0.98 - 1.00) Serodia - Lien Serodia - Lien 0.96 (0.90 - 0.99) Rapid care - Mutocheluh 0.99 (0.96 - 1.00) Serodia Oh Serodia - Oh 0.96 (0.91 - 0.99) SD Bioline - Upretti Serodia 1.00 (0.99 - 1.00) Sato Serodia - Sato 0.96 (0.91 - 0.98) Serodia - Lien 1.00 (0.98 - 1.00) Mvere SimpliRed - Mvere SimpliRed 0.93 (0.68 - 1.00) SerodiaLAB - Oh - Honge 1.00 (0.96 - 1.00) VEDA - Honge (H+) VEDA (H+) LAB 0.62 (0.51 - 0.73) VIKIA - Bottero VIKIA -Bottero 0.96 (0.90 - 0.99) Serodia Sato 1.00 (0.98 - 1.00) VIKIA - Geretti (H+) VIKIA - Geretti (H+) 0.71 (0.62 - 0.78) SimpliRed - Mvere 1.00 (0.98 - 1.00) VIKIA (Screen) VIKIA - Njai (Screen) 0.90 (0.79 - 0.96) VEDA LAB - Honge (H+) - Njai 0.99 (0.98 - 1.00) Virucheck Abraham CC Virucheck - Abraham CC 1.00 0.90 (0.68 - 0.99) VIKIA - Bottero (1.00 - 1.00) Abraham CS Virucheck - Abraham CS 1.00 0.79 (0.54 - 0.94) VIKIA - Geretti (H+) Virucheck (0.99 - 1.00) Virucheck Randrianirina Virucheck - Randrianirina 0.96 (0.89 - 0.99) VIKIA - Njai (Screen) 1.00 (0.99 - 1.00) Wondfo - Mutocheluh Wondfo - Mutocheluh 0.59 (0.36 - 0.79) Virucheck - Abraham CC Virucheck - Abraham CS Virucheck - Randrianirina Wondfo - Mutocheluh 1.00 0.97 0.98 0.99 (0.88 - 1.00) (0.95 - 0.98) (0.94 - 1.00) (0.96 - 1.00)

Abon Accur ACON ACON ACON AcullAMRA Binax Binax Binax Binax Binax Binax Biotec Core Corte Cypre Daina Daina Deter Deter Deter Deter Deter Deter Deter Deter Deter Deter Deter Deter Dipsti DRW DRW DRW DRW DRW Esplin Gene Hepac Hepac Hexag Intec Nanos Onec Onsite QUIC Quick Quick Rapid SD Bi Serod Serod Serod Simpl VEDA VIKIA VIKIA VIKIA Viruch Viruch Viruch Wond

Specificity Pooled Sensitivity Pooled = 0.90 (0.89 to 0.91)= 0.99 (0.99 to 1.00) Chi-square == 651.98; Chi-square = 932.08; df = 62 (p 0.0000)df = 62 (p = 0.0000) (I-square) = 93.3 % (I-square) = 90.5 % .4 0.8 0.6 0.8 1 Inconsistency 1 Inconsistency Specificity

0.2

0.4 0.6 Sensitivity

0.8

Pooled Sensitivity = 0.90 (0.89 to 0.91) Chi-square = 932.08; df = 62 (p = 0.0000) Inconsistency 10 0.2 (I-square) 0.4 = 93.3 0.6% 0.8 Specificity

Poole Chi-sq 1 Incon

0.8

Pooled Specificity = 0.99 (0.99 to 1.00) Chi-square = 651.98; df = 62 (p = 0.0000) 1 Inconsistency (I-square) = 90.5 %

Page | 208

Positive LR (95% CI) Abon - Mutocheluh Accurate - Khan ACON - Bjoerkvoll (Camb) ACON - Bjoerkvoll (Viet) ACON - Erhabor Acull-Tell - Mutocheluh AMRAD - Ola Binax - Akanmu (BD) Binax - Akanmu (CLD) Binax - Clement Binax - Lau (Fresh S) Binax - Lau (Frozen S) Binax - Lau (WB) Biotec - Ola Core TM - Mutocheluh Cortez - Chameera Cypress - Randrianirina Dainascreen - Lien Dainascreen - Sato Determine - Bottero Determine - Davies (H+) Determine - Franzeck (H+) Determine - Geretti (H+) Determine - Hoffman (H+) Determine - Lien Determine - Lin (China) Determine - Lin (Guinea) Determine - Njai (CHB) Determine - Njai (Screen) Determine - Nyirendra Determine - Randrianirina Dipstick (PATH) - Mvere DRW - Lin (China) DRW - Lin (Guinea) DRW v2.0 - Chevaliez (Hep) DRW v2.0 - Chevaliez (Preg) DRW v2.0 - Chevaliez CC Espline - Njai (CHB) Genedia - Oh Hepacard - Kaur Hepacard - Raj Hexagon - Randrianirina Intec - Liu Nanosign - Gish (H+) Onecheck - Khan Onsite - Chameera QUICK PROFILE - Bottero QuickChaser - Abraham CC QuickChaser - Abraham CS Rapid care - Mutocheluh SD Bioline - Upretti Serodia - Lien Serodia - Oh Serodia - Sato SimpliRed - Mvere VEDA LAB - Honge (H+) VIKIA - Bottero VIKIA - Geretti (H+) VIKIA - Njai (Screen) Virucheck - Abraham CC Virucheck - Abraham CS Virucheck - Randrianirina Wondfo - Mutocheluh 64.00 (8.69 - 471.26) 9.50 (1.37 - 65.71) 517.87 (129.54 - 2,070.26) 434.51 (108.57 - 1,739.03) 10.46 (5.70 - 19.19) 69.82 (9.55 - 510.36) 9.32 (0.67 - 129.54) 3.72 (0.34 - 41.10) 59.40 (3.74 - 944.36) 268.83 (67.41 - 1,072.16) 1,811.52 (113.19 - 28,991.24) 1,448.64 (90.62 - 23,156.72) 1,051.25 (65.81 - 16,793.27) 4.12 (1.39 - 12.18) 32.00 (7.60 - 134.68) 53.67 (3.14 - 917.12) 26.35 (10.06 - 69.00) 422.20 (26.49 - 6,728.17) 608.08 (38.12 - 9,699.86) 4,498.21 (281.12 - 71,976.88) 100.04 (6.34 - 1,578.57) 467.38 (29.27 - 7,463.70) 966.70 (60.39 - 15,473.86) 174.94 (64.74 - 472.74) 422.20 (26.49 - 6,728.17) 959.01 (60.07 - 15,311.15) 753.59 (47.21 - 12,030.12) 14.29 (3.74 - 54.54) 1,180.48 (73.85 - 18,870.34) 1.82 (1.25 - 2.67) 214.02 (13.47 - 3,400.91) 348.00 (21.75 - 5,568.55) 120.60 (45.44 - 320.04) 771.27 (48.32 - 12,311.21) 74.50 (29.76 - 186.52) 73.38 (30.62 - 175.83) 43.75 (27.08 - 70.69) 17.85 (4.63 - 68.81) 197.32 (12.43 - 3,133.39) 4,996.94 (312.36 - 79,938.56) 73.30 (39.56 - 135.82) 26.05 (9.95 - 68.23) 73.20 (4.61 - 1,163.40) 34.14 (14.80 - 78.75) 21.03 (1.34 - 329.93) 69.00 (4.22 - 1,129.19) 347.25 (186.27 - 647.36) 54.62 (3.48 - 857.69) 146.05 (35.99 - 592.65) 69.82 (9.55 - 510.36) 642.22 (40.07 - 10,293.22) 404.24 (25.36 - 6,443.55) 193.30 (12.17 - 3,069.96) 290.53 (41.05 - 2,056.47) 348.00 (21.75 - 5,568.55) 75.01 (23.99 - 234.59) 1,853.68 (463.49 - 7,413.61) 986.53 (61.64 - 15,789.02) 374.40 (52.77 - 2,656.47) 54.62 (3.48 - 857.69) 25.07 (13.71 - 45.82) 52.10 (13.19 - 205.84) 75.64 (10.41 - 549.47)

0.01

1 Positive LR

Random Effects Model Pooled Positive LR = 117.54 (67.69 to 204.12) Cochran-Q = 658.59; df = 62 (p = 0.0000) 100.0 Inconsistency (I-square) = 90.6 % Tau-squared = 3.8937

Page | 209

Negative LR (95% CI) Abon - Mutocheluh Accurate - Khan ACON - Bjoerkvoll (Camb) ACON - Bjoerkvoll (Viet) ACON - Erhabor Acull-Tell - Mutocheluh AMRAD - Ola Binax - Akanmu (BD) Binax - Akanmu (CLD) Binax - Clement Binax - Lau (Fresh S) Binax - Lau (Frozen S) Binax - Lau (WB) Biotec - Ola Core TM - Mutocheluh Cortez - Chameera Cypress - Randrianirina Dainascreen - Lien Dainascreen - Sato Determine - Bottero Determine - Davies (H+) Determine - Franzeck (H+) Determine - Geretti (H+) Determine - Hoffman (H+) Determine - Lien Determine - Lin (China) Determine - Lin (Guinea) Determine - Njai (CHB) Determine - Njai (Screen) Determine - Nyirendra Determine - Randrianirina Dipstick (PATH) - Mvere DRW - Lin (China) DRW - Lin (Guinea) DRW v2.0 - Chevaliez (Hep) DRW v2.0 - Chevaliez (Preg) DRW v2.0 - Chevaliez CC Espline - Njai (CHB) Genedia - Oh Hepacard - Kaur Hepacard - Raj Hexagon - Randrianirina Intec - Liu Nanosign - Gish (H+) Onecheck - Khan Onsite - Chameera QUICK PROFILE - Bottero QuickChaser - Abraham CC QuickChaser - Abraham CS Rapid care - Mutocheluh SD Bioline - Upretti Serodia - Lien Serodia - Oh Serodia - Sato SimpliRed - Mvere VEDA LAB - Honge (H+) VIKIA - Bottero VIKIA - Geretti (H+) VIKIA - Njai (Screen) Virucheck - Abraham CC Virucheck - Abraham CS Virucheck - Randrianirina Wondfo - Mutocheluh 0.50 0.53 0.07 0.18 0.02 0.46 0.08 0.09 0.10 0.00 0.08 0.06 0.05 0.48 0.51 0.42 0.03 0.00 0.00 0.07 0.02 0.06 0.31 0.25 0.00 0.01 0.06 0.05 0.12 0.64 0.03 0.09 0.01 0.04 0.00 0.04 0.05 0.06 0.02 0.07 0.18 0.05 0.50 0.27 0.49 0.25 0.10 0.12 0.23 0.46 0.06 0.05 0.04 0.04 0.09 0.38 0.04 0.29 0.10 0.12 0.22 0.04 0.41 (0.33 (0.38 (0.03 (0.13 (0.00 (0.29 (0.02 (0.03 (0.01 (0.00 (0.02 (0.02 (0.02 (0.31 (0.33 (0.16 (0.01 (0.00 (0.00 (0.03 (0.00 (0.01 (0.24 (0.15 (0.00 (0.00 (0.03 (0.02 (0.07 (0.43 (0.01 (0.02 (0.00 (0.02 (0.00 (0.01 (0.01 (0.03 (0.01 (0.03 (0.07 (0.02 (0.43 (0.13 (0.35 (0.06 (0.05 (0.04 (0.11 (0.29 (0.00 (0.02 (0.02 (0.02 (0.02 (0.28 (0.01 (0.23 (0.05 (0.04 (0.09 (0.02 (0.25 0.77) 0.74) 0.14) 0.26) 0.28) 0.72) 0.36) 0.26) 1.41) 0.02) 0.25) 0.16) 0.13) 0.74) 0.77) 1.09) 0.10) 0.07) 0.05) 0.20) 0.30) 0.27) 0.40) 0.43) 0.07) 0.05) 0.10) 0.11) 0.20) 0.94) 0.09) 0.43) 0.04) 0.08) 0.04) 0.26) 0.33) 0.12) 0.07) 0.18) 0.43) 0.12) 0.58) 0.57) 0.68) 1.01) 0.18) 0.39) 0.49) 0.72) 0.82) 0.11) 0.09) 0.09) 0.43) 0.51) 0.11) 0.38) 0.21) 0.39) 0.52) 0.12) 0.68)

0.01

1 Negative LR

Random Effects Model Pooled Negative LR = 0.10 (0.07 to 0.14) Cochran-Q = 1298.04; df = 62 (p = 0.0000) 100.0 Inconsistency (I-square) = 95.2 % Tau-squared = 1.7188

Page | 210

Key brands of RDT and reference standards used in studies Study Mvere, 1996 Test brand (manufacturer) Dipstick (PATH) SimpliRed Sato, 1996 Dainascreen HBsAG Serodia HBsAg Abraham, 1998 QuickChaser Virucheck Oh, 1999 Genedia Serodia Kaur, 2000 Lien, 2000 Hepacard Dainascreen Determine HBsAg Serodia Raj, 2001 Hepacard EIA, Auszyme MEIA, AxSYM v2 Clement, 2002 Lau, 2003 Akanmu, 2006 Bjoerkvoll, 2010 Lin, 2008 BinaxNOW BinaxNOW BinaxNOW ACON Determine HBsAg DRW HBsAg Nyirendra, 2008 Determine HBsAG EIA, Bioelisa Neutralisation (positives) Randrianirina, 2008 Cypress Determine HBsAg Hexagon Virucheck Ola, 2009 AMRAD GWHB Biotec Latex Davies, 2010 Geretti, 2010 Determine Determine VIKIA Khan, 2010 Accurate Onecheck EIA, Biokit; Neutralisation (for all positives) CMIA, Architect/ Liason EIA, Murex v3 ELISA, 2 generation nd

Reference test type, brand (manufacturer) EIA, Auszyme

EIA, Auszyme

EIA, Auszyme or Hepanostika

EIA, Cobas Core

EIA, Ortho 3 generation EIA, Monolisa MEIA for discordant

MEIA, AxSYM v2 EIA, ETI-MAK2 ELISA, Monolisa EIA, Monolisa Ultra* EIA, Hepanostika Ultra

EIA, AxSYM

ELISA, Wellcozyme Kit

Page | 211

Study Hoffman, 2012 Bottero, 2013

Test brand (manufacturer) Determine Determine QUICK PROFILE VIKIA

Reference test type, brand (manufacturer) ELISA, AxSYM ELISA, Monolisa Ultra Neutralisation (for all positive)

Franzeck, 2013

Determine HBsAg

EIA, Murex v3 Neutralisation (for all positives)

Chameera, 2013

Cortez Onsite

EIA, Surase B-96 (TMB)

Chevaliez, 2014 Erhabor, 2014 Gish, 2014 Honge, 2014 Liu, 2014 Upretti, 2014 Mutocheluh, 2014

DRW v 2 HBsAg ACON Nanosign VEDA LAB Intec One Step SD Bioline Abon Acull-Tell Core-TM Rapid care Wondfo

CMIA, Architect ELISA, HBsAg Ultra EIA, Quest Diagnostics CLIA, Architect CMIA, Architect EIA, Surase B-96 (TMB) ELISA, Human Gesellschaft

Njai, 2015

Determine HBsAg VIKIA Espline

EIA (DBS), AxSym Neutralisation CMIA (quantitative), Architect

Page | 212

Fig. 4. Forest plots, EIA vs EIA, ordered by [Test, Author]**

Specificity (95% CI) Sensitivity (95% CI) Architect -0.98 Geretti H+ - 1.00) 1.00 (0.99 Architect - Geretti H+ (0.94 Cobas - Liu (0.95 Cobas - Liu 0.97 (0.93 - 0.99) 1.00 KHB - Peng (0.82 KHB - Peng 0.74 (0.69 - 0.79) 0.88 Specificity (95% CI) Liaison - Geretti (0.99 Liaison Ultra - Geretti H+ Ultra 0.97 (0.93 H+ - 0.99) 0.99 Architect - Geretti H+ 1.00 - 1.00) Monolisa Ol (0.99 (0.79 Monolisa Ol 1.00 (0.95 - 1.00) 0.91 Cobas - Liu 1.00 (0.95 - 1.00) Monolisa UltraViet (0.78 Monolisa Ultra- Viet 0.99 (0.93 - 1.00) 0.91 Murex v3.0 - (0.82 Geretti H+ (0.98 Murex v3.0 - Geretti H+ 0.99 (0.95 - 1.00) 0.99 KHB - Peng 0.88 - 0.93) Wantai - Liu (0.95 Wantai - Liu - Geretti 0.78 (0.71 - 0.84) 1.00 Liaison Ultra H+ 0.99 (0.99 - 1.00) Monolisa - Ol 0.91 (0.79 - 0.98) Monolisa Ultra- Viet 0.91 (0.78 - 0.97) Murex v3.0 - Geretti H+ 0.99 (0.98 - 1.00) Wantai - Liu 1.00 (0.95 - 1.00) Pooled Specificity = 0.98 (0.98 to 0.99) Pooled Sensitivity = 0.89 (0.87 to 0.91) Chi-square df = 7 (p = 0.0000) Chi-square = 166.37; df = 7 (p = = 88.56; 0.0000) (I-square) = 95.8 % (I-square) = 92.1 % 0.4 0.6 0.8 1 Inconsistency 0.8 1 Inconsistency Specificity Pooled Specificity = 0.98 (0.98 to 0.99) Chi-square = 88.56; df = 7 (p = 0.0000) 0.8 1 Inconsistency (I-square) = 92.1 % 1.00) 1.00) 0.93) 1.00) 0.98) 0.97) 1.00) 1.00)

Sensitivity (95% CI) Architect - Geretti H+ Cobas - Liu KHB - Peng Liaison Ultra - Geretti H+ Monolisa - Ol Monolisa Ultra- Viet Murex v3.0 - Geretti H+ Wantai - Liu 0.98 0.97 0.74 0.97 1.00 0.99 0.99 0.78 (0.94 (0.93 (0.69 (0.93 (0.95 (0.93 (0.95 (0.71 1.00)Architect - Geretti H+ 0.99)Cobas - Liu 0.79)KHB - Peng 0.99)Liaison Ultra - Geretti H+ 1.00)Monolisa - Ol 1.00)Monolisa Ultra- Viet 1.00)Murex v3.0 - Geretti H+ 0.84)Wantai - Liu

Specificity (95% CI) 1.00 1.00 0.88 0.99 0.91 0.91 0.99 1.00 (0.99 (0.95 (0.82 (0.99 (0.79 (0.78 (0.98 (0.95 1.00) 1.00) 0.93) 1.00) 0.98) 0.97) 1.00) 1.00)

2 .6 y

0.2

0.4 0.6 Sensitivity

0.8

Pooled Sensitivity = 0.89 (0.87 to 0.91) Chi-square = 166.37; df = 7 (p = 0.0000) (I-square) = 95.8 1 0Inconsistency 0.2 0.4 0.6 % 0.8 Specificity

Pooled Specificity = 0.98 (0.98 to 0.99) Chi-square = 88.56; df = 7 (p = 0.0000) 1 Inconsistency (I-square) = 92.1 %

0.6 city

**H = HIV positive

Key – Types of EIA and reference standards used in studies Study Geretti Test type, brand (manufacturer) EIA, Murex v.3.0 (Abbott) CMIA, Architect (Abbott) CMIA, Liaison Ultra (Diasorin) Liu ELISA, Wantai (Beijing Wantai) ECLIA, Cobas e601 (Roche) CMIA, Architect (Abbott) Reference CMIA and EIA (agreement) or neutralization

Page | 213

Ol Viet Peng

ELISA, Monolisa (bioRad) EIA, Monolisa Ultra (bioRad) ELISA, KHB (Kehua Bio-engineeering Co)

CMIA, Architect (Abbott) CMIA, Architect (Abbott) CMIA, Architect (Abbott)

Positive LR (95% CI) Architect - Geretti H+ Cobas - Liu KHB - Peng Liaison Ultra - Geretti H+ Monolisa - Ol Monolisa Ultra- Viet Murex v3.0 - Geretti H+ Wantai - Liu 227.68 (73.59 - 704.41) 138.80 (8.76 - 2,198.11) 6.22 (4.14 - 9.35) 169.51 (63.77 - 450.57) 10.15 (4.22 - 24.43) 10.61 (4.17 - 26.99) 137.61 (57.44 - 329.64) 112.40 (7.09 - 1,781.70)

Negative LR (95% CI) Architect - Geretti H+ Cobas - Liu KHB - Peng Liaison Ultra - Geretti H+ Monolisa - Ol Monolisa Ultra- Viet Murex v3.0 - Geretti H+ Wantai - Liu 0.02 0.04 0.30 0.03 0.01 0.01 0.01 0.22 (0.01 (0.02 (0.25 (0.01 (0.00 (0.00 (0.00 (0.17 0.07) 0.08) 0.36) 0.08) 0.12) 0.10) 0.06) 0.29)

0.01

1 Positive LR

Random Effects Model Pooled Positive LR = 46.77 (12.86 to 170.03) Cochran-Q = 100.82; df = 7 (p = 0.0000) 100.0 Inconsistency (I-square) = 93.1 % Tau-squared = 2.9510

0.01

1 Negative LR

Random Effects Model Pooled Negative LR = 0.04 (0.01 to 0.13) Cochran-Q = 238.26; df = 7 (p = 0.0000) 100.0 Inconsistency (I-square) = 97.1 % Tau-squared = 2.4644

Abbreviations used in Forest plots +

*Camb: Cambodia; Viet: Viet Nam; BD: blood donor study; CLD: chronic liver disease study; Fresh S: fresh serum; Frozen S: frozen serum; WB: whole blood; H : HIV positive; CHB: chronic hepatitis B cohort; Screen: general screen cohort; Hep: acute hepatitis cohort; Preg: antenatal cohort; CC: case–control study; CS: cross-sectional study **H : HIV positive +

Page | 214

Table 3. Summary pooled diagnostic accuracy of HBsAg assays using EIA and NAT reference standards Reference Index test Pooled clinical accuracy Likelihood ratios (REM) 5

Heterogeneity (Tau-squared)

Sen (95% CI)

Spec (95% CI) 99.5 (99.4–99.5) 98.4 (97.8–98.8) 98.1 (97.0–98.9) 86.1 (83.8–88.2)

PLR (95% CI) 118 (67.7–204) 46.8 (12.9–170) 39.4 (22.1–70.2) 7.23 (4.44–11.8)

NLR (95% CI) 0.095 (0.067–0.136) 0.041 (0.013–0.134) 0.051 (0.009–0.27) 0.296 (0.192–-0.458)

PLR

NLR

EIA

RDT

90.0 (89.1–90.8)

3.89

1.72

EIA

88.9 (87.0–90.6)

2.95

2.46

NAT

RDT

93.3 (91.3–94.9)

0.22

6.26

EIA

75.7 (72.1–79.1)

0.37

0.33

*n = number of data points

REM : Random effects model

Page | 215

Table 4. Summary pooled diagnostic accuracy of HBsAg assays in patients with known HIV status Test type HIV status n Sen (95% CI) RDT HIV +6

Pooled clinical accuracy Spec (95% CI) 99.8 (99.5–99.9) 99.6 (99.0–99.9) 99.4 (99.0–99.7) PLR

Likelihood ratios (REM) NLR (95% CI) 0.29 (0.22–0.38) 0.08 (0.05–0.13) 0.02 (0.01–0.04)

Heterogeneity (Tau-squared) PLR NLR

(95% CI) 193 (77.4–497) 79.5 (11.6–545) 167 (95.1–294)

72.3 (67.9–76.4)

0.384

0.0059

HIV–

92.6 (89.8–94.8)

2.967

0.080

EIA

HIV

97.9 (96.0–99.0)

<0.005

HIV–

Three studies on ART-naive patients (Hoffman, Davies, Franzeck); two studies from single article (Geretti) in patients who included 1/3 on lamivudine; with one study (Honge) on a mixture

Page | 216

9.1.1.Table 5. Summary pooled diagnostic accuracy of rapid HBsAg assays stratified by study, patient, index and reference tests Sub-analysis n Study Pre 2005 Post 2005 Case–control Patient Blood donors HIV+ HIV– Index test Whole blood Determine BinaxNOW VIKIA Serodia Reference test CMIA 19 44 21 19 6 4 11 12 6 3 3 9 Pooled clinical accuracy Sen (95% CI) 96.9 (96.0–97.7) 86.4 (85.2–87.5) 96.7 (96.0–97.3) 91.6 (90.1–92.9) 72.3 (67.9 –76.4) 92.6 (89.8–94.8) 91.7 (89.1–93.9) 90.8 (88.9–92.4) 97.6 (96.2–98.6) 82.5 (77.5–86.7) 82.5 (77.5–86.7) 80.4 (77.9–82.6) Spec (95% CI) 99.7 (99.6–99.8) 99.4 (99.2–99.5) 99.3 (99.0–99.5) 99.5 (99.3–99.7) 99.8 (99.5–99.9) 99.6 (99.0–99.9) 99.9 (99.8–99.9) 99.1 (98.9–99.4) 100 (99.7–100) 99.9 (99.8–100) 99.9 (99.8–100) 99.0 (99.6–99.3) Likelihood ratios (REM) LR+ (95% CI) 266 (106–665) 84.6 (43.6–165) 105 (48.0–230) 89.2 (32.8–243) 193 (77.4–497) 79.5 (11.6–545) 347 (158–762) 239 (17.1–33300) 221 (36.1–1350) 1070 (376–3060) 285 (71.4–1140) 58.5 (31.3–109) LR(95% CI) 0.056 (0.033–0.095) 0.126 (0.087–0.183) 0.028 (0.010–0.076) 0.106 (0.055–0.204) 0.29 (0.22–0.38) 0.08 (0.05-0.13) 0.089 (0.058–0.136) 0.077 (0.035–0.168) 0.045 (0.016–0.128) 0.108 (0.026–0.458) 0.045 (0.029–0.069) 0.141 (0.074–0.268) 2.72 4.10 2.23 3.82 0.384 2.97 0.81 20.2 3.53 <0.005 <0.005 0.44 Heterogeneity (Tau-squared) PLR 0.91 1.27 4.86 1.86 0.0059 0.080 0.24 1.56 1.20 1.472 <0.005 0.73 NLR

Page | 217

n = number of data points

Page | 218

Table 6. Summary pooled diagnostic accuracy of HBsAg assays by brand EIA reference Type RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT Brand name Abon Accurate ACON Acull–Tell AMRAD Atlas BINAX Blue Cross Biotec Core TM Cortez Cypress Dainascreen Determine DIMA Dipstick (PATH) DRW DRW v2 Espline Genedia Hepacard Hexagon Intec Nanosign Onecheck Onsite Quick Profile QuickChaser Rapid Care SD Bioline Serodia SimpliRed VEDA Lab 93.3 (68.1–99.8) 98.1 (96.1–99.2) 99.3 (97.4–99.9) 93.9 (89.1–97.1) 98.0 (94.3–99.6) 90.5 (82.1–95.8) 95.6 (89.1–98.8) 50.8 (43.3–58.4) 73.7 (48.8–90.9) 52.6 (35.8–69.0) 80.0 (28.4–99.5) 90.5 (82.1–95.8) 83.3 (68.6–93.0) 54.5 (32.2–75.6) 100 (63.1–100) 95.8 (93.4–97.5) 93.3 (68.1-99.8) 62.3 (50.6-73.1) 100 (98.1–100) 99.5 (98.8–99.9) 98.3 (97.5–98.9) 94.7 (82.3–99.4) 100 (96.4–100) 99.7 (99.5–99.9) 96.4 (90.9–99.0) 100 (94.9–100) 97.8 (95.4–99.2) 100 (82.4–100) 100 (92.1–100) 99.7 (99.5–99.9) 99.5 (98.2–99.9) 99.2 (95.7–100) 100 (98.9–100) 99.8 (99.1–100) 100 (98.1-100) 99.2 (97.6-99.8) 99.2 (95.4–100) 97.5 (92.9–99.5) 58.8 (40.7–75.4) 50.0 (28.2–71.8) 60.0 (14.7–94.7) 96.7 (90.7–99.3) 100 (98.7–100) 90.8 (88.9–92.4) 85.7 (63.7–97.0) 98.4 (94.5–99.8) 100 (92.1–100) 96.3 (90.9–99.0) 100 (99.3–100) 99.1 (98.9–99.4) 98.3 (94.1–99.8) 99.2 (95.4–100) 79.7 (73.1–85.3) 97.2 (94.0–99.0) 97.6 (96.2–98.6) 99.9 (99.7–100) 99.2 (95.4–100) 98.3 (94.1–99.8) Sen (95% CI) 50.0 (28.2–71.8) 50.0 (33.4–66.6) 88.0 (83.4–91.7) 54.5 (33.2–75.6) 95.2 (76.2–99.9) Spec (95% CI) 99.2 (95.7–100) 94.7 (74.0–99.9) 99.4 (99.0–99.7) 99.2 (95.7–100) 100 (39.8–100) 97.5 (92.9–99.5) 97.5 (92.9–99.5) 92.9 (87.3–96.5) 99.1 (96.6–99.9) NAT reference Sen (95% CI) Spec (95% CI)

Page | 219

RDT RDT RDT EIA EIA EIA EIA EIA EIA EIA EIA EIA EIA EIA EIA EIA EIA

VIKIA Virucheck Wondfo ADVIA Architect AxSym AxSym v2 Cobas Elecsys KHB Liaison Liaison Ultra Monolisa Monolisa Ultra Murex v3 VIDAS Ultra Wantai

82.5 (77.5-86.7) 92.3 (86.3-96.2) 59.1 (36.4-79.3)

99.9 (99.8-100) 97.3 (95.5-98.5) 99.2 (95.7-100) 77.4 (65.0–87.1) 97.9 (92.6–99.7)

97.9 (93.9–99.6)

99.6 (98.7–99.9) 56.6 (44.7–67.9) 77.4 (67.0–85.8) 86.8 (81.5–90.9) 75.0 (66.1–82.6)

96.6 (92.8–98.8)

100 (94.9–100) 63.4 (55.2–71.0) 95.8 (92.4–98.0)

73.8 (68.7–78.6)

88.1 (82.4–92.5) 100 (97.6–100) 70.0 (63.1–76.3)

97.1 (92.8–99.2) 100 (95.2–100) 98.7 (92.9–100) 98.6 (94.9–99.8)

99.4 (98.5–99.8) 91.1 (78.8–97.5) 90.7 (77.9–97.4) 99.3 (98.3–99.8) 69.0 (58.0–78.7) 94.0 (88.0–97.5)

78.2 (71.4–84.0)

100 (94.9–100)

Pooled results and I2 (heterogeneity) Test ACON Cortez ADVIA AxSym Elecsys ACON Binax Dainascreen Determine DRW DRW v2 Hepacard Quickchaser Serodia VIKIA Virucheck Ref NAT NAT NAT NAT NAT EIA EIA EIA EIA EIA EIA EIA EIA EIA EIA EIA Studies, n 2 2 1 1 2 2 3 2 10 1 1 2 1 3 3 2 Data, nd 2 3 2 2 2 3 6 2 12 2 3 2 2 3 3 3 I Sen, % 96.0 97.5 0.0 95.8 0.0 86.8 77.3 0.0 92.9 76.2 71.5 51.3 20.7 0.0 93.4 61.0 2

I Spec, % 0.0 0.0 0.0 89.1 49.2 95.2 25.3 0.0 97.0 79.2 0.0 96.6 0.0 0.0 7.3 11.9

Page | 220

GRADE Question: Should RDTs be used to diagnose HBsAg in HIV-negative individuals? Sensitivity Specificity Outcome No. of studies (no. of patients)

0.88–0.95 0.93–1.00 Study design Risk of bias

Prevalences

20%

Factors that may decrease quality of evidence Indirectness Inconsistency Imprecision Publication bias None

Effect per 1000 patients/year pre-test probability of 5% 44–48 pre-test probability of 20% 176–190

Test accuracy QoE

4 studies 997 patients

Cross-sectional (cohort type accuracy study)

Serious

Not serious

Serious

⨁ ◯◯◯ Very low 1234

2–6

10–24

4 studies 997 patients

Cross-sectional (cohort type accuracy study)

Serious

Not serious

Serious

Not 5 serious

None

884–950

744–800

⨁ ⨁ ◯◯ Low 1235

0–66

0 –56

Downgraded by one for risk of bias: all studies were prospective cohort studies , although one was assessed as high risk of bias because patients were pre-selected based from known chronic hepatitis B patients. Although study was not specifically designed in HIV-negative patients, clear testing and results were included. Downgraded by one for inconsistency: unexplained heterogeneity may arise from differences between studies in specimen condition (serum, whole blood), specimen processing (field vs laboratory), reference tests (CMIA; EIA on dried blood spots) and study population (e.g. known chronic hepatitis B patients, general community screen). Downgraded by one for imprecision: confidence intervals extend below 90% accuracy, with tau-squared for PLR >1 (indicating substantial heterogeneity).

Question: Should RDTs be used to diagnose HBsAg in HIV-positive individuals? Sensitivity 0.72 (95% CI: 0.68–0.76) Prevalences 5% 20% Page | 221

Specificity Outcome

1.00 (95% CI: 0.99–1.00) No. of studies (no. of patients) Study design Risk of bias Factors that may decrease quality of evidence Indirectness Inconsistency Imprecision Publication bias None Effect per 1000 patients/year pre-test pre-test probability probability of 5% of 20% 36 (34–38) 145 (136–153) ⨁ ◯◯◯ Very low 14 (12–16) 55 (47–64) 1234

Test accuracy QoE

5 studies 2566 patients

Cross-sectional (cohort type accuracy study)

Serious

Not serious Serious

Serious

5 studies 2566 patients

Cross-sectional (cohort type accuracy study)

Serious

Not serious Not serious

Not 6 serious

None

948 (945–949)

798 (796–799)

⨁ ⨁ ⨁ ◯ Moderate 1256

2 (1–5)

2 (1–4)

Downgraded by one for risk of bias: all studies were prospective cohort studies of consecutive patients. Studies used different specimens (serum, 2; capillary whole blood, 1; venous whole blood, 1), reference standards (CMIA, EIA confirmed by neutralization), and had patients with different ART status (four studies ART naive). Not downgraded for indirectness: all studies performed in cohorts of consecutive patients in Tanzania , Ghana , Malawi , South Africa and Bissau . Downgraded by one for inconsistency with sensitivities ranging from 62% to 100%: unexplained heterogeneity may arise from differences between studies in specimen type, specimen processing and 19, 20 36 38 study population. Two Studies had very high sensitivities (100%, 96%) while remainder ‒ had low sensitivities (range 62-70%). Tau-squared <1 for studies Downgraded by one for imprecision: confidence intervals 67.9–76.4%. Two studies 19, 20 20 36 19 37 38

had very high sensitivities (100%, 96%) while remainder

36‒38

had low sensitivities (range 62–70%).

Not downgraded for inconsistency: specificities ranged from 99% to 100%, with tau-squared <1

6. Not downgraded for imprecision: narrow confidence interval

Page | 222

Question: Should Determine HBsAg be used to diagnose HBsAg in a global setting? Sensitivity Specificity Outcome

0.91 (95% CI: 0.89 to 0.92) 0.99 (95% CI: 0.99 to 0.99) No. of studies (no. of patients) Study design Risk of bias

Prevalences

20%

Factors that may decrease quality of evidence Indirectness Inconsistency Imprecision Publication bias None

Effect per 1000 patients/year pre-test probability of 5% 45 (44–46) pre-test probability of 20% 182 (178–185)

Test accuracy QoE

12 studies 7552 patients

Cohort & case–control type studies1

Serious

Not serious

Very serious2

Not serious

⨁ ◯◯◯ Very low2

5 (4–6)

18 (15–22)

12 studies 7552 patients

Cohort & case–control type studies

Serious

Not serious

Serious3

Not serious

None

941 (940–944)

793 (791–795)

⨁ ⨁ ◯◯ Low3

9 (6–10)

7 (5–9)

1. 2. 3.

Lin , Lien and Randrianirina used a case–control design Significant heterogeneity across studies for sensitivity; tau-squared 20.2 Heterogeneity exists, but with lower clinical impact; tau-squared 1.56

Page | 223

6. Discussion Study findings Diagnostic accuracy of HBsAg assays using immunoassay reference Overall, the diagnostic accuracy of 33 RDTs and 8 EIAs were assessed against an EIA reference standard. Total numbers of patients included 36,131 (RDT vs EIA, figure 3) and 3751 (EIA vs EIA, figure 4). Both RDTs and EIAs had similar sensitivity and specificity compared to an EIA reference standard (Table 3). Clinical sensitivity estimates for both RDTs and EIAs were characterized by statistical heterogeneity, whereas specificity estimates were less heterogeneous (Figures 3 and 4). This applied across brands (Table 6). Heterogeneity can be caused by the use of different reference standards assays, clinical subgroups within the study population, age (children versus adults), patient status and stage of disease. Compared to previous systematic reviews, the pooled clinical sensitivity 90.0% (95% CI: 89.1–90.8) and specificity 99.5% (95% CI: 99.4–99.5) is slightly inferior for RDTs compared to an EIA reference standard (Table 3). In particular Results were very heterogeneous in terms of sensitivity (Table 7). Khuroo et al.11 reported 96.7% sensitivity (95% CI: 95.3, 97.7) and 99.7% specificity (95% CI: 98.6, 99.9). Studies included conference abstracts and studies using reference panels. Shivkumar et al.60 reported a pooled sensitivity 98.2% (95% CI: 94.7, 99.9) and pooled specificity 99.9% (95% CI: 99.3, 100). Table 7. Summary pooled diagnostic accuracy of HBsAg assays compared to previous reviews. Review, year Index Test PICO 1, 2015 RDT EIA Hwang, 2008 Shivkumar, 2012 Khuroo, 2014 RDT RDT RDT n 63 8 Clinical Accuracy Sen (95% CI) 90.0 (89.1–90.8) 88.9 (87.0–90.6) 98.1 (97.7–98.5) 94.8 (90.1–98.2) 97.1 (96.1–7.9) Spec (95% CI) 99.5 (99.4–99.5) 98.4 (97.8–98.8) 99.6 (99.2–99.9) 99.5 (99.0–100) 99.9 (99.8–00)

*Sen : sensitivity; Spec : specificity; CI : confidence interval; RDT : rapid diagnostic test; EIA : enzyme immunoassay; + – LR : positive likelihood ratio; LR : negative likelihood ratio; REM : random effects model

When comparing EIAs to newer CMIA (chemiluminescent assays), two standard ELISA/ EIA based assays manufactured in markets with transitioning economies appeared to perform poorly compared to other assays.4 Of note, data exists for one assay (KHB) using different signal cut-off ratio’s (S/CO), with improved sensitivity but worse specificity when using the grey zone (S/CO 0.2–0.99); sensitivity 96.2%, specificity 70.6% compared to sensitivity of 73.8% and specificity of 88.1%.45 Page | 224

Results for EIAs are more specific but less sensitive when used in conjunction with manufacturers’ neutralisation assays. In one study,56 Liason HBsAg (Diasorin) had sensitivity 100% and specificity 70% compared to NAT, with improved overall accuracy using with a neutralisation assay (sensitivity 96%, specificity 100%). This shows the critical need to use the neutralization step to confirm any HBsAg reactivity observed upon initial testing. Diagnostic accuracy of HBsAg assays using NAT reference The specificity of RDTs and EIA was lower when using a NAT reference; this is understandable as the serological assays detect HBsAg whereas NAT detects HBV DNA. The viral kinetics of HBV DNA and HBsAg are not identical. One limitation of HBV DNA as a diagnostic assay would be that those on anti-viral treatment, including antiretroviral therapy for HIV (containing lamivudine or tefonovir) might be non-detectable HBV DNA in the presence of HBsAg. The pooled clinical sensitivity of RDTs was 93.3% (95% CI 91.3–94.9) and significantly higher than EIAs 75.7% (95% CI 72.1–79.1) using a NAT reference. It is important to note that study characteristics varied, with one laboratory based case-control study with pre-selected patients showing particularly good results for RDTs vs NAT and providing 6 data points in this analysis.47 The remaining studies in this sub-analysis were all performed in field settings in resource-limited settings, with poor sensitivity (38–60%). Sub-analyses HIV Our results showed that RDTs may be less sensitive in HIV-positive patients. There was still heterogeneity in terms of results, with one otherwise good quality review finding that Determine was 96 % sensitive (95% CI: 80,100) and 100% specific (95% CI: 99,100) in this cohort.20 The difficulty of accurate diagnosis in HIV patients is possibly explained by an increased incidence of hepatitis B and in particular occult hepatitis B in this cohort. In Sudanese HIV-positive ART naïve patients, 27% had detectable HBV DNA, with occult hepatitis B in 15%.61 Among 495 treatment naïve, HIV-infected adults in Cote-d’Ivoire, 13% were HBsAg positive, 42% isolated anti-HBc positive, and 10% occult hepatitis B only detected by NAT.62 Median HBV DNA level was lower in those with occult HBV compared to those with CHB. Immune pressure has also been hypothesized to contribute, with Geretti et al. noting that discrepant results for RDTs were all mutants in their study. The overlapping surface and polymerase genes in the HBV genome could imply that RT inhibitors (e.g. lamivudine) can lead to the emergence of variants carrying mutations of both the polymerase and surface genes, hence avoiding detection by standard HBsAg assays. Blood Donors Pre-transfusion screening of blood donations is a major public health challenge in resourcelimited settings, where prevalence rates for TTIs (transfusion-transmissible infections) are significant. Screening of individuals with RDTs pre-donation have been adopted in areas with insufficient laboratory capacity. Page | 225

Diagnostic accuracy in our review was similar in blood donors compared to the overall pooled estimates of RDTs, with sensitivity 91.6 % (95% CI: 90.1, 92.9) and specificity 99.5% (95% CI 99.3, 99.7). Results were very heterogeneous, with one study in particular having very low sensitivity (~50–60). Two studies comparing EIAs against EIAs in blood donors had higher accuracy, with pooled sensitivity 99.3 % (95% CI: 96.4, 100) and specificity 90.9 % (95% CI 82.9, 96.0). A recent multinational assessment accuracy of TTI screening in Africa using both RDTs and EIAs on an external quality assessment panel found poor overall sensitivity (75.6%) and specificity (94.5%) for HBsAg detection.63 This was driven by very poor clinical sensitivity (47.4%) of HBsAg RDTs, which was lower than that for HCV (63.7%) and HIV (72.4%) in this population. This can be explained by their lower analytical sensitivity, difficulties in transport and quality assurance, in addition to often studies being performed on smaller scales. In a Nigerian blood donor study, 10% of 113 HBsAg-negative repeat donors using RDTs were found to have quantifiable HBV DNA.57 These patients either had acute infection or occult chronic infection. In a recent systematic review of studies evaluating RDTs for infectious disease blood screening in Africa, there was again significant variability in performance.64 RDTs for HBsAg detection were again identified for suboptimal sensitivities, with questionable suitability, especially in high prevalence regions. High false negatives could be due to operator error, low HBsAg levels, assay degeneration or lot variation. Whole blood For rapid diagnostic tests, accuracy using whole blood (capillary and venous) was marginally superior to serum. The accuracy was comparable to that of EIAs using serum; data from the eight studies (eleven data points) is also less heterogenous (Annex 9.3.2). The significantly lower sensitivity of RDTs using plasma is possibly explained by the nature of the studies; one was in a population of blood donors, while the other was initially designed to assess the accuracy of RDTs in determining HBV and HCV prevalence in a Malawian population with high HIV-co-infection rates.41 The authors of the latter study hypothesised that local operational problems or unexpected technical issues were the reason for poor performance in resourcelimited setting. Others have also since suggested that the high HIV-co-infection rate could have contributed, with suppression of HBV replication using lamivudine containing regiments potentially hindering affecting detection by RDT. Study setting – field Some heterogeneity is explained by location where RDTs are performed. Two different studies from Malawi, in a predominantly HIV-positive cohort41 or entirely HIV-positive cohort19 produced very different results for the same test (Determine HBsAg). Pooled sensitivity 56% (95% CI: 38, 73) and specificity 69% (95% CI: 62, 76) were much lower in the field study compared to the sensitivity 100% (95% CI: 86, 100) and specificity 100% (95% CI: 93, 100) in the study where samples were returned to the UK.

Page | 226

Analytical sensitivity of different assays as a source of heterogeneity We were unable to explain heterogeneity of results using different assays. Very few rapid tests meet required analytical LOD (0.130 IU/mL) required by regulatory authorities, but because of insufficient data in studies we were unable to stratify using LOD as a source of heterogeneity. This is important as it has been suggested that false-negative HBsAg RDTs are associated with lower HBsAg levels, low viral load, HBsAg mutants, or specific genotypes, in addition to ART exposure where lamivudine and tenofovir are used. (16, 25, 36, 65) In a recent fields studies in the Gambia,40 the range of serum HBsAg levels quantified by CMIA that showed reactivity with RDTs in the field was 26.5–27, 320 IU/mL, with a statistically significant (P = 0.0002) difference in median HBsAg level (875 IU/mL) compared to false negatives using RDT’s (median 1.2; range 0.8–25.5 IU/mL). Interestingly significantly more false-negatives were female (P = 0.05) with lower median ALT levels (P = 0.01). The laboratory-based study from the same publication in a chronic hepatitis B cohort found a higher range of HBsAg levels (2.8–124,925 IU/mL) in those testing positive with RDTs, with a significant difference in median HBsAg levels (7, 482 vs 0.40 IU/mL; P <0.0001) and median ALT (P = 0.01) between true-positives and false negatives. The lower limit of detection may be explained by differences in methodology, given the setting (laboratory vs field), reference test (CMIA vs ELISA) and sample type (dried blood spots vs serum). This suggests that subjects with false negatives have lower HBsAg levels and inactive disease compared to true-positives, minimising the impact of reduced sensitivity. Unfortunately, in the single study identified also assessing LSM, 17% (4/23) subjects with false negative results had evidence of fibrosis and would require antiviral therapy. Another study (Bottero et al)16 also found significantly lower median HBsAg in false negatives vs true positives. [19.5 vs 2351 IU/mL; p=0.0001], with only 4 false negative having HBsAg >10 IU/mL. HBV DNA was usually below 200 IU/mL. False positives occurred in vaccinate patients (n=7), and one patient with resolved infection and anti-HBs titre. Interestingly ALL false negatives were HBcAb positive. Data exists from large studies of analytical sensitivity using reference, seroconversion, mutant panels.53, 65, 66 These include specimens from individuals with low antigenaemia, such as early infection. Unfortunately, studies of analytical sensitivity of EIAs are conflicting. One study suggested that 9 out of 10 EIAs were able to detect HBsAg levels as low as 0.2 IU/mL irrespective of genotype.67 Another comparing newer EIAs (Advia Centaur; Monolisa Ultra; Liasion; Vidas Ultra) using reference and mutant panels found a lower limit of detection <350IU/mL, but with varying sensitivity for mutant detection (37.1%–91.4%).68 The authors hypothesised that the lack of detection was due to epitope recognition of the anti-HBs assay reagents in the capture phase and in conjugates. Another study assessing 13 different assays with mutant panels found a range of LOD (0.011-0.096 IU/mL) and sensitivity (63%–98%) in mutants. Another study found comparable analytical sensitivity between four EIAs but significant differences in detection of mutants between assays.69 One blood donor study in China found a significant difference in sensitivities and mutant detection capabilities amongst assays used by blood banks, with the urgent recommendation of a list of high sensitivity assays for blood bank screening.70 Page | 227

In blood donors, studies have found some correlation between HBsAg levels (IU/mL) and NAT (copies/mL).71 The obvious benefit of more sensitive assays with lower limits of detection would be improved detection of those with occult hepatitis B or in the early window period of sero-conversion. It has been suggested that utilizing “grey zones” in EIAs could improve sensitivity and allow combination of tests to develop of economic testing strategies.45 Sensitivity improved from 76–88% to 96–97%, with a further increase to 99% when combining the use of two EIAs. Studies looking at RDTs using clinical panels have found sometimes conflicting results. One study found equivalent specificities but significant differences in assay sensitivity between Uni-Gold™ HBsAg and Determine HBsAg.72 Interestingly, studies in Cambodia and Viet Nam by the same group produced different results for sensitivity, suggesting uncontrolled variables, such as prozone effect and genotype variations. The prozone effect may explain why some true positives turned out negative with rapid tests. Given that specificity is excellent but sensitivity is low suggests that this is genuine poor performance. Study strengths and limitations Significant strengths of this meta-analysis include the global evidence base, rigorous prespecified protocol incorporating numerous major scientific databases, in addition to review of the related literature, notably occult hepatitis B and the impact of NAT. We included studies performed in a range of settings, with a diverse population. We only included studies with bivariate data, to minimise bias, measuring clinical sensitivity which are more applicable. We also included evaluations of both RDTs and EIAs, and as such are able to provide a more comprehensive meta-analysis. Comparing RDTs with EIAs, we were able to identify an additional 11 studies not found in previous reviews. Incorporation bias was unlikely as all participants received both index and confirmatory tests independently. As all studies administered the same reference standard to all patients, which reduces risk of verification bias. Our study also excluded articles deemed to be high risk of bias or less applicable, such as conference abstracts or reference panels; reference panels, included in previous reviews, have higher accuracy than that of tests used in the field but are not as useful in guiding policy. Accuracy on seroconversion panels do not necessarily reflect the antibody or antigen spectrum in the populations studied. Our study, did, however have a number of limitations. First, we only included studies in English, which potentially introduces publication (language) bias. We will identify relevant studies in non-English languages from reference lists to address whether this contributed to a substantial bias or not. Second, a significant proportion of studies were case–control in design or used preselected cohorts, which would bias results. For example, in well-conducted studies from the Gambia,40 those in the community setting had a smaller range of HBsAg levels (26.5 – 27, 320 IU/mL) than the study conducted in chronic hepatitis B patients (2.8–124,925 IU/mL). This is one reason for the reduced sensitivity of the same test (Determine HBsAg) in the screening cohort (88%; 95% CI: 81, 94) compared to the chronic hepatitis cohort (95%; 95% CI:

Page | 228

90, 98) from the same community. Third, some analyses were based on a small number of studies, which included few positive samples. There are number of technical and patient factors that could impact accuracy, which cannot be addressed based on the currently available literature. Specific to hepatitis B diagnosis, we were unable to correlate the heterogeneity of sensitivity with different stages, severity and genotypes infection. This was due to insufficient information in studies, principally additional serology such as HBeAg, anti-HBc IgM, anti-HBc total and anti-HBs antibodies. Genetic information has long been suspected to impact on diagnostic accuracy,67‒ 70, 73, 74 although a recent study of analytical performance found no difference in the detection of mutants.70, 73‒75 It should be noted that this study was conducted by authors with significant conflicts of interest. Mutants themselves are also rapidly evolving, such that the prevalence and type of specific mutants cannot be determined based on historical data, making studies difficult to organize. Finally, occult hepatitis should also be considered. The addition of NAT would be useful to stratify patients’ results, but the lack of sufficient of quality studies (Annex 9.2) is testament to the challenges in conducting advanced laboratory based studies in areas of high disease prevalence. Reference standards are also imprecise, resulting in overdiagnosis of clinically insignificant disease, and underestimation of diagnostic accuracy of clinically relevant disease. The natural history of hepatitis B, notably progression and infectiousness, is being investigated for correlations of quantitative HBsAg, HbeAg and HBV DNA levels. Low levels of either antigen may not be significant clinically. Short-term spontaneous fluctuations in DNA and HBsAg are recognized in those with chronic hepatitis B and add extra challenges to accurate diagnosis with a “gold standard”.76 Statistical heterogeneity was an obvious issue as is often observed in diagnostic accuracy reviews. Although we performed stratified analyses to identify potential sources, none fully explained the heterogeneity observed. Firstly, studies evaluated different RDTs, with rapid changes in technology for both EIAs and RDTs meaning that analytical sensitivity is variable among the assays evaluated. Although we pooled based on some RDT brands (Determine HBsAg, BinaxNOW) there were insufficient studies and this pooling did not entirely account for heterogeneity. Another potential confounder is changes in manufacturing processes, including components used to manufacture assays Determine, as one example, has been produced by Abbott, Inverness Medical and Alere Medical Co. Ltd; as the test has been commercially available for over 10 years, there will undoubtedly be minor product changes.

7. Conclusion WHO has emphasized the importance of timely global testing, prevention and treatment of hepatitis B, with predictions of an increasing prominence as a cause of death globally in years to come. Although RDTs have limitations, many of which can be addressed through improved training and quality assurance systems, they are frequently the only viable option for infectious screening in resource-limited settings. Therefore, additional studies and specific guidelines regarding the use of RDTs in the context of blood safety and patient screening are needed. In terms of global uptake, lower costs of these assays and ease of use across a variety of endemic settings is crucial to achieving goals for control of hepatitis. Worldwide, a Page | 229

significant proportion of countries are unable to afford quality-assured laboratory-based testing with enzyme immunoassays; the use of NAT to further reduce the window period of infection and detect occult hepatitis is beyond reach in many settings at present. This meta-analysis, along with others, suggests that assays for detection of HBsAg including RDTs and enzyme immunoassays have the potential to contribute significantly to the control of hepatitis B globally in endemic areas, which are often include low resource remote regions. Other benefits of RDTS include easy storage, small sample volumes required with minimal staff training or additional equipment. Unfortunately, with current issues with poor clinical and analytical sensitivity and potential difficulties in detection of occult hepatitis B and mutant variants, a number of cases would be missed. There is also concern that sensitivity is significantly reduced in HIV-positive patients. There are numerous difficulties in conducting systematic reviews of the performance of in vitro diagnostics, particularly in resource-limited settings. There is a significant variation in terms of quality of studies, most with key parameters missing. Further promotion of current accepted standards to performing and reporting studies of diagnostic accuracy globally can help improve the evidence base currently available. Further high quality studies are desperately needed to assess the accuracy in a variety of settings and support the growing evidence base for RDTs. Specifically, further studies looking at the impact of different geographic locations and mutant phenotypes would be invaluable. From included studies, excellent robust specificity of all assays is reassuring in terms of ensuring cost–effective initiation of algorithms for further investigation and treatment. Significant heterogeneity and suboptimal sensitivity of RDTs has to be taken into consideration as country control programmes consider the trade-off between affordability, accuracy and accessibility (i.e. ease of use in all levels of the health-care system). The weighting of these three factors are country specific and could be modelled.

8. References 1. Ott JJ, Stevens GA, Groeger J, Wiersma ST. Global epidemiology of hepatitis B virus infection: new estimates of age-specific HBsAg seroprevalence and endemicity. Vaccine. 2012;30(12):2212–9. Ganem D, Prince AM. Hepatitis B virus infection – natural history and clinical consequences. N Engl J Med. 2004;350(11):1118–29. Fattovich G, Stroffolini T, Zagni I, Donato F. Hepatocellular carcinoma in cirrhosis: incidence and risk factors. Gastroenterology. 2004;127(5 Suppl 1):S35–50. Lozano R, Naghavi M, Foreman K, Lim S, Shibuya K, Aboyans V, et al. Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380(9859):2095–128. Guidelines for the prevention, care, and treatment of persons with chronic hepatitis B infection. Geneva: WHO; 2015.

2. 3. 4.

Page | 230

Fabrizi F, Martin P, Dixit V, Bunnapradist S, Dulai G. Meta-analysis: the effect of age on immunological response to hepatitis B vaccine in end-stage renal disease. Aliment Pharmacol Ther. 2004;20(10):1053–62. Thompson Coon J, Rogers G, Hewson P, Wright D, Anderson R, Cramp M, et al. Surveillance of cirrhosis for hepatocellular carcinoma: systematic review and economic analysis. Health Technol Assess. 2007;11(34):1–206. Qin XK, Li P, Han M, Liu JP. Xiaochaihu Tang for treatment of chronic hepatitis B: a systematic review of randomized trials. [Article in Chinese]. Zhong Xi Yi Jie He Xue Bao. 2010;8(4):312–20. Hwang SH, Oh HB, Choi SE, Kim HH, Chang CL, Lee EY, et al. [Meta-analysis for the pooled sensitivity and specificity of hepatitis B surface antigen rapid tests.] [Article in Korean]. Korean J Lab Med. 2008;28(2):160–8.

10. Shivkumar S, Peeling R, Jafari Y, Joseph L, Pai NP. Rapid point-of-care first-line screening tests for hepatitis B infection: a meta-analysis of diagnostic accuracy (1980–2010). Am J Gastroenterol. 2012;107(9):1306–13. 11. Khuroo NS, Khuroo MS. Accuracy of rapid point-of-care diagnostic tests for hepatitis B surface antigen – a systematic review and meta-analysis. J Clin Exp Hepatol. 2014;4(3):226–40. 12. Whiting PF, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155(8):529–36. 13. Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig LM, et al. The STARD statement for reporting studies of diagnostic accuracy: explanation and elaboration. Ann Intern Med. 2003;138(1):W1–W12. 14. Abraham P, Sujatha R, Raghuraman S, Subramaniam T, Sridharan G. Evaluation of two immunochromatographic assays in relation to 'RAPID' Screening of HBsAg. Indian J Med Microbiol. 1998;16:23–5. 15. Akanmu AS, Esan OA, Adewuyi JO, Davies AO, Okany CC, Olatunji RO, et al. Evaluation of a rapid test kit for detection of HBsAg/eAg in whole blood: a possible method for predonation testing. Afr J Med Med Sci. 2006;35(1):5–8. 16. Bottero J, Boyd A, Gozlan J, Lemoine M, Carrat F, Collignon A, et al. Performance of rapid tests for detection of HBsAg and anti-HBsAb in a large cohort, France. . J Hepatol. 2013;58(3):473–8. 17. Chameera EWS, Noordeen F, Pandithasundara H, Abeykoon AMSB. Diagnostic efficacy of rapid assays used for the detection of hepatitis B virus surface antigen. Sri Lankan J Infect Dis. 2013;3(2):21–7. 18. Clement F, Dewint P, Leroux-Roels G. Evaluation of a new rapid test for the combined detection of hepatitis B virus surface antigen and hepatitis B virus e antigen. J Clin Microbiol. 2002;40(12):4603–6.

Page | 231

19. Davies J, van Oosterhout JJ, Nyirenda M, Bowden J, Moore E, Hart IJ, et al. Reliability of rapid testing for hepatitis B in a region of high HIV endemicity. Trans R Soc Trop Med Hyg. 2010;104(2):162–4. 20. Franzeck FC, Ngwale R, Msongole B, Hamisi M, Abdul O, Henning L, et al. Viral hepatitis and rapid diagnostic test based screening for HBsAg in HIV-infected patients in rural Tanzania. PLoS One. 2013;8(3):e58468. 21. Kaur H, Dhanao J, Oberoi A. Evaluation of rapid kits for detection of HIV, HBsAg and HCV infections. Indian J Med Sci. 2000;54(10):432–4. 22. Khan J, Lone D, Hameed A. Evaluation of the performance of two rapid immunochromatographic tests for detection of Hepatitis B surface antigen and anti HCV antibodies using ELISA tested samples. . AKEMU. 2010;16(1 S1):84–7. 23. Lau DT, Ma H, Lemon SM, Doo E, Ghany MG, Miskovsky E, et al. A rapid immunochromatographic assay for hepatitis B virus screening. J Viral Hepat. 2003;10(4):331–4. 24. Lien TX, Tien NTK, Chanpong GF, Cuc CT, Yen VT, Soderquist R, et al. Evaluation of rapid diagnostic tests for the detection of human immunodeficiency virus types 1 and 2, hepatitis B surface antigen, and syphilis in Ho Chi Minh City, Vietnam. Am J Trop Med Hyg. 2000;62(2):301–9. 25. Lin YH, Wang Y, Loua A, Day GJ, Qiu Y, Nadala EC, Jr., et al. Evaluation of a new hepatitis B virus surface antigen rapid test with improved sensitivity. J Clin Microbiol. 2008;46(10):3319–24. 26. Mvere D, Constantine NT, Katsawde E, Tobaiwa O, Dambire S, Corcoran P. Rapid and simple hepatitis assays: encouraging results from a blood donor population in Zimbabwe. Bull World Health Organ. 1996;74(1):19–24. 27. Oh J, Kim TY, Yoon HJ, Min HS, Lee HR, Choi TY. Evaluation of Genedia® HBsAg Rapid and Genedia® anti-HBs Rapid for the screening of HBsAg and Anti-HBs. Korean J Clin Pathol. 1999;19:114–7. 28. Ola SO, Otegbayo JA, Yakubu A, Aje AO, Odaibo GN, Shokunbi W. Pitfalls in diagnosis of Hepatitis B virus infection among adults nigerians. Niger J Clin Pract. 2009;12(4):350–4. 29. Raj AA, Subramaniam T, Raghuraman S, Abraham P. Evaluation of an indigenously manufactured rapid immunochromatographic test for detection of HBsAg. Indian J Pathol Microbiol. 2001;44(4):413–4. 30. Randrianirina F, Carod JF, Ratsima E, Chretien JB, Richard V, Talarmin A. Evaluation of the performance of four rapid tests for detection of hepatitis B surface antigen in Antananarivo, Madagascar. J Virol Methods. 2008;151(2):294–7. 31. Sato K, Ichiyama S, Iinuma Y, Nada T, Shimokata K, Nakashima N. Evaluation of immunochromatographic assay systems for rapid detection of hepatitis B surface antigen and antibody, Dainascreen HBsAg and Dainascreen Ausab. J Clin Microbiol. 1996;34(6):1420–2. 32. Bjoerkvoll B, Viet L, Ol HS, Lan NT, Sothy S, Hoel H, et al. Screening test accuracy among potential blood donors of HBsAg, anti-HBc and anti-HCV to detect hepatitis B and C virus Page | 232

infection in rural Cambodia and Vietnam. Southeast Asian J Trop Med Public Health. 2010;41(5):1127–35. 33. Chevaliez S, Challine D, Naija H, Luu TC, Laperche S, Nadala L, et al. Performance of a new rapid test for the detection of hepatitis B surface antigen in various patient populations. J Clin Virol. 2014;59(2):89–93. 34. Erhabor O, Kwaifa I, Bayawa A, Isaac Z, Dorcas IaS I. Comparison of ELISA and rapid screening techniques for the detection of HBsAg among blood donors in Usmanu Danfodiyo University Teaching Hospital Sokoto, North Western Nigeria. J Blood Lymph. 2013;4(2). 35. Gish RG, Gutierrez JA, Navarro-Cazarez N, Giang K, Adler D, Tran B, et al. A simple and inexpensive point-of-care test for hepatitis B surface antigen detection: serological and molecular evaluation. . J Viral Hepat. 2014;21(12):905–8. 36. Geretti AM, Patel M, Sarfo FS, Chadwick D, Verheyen J, Fraune M, et al. Detection of highly prevalent hepatitis B virus coinfection among HIV-seropositive persons in Ghana. J Clin Microbiol. 2010;48(9):3223–30. 37. Hoffmann CJ, Dayal D, Cheyip M, McIntyre JA, Gray GE, Conway S, et al. Prevalence and associations with hepatitis B and hepatitis C infection among HIV-infected adults in South Africa. Int J STD AIDS. 2012;23(10):e10–3. 38. Honge BL, Jespersen S, Te DS, da Silva ZJ, Laursen AL, Krarup H, et al. Hepatitis B virus surface antigen and anti-hepatitis C virus rapid tests underestimate hepatitis prevalence among HIV-infected patients. HIV Med. 2014;15(9):571–6. 39. Mutocheluh M, Owusu M, Kwofie TB, Akadigo T, Appau E, Narkwa PW. Risk factors associated with hepatitis B exposure and the reliability of five rapid kits commonly used for screening blood donors in Ghana. BMC Res Notes. 2014;7:873. 40. Njai HF, Shimakawa Y, Sanneh B, Ferguson L, Ndow G, Mendy M, et al. Validation of rapid point-of-care (POC) tests for detection of hepatitis B surface antigen in field and laboratory settings in the Gambia, Western Africa. J Clin Microbiol. 2015;53(4):1156–63. 41. Nyirenda M, Beadsworth MB, Stephany P, Hart CA, Hart IJ, Munthali C, et al. Prevalence of infection with hepatitis B and C virus and coinfection with HIV in medical inpatients in Malawi. J Infect. 2008;57(1):72–7. 42. Upreti SR, Gurung S, Patel M, Dixit SM, Krause LK, Shakya G, et al. Prevalence of chronic hepatitis B virus infection before and after implementation of a hepatitis B vaccination program among children in Nepal. Vaccine. 2014;32(34):4304–9. 43. Liu C, Chen T, Lin J, Chen H, Chen J, Lin S, et al. Evaluation of the performance of four methods for detection of hepatitis B surface antigen and their application for testing 116,455 specimens. J Virol Methods. 2014;196:174–8. 44. Ol HS, Bjoerkvoll B, Sothy S, Van Heng Y, Hoel H, Husebekk A, et al. Prevalence of hepatitis B and hepatitis C virus infections in potential blood donors in rural Cambodia. Southeast Asian J Trop Med Public Health. 2009;40(5):963–71.

Page | 233

45. Peng J, Cheng L, Yin B, Guan Q, Liu Y, Wu S, et al. Development of an economic and efficient strategy to detect HBsAg: application of "gray-zones" in ELISA and combined use of several detection assays. Clin Chim Acta. 2011;412(23–24):2046–51. 46. Viet L, Lan NT, Ty PX, Bjorkvoll B, Hoel H, Gutteberg T, et al. Prevalence of hepatitis B & hepatitis C virus infections in potential blood donors in rural Vietnam. Indian J Med Res. 2012;136(1):74–81. 47. Ansari MHK, Omrani MD, Movahedi V. Comparative evaluation of immunochromatographic rapid diagnostic tests (strip and device) and PCR methods for detection of human hepatitis B surface antigens. Hepat Mon. 2007;7(2):87–91. 48. Nakata JM, Johnson JM. Evaluation of rapid immunoconcentration assay for HBsAg in a prison inmate population. Annual Meeting of the American Society for Microbiology, Anaheim. 1990. 49. Palmer C, Cuadrado R, Koenig E. Multicenter evaluation of the determine[TM] Rapid tests for the diagnosis of HIV, Hepatitis B Surface Antigen, and Syphilis. Interscience Conference on Antimicrobial Agents and Chemotherapy, Ft Lauderdale. 1999. 50. Torane V, Shastri J. Comparison of ELISA and rapid screening tests for the diagnosis of HIV, hepatitis B and hepatitis C among healthy blood donors in a tertiary care hospital in Mumbai. Indian J Med Microbiol. 2008;26(3):284–5. 51. Cha YJ, Yang JS, Chae SL. [Evaluation of indigenously manufactured immunochromatographic assay systems for rapid detection of hepatitis B surface antigen and antibody.] [Article in Korean]. Korean J Lab Med. 2006;26:52–7. 52. Whang DH, Um TH. [Comparison of immunochromatography assays and quantitative immunoassays for detecting HBsAg and anti-HBs.] [Article in Korean]. Korean J Lab Med. 2005;25:186–91. 53. Maity S, Nandi S, Biswas S, Sadhukhan SK, Saha MK. Performance and diagnostic usefulness of commercially available enzyme linked immunosorbent assay and rapid kits for detection of HIV, HBV and HCV in India. Virol J. 2012;9:290. 54. Lukhwareni A, Burnett RJ, Selabe SG, Mzileni MO, Mphahlele MJ. Increased detection of HBV DNA in HBsAg-positive and HBsAg-negative South African HIV/AIDS patients enrolling for highly active antiretroviral therapy at a Tertiary Hospital. J Med Virol. 2009;81(3):406–12. 55. Mphahlele MJ, Lukhwareni A, Burnett RJ, Moropeng LM, Ngobeni JM. High risk of occult hepatitis B virus infection in HIV-positive patients from South Africa. J Clin Virol. 2006;35(1):14–20. 56. Khadem-Ansari MH, Omrani MD, Rasmi Y, Ghavam A. Diagnostic validity of the chemiluminescent method compared to polymerase chain reaction for hepatitis B virus detection in the routine clinical diagnostic laboratory. Adv Biomed Res. 2014;3:116. 57. Nna E, Mbamalu C, Ekejindu I. Occult hepatitis B viral infection among blood donors in South-Eastern Nigeria. Pathog Glob Health. 2014;108(5):223–8. 58. Olinger CM, Weber B, Otegbayo JA, Ammerlaan W, van der Taelem-Brule N, Muller CP. Hepatitis B virus genotype E surface antigen detection with different immunoassays and Page | 234

diagnostic impact of mutations in the preS/S gene. Med Microbiol Immunol. 2007;196(4):247–52. 59. Seremba E, Ocama P, Opio CK, Kagimu M, Yuan HJ, Attar N, et al. Validity of the rapid strip assay test for detecting HBsAg in patients admitted to hospital in Uganda. J Med Virol. 2010;82(8):1334–40. 60. Shivkumar S, Peeling R, Jafari Y, Joseph L, Pai NP. Rapid point-of-care first-line screening tests for hepatitis B infection: a meta-analysis of diagnostic accuracy (1980–2010). Am J Gastroenterol. 2012;107(9):1306–13. 61. Mudawi H, Hussein W, Mukhtar M, Yousif M, Nemeri O, Glebe D, et al. Overt and occult hepatitis B virus infection in adult Sudanese HIV patients. Int J Infect Dis. 2014;29:65–70. 62. N'Dri-Yoman T, Anglaret X, Messou E, Attia A, Polneau S, Toni T, et al. Occult HBV infection in untreated HIV-infected adults in Cote d'Ivoire. Antivir Ther. 2010;15(7):1029– 34. 63. Laperche S, Francophone African Group for Research in Blood T. Multinational assessment of blood-borne virus testing and transfusion safety on the African continent. Transfusion. 2013;53(4):816–26. 64. Pruett CR, Vermeulen M, Zacharias P, Ingram C, Tayou Tagny C, Bloch EM. The Use of rapid diagnostic tests for transfusion infectious screening in Africa: a literature review. Transfus Med Rev. 2015;29(1):35–44. 65. Scheiblauer H, El-Nageh M, Diaz S, Nick S, Zeichhardt H, Grunert HP, et al. Performance evaluation of 70 hepatitis B virus (HBV) surface antigen (HBsAg) assays from around the world by a geographically diverse panel with an array of HBV genotypes and HBsAg subtypes. [Erratum appears in Vox Sang. 2010 May;98(4):581]. Vox Sang. 2010;98(3 Pt 2):403–14. 66. Scheiblauer H, Soboll H, Nick S. Evaluation of 17 CE-marked HBsAg assays with respect to clinical sensitivity, analytical sensitivity, and hepatitis B virus mutant detection. J Med Virol. 2006;78 Suppl 1:S66–70. 67. Mizuochi T, Okada Y, Umemori K, Mizusawa S, Yamaguchi K. Evaluation of 10 commercial diagnostic kits for in vitro expressed hepatitis B virus (HBV) surface antigens encoded by HBV of genotypes A to H. J Virol Methods. 2006;136(1–2):254–6. 68. Ly TD, Servant-Delmas A, Bagot S, Gonzalo S, Ferey MP, Ebel A, et al. Sensitivities of four new commercial hepatitis B virus surface antigen (HBsAg) assays in detection of HBsAg mutant forms. J Clin Microbiol. 2006;44(7):2321–6. 69. Moerman B, Moons V, Sommer H, Schmitt Y, Stetter M. Evaluation of sensitivity for wild type and mutant forms of hepatitis B surface antigen by four commercial HBsAg assays. Clin Lab. 2004;50(3–4):159–62. 70. Zhang R, Wang L, Li J. Hepatitis B virus transfusion risk in China: proficiency testing for the detection of hepatitis B surface antigen. Transfus Med. 2010;20(5):322–8. 71. Martin LA, Stramer SL, Kuhns MC, Schlauder GG. Correlation of improved hepatitis B surface antigen detection limits with hepatitis B virus DNA nucleic acid test yield in blood donations. Transfusion. 2012;52(10):2201–8. Page | 235

72. Austin PM. Comparative sensitivities and specificities of two rapid HBsAg detection methods and their relationship to a third generation commercial enzyme immunoassay. NZ J Med Lab Science. 2005;59(2):38–40. 73. Kuhns MC, McNamara AL, Holzmayer V, Lou SC, Busch MP. Frequency of diagnostically significant hepatitis B surface antigen mutants. J Med Virol. 2007;79(Suppl. 1):S42–S6. 74. Servant-Delmas A, Mercier-Darty M, Ly TD, Wind F, Alloui C, Sureau C, et al. Variable capacity of 13 hepatitis B virus surface antigen assays for the detection of HBsAg mutants in blood samples. J Clin Virol. 2012;53(4):338–45. 75. Hirzel C, Pfister S, Gorgievski-Hrisoho M, Wandeler G, Zuercher S. Performance of HBsAg point-of-care tests for detection of diagnostic escape-variants in clinical samples. J Clin Virol. 2015;69:33–5. 76. Maylin S, Sire JM, Mbaye PS, Simon F, Sarr A, Evra ML, et al. Short-term spontaneous fluctuations of HBV DNA levels in a Senegalese population with chronic hepatitis B. BMC Infect Dis. 2015;15(1):154.

Page | 236

9. Annexes 9.2. 9.1 Search strategy Ovid Medline search strategy Searched on 20 April 2015 from 1946 – April week 2 2015. 1 2 3 4 5 6 7 8 9 10 11 12 13 Hepatitis, Viral, Human/ (10382) Hepatitis Viruses/ (1363) Hepatitis Antibodies/ (5082) exp Hepadnaviridae Infections/ (47484) Hepatitis B Antibodies/ (8638) Hepatitis B virus/ (20604) Hepadnaviridae/ (192) Hepatitis B Surface Antigens/ (17007) (heptatitis-b or hep-b or (hepatitis adj5 b) or (hep adj5 b) or hbv).ti,ab. (64488) hbsag.ti,ab. (15146) or/1-10 [HEPATITIS B] (87943) exp Reagent Kits, Diagnostic/ (17747) ((rapid or point of care or near patient or poc or poct or bedside) adj5 (test or tests or testing or detect* or diagnos* or screen* or kit or kits or assay* or device*)).ti,ab. (63080) (radt or radts or rdt or rdts).ti,ab. (909) rapid test*.ti,ab. (3400) exp Enzyme-Linked Immunosorbent Assay/ (127391) Immunoassay/ (23237) Immunoenzyme Techniques/ (64864) (enzyme-linked immunosorbent assay or ELISA).ti,ab. (139374) (enzyme adj2 (immunoassay* or immuno-assay* or immunosorbent)).ti,ab. (83849) ((antigen* or antibod*) adj3 detect*).ti,ab. (59427) or/12-21 [RAPID DIAGNOSTIC TESTS] (394724) exp "Sensitivity and Specificity"/ (435087) (diagnos* accura* or sensitiv* or specific* or valid*).ti,ab. (3016884) roc curve.ti,ab. (10226) positive predictive value.ti,ab. (25496) negative predictive value.ti,ab. (20415) or/23-27 [DIAGNOSTIC ACCURACY] (3235789) 11 and 22 and 28 (3103) Humans/ (13846846) Animals/ (5442465) 30 and 31 (1513142) 31 not 32 [ALL ANIMAL STUDIES WHICH DO NOT INCLUDE COMPARISON WITH HUMANS] (3929323) 29 not 33 (2856) Page | 237

14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34

limit 34 to english language (2345)

Ovid Embase search strategy Searched on 20 April 2015 from 1947 – 2015 April 17. 1 2 3 4 5 6 7 8 9 10 hepatitis virus/ (4410) hepatitis antibody/ (2216) exp hepadnaviridae/ (42214) hepatitis B surface antigen/ (27312) (heptatitis-b or hep-b or (hepatitis adj5 b) or (hep adj5 b) or hbv).ti,ab. (94627) hbsag.ti,ab. (22290) or/1-6 [HEPATITIS B] (111321) exp diagnostic kit/ (13384) "point of care testing"/ (5530) ((rapid or point of care or near patient or poc or poct or bedside) adj5 (test or tests or testing or detect* or diagnos* or screen* or kit or kits or assay* or device*)).ti,ab. (88003) (radt or radts or rdt or rdts).ti,ab. (1652) rapid test*.ti,ab. (5049) enzyme linked immunosorbent assay/ (229634) immunoassay/ (48491) enzyme immunoassay/ (36845) enzyme linked immunospot assay/ (6789) enzyme multiplied immunoassay technique/ (768) (enzyme-linked immunosorbent assay or ELISA).ti,ab. (204052) (enzyme adj2 (immunoassay* or immuno-assay* or immunosorbent)).ti,ab. (98704) antigen detection/ (18155) antibody detection/ (34389) ((antigen* or antibod*) adj3 detect*).ti,ab. (76053) or/8-22 [RAPID DIAGNOSTIC TESTS] (525203) "sensitivity and specificity"/ (221828) diagnostic accuracy/ (189329) (diagnos* accura* or sensitiv* or specific* or valid*).ti,ab. (4113104) roc curve.ti,ab. (20232) positive predictive value.ti,ab. (37541) negative predictive value.ti,ab. (31612) or/24-29 [DIAGNOSTIC ACCURACY] (4280338) 7 and 23 and 30 (4018) human/ (15785497) animal/ (1646303) 32 and 33 (404532) 33 not 34 [ALL ANIMAL STUDIES WHICH DO NOT INCLUDE COMPARISON WITH HUMANS] (1241771) Page | 238

11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

36 37

31 not 35 (3963) limit 36 to english language (3344)

Page | 239

Web of Science Search was conducted on the Science Citation Index Expanded (1970–20 April 2015) and the Conference Proceedings Citation Index-Science (1990–20 April 2015). 1 2 3 4 TOPIC: ("hepatitis-b" OR "hep-b" OR (hepatitis near/5 b) OR (hep near/5 b) OR hbv) (79,505) TOPIC: (hbsag) (12,160) #2 OR #1 (81,526) TOPIC: ((rapid near/5 test) or (rapid near/5 tests) or (rapid near/5 testing) or (rapid near/5 detect*) or (rapid near/5 diagnos*) or (rapid near/5 screen*) or (rapid near/5 kit) or (rapid near/5 kits) or (rapid near/5 assay*) or (rapid near/5 device*)) (77,863) TOPIC: (("point of care" near/5 test) or ("point of care" near/5 tests) or ("point of care" near/5 testing) or ("point of care" near/5 detect*) or ("point of care" near/5 diagnos*) or ("point of care" near/5 screen*) or ("point of care" near/5 kit) or ("point of care" near/5 kits) or ("point of care" near/5 assay*) or ("point of care" near/5 device*)) (5,974) TOPIC: (("near patient" near/5 test) or ("near patient" near/5 tests) or ("near patient" near/5 testing) or ("near patient" near/5 detect*) or ("near patient" near/5 diagnos*) or ("near patient" near/5 screen*) or ("near patient" near/5 kit) or ("near patient" near/5 kits) or ("near patient" near/5 assay*) or ("near patient" near/5 device*)) (423) TOPIC: ((poc near/5 test) or (poc near/5 tests) or (poc near/5 testing) or (poc near/5 detect*) or (poc near/5 diagnos*) or (poc near/5 screen*) or (poc near/5 kit) or (poc near/5 kits) or (poc near/5 assay*) or (poc near/5 device*)) (866) TOPIC: ((poct near/5 test) or (poct near/5 tests) or (poct near/5 testing) or (poct near/5 detect*) or (poct near/5 diagnos*) or (poct near/5 screen*) or (poct near/5 kit) or (poct near/5 kits) or (poct near/5 assay*) or (poct near/5 device*)) (522) TOPIC: ((bedside near/5 test) or (bedside near/5 tests) or (bedside near/5 testing) or (bedside near/5 detect*) or (bedside near/5 diagnos*) or (bedside near/5 screen*) or (bedside near/5 kit) or (bedside near/5 kits) or (bedside near/5 assay*) or (bedside near/5 device*)) (2,705) TOPIC: (radt or radts or rdt or rdts) (1,406) TOPIC: ("rapid test*") (3,783) TOPIC: ("enzyme-linked immunosorbent assay" or ELISA) (141,435) TOPIC: ((enzyme near/2 immunoassay*) or (enzyme near/2 immuno-assay*) or (enzyme near/2 immunosorbent)) (85,660) TOPIC: ((antigen* near/3 detect*) or (antibod* near/3 detect*)) (56,976) #14 OR #13 OR #12 OR #11 OR #10 OR #9 OR #8 OR #7 OR #6 OR #5 OR #4 (286,936) TOPIC: ("diagnos* accura*" or sensitiv* or specific* or valid*) (4,557,124) TOPIC: ("roc curve") (12,767) TOPIC: ("positive predictive value") (23,706) TOPIC: ("negative predictive value") (18,947) #19 OR #18 OR #17 OR #16 (4,566,667) #20 AND #15 AND #3 (1,789) #20 AND #15 AND #3 Refined by: LANGUAGES: ( ENGLISH ) (1,720)

10 11 12 13 14 15 16 17 18 19 20 21 22

Page | 240

Scopus Search was conducted on 20 April 2015. TITLE-ABS-KEY (("heptatitis-b" OR "hep-b" OR (hepatitis W/5 b) OR (hep W/5 b) OR hbv OR hbsag) AND (((rapid OR "point of care" OR "near patient" OR poc OR poct OR bedside) W/5 (tests OR test OR testing OR detect* OR diagnos* OR screen* OR kit OR kits OR assay* OR device*)) OR radt OR radts OR rdt OR rdts OR "rapid test*" OR "enzyme-linked immunosorbent assay" OR elisa OR (enzyme W/2 (immunoassay* OR immuno-assay* OR immunosorbent)) OR ((antibod* OR anigen*) W/3 detect*)) AND ("diagnos* accura*" OR sensitiv* OR specific* OR valid* OR "roc curve" OR "positive predictive value" OR "negative predictive value")) AND (LIMIT-TO (LANGUAGE, "English")) (3,605)

Cochrane Central Register of Controlled Trials, Wiley The search was run on 20 April 2015. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 MeSH descriptor: [Hepatitis, Viral, Human] this term only MeSH descriptor: [Hepatitis Viruses] this term only MeSH descriptor: [Hepatitis Antibodies] this term only MeSH descriptor: [Hepadnaviridae Infections] explode all trees MeSH descriptor: [Hepatitis B Antibodies] this term only MeSH descriptor: [Hepatitis B virus] this term only MeSH descriptor: [Hepadnaviridae] this term only MeSH descriptor: [Hepatitis B Surface Antigens] explode all trees "hepatitis-b":ti,ab,kw (Word variations have been searched) "hep-b":ti,ab,kw (Word variations have been searched) hepatitis near/5 b:ti,ab,kw (Word variations have been searched) hep near/5 b:ti,ab,kw (Word variations have been searched) hbv:ti,ab,kw (Word variations have been searched) hbsag:ti,ab,kw (Word variations have been searched) #1 or #2 or #3 or #4 or #5 or #6 or #7 or #8 or #9 or #10 or #11 or #12 or #13 or #14 MeSH descriptor: [Reagent Kits, Diagnostic] explode all trees (rapid or "point of care" or "near patient" or poc or poct or bedside) near/5 (test or tests or testing or detect* or diagnos* or screen* or kit or kits or assay* or device*):ti,ab,kw (Word variations have been searched) radt or radts or rdt or rdts:ti,ab,kw (Word variations have been searched) "rapid test*":ti,ab,kw (Word variations have been searched) MeSH descriptor: [Enzyme-Linked Immunosorbent Assay] explode all trees enzyme near/2 (immunoassay* or immuno-assay* or immunosorbent):ti,ab,kw (Word variations have been searched) (antigen* or antibod*) near/3 detect*:ti,ab,kw (Word variations have been searched) MeSH descriptor: [Immunoassay] this term only MeSH descriptor: [Immunoenzyme Techniques] this term only

18 19 20 21 22 23 24

Page | 241

25 26 27 28 29 30 31 32 33

"enzyme-linked immunosorbent assay" or ELISA:ti,ab,kw (Word variations have been searched) #16 or #17 or #18 or #19 or #20 or #21 or #22 or #23 or #24 or #25 MeSH descriptor: [Sensitivity and Specificity] explode all trees diagnos* accura* or sensitiv* or specific* or valid*:ti,ab,kw (Word variations have been searched) "roc curve":ti,ab,kw (Word variations have been searched) "positive predictive value":ti,ab,kw (Word variations have been searched) "negative predictive value":ti,ab,kw (Word variations have been searched) #27 or #28 or #29 or #30 or #31 #15 and #26 and #32

The search found 64 trials. Literatura Latino-Americana e do Caribe em Ciências da Saúde (LILACS) (BIREME interface) LILACS was searched on 20 April 2015 ("hepatitis b" or "hep b" or "hbv" or "hbsag") and ("rapid test$" or "point of care test$" or "near patient test$" or "poc test$" or poct or "bedside test$" or "rapid detect$" or "point of care detect$" or "near patient detect$" or "poc detect$" or "bedside detect$" or "rapid diagnos$" or "point of care diagnos$" or "near patient diagnos$" or "poc diagnos$" or "bedside diagnos$" or "rapid screen$" or "point of care screen$" or "near patient screen$" or "poc screen$" or "bedside screen$" or "rapid kit$" or "point of care kit$" or "near patient kit$" or "poc kit$" or "bedside kit$" or "rapid assay$" or "point of care assay$" or "near patient assay$" or "poc assay$" or "bedside assay$" or "rapid device$" or "point of care device$" or "near patient device$" or "poc device$" or "bedside device$" or radt or radts or rdt or rdts or "enzyme-linked immunosorbent assay" or "antigen$ detect$" or "antibod$ detect$" or elisa or immunoassay or immunoenzyme or "immuno-assay") and ("diagnos$ accura$" or sensitiv$ or specific$ or valid$ or "roc curve" or "positive predictive value" or "negative predictive value") (33)

WHO Global Index Medicus The database was searched on 22 April 2015. SUBJECT: (("Hepatitis, Viral, Human" OR "Hepatitis Viruses" OR "Hepatitis B virus" OR "Hepatitis Antibodies" OR "Hepadnaviridae Infections" OR "Hepatitis B Antibodies" OR "Hepatitis B Virus" OR "Hepadnaviridae" OR "Hepatitis B Surface Antigens") AND ("Reagent Kits, Diagnostic" OR "Enzyme-Linked Immunosorbent Assay" OR "Immunoassay" OR "Immunoenzyme Techniques") AND ("Sensitivity and Specificity")) (478)

Page | 242

Summary data for studies assessing diagnostic accuracy against a NAT-reference standard Table 8. Study characteristics – RDT/ EIA vs NAT Study [Author, Year] Ansari, 2007 Location [Country, City] Iran, Urumieh 240 Sample size Study design CC Hospital patients S Setting Sample Test under evaluation [Type, Brand] RDT, ACON RDT, Atlas RDT, Blue Cross RDT, Cortez RDT, DIMA RDT, Intec Khadem-Ansari, 2014 Lukhwareni, 2009 Iran, Urumieh 350 CC – CSQ Hospital patients – referred as ?HBV HIV cohort – pre ART HIV cohort S ChLIA, Liaison Rt-PCR Reference test [Type, Brand] qPCR

South Africa

192

ChLIA, Elecsys

qPCR

Mphahlele, 2006

South Africa

167 (HIV+) 128 (HIV–)

EIA, AxSYM

Nested PCR

Nna, 2014

Nigeria

113

Blood donors (repeat) Hospital patients – liver disease, HIV

RDT, ACON

Nested PCR; qPCR for positive

Olinger, 2007

Nigeria, Ibadan

200

MEIA, AxSYM v2 ChLIA, Elecsys ELFA, VIDAS Ultra

rtPCR and nested PCR

Seremba, 2010

Uganda

74 (HIV-) 83 (HIV+)

CS – CSQ

Hospital patients ED, including HIV

RDT, Cortez EIA, ADVIA

PCR

qPCR: quantitative PCR; rtPCR: realtime PCR; ChLIA: chemiluminescent immunoassay; CMIA: chemiluminescent microparticle enzyme immunoassay; ECLIA: electrochemiluminescent immunoassay; EIA: enzyme immunoassay; ELFA: enzyme-linked fluorescent assay; ELISA: enzyme-linked immunosorbent assay; MEIA: microparticle enzyme immunoassay; RDT: rapid diagnostic test; rtPCR: real-time PCR; CC: case–control; CS: cross-sectional; CSQ: consecutive patients; LB: lab-based study

Page | 243

Table 9. Summary pooled diagnostic accuracy of HBsAg assays compared to NAT reference Test type Sen (95% CI) RDT EIA 93.3 (91.3–94.9) 75.7 (72.1–79.1) NAT reference Spec (95% CI) 98.1 (97.0–98.9) 86.1 (83.8–88.2)

*Sen: sensitivity; Spec: specificity; CI: confidence interval; RDT: rapid diagnostic test; EIA: enzyme immunoassay

Table 10. Summary pooled diagnostic accuracy of HBsAg assays Test type HIV status n RDT HIV HIV EIA HIV HIV +

NAT reference Sen (95% CI) 37.5 (22.7–54.2) 57.1 (41.0–72.3) 57.9 (49.8–65.6) 83.3 (69.8–92.5) Spec (95% CI) 97.7 (87.7–99.9) 97.2 (93.1–99.2) 95.8 (92.7–97.8) 85.7 (79.2–90.8)

1 2 3 2

–

Page | 244

Fig. 5. Forest plots, RDT vs NAT, ordered by [Test, Author] Sensitivity (95% CI) Specificity (95% CI) 1.00) 1.00) 0.99) 1.00) 1.00) 0.99) 1.00) 1.00) 0.99)

Sensitivity (95% CI) ACON - Ansari ACON - Nna Atlas - Ansari Blue Cross - Ansari Cortez - Ansari Cortez - Seremba (HIV-) Cortez - Seremba (HIV+) DIMA - Ansari Intec - Ansari 0.98 0.60 0.98 0.99 0.98 0.55 0.38 0.98 0.99 (0.94 (0.36 (0.93 (0.95 (0.94 (0.32 (0.23 (0.94 (0.95 -

ACON - Ansari ACON - 0.98 Ansari (0.94 - 1.00) 0.99 (0.95 ACON - Nna ACON - 0.60 Nna (0.36 - 0.81) 0.99 (0.94 Atlas - Ansari 0.98 (0.93 - 0.99) 0.98 (0.93 Atlas - Ansari Specificity (95% CI) - Ansari Blue Cross - Ansari Blue Cross 0.99 (0.95 - 1.00) 0.98 (0.94 ACON --Ansari 0.99 (0.95 - 1.00) Cortez - 0.98 Ansari Cortez Ansari (0.94 - 1.00) 0.98 (0.94 ACON --Nna 0.99 (0.94 -(HIV-) 1.00) Cortez - 0.55 Seremba Cortez Seremba (HIV-) (0.32 - 0.76) 0.94 (0.84 Atlas - Ansari 0.98 (0.93 -(HIV+) 0.99) Cortez - 0.38 Seremba Cortez - Seremba (HIV+) (0.23 - 0.54) 0.98 (0.88 Blue Cross - Ansari 0.98 - 1.00) DIMA Ansari DIMA - Ansari 0.98(0.94 (0.94 - 1.00) 0.99 (0.95 Cortez Ansari - 1.00) Intec - 0.98 Ansari Intec -Ansari 0.99(0.94 (0.95 - 1.00) 0.98 (0.93 Cortez - Seremba (HIV-) 0.94 (0.84 - 0.99) Cortez - Seremba (HIV+) 0.98 (0.88 - 1.00) DIMA - Ansari 0.99 (0.95 - 1.00) Intec - Ansari 0.98 (0.93 - 0.99) Pooled Specificity Pooled Sensitivity = 0.93 (0.91 to 0.95) = 0.98 (0.97 to 0.99) 5.50; df = 8 (p = 0.7026) Chi-square = 173.29;Chi-square df = 8 (p = = 0.0000) Inconsistency = 95.4 % (I-square) = 0.0 % 4 0.8 (I-square) 1 Inconsistency 1 0.6 pecificity Pooled Specificity = 0.98 (0.97 to 0.99) Chi-square = 5.50; df = 8 (p = 0.7026) 8 1 Inconsistency (I-square) = 0.0 %

1.00) ACON - Ansari 0.81) ACON - Nna 0.99) Atlas - Ansari 1.00) Blue Cross - An 1.00) Cortez - Ansari 0.76) Cortez - Seremb 0.54) Cortez - Seremb 1.00) DIMA - Ansari 1.00) Intec - Ansari

0.2

0.4 0.6 Sensitivity

0.8

Pooled Sensitivity = 0.93 (0.91 to 0.95) Chi-square = 173.29; df = 8 (p = 0.0000) Inconsistency 10 0.2 (I-square) 0.4 = 95.4 0.6 % 0.8 Specificity

Pooled Specific Chi-square = 5. 1 Inconsistency (I

Key – Types of RDT and reference standards used in studies Study Ansari Test brand (manufacturer) ACON (Acon laboratories) Atlas (William James House) Blue Cross (Blue Cross Inc.) Cortez (Cortez diagnostics) DIMA (Geseeschaft fur Diagnostika mbH) Intex (Intec Products Inc.) Reference test type, brand (manufacturer) QPCR Roto-GENE 3000 Research (Corbet real time PCR) and kit artus (Hamburg)

Page | 245

Nna

ACON (Acon laboratories)

Nested PCR QPCR for positive

Seremba

Cortez (Cortez diagnostics)

b-DNA (Versant); PCR, Amplicor for discrepant

Page | 246

Fig. 6. Forest plots, EIA vs NAT, ordered by [Test, Author]

Sensitivity (95% CI) Specificity (95% CI) 1.00) 1.00) 0.87) 0.97) 0.83) 0.99) 0.98) 0.76) 0.98)

Sensitivity (95% CI) ADVIA - Seremba (HIV-) ADVIA - Seremba (HIV+) AxSym - Mphahlele (HIV-) AxSym - Mphahlele (HIV+) AxSym v2 - Olinger Elecsys - Lukhwareni (HIV+) Elecsys - Olinger Liaison - Khadem-Ansari VIDAS ULTRA - Olinger 0.73 0.80 0.92 0.38 0.77 0.59 0.67 1.00 0.69 (0.50 (0.64 (0.75 (0.25 (0.67 (0.47 (0.56 (0.98 (0.58 0.89) 0.91) 0.99) 0.53) 0.86) 0.71) 0.77) 1.00) 0.79)

ADVIA - Seremba (HIV-) 0.73 (0.50 - 0.89) 0.98 (0.90 ADVIA - Seremba (HIV-) ADVIA - Seremba (HIV+) 0.80 (0.64 - 0.91) 0.98 (0.88 ADVIA - Seremba (HIV+) AxSym - Mphahlele (HIV-) 0.92 (0.75 - 0.99) AxSym -Specificity Mphahlele (HIV-) (95% CI) 0.79 (0.70 AxSym - Mphahlele (HIV+) (0.25 - 0.53) 0.93 (0.87 AxSym - 0.38 Mphahlele (HIV+) ADVIA v2 - Seremba 0.98 (0.90 - 1.00)0.75 (0.66 AxSym - Olinger(HIV-) 0.77 (0.67 - 0.86) AxSym v2 - Olinger ADVIA Seremba (HIV+) 0.98 (0.88 - 1.00)0.98 (0.93 Elecsys - Lukhwareni (HIV+) -0.59 (0.47 - (HIV+) 0.71) Elecsys Lukhwareni AxSym (HIV-) -0.67 0.79 (0.56 (0.70 - 0.87)0.94 (0.88 Elecsys - Mphahlele Olinger - 0.77) Elecsys Olinger AxSym - Khadem-Ansari MphahleleLiaison (HIV+) - 1.00 0.93 (0.98 (0.87 - 0.97)0.70 (0.63 Liaison - 1.00) Khadem-Ansari AxSym v2 - Olinger 0.75 (0.66 - 0.83)0.94 (0.88 VIDAS ULTRA - Olinger 0.69 (0.58 - 0.79) VIDAS ULTRA Olinger Elecsys - Lukhwareni (HIV+) 0.98 (0.93 - 0.99) Elecsys - Olinger 0.94 (0.88 - 0.98) Liaison - Khadem-Ansari 0.70 (0.63 - 0.76) VIDAS ULTRA - Olinger 0.94 (0.88 - 0.98) Pooled Sensitivity = Pooled 0.76 (0.72 to 0.79) = 0.86 (0.84 to 0.88) Specificity Chi-square = 135.22; df = 8 (p == 0.0000) Chi-square 104.27; df = 8 (p = 0.0000) Inconsistency (I-square) = 94.1 % (I-square) = 92.3 % 1 0.6 0.8 1 Inconsistency ecificity Pooled Specificity = 0.86 (0.84 to 0.88) Chi-square = 104.27; df = 8 (p = 0.0000) 1 Inconsistency (I-square) = 92.3 %

ADVIA - Se ADVIA - Se AxSym - M AxSym - M AxSym v2 Elecsys - L Elecsys - O Liaison - K VIDAS ULT

0.2

0.4 0.6 Sensitivity

0.8

Pooled Sensitivity = 0.76 (0.72 to 0.79) Chi-square = 135.22; df = 8 (p = 0.0000) Inconsistency 1 0 0.2 (I-square) 0.4 = 94.1 0.6 % 0.8 Specificity

Pooled Sp Chi-square 1 Inconsiste

Key – Types of EIA and reference standards used in studies Study Seremba Test type, brand (manufacturer) EIA, ADVIA Centaur (Siemens) Reference test type, brand (manufacturer) b-DNA (Versant); PCR, Amplicor for discrepant Mphahlele EIA, AxSym (Abbott) Nested PCR (in house); positive quantified with COBAS Amplicor () Olinger MEIA, AxSym v2 (Abbott) ELFA, VIDAS Ultra (Biomérieux) ChLIA, Elecsys (Roche) Lukhwareni ChLIA, Elecsys (Roche) Nested PCR, High Pure Viral Nucleic Acid assay (Roche) RT-PCR

Page | 247

Q-PCR, COBAS TaqMan HBV Test 48 assay () Khadem- Ansari ChLIA, Liasison (Diasorin) RT-PCR, Robogene (Corbett)

Page | 248

Summary Receiver Operating Characteristic (SROC) curves Fig. 7. SROC curves for studies comparing RDTs with EIAs Sensitivity 1 SROC Curve

0.9

0.8

Symmetric SROC AUC = 0.9944 SE(AUC) = 0.0025 Q* = 0.9704 SE(Q*) = 0.0078

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.2

0.4 1-specificity

0.6

0.8

Page | 249

Fig. 8. SROC curves for studies comparing RDTs with NATs Sensitivity 1 SROC Curve

0.9

0.8

Symmetric SROC AUC = 0.9974 SE(AUC) = 0.0012 Q* = 0.9816 SE(Q*) = 0.0053

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.2

0.4 1-specificity

0.6

0.8

Page | 250

Fig. 9. SROC curves for studies comparing EIAs with EIAs Sensitivity 1 SROC Curve

0.9

0.8

Symmetric SROC AUC = 0.9953 SE(AUC) = 0.0037 Q* = 0.9735 SE(Q*) = 0.0124

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.2

0.4 1-specificity

0.6

0.8

Page | 251

Fig. 10. SROC curves for studies comparing EIAs with NAT Sensitivity 1 SROC Curve

0.9

0.8

Symmetric SROC AUC = 0.9379 SE(AUC) = 0.0217 Q* = 0.8748 SE(Q*) = 0.0271

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.2

0.4 1-specificity

0.6

0.8

Forest plots of sub-analyses Forest plots, analysed by HIV status

Fig. 11. Forest plots of RDTs vs EIAs in HIV-positive patients Sensitivity (95% CI) Determine - Davies Determine - Franzeck Determine - Geretti Determine - Hoffman VEDA LAB - Honge VIKIA - Geretti 1.00 0.96 0.69 0.75 0.62 0.71 (0.86 (0.80 (0.61 (0.59 (0.51 (0.62 1.00) 1.00) 0.77) 0.87) 0.73) 0.78)

Specificity (95% CI) Determine - Davies Determine - Franzeck Determine - Geretti Determine - Hoffman VEDA LAB - Honge VIKIA - Geretti 1.00 1.00 1.00 1.00 0.99 1.00 (0.93 (0.99 (0.99 (0.99 (0.98 (0.99 1.00) 1.00) 1.00) 1.00) 1.00) 1.00)

0.2

0.4 0.6 Sensitivity

0.8

Pooled Sensitivity = 0.72 (0.68 to 0.76) Chi-square = 30.49; df = 5 (p = 0.0000) 1 Inconsistency (I-square) = 83.6 %

0.2

0.4 0.6 Specificity

0.8

Pooled Specificity = 1.00 (1.00 to 1.00) Chi-square = 12.45; df = 5 (p = 0.0291) 1 Inconsistency (I-square) = 59.8 %

Fig. 13. Forest plots, RDTs vs EIAs in HIV-negative patients Sensitivity (95% CI) Determine - Njai (CHB) Determine - Njai (Screen) Espline - Njai (CHB) VIKIA - Njai (Screen) 0.95 0.88 0.94 0.90 (0.90 - 0.98) (0.81 - 0.94) (0.89 - 0.97) (0.79 - 0.96)

Specificity (95% CI) Determine - Njai (CHB) Determine - Njai (Screen) Espline - Njai (CHB) VIKIA - Njai (Screen) 0.93 1.00 0.95 1.00 (0.78 - 0.99) (0.99 - 1.00) (0.82 - 0.99) (0.99 - 1.00)

0.2

0.4 0.6 Sensitivity

0.8

Pooled Sensitivity = 0.93 (0.90 to 0.95) Chi-square = 4.74; df = 3 (p = 0.1916) = 36.8 1 Inconsistency (I-square) 0 0.2% 0.4 0.6 Specificity

0.8

Pooled Specificity = 1.00 (0.99 to 1.00) Chi-square = 19.96; df = 3 (p = 0.0002) 1 Inconsistency (I-square) = 85.0 %

Page | 252

Fig. 14. Forest plots, EIAs vs EIAs in HIV-positive patients Specificity (95% CI) Architect - Geretti Liaison Ultra - Geretti Murex v3.0 - Geretti 1.00 0.99 0.99 (0.99 - 1.00) (0.99 - 1.00) (0.98 - 1.00)

0.2

0.4 0.6 Specificity

0.8

Pooled Specificity = 0.99 (0.99 to 1.00) Chi-square = 0.51; df = 2 (p = 0.7756) 1 Inconsistency (I-square) = 0.0 %

Sensitivity (95% CI) Architect - Geretti Liaison Ultra - Geretti Murex v3.0 - Geretti 0.98 0.97 0.99 (0.94 - 1.00) (0.93 - 0.99) (0.95 - 1.00)

0.2

0.4 0.6 Sensitivity

0.8

Pooled Sensitivity = 0.98 (0.96 to 0.99) Chi-square = 0.69; df = 2 (p = 0.7067) 1 Inconsistency (I-square) = 0.0 %

Fig. 15. Forest plots, EIAs vs NAT in HIV-positive patients Sensitivity (95% CI) ADVIA - Seremba AxSym - Mphahlele Elecsys - Lukhwareni 0.80 0.38 0.59 (0.64 - 0.91) (0.25 - 0.53) (0.47 - 0.71)

0.2

0.4 0.6 Sensitivity

0.8

Pooled Sensitivity = 0.58 (0.50 to 0.66) Chi-square = 16.84; df = 2 (p = 0.0002) 1 Inconsistency (I-square) = 88.1 %

Specificity (95% CI) ADVIA - Seremba AxSym - Mphahlele Elecsys - Lukhwareni 0.98 0.93 0.98 (0.88 - 1.00) (0.87 - 0.97) (0.93 - 0.99)

0.2

0.4 0.6 Specificity

0.8

Pooled Specificity = 0.96 (0.93 to 0.98) Chi-square = 3.27; df = 2 (p = 0.1953) 1 Inconsistency (I-square) = 38.8 %

Fig. 16. Forest plots, EIAs vs NAT in HIV-negative patients Sensitivity (95% CI) AxSym - Mphahlele (HIV-) ADVIA - Seremba (HIV-) 0.92 0.73 (0.75 - 0.99) (0.50 - 0.89)

0.2

0.4 0.6 Sensitivity

0.8

Pooled Sensitivity = 0.83 (0.70 to 0.93) Chi-square = 3.37; df = 1 (p = 0.0664) 1 Inconsistency (I-square) = 70.3 %

Specificity (95% CI) AxSym - Mphahlele (HIV-) ADVIA - Seremba (HIV-) 0.79 0.98 (0.70 - 0.87) (0.90 - 1.00)

0.2

0.4 0.6 Specificity

0.8

Pooled Specificity = 0.86 (0.79 to 0.91) Chi-square = 12.71; df = 1 (p = 0.0004) 1 Inconsistency (I-square) = 92.1 %

Page | 253

Fig. 17. Forest plots, RDTs vs NAT in HIV-negative patients Sensitivity (95% CI) ACON - Nna Cortez - Seremba (HIV-) 0.60 0.55 (0.36 - 0.81) (0.32 - 0.76)

0.2

0.4 0.6 Sensitivity

0.8

Pooled Sensitivity = 0.57 (0.41 to 0.72) Chi-square = 0.13; df = 1 (p = 0.7211) 1 Inconsistency (I-square) = 0.0 %

Specificity (95% CI) ACON - Nna Cortez - Seremba (HIV-) 0.99 0.94 (0.94 - 1.00) (0.84 - 0.99)

0.2

0.4 0.6 Specificity

0.8

Pooled Specificity = 0.97 (0.93 to 0.99) Chi-square = 2.62; df = 1 (p = 0.1056) 1 Inconsistency (I-square) = 61.8 %

Forest plots, analysed by sample type

Fig. 18. Forest plots, RDTs vs EIA in whole blood Sensitivity (95% CI) Binax - Akanmu Binax - Lau Determine - Njai GWHB - GWHB VIKIA - Njai Determine - Hoffman Determine - Bottero Determine - Franzeck Determine - Lien QUICK PROFILE - Bottero VIKIA - Bottero 1.00 0.96 0.88 0.95 0.90 0.75 0.94 0.96 1.00 0.90 0.96 (0.72 - 1.00) (0.89 - 0.99) (0.81 - 0.94) (0.76 - 1.00) (0.79 - 0.96) (0.59 - 0.87) (0.82 - 0.99) (0.80 - 1.00) (0.79 - 1.00) (0.82 - 0.96) (0.90 - 0.99)

0.2

0.4 0.6 Sensitivity

0.8

Pooled Sensitivity = 0.92 (0.89 to 0.94) Chi-square = 23.13; df = 10 (p = 0.0103) 1 Inconsistency (I-square) = 56.8 %

Specificity (95% CI) Binax - Akanmu Binax - Lau Determine - Njai GWHB - GWHB VIKIA - Njai Determine - Hoffman Determine - Bottero Determine - Franzeck Determine - Lien QUICK PROFILE - Bottero VIKIA - Bottero 0.99 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 (0.96 - 1.00) (0.99 - 1.00) (0.99 - 1.00) (0.40 - 1.00) (0.99 - 1.00) (0.99 - 1.00) (1.00 - 1.00) (0.99 - 1.00) (0.97 - 1.00) (1.00 - 1.00) (1.00 - 1.00)

0.2

0.4 0.6 Specificity

0.8

Pooled Specificity = 1.00 (1.00 to 1.00) Chi-square = 22.85; df = 10 (p = 0.0113) 1 Inconsistency (I-square) = 56.2 %

Page | 254

Forest plots, analysed by study design

Fig. 19. Forest plots, RDTs vs EIA in case–control studies Sensitivity (95% CI) Abraham - Virucheck Chevaliez Clement - Binax Erhabor - ACON Khan - Accurate Khan - Onecheck Lien - Dainascreen Lien - Determine Lien - Serodia Lin - Determine Lin - Determine Lin - DRW Lin - DRW Oh - Genedia Oh - Serodia Randrianirina - Cypress Randrianirina - Determine Randrianirina - Hexagon Randrianirina - Virucheck Sato - Dainascreen Sato - Serodia 0.90 1.00 1.00 1.00 0.50 0.53 1.00 1.00 0.96 0.99 0.94 0.99 0.97 0.98 0.96 0.97 0.98 0.96 0.96 1.00 0.96 (0.68 (0.98 (0.99 (0.88 (0.33 (0.36 (0.97 (0.97 (0.90 (0.96 (0.90 (0.97 (0.93 (0.94 (0.91 (0.91 (0.92 (0.89 (0.89 (0.98 (0.91 0.99) 1.00) 1.00) 1.00) 0.67) 0.69) 1.00) 1.00) 0.99) 1.00) 0.97) 1.00) 0.99) 1.00) 0.99) 0.99) 1.00) 0.99) 0.99) 1.00) 0.98)

0.2

0.4 0.6 Sensitivity

0.8

Pooled Sensitivity = 0.97 (0.96 to 0.97) Chi-square = 230.85; df = 20 (p = 0.0000) 1 Inconsistency (I-square) = 91.3 %

Specificity (95% CI) Abraham - Virucheck Chevaliez Clement - Binax Erhabor - ACON Khan - Accurate Khan - Onecheck Lien - Dainascreen Lien - Determine Lien - Serodia Lin - Determine Lin - Determine Lin - DRW Lin - DRW Oh - Genedia Oh - Serodia Randrianirina - Cypress Randrianirina - Determine Randrianirina - Hexagon Randrianirina - Virucheck Sato - Dainascreen Sato - Serodia 1.00 0.99 0.99 0.91 0.95 1.00 1.00 1.00 1.00 1.00 1.00 0.99 1.00 1.00 1.00 0.96 1.00 0.96 0.98 1.00 1.00 (0.88 (0.97 (0.98 (0.84 (0.74 (0.82 (0.98 (0.98 (0.98 (0.99 (0.99 (0.98 (0.99 (0.96 (0.96 (0.91 (0.97 (0.91 (0.94 (0.99 (0.98 1.00) 1.00) 1.00) 0.96) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 0.99) 1.00) 0.99) 1.00) 1.00) 1.00)

0.2

0.4 0.6 Specificity

0.8

Pooled Specificity = 0.99 (0.99 to 1.00) Chi-square = 85.62; df = 20 (p = 0.0000) 1 Inconsistency (I-square) = 76.6 %

9.3.4 Forest plots, analysed by study setting

Page | 255

Fig. 20. Forest plots, RDTs vs EIA in blood donors Sensitivity (95% CI) Abon - Mutocheluh ACON - Bjoerkvoll (Camb) ACON - Bjoerkvoll (Viet) ACON - Erhabor Acull-Tell - Mutocheluh AMRAD - Ola Binax - Akanmu Biotec - Ola Core TM - Mutocheluh Determine - Lin (China) Determine - Lin (Guinea) Dipstick (PATH) - Mvere DRW - Lin (China) DRW - Lin (Guinea) Genedia - Oh Rapid care - Mutocheluh Serodia - Oh SimpliRed - Mvere Wondfo - Mutocheluh 0.50 0.93 0.82 1.00 0.55 0.95 0.94 0.59 0.50 0.99 0.94 0.93 0.99 0.97 0.98 0.55 0.96 0.93 0.59 (0.28 (0.86 (0.74 (0.88 (0.32 (0.76 (0.87 (0.41 (0.28 (0.96 (0.90 (0.68 (0.97 (0.93 (0.94 (0.32 (0.91 (0.68 (0.36 0.72) 0.98) 0.88) 1.00) 0.76) 1.00) 0.97) 0.75) 0.72) 1.00) 0.97) 1.00) 1.00) 0.99) 1.00) 0.76) 0.99) 1.00) 0.79)

0.2

0.4 0.6 Sensitivity

0.8

Pooled Sensitivity = 0.92 (0.90 to 0.93) Chi-square = 229.22; df = 18 (p = 0.0000) 1 Inconsistency (I-square) = 92.1 %

Specificity (95% CI) Abon - Mutocheluh ACON - Bjoerkvoll (Camb) ACON - Bjoerkvoll (Viet) ACON - Erhabor Acull-Tell - Mutocheluh AMRAD - Ola Binax - Akanmu Biotec - Ola Core TM - Mutocheluh Determine - Lin (China) Determine - Lin (Guinea) Dipstick (PATH) - Mvere DRW - Lin (China) DRW - Lin (Guinea) Genedia - Oh Rapid care - Mutocheluh Serodia - Oh SimpliRed - Mvere Wondfo - Mutocheluh 0.99 1.00 1.00 0.91 0.99 1.00 1.00 0.86 0.98 1.00 1.00 1.00 0.99 1.00 1.00 0.99 1.00 1.00 0.99 (0.96 (0.99 (0.99 (0.84 (0.96 (0.40 (0.03 (0.64 (0.94 (0.99 (0.99 (0.98 (0.98 (0.99 (0.96 (0.96 (0.96 (0.98 (0.96 1.00) 1.00) 1.00) 0.96) 1.00) 1.00) 1.00) 0.97) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00) 1.00)

0.2

0.4 0.6 Specificity

0.8

Pooled Specificity = 1.00 (0.99 to 1.00) Chi-square = 78.41; df = 18 (p = 0.0000) 1 Inconsistency (I-square) = 77.0 %

Page | 256

Forest plots, analysed by study year

Fig. 21. Forest plots, RDTs vs EIA for studies after 2005 Sensitivity (95% CI) Abon - Mutocheluh Accurate - Khan ACON - Bjoerkvoll (Camb) ACON - Bjoerkvoll (Viet) ACON - Erhabor Acull-Tell - Mutocheluh AMRAD - Ola Binax - Akanmu (BD) Binax - Akanmu (CLD) Biotec - Ola Core TM - Mutocheluh Cortez - Chameera Cypress - Randrianirina Determine - Bottero Determine - Davies (H+) Determine - Franzeck (H+) Determine - Geretti (H+) Determine - Hoffman (H+) Determine - Lin (China) Determine - Lin (Guinea) Determine - Njai (CHB) Determine - Njai (Screen) Determine - Nyirendra Determine - Randrianirina DRW - Lin (China) DRW - Lin (Guinea) DRW v2.0 - Chevaliez (Hep) DRW v2.0 - Chevaliez (Preg) DRW v2.0 - Chevaliez CC Espline - Njai (CHB) Hexagon - Randrianirina Intec - Liu Nanosign - Gish (H+) Onecheck - Khan Onsite - Chameera QUICK PROFILE - Bottero Rapid care - Mutocheluh SD Bioline - Upretti VEDA LAB - Honge (H+) VIKIA - Bottero VIKIA - Geretti (H+) VIKIA - Njai (Screen) Virucheck - Randrianirina Wondfo - Mutocheluh 0.50 0.50 0.93 0.82 1.00 0.55 0.95 0.94 1.00 0.59 0.50 0.60 0.97 0.94 1.00 0.96 0.69 0.75 0.99 0.94 0.95 0.88 0.56 0.98 0.99 0.97 1.00 0.96 0.95 0.94 0.96 0.51 0.74 0.53 0.80 0.90 0.55 1.00 0.62 0.96 0.71 0.90 0.96 0.59 (0.28 (0.33 (0.86 (0.74 (0.88 (0.32 (0.76 (0.87 (0.40 (0.41 (0.28 (0.15 (0.91 (0.82 (0.86 (0.80 (0.61 (0.59 (0.96 (0.90 (0.90 (0.81 (0.38 (0.92 (0.97 (0.93 (0.98 (0.81 (0.76 (0.89 (0.89 (0.43 (0.49 (0.36 (0.28 (0.82 (0.32 (0.63 (0.51 (0.90 (0.62 (0.79 (0.89 (0.36 0.72) 0.67) 0.98) 0.88) 1.00) 0.76) 1.00) 0.97) 1.00) 0.75) 0.72) 0.95) 0.99) 0.99) 1.00) 1.00) 0.77) 0.87) 1.00) 0.97) 0.98) 0.94) 0.73) 1.00) 1.00) 0.99) 1.00) 1.00) 1.00) 0.97) 0.99) 0.58) 0.91) 0.69) 0.99) 0.96) 0.76) 1.00) 0.73) 0.99) 0.78) 0.96) 0.99) 0.79)

0.2

0.4 0.6 Sensitivity

0.8

Pooled Sensitivity = 0.86 (0.85 to 0.88) Chi-square = 680.76; df = 43 (p = 0.0000) 1 Inconsistency (I-square) = 93.7 %

Page | 257

Specificity (95% C Abon - Mutocheluh Accurate - Khan ACON - Bjoerkvoll (Camb) ACON - Bjoerkvoll (Viet) ACON - Erhabor Acull-Tell - Mutocheluh AMRAD - Ola Binax - Akanmu (BD) Binax - Akanmu (CLD) Biotec - Ola Core TM - Mutocheluh Cortez - Chameera Cypress - Randrianirina Determine - Bottero Determine - Davies (H+) Determine - Franzeck (H+) Determine - Geretti (H+) Determine - Hoffman (H+) Determine - Lin (China) Determine - Lin (Guinea) Determine - Njai (CHB) Determine - Njai (Screen) Determine - Nyirendra Determine - Randrianirina DRW - Lin (China) DRW - Lin (Guinea) DRW v2.0 - Chevaliez (Hep) DRW v2.0 - Chevaliez (Preg) DRW v2.0 - Chevaliez CC Espline - Njai (CHB) Hexagon - Randrianirina Intec - Liu Nanosign - Gish (H+) Onecheck - Khan Onsite - Chameera QUICK PROFILE - Bottero Rapid care - Mutocheluh SD Bioline - Upretti VEDA LAB - Honge (H+) VIKIA - Bottero VIKIA - Geretti (H+) VIKIA - Njai (Screen) Virucheck - Randrianirina Wondfo - Mutocheluh 0.99 0.95 1.00 1.00 0.91 0.99 1.00 1.00 1.00 0.86 0.98 1.00 0.96 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.93 1.00 0.69 1.00 0.99 1.00 0.99 0.99 0.98 0.95 0.96 1.00 0.98 1.00 1.00 1.00 0.99 1.00 0.99 1.00 1.00 1.00 0.98 0.99

(0.96 - 1.0 (0.74 - 1.0 (0.99 - 1.0 (0.99 - 1.0 (0.84 - 0.9 (0.96 - 1.0 (0.40 - 1.0 (0.03 - 1.0 (0.89 - 1.0 (0.64 - 0.9 (0.94 - 1.0 (0.92 - 1.0 (0.91 - 0.9 (1.00 - 1.0 (0.93 - 1.0 (0.99 - 1.0 (0.99 - 1.0 (0.99 - 1.0 (0.99 - 1.0 (0.99 - 1.0 (0.78 - 0.9 (0.99 - 1.0 (0.62 - 0.7 (0.97 - 1.0 (0.98 - 1.0 (0.99 - 1.0 (0.97 - 1.0 (0.97 - 1.0 (0.97 - 0.9 (0.82 - 0.9 (0.91 - 0.9 (0.95 - 1.0 (0.95 - 0.9 (0.82 - 1.0 (0.92 - 1.0 (1.00 - 1.0 (0.96 - 1.0 (0.99 - 1.0 (0.98 - 1.0 (1.00 - 1.0 (0.99 - 1.0 (0.99 - 1.0 (0.94 - 1.0 (0.96 - 1.0

0.2

0.4 0.6 Specificity

0.8

Pooled Specificity = 0.99 (0.99 to 0.99) Chi-square = 548.27; df = 43 (p = 0.0000) 1 Inconsistency (I-square) = 92.2 % Page | 258

Positive LR (95% CI) Abon - Mutocheluh Accurate - Khan ACON - Bjoerkvoll (Camb) ACON - Bjoerkvoll (Viet) ACON - Erhabor Acull-Tell - Mutocheluh AMRAD - Ola Binax - Akanmu (BD) Binax - Akanmu (CLD) Biotec - Ola Core TM - Mutocheluh Cortez - Chameera Cypress - Randrianirina Determine - Bottero Determine - Davies (H+) Determine - Franzeck (H+) Determine - Geretti (H+) Determine - Hoffman (H+) Determine - Lin (China) Determine - Lin (Guinea) Determine - Njai (CHB) Determine - Njai (Screen) Determine - Nyirendra Determine - Randrianirina DRW - Lin (China) DRW - Lin (Guinea) DRW v2.0 - Chevaliez (Hep) DRW v2.0 - Chevaliez (Preg) DRW v2.0 - Chevaliez CC Espline - Njai (CHB) Hexagon - Randrianirina Intec - Liu Nanosign - Gish (H+) Onecheck - Khan Onsite - Chameera QUICK PROFILE - Bottero Rapid care - Mutocheluh SD Bioline - Upretti VEDA LAB - Honge (H+) VIKIA - Bottero VIKIA - Geretti (H+) VIKIA - Njai (Screen) Virucheck - Randrianirina Wondfo - Mutocheluh

64.00 (8.69 - 471.26) 9.50 (1.37 - 65.71) 517.87 (129.54 - 2,070.26) 434.51 (108.57 - 1,739.03) 10.46 (5.70 - 19.19) 69.82 (9.55 - 510.36) 9.32 (0.67 - 129.54) 3.72 (0.34 - 41.10) 59.40 (3.74 - 944.36) 4.12 (1.39 - 12.18) 32.00 (7.60 - 134.68) 53.67 (3.14 - 917.12) 26.35 (10.06 - 69.00) 4,498.21 (281.12 - 71,976.8 100.04 (6.34 - 1,578.57) 467.38 (29.27 - 7,463.70) 966.70 (60.39 - 15,473.86) 174.94 (64.74 - 472.74) 959.01 (60.07 - 15,311.15) 753.59 (47.21 - 12,030.12) 14.29 (3.74 - 54.54) 1,180.48 (73.85 - 18,870.34 1.82 (1.25 - 2.67) 214.02 (13.47 - 3,400.91) 120.60 (45.44 - 320.04) 771.27 (48.32 - 12,311.21) 74.50 (29.76 - 186.52) 73.38 (30.62 - 175.83) 43.75 (27.08 - 70.69) 17.85 (4.63 - 68.81) 26.05 (9.95 - 68.23) 73.20 (4.61 - 1,163.40) 34.14 (14.80 - 78.75) 21.03 (1.34 - 329.93) 69.00 (4.22 - 1,129.19) 347.25 (186.27 - 647.36) 69.82 (9.55 - 510.36) 642.22 (40.07 - 10,293.22) 75.01 (23.99 - 234.59) 1,853.68 (463.49 - 7,413.61 986.53 (61.64 - 15,789.02) 374.40 (52.77 - 2,656.47) 52.10 (13.19 - 205.84) 75.64 (10.41 - 549.47)

0.01

1 Positive LR

Random Effects Model Pooled Positive LR = 84.66 (43.55 to 164.55) Cochran-Q = 534.07; df = 43 (p = 0.0000) 100.0 Inconsistency (I-square) = 91.9 % Tau-squared = 4.0986

Page | 259

Negative LR (95% CI) Abon - Mutocheluh Accurate - Khan ACON - Bjoerkvoll (Camb) ACON - Bjoerkvoll (Viet) ACON - Erhabor Acull-Tell - Mutocheluh AMRAD - Ola Binax - Akanmu (BD) Binax - Akanmu (CLD) Biotec - Ola Core TM - Mutocheluh Cortez - Chameera Cypress - Randrianirina Determine - Bottero Determine - Davies (H+) Determine - Franzeck (H+) Determine - Geretti (H+) Determine - Hoffman (H+) Determine - Lin (China) Determine - Lin (Guinea) Determine - Njai (CHB) Determine - Njai (Screen) Determine - Nyirendra Determine - Randrianirina DRW - Lin (China) DRW - Lin (Guinea) DRW v2.0 - Chevaliez (Hep) DRW v2.0 - Chevaliez (Preg) DRW v2.0 - Chevaliez CC Espline - Njai (CHB) Hexagon - Randrianirina Intec - Liu Nanosign - Gish (H+) Onecheck - Khan Onsite - Chameera QUICK PROFILE - Bottero Rapid care - Mutocheluh SD Bioline - Upretti VEDA LAB - Honge (H+) VIKIA - Bottero VIKIA - Geretti (H+) VIKIA - Njai (Screen) Virucheck - Randrianirina Wondfo - Mutocheluh 0.50 0.53 0.07 0.18 0.02 0.46 0.08 0.09 0.10 0.48 0.51 0.42 0.03 0.07 0.02 0.06 0.31 0.25 0.01 0.06 0.05 0.12 0.64 0.03 0.01 0.04 0.00 0.04 0.05 0.06 0.05 0.50 0.27 0.49 0.25 0.10 0.46 0.06 0.38 0.04 0.29 0.10 0.04 0.41 (0.33 - 0.77) (0.38 - 0.74) (0.03 - 0.14) (0.13 - 0.26) (0.00 - 0.28) (0.29 - 0.72) (0.02 - 0.36) (0.03 - 0.26) (0.01 - 1.41) (0.31 - 0.74) (0.33 - 0.77) (0.16 - 1.09) (0.01 - 0.10) (0.03 - 0.20) (0.00 - 0.30) (0.01 - 0.27) (0.24 - 0.40) (0.15 - 0.43) (0.00 - 0.05) (0.03 - 0.10) (0.02 - 0.11) (0.07 - 0.20) (0.43 - 0.94) (0.01 - 0.09) (0.00 - 0.04) (0.02 - 0.08) (0.00 - 0.04) (0.01 - 0.26) (0.01 - 0.33) (0.03 - 0.12) (0.02 - 0.12) (0.43 - 0.58) (0.13 - 0.57) (0.35 - 0.68) (0.06 - 1.01) (0.05 - 0.18) (0.29 - 0.72) (0.00 - 0.82) (0.28 - 0.51) (0.01 - 0.11) (0.23 - 0.38) (0.05 - 0.21) (0.02 - 0.12) (0.25 - 0.68)

0.01

1 Negative LR

Random Effects Model Pooled Negative LR = 0.13 (0.09 to 0.18) Cochran-Q = 866.48; df = 43 (p = 0.0000) 100.0 Inconsistency (I-square) = 95.0 % Tau-squared = 1.2712

Page | 260

Fig. 22. Forest plots, RDTs vs EIA for studies before 2005 Sensitivity (95% CI) Binax - Clement Binax - Lau (Fresh S) Binax - Lau (Frozen S) Binax - Lau (WB) Dainascreen - Lien Dainascreen - Sato Determine - Lien Dipstick (PATH) - Mvere Genedia - Oh Hepacard - Kaur Hepacard - Raj QuickChaser - Abraham CC QuickChaser - Abraham CS Serodia - Lien Serodia - Oh Serodia - Sato SimpliRed - Mvere Virucheck - Abraham CC Virucheck - Abraham CS 1.00 0.94 0.95 0.96 1.00 1.00 1.00 0.93 0.98 0.93 0.83 0.90 0.77 0.96 0.96 0.96 0.93 0.90 0.79 (0.99 (0.79 (0.86 (0.89 (0.97 (0.98 (0.97 (0.68 (0.94 (0.84 (0.61 (0.68 (0.55 (0.90 (0.91 (0.91 (0.68 (0.68 (0.54 1.00) 0.99) 0.99) 0.99) 1.00) 1.00) 1.00) 1.00) 1.00) 0.98) 0.95) 0.99) 0.92) 0.99) 0.99) 0.98) 1.00) 0.99) 0.94)

0.2

0.4 0.6 Sensitivity

0.8

Pooled Sensitivity = 0.97 (0.96 to 0.98) Chi-square = 83.02; df = 18 (p = 0.0000) 1 Inconsistency (I-square) = 78.3 %

Page | 261

Specificity (95% CI) Binax - Clement Binax - Lau (Fresh S) Binax - Lau (Frozen S) Binax - Lau (WB) Dainascreen - Lien Dainascreen - Sato Determine - Lien Dipstick (PATH) - Mvere Genedia - Oh Hepacard - Kaur Hepacard - Raj QuickChaser - Abraham CC QuickChaser - Abraham CS Serodia - Lien Serodia - Oh Serodia - Sato SimpliRed - Mvere Virucheck - Abraham CC Virucheck - Abraham CS 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.99 1.00 0.99 1.00 1.00 1.00 1.00 1.00 0.97 (0.99 - 1.00) (1.00 - 1.00) (1.00 - 1.00) (0.99 - 1.00) (0.98 - 1.00) (0.99 - 1.00) (0.98 - 1.00) (0.98 - 1.00) (0.96 - 1.00) (1.00 - 1.00) (0.98 - 0.99) (0.88 - 1.00) (0.98 - 1.00) (0.98 - 1.00) (0.96 - 1.00) (0.98 - 1.00) (0.98 - 1.00) (0.88 - 1.00) (0.95 - 0.98)

0.2

0.4 0.6 Specificity

0.8

Pooled Specificity = 1.00 (1.00 to 1.00) Chi-square = 88.17; df = 18 (p = 0.0000) 1 Inconsistency (I-square) = 79.6 %

Positive LR (95% CI) Binax - Clement Binax - Lau (Fresh S) Binax - Lau (Frozen S) Binax - Lau (WB) Dainascreen - Lien Dainascreen - Sato Determine - Lien Dipstick (PATH) - Mvere Genedia - Oh Hepacard - Kaur Hepacard - Raj QuickChaser - Abraham CC QuickChaser - Abraham CS Serodia - Lien Serodia - Oh Serodia - Sato SimpliRed - Mvere Virucheck - Abraham CC Virucheck - Abraham CS

268.83 (67.41 - 1,072. 1,811.52 (113.19 - 28, 1,448.64 (90.62 - 23,1 1,051.25 (65.81 - 16,7 422.20 (26.49 - 6,728. 608.08 (38.12 - 9,699. 422.20 (26.49 - 6,728. 348.00 (21.75 - 5,568. 197.32 (12.43 - 3,133. 4,996.94 (312.36 - 79, 73.30 (39.56 - 135.82) 54.62 (3.48 - 857.69) 146.05 (35.99 - 592.65 404.24 (25.36 - 6,443. 193.30 (12.17 - 3,069. 290.53 (41.05 - 2,056. 348.00 (21.75 - 5,568. 54.62 (3.48 - 857.69) 25.07 (13.71 - 45.82)

0.01

1 Positive LR

Random Effects Model Pooled Positive LR = 265.53 (106.10 to 664.50) Cochran-Q = 88.07; df = 18 (p = 0.0000) 100.0 Inconsistency (I-square) = 79.6 % Tau-squared = 2.7236

Page | 262

Negative LR (95% CI) Binax - Clement Binax - Lau (Fresh S) Binax - Lau (Frozen S) Binax - Lau (WB) Dainascreen - Lien Dainascreen - Sato Determine - Lien Dipstick (PATH) - Mvere Genedia - Oh Hepacard - Kaur Hepacard - Raj QuickChaser - Abraham CC QuickChaser - Abraham CS Serodia - Lien Serodia - Oh Serodia - Sato SimpliRed - Mvere Virucheck - Abraham CC Virucheck - Abraham CS 0.00 0.08 0.06 0.05 0.00 0.00 0.00 0.09 0.02 0.07 0.18 0.12 0.23 0.05 0.04 0.04 0.09 0.12 0.22 (0.00 - 0.02) (0.02 - 0.25) (0.02 - 0.16) (0.02 - 0.13) (0.00 - 0.07) (0.00 - 0.05) (0.00 - 0.07) (0.02 - 0.43) (0.01 - 0.07) (0.03 - 0.18) (0.07 - 0.43) (0.04 - 0.39) (0.11 - 0.49) (0.02 - 0.11) (0.02 - 0.09) (0.02 - 0.09) (0.02 - 0.43) (0.04 - 0.39) (0.09 - 0.52)

0.01

1 Negative LR

Random Effects Model Pooled Negative LR = 0.06 (0.03 to 0.09) Cochran-Q = 74.14; df = 18 (p = 0.0000) 100.0 Inconsistency (I-square) = 75.7 % Tau-squared = 0.9133

9.3. 9.5 List of tests available Manufacturer Abbott Abbott Abbott Abbott Abbott Abbott Adaltis Siemens Siemens Test AxSym HBsAg v2 Architect HBsAg Architect HBsAg Qualitative Architect HBsAg Qualitative II Auszyme Monoclonal PRISM HBsAg EIAgen HBsAg Kit ADVIA Centaur HBsAg ADVIA Centaur HBsAg II ChLIA EIA ChLIA ChLIA Test type MEIA CMIA (Quant) CMIA CMIA

Page | 263

BIOKIT bioMerieux bioMerieux bioMerieux bioMerieux BIO-RAD BIO-RAD BIO-RAD BIO-RAD Dade Behring Dade Behring Diasorin Diasorin (prev Abbott) Diasorin Diasorin General Biologicals Corp. MBS SRL Medical Biological Services Ortho Ortho Roche Roche Roche Siemens Medical Solutions

Bioelisa HBsAg 3.0 Hepanostika Uni-Form II VIKIA HBsAg Kit VIDAS HBsAg VIDAS HBsAg Ultra Genscreen HBsAg 3.0 MONOLISA HBsAg Ultra MONOLISA HBsAg Plus MONOLISA AgHBs (2 Gen) Enzygnost HBsAg EIA 5.0 Enzygnost HBsAg EIA 6.0 ETI-MAK 4 HBsAg EIA Murex HBsAg Version 3 Liason HBsAg Liason XL MUREX HBsAg SURASE B-96, TMB HBsAg One Step EIA EIA EIA - ? not available ELISA ChLIA ChLIA (Quant) ELISA ELISA nd

Microelisa

ELFA ELFA

EIA/ ELISA EIA

HBsAg ELISA Test System 3 Vitros ECi HBsAg Cobas Core HBsAg II Elecsys HBsAg Elecsys II HBsAg Immulite 2000 HBsAg EIA ECLIA ECLIA ELISA

Page | 264

Diagnostics Beijing Wantai Biological Pharmacy Enterprise Co., Ltd Hepatitis B Virus Surface Antigen (HBsAg) ELISA ELISA

Page | 265

Annex 5.4 PICO 2 - How to test (HCV) Diagnostic accuracy of tests to detect hepatitis C antibody: a meta-analysis and review of the literature

Weiming Tang,* MD, PhD (weimingtangscience@gmail.com), Wen Chen,* PhD (chenw43@mail.sysu.edu.cn), Ali Amini, MBBS, Debi Boeras, MPH, Jane Falconer, MS, Helen Kelly, MD, Rosanna Peeling, PhD (Team lead), Olivia Varsaneux, MBBS, Joseph Tucker, MD, PhD London School of Hygiene and Tropical Medicine team *Co-leaders of this review

September 2015

Page | 266

1. Executive summary Background: Although direct acting antivirals (DAAs) have led to sustained virological response in greater than 90% of all individuals treated for hepatitis C virus (HCV), most HCVinfected individuals remain undiagnosed and untreated. Enzyme immunoassays have been used to detect exposure to HCV but access to these laboratory-based assays has been a barrier in reaching at-risk populations for testing and treatment. Rapid diagnostic tests (RDTs) to detect HCV antibody (HCV Ab) are commercially available and may be useful in decentralizing HCV screening outside of laboratory settings. The purpose of this work was to review the peer-reviewed literature and determine the diagnostic accuracy of available assays in detecting antibodies to HCV as a biomarker of exposure. Method: We used the PRISMA guidelines and Cochrane guidance to develop our search protocol. The search strategy was registered in PROSPERO (CRD42015023567). A literature search was conducted focused on hepatitis C, diagnostic tests and diagnostic accuracy among eight databases. Studies were included if they evaluated an assay to determine the sensitivity and specificity of HCV Ab in humans. Reference standards included enzyme immunoassay (EIA), immunoblot (e.g. recombinant immunoblot assay), and/or nucleic acid testing (NAT). Two reviewers independently extracted data and performed a quality assessment of the studies using the QUADAS tool. Results: A total of 52 studies were included that included 52 273 unique test measurements. Based on five studies, the pooled RDT sensitivity and specificity were 0.98 (95% CI 0.98–1.00) and 1.00 (95% CI 1.00–1.00) compared to an EIA reference standard. High HCV Ab RDT sensitivity and specificity were observed across screening populations (general population, key populations, hospital patients) using different reference standards (EIA, NAT, immunoblot). Limiting the RDT analysis to studies published in the past five or ten years did not change the results. There were insufficient studies to undertake subanalyses based on HIV coinfection. Oral HCV Ab RDTs had excellent sensitivity and specificity compared to blood reference tests, respectively, at 0.94 (95% CI 0.93–0.96) and 1.00 (95% CI 1.00–1.00). Among studies that assessed individual oral RDT tests, the eight studies that examine OraQuick ADVANCE® had a slightly higher sensitivity (0.98, 95% CI 0.97–0.98) compared to the six studies that examined other brands and found a pooled sensitivity of 0.88 (95% CI 0.84-0.92). Conclusions: RDTs, including oral tests, have excellent sensitivity and specificity compared to laboratory-based methods for HCV antibody detection across a wide range of settings. Although the sensitivity of the HCV Ab RDT decreases in low- and middle-income country (LMIC) contexts, this would still be an important public health tool for screening purposes. Oral HCV Ab RDTs had good sensitivity and specificity compared to blood reference standards and may be particularly useful in contexts where the use of blood-based tests may be challenging.

Page | 267

2. Background Hepatitis C is a liver disease caused by the hepatitis C virus (HCV) that causes acute and chronic infection.1, 2 An estimated 130–150 million people have chronic hepatitis C infection worldwide, leading to 350 000–500 000 deaths per year.1–4 The introduction of direct-acting antivirals (DAAs) has led to sustained virological response (SVR) in more than 90% of all HCVinfected individuals.5, 6 DAAs are now recommended by WHO and many other HCV treatment guidelines.1 DAAs will not only improve SVR rates, but also may simplify HCV management algorithms and allow smaller health facilities to manage HCV-infected individuals.7 Despite the availability of effective treatment, most HCV-infected individuals remain undiagnosed and untreated.4 As a result, approximately 15–30% of individuals with chronic HCV infection progress to cirrhosis, leading to end-stage liver disease and hepatocellular carcinoma.1, 2 In April 2014, WHO published the guidelines for the screening, care and treatment of individuals with HCV infection.8 These guidelines included recommendations on who to screen for HCV and how to confirm HCV infection, but not which tests are optimal for initial screening. The World Health Assembly has passed several resolutions highlighting the importance of prevention and control of viral hepatitis for global health. Advances in HCV detection technology create new opportunities for enhancing screening, referral, and treatment. Previous systematic reviews on hepatitis C infection have focused on treatment response,9,10 clinical complications11 and epidemiology.12, 13 Two systematic reviews on hepatitis C testing focused on evaluating point-of-care (POC) tests compared to EIAs and other reference tests.14, 15 This review extends previous reviews by including new studies, including a subanalysis focused on oral tests, and including studies that evaluated immunoassays using a NAT reference standard.

3. Objectives The purpose of this review was to identify quantitative evidence on the sensitivity and specificity of rapid diagnostic tests used to detect HCV Ab, synthesize the evidence, and inform models to estimate cost–effectiveness of different strategies for testing.

PICO 2 P I

Among persons identified for hepatitis C testing, what is the diagnostic accuracy of available assays for detecting HCVAb? Persons identified for HCV testing Rapid diagnostic tests and enzyme immunoassays for HCsAg detection 1) 2) 3) 4) EIA (with subanalysis based on the last ten years) NAT (nucleic acid testing) Immunoblot or similar assay A combination of 1,2,3 above

Diagnostic accuracy (Sensitivity, Specificity, Positive predictive value, Negative predictive value, TN, TP, FN, and FP).

Page | 268

4. Methods Search strategy and identification of studies We included observational and randomized control trial (RCT) studies that provide original data from patient specimens from cross-sectional or case–control studies. Literature search strategies were developed by a medical librarian with expertise in designing systematic review searches. Our search algorithm consisted of the following components: hepatitis C, diagnostic tests, and diagnostic accuracy (see annex 1). We searched MEDLINE (OVID interface, 1946 onwards), EMBASE (OVID interface, 1947 onwards), the Cochrane Central Register of Controlled Trials (Wiley interface, current issue), Science Citation Index Expanded (Web of Science interface, 1970 onwards), Conference Proceedings Citation Index-Science (Web of Science interface, 1990 onwards), SCOPUS (1960 onwards), Literatura Latino-Americana e do Caribe em Ciências da Saúde (LILACS) (BIREME interface) and WHO Global Index Medicus. The search was supplemented by searching for ongoing studies in WHO’s International Clinical Trials Registry. The literature search was limited to the English language and human subjects. In addition to searching databases, we contacted individual researchers and authors of major trials to address whether any relevant manuscripts are in preparation or in press. The references of published articles found in the above databases were searched for additional pertinent materials. Study selection proceeded in three stages: (1) titles/abstracts were screened by a single reviewer according to standard inclusion and exclusion criteria; (2) full manuscripts were obtained and assessed against inclusion criteria. Papers were accepted or rejected and reasons for rejection were specified; (3) two independent reviewers assessed each manuscript. Differences were resolved by a third independent reviewer. Selection criteria The inclusion criteria included the following: primary purpose is HCV Ab test evaluation, reported sensitivity and specificity of HCV Ab test kits, and studies published before May 2015. Studies that only reported sensitivity or specificity, conference abstracts, comments or review papers, panel studies, and studies that only used reference assay for positive samples were excluded. Data extraction Information on the following variables were extracted from each individual study: first author, total sample size, country (and city) of sampling, sample type (oral fluid, finger-prick, venous blood, etc.), POC (defined as being able to give a result within 60 minutes and having the results to guide clinical management in the same encounter), eligibility criteria, reference standard, manufacturer, raw cell numbers (true positives, false negatives, false positives, true negatives), antibody–antigen combo (yes or no), sources of funding, reported conflicts of interest. Assessment of methodological quality

Page | 269

Study quality was evaluated using the QUADAS-2 tool16 and the STARD checklist.17 QUADAS includes domains to evaluate bias in the following categories: risk of bias (patient selection, index test, reference standard, flow and timing); applicability concerns (patient selection, index test, reference standard).

Data analysis and synthesis Data synthesis Data were extracted to construct 2 × 2 tables. According to the test results of reference standard, anti-HCV positive and negative were defined. By comparing with reference standard results, the index test results were categorized as a true positive, a false positive, a false negative, or a true negative. Indeterminate test results were not included in pooled analyses. Statistical analysis To estimate test accuracy, we calculated sensitivity and specificity for each study and pooled statistics, along with 95% confidence intervals (CIs). We pooled test estimates using the DerSimonian–Laird method, a bivariate random effect model. We did further sub-analysis based on reference standard (EIA alone; NAT or immunoblot; EIA, NAT, or immunoblot), brand, and combination test. We performed all statistical analysis (including heterogeneity) using the software R and RevMan 5.3.

5. Results Study selection A total of 11 163 citations were identified and 6163 duplicates were removed. Each of the 5000 titles was examined. A total of 52 research studies were included in the final analysis (Fig. 1 below).8, 18-68 Of the 52 studies, 32 studies evaluated the accuracy of 30 different RDTs, of which 5 evaluated RDTs compared to EIA alone, 13 compared RDT results to NAT or immunoblot, and 15 focused on evaluating RDT by comparing with the results of EIA or immunoblot or NAT. Twelve studies evaluated the diagnostic accuracy of oral fluid RDTs. There were insufficient data to undertake a subanalysis based on HIV coinfection or other coinfections.

Page | 270

Fig. 1. PRISMA flow diagram outlining study selection examining diagnostic accuracy of HCV antibody tests

Page | 271

Study characteristics Of the 52 included studies, 9 were published before 2010, 12 reported evaluation using oral fluid samples, and 34 were reported performance of POC tests (Table 1). Of the 52 studies, 41 different brands of testing kits were evaluated (Table 1).

Table 1. Characteristics of studies focused on evaluating diagnostic accuracy of HCV antibody tests First author Al-Tahish et al. Year 2013 Settings Egypt Sample type Venous blood Manufacturer HCV one step test device (ACON Laboratories, USA), Fourth- generation HCV TRI_DOT (J. Mitra Co, India) and ImmunoComb II HCV (Inverness Medical Innovations, USA) Ortho Clinical Diagnostics (Raritan, NJ, USA) Study type Cross-sectional Sample size 100 POC (Y/N) Y Reference standard PCR

Bonacini et al.

2001

USA

Venous blood

Cross-sectional

222

Chiron Immunoblot HCV 3.0 SIA Immunoblot PCR

Buti et al. Caudai et al.

2000 1998

Spain USA

Serum Serum or plasma samples Oral fluids and serum Venous blood

Not available ELISA 2nd generation Abbott Laboratories, Abbott park, IL, USA) OraQuick (OraSure Technologies, PA USA)

Cross-sectional Cross-sectional

188 682

Y N

Cha et al.

2013

Korea

Case–control

437

PCR

Croom et al

2006

Austria

Monolisa anti-HCV PLUSVersion 2 EIA (Bio-Rad, France) Rapid Test Bioeasy (Standard Diagnostics, Yongin, ® Korea) and Imuno-Rapido HCV (Wama Diagnostica, São Carlos, Brazil). TRI DOT (J. MITRA &Co. Ltd., New Delhi, India) ®

Cross-sectional

182

EIA

da Rosa et al.

2013

Brazil

Serum

Cross-sectional

307

Architect HCV, PCR

Daniel et al.

2005

India

Serum

Cross-sectional

2590

EIA, Immunoblot, PCR

Page | 272

First author Denoyel et al.

Year 2004

Settings France and Germany USA USA Turkey Germany

Sample type Serum or plasma samples Blood Oral fluid Plasma specimens Plasma specimens

Manufacturer AxSYM HCV 3.0 (other information is not available)

Study type Cross-sectional

Sample size 5700

POC (Y/N) N

Reference standard Immunoblot

Dokubo et al. Drobnik et al. Eroglu et al. Feucht et al.

2014 2011 2000 1995

HCV Version 3.0 ELISA (Ortho ) OraQuick (OraSure Technologies, PA USA) ELISA v3.0(Ortho ) Abbott HCV second-generation enzyme immunoassay (other information is not available) OraQuick (OraSure Technologies, PA USA) DPP HIV-HCV-Syphilis Assay (Chembio Diagnostic Systems, Inc. , Medford, NY) OraQuick (OraSure Technologies, PA USA) ®

Cross-sectional Cross-sectional Cross-sectional Cross-sectional

132 484 160 262

N Y N N

PCR EIA, Immunoblot PCR Immunoblot

Gao et al. Hess et al.

2014 2014

USA USA

Serum Whole blood

Cross-sectional Cross-sectional

289 948

Y Y

EIA EIA

Hui et al.

2002

Hong Kong, China Saudi Arabia Brazil

Whole blood

Cross-sectional

197

EIA

Ibrahim et al. Ivantes et al.

2015 2010

Oral fluid Whole blood

OraQuick (OraSure Technologies, PA USA) HCV Rapid Test Bioeasy (Bioeasy Diagnostica Ltda, Minas Gerais, Brazil) Chembio DPP HCV test (Chembio Diagnostic Systems,USA) and Rapid HIV/HCV antibody test (Medmira Laboratories, Canada) Toyo anti-HCV test (Turklab, Izmir, Turkey) HCV Bidot (J. Mitra Co., India) GENEDIA® HCV Rapid LF (Green Cross medical

Case–control Cross-sectional

160 71

Y Y

PCR CLIA

Jewett et al.

2012

USA

Oral fluids and serum

Cross-sectional

407

Immunoblot/NAT

Kant et al. Kaur et al. Kim et al.

2012 2000 2013

Germany India Republic of

Whole blood Serum Serum

Cross-sectional Cross-sectional Case–control

185 2754 200

Y Y Y

Architect HCV EIA 3rd generation Immunoblot

Page | 273

First author

Year

Settings Korea

Sample type

Manufacturer science corp., Korea)

Study type

Sample size

POC (Y/N)

Reference standard

Kosack et al.

2014

Germany

Serum

The ImmunoFlow HCV test (Core Diagnostics,United Kingdom) Beijing United Biomedical, Ortho Clinical Diagnostics, General Biologicals; other information is not avaliable cEIA: the Monolisa® HCV-Ag-Ab-ULTRA (Bio-Rad, Marnes-la-Coquette, France) OraQuick (OraSure Technologies, PA USA)

Cross-sectional

Immunoblot

Lakshmi et al.

2007

India

Blood

Cross-sectional

PCR

Larrat et al.

2012

France

FSB (finger-stick blood) and oral fluid Oral fluid, whole blood Serum, plasma, venous blood, finger-stick blood and oral fluid Oral fluid Serum

Case–control

201

PCR

Lee et al.

2010

USA

Cross-sectional

572

EIA, Immunoblot

Lee et al.

2011

USA

OraQuick (OraSure Technologies, PA USA)

Cross-sectional

2183

EIA, Immunoblot, PCR

Lee et al. Maity et al.

2011 2012

USA India

OraQuick (OraSure Technologies, PA USA) J Mitra & Co. Pvt Ltd, SPAN Diagnostics Ltd. and Standard Diagnostics, INC, other information is not available Anti-HCV Ab rapid test (1st IRP 75/537 by Thema Ricerca, WHO Geneva) HCV-SPOT (Genelabs Diagnostics, Singapore)

Cross-sectional Case–control

2180 or 2178 100

Y Y

EIA EIA

Montebugnoil et al. Mvere et al.

1999

Italy

Whole blood

Case–control

100

EIA, Immunoblot

1996

Zimbabwe

Serum

Cross-sectional

206

EIA 2nd generation, INNO-LIA HCV Ab III

Page | 274

First author Nalpas et al.

Year 1992

Settings France

Sample type Serum

Manufacturer Ortho Diagnostics, other information is not available ImmunoComb® II HCV assay (Orgenics Ltd, Not reported manufacturer located country); ImmunoComb® II HCV assay (Orgenics Ltd, Not reported manufacturer located country) Monoelisa HCV Ag/Ab ultra microplate EIA (BioRad, France) OraQuick (OraSure Technologies, PA USA); CORE (CORE Diagnostics, United Kingdom); Axiom (Axiom Diagnostics, Burstadt,Germany ); FirstVue (AT First Diagnostic, Woodbury,NY, USA) and Instant View Cassette (Alfa Scientific Designs, Poway) AxSYM (Abbott laboratories, other information is not available)

Study type Cross-sectional

Sample size 62

POC (Y/N) N

Reference standard PCR

Njouom et al.

2006

Cameroon

Plasma

Cross-sectional

329

EIA 3rd generation, PCR

Nyirenda et al.

2008

Malawi

Serum

Cross-sectional

202

EIA

O'Connell et al.

2013

USA

Plasma, whole blood (normal) and whole blood (cold storge)

Case–control

674 or 168

EIA, Immunoblot, and when available viral load)

O'Flynn et al.

1997

Ireland, Germany, UK

Plasma and serum

Case–control

5554, 1421 or 643

ABBOTT MATRIX HCV, Chiron Immunoblot HCV 2.0 or 3.0 Immunoblot HCV 3.0 and Cobas Ampliprep/ Taqman HCV RNA EIA 2nd generation or Immunoblot Immunoblot 2.0

Park et al.

2012

Korea

Serum

Vitros anti-HCV assay kits (Ortho-Clinical Diagnostics, Buckinghamshire, UK) and Elecsys (Roche Diagnostics GmbHMannheim, Germany)

Cross-sectional

1008

Poovorawari et al. 1994

Thailand

Serum

HCV-SPOT assay (Genelabs Diagnostics Pty Ltd, Singapore) C100-3 HCV EIA (Abbott laboratories, other information is not available) MATRIX hcv2 (Abbott laboratories, other information is not available)

Cross-sectional

192

Prayson et al.

1993

USA

Serum

Cross-sectional

123

Rihn et al.

2000

France

Serum

Cross-sectional

146

PCR

Page | 275

First author Scalioni Lde et al.

Year 2014

Settings Brazil

Sample type Serum, whole blood and oral fluid

Manufacturer WAMA Imuno-Rápido HCV Kit (WAMA Diagnóstica, Brazil); Bioeasy HCV Rapid Test, (Bioeasy Diagnóstica Ltd, Brazil) and OraQuick (OraSure Technologies, PA USA) Multiplo Rapid HIV/HCV Antibody Test (MedMira, Canada); Chembio DPP HCV test (Chembio Diagnostic Systems, USA) and OraQuick (OraSure Technologies, PA USA) Multiplo Rapid HIV/HCV Antibody Test (MedMira, Canada); Chembio DPP HCV test (Chembio Diagnostic Systems, USA)

Study type Cross-sectional

Sample size 194 or 172

POC (Y/N) Y

Reference standard PCR

Smith et al.

2011

USA

Whole blood, oral fluid

Cross-sectional

476, 385, 432, 549 or 266

MEIA/EIA/CLIA, Immunoblot

Smith et al.

2011

USA

Oral fluid and blood

Cross-sectional

1081

Chiron Immunoblot HCV 3.0 SIA; Bayer Advia Centaur HCV Chemiluminescent immunoassay INNO-LIA (Innogenetics, Ghent, Belgium), NAT EIA

Sommese et al.

2014

Italy

Blood

CMIA assays (Abbott Diagnostics, Wiesbaden, Germany)

Cross-sectional

17894

Tagny et al.

2014

Cameron

Plasma

HCV Ag/Ab combination assay (Monolisa HCV AgAb Ultra, BioRad, Marnes La Coquette, France) Abbott HCV EIA 3.0 (Abbott laboratories, Murex anti-HCV VK47 (Murex Diagnostic) and Ortho HCV 3.0 elisa (Ortho Diagnostic Systems; other information is not avaliable Monolisa anti-HCV new antigens (Sanofi Diagnostics Pasteur), Abbott HCV EIA 3.0 (Abbott laboratories); other information is not available

Cross-sectional

1998

Vrielink et al.

1996

Netherlands

Blood

Cross-sectional

403, 212, 253 03 1055

PCR

Vrielink et al.

1995

Netherlands

Blood

Cross-sectional

403, 212, 253

PCR

Page | 276

First author Yang et al.

Year 2011

Settings China

Sample type Serum

Manufacturer AxSYM HCV 3.0 (Abbott Laboratories), Murex Ag/Ab test (Abbott Laboratories); other information is not available Elecsys anti-HCV II (Roche Diagnostics GmbH), Architect anti-HCV (Abbott) and Vitros antiHCV(Ortho-Clinical Diagnostics), other information is not available ImmunoComb II HCV (Inverness Medical Innovations, USA) Elecsys Anti-HCV II assay (Roche Diagnostics GmbH, other information is not avaliable) ®

Study type Case–control

Sample size 101 or 100

POC (Y/N) N

Reference standard HCV RNA test (COBAS AMPLICOR Hepatitis C Virus Test, version 2.0 IMMUNOBLOT 3.0 test or the Realtime HCV RNA assay

Yang et al.

2013

China

Serum

Cross-sectional

859 or 167

Yarri et al.

2006

Israel

Serum and oral fluid Serum

Cross-sectional

PCR

Yoo et al.

2015

South Korea; China; China/Taiwa n; Thailand; Australia; Malaysia; Indonesia

Cross-sectional

7726

1 or more of the following comparator assays at 9 centers: ARCHITECTTM ® Anti-HCV; Serodia -HCV Particle Agglutination; ® Vitros ECi Anti® HCV; Elecsys Anti-HCV; ® ADVIA Centaur HCV; ® ® InTec HCV EIA; or Livzon Anti-HCV. EIA, PCR

Yuen et al.

2001

China

Serum

SM-HCV Rapid Test (SERO-Med Laborspezialita¨ten GmbH, Eichsta ¨tt, Germany)

Case–control

290

Page | 277

Assessment of the quality of the studies All studies used cross-sectional or case–control design. Risk of bias in patient selection, index test, or reference standard was assessed using QUADAS-2 (Table 2). Among the included studies, 21 have at least one category that was considered high risk. Risk of bias in patient selection usually came from a poor description of patient selection and clinical scenario. Bias in the index test was primarily due to a lack of reported blinding while reading test results. Bias in the reference standard was due to the use of multiple reference standards (EIA, NAT, and/or immunoblot). Bias in the flow and timing was primarily due to a lack of reported details. Table 2. Quality assessment by QUADAS-2 of the included studies Reports Patient selection Al-Tahish et al. Bonacini et al. Buti et al. Caudai et al. Cha et al. Croom et al. da Rosa et al. Daniel et al. Denoyel et al. Drobnik et al. Eroglu et al. Feucht et al. Gao et al. Hess et al. Hui et al. Ivantes et al. Jewett et al. Dokuboa et al. Kant et al. Kaur et al. Kim et al. Kosack et al. Lakshmi et al. Larrat et al. Lee et al. 2013 2001 2000 1998 2013 2006 2013 2005 2004 2011 2000 1995 2014 2014 2002 2010 2012 2014 2012 2000 2013 2014 2007 2012 2010 UC HR UC HR HR LR HR LR UC LR LR HR LR LR HR LR LR UC HR LR UC HR UC LR LR Bias assessment/Risk of bias Index test LR LR UC LR LR LR UC LR LR UC LR LR LR HR LR UC LR LR UC UC LR LR LR LR UC Reference standard LR LR LR LR LR LR LR LR LR LR LR LR LR LR HR HR LR LR HR HR LR LR LR LR LR Flow and timing LR LR LR LR LR UC LR LR LR UC LR LR HR LR LR LR LR LR LR LR LR LR UC LR LR Acceptability concerns Patient selection LR LR HR UC UC LR HR LR UC LR LR HR LR LR HR LR LR UC HR LR UC HR HR LR LR Index test LR LR LR LR LR LR UC LR LR UC LR LR LR HR LR LR LR LR UC LR LR LR LR LR UC Reference standard LR LR LR LR LR LR LR LR HR LR LR LR LR LR HR HR LR LR HR LR LR LR LR LR LR

Page | 278

Lee et al. Maity et al. Montebugnoil et al. Mvere et al. Nalpas et al. Njouom et al. Nyirenda et al. O'Connell et al. O'Flynn et al. Park et al. Poovorawari et al. Prayson et al. Rihn et al. Scalioni et al. Smith et al. Smith et al. Sommese et al. Lee et al. Ibrahim et al. Tagny et al. Vrielink et al. Vrielink et al. Yang et al. Yang et al. Yarri et al. Yoo Yuen et al.

2011 2012 1999 1996 1992 2006 2008 2013 1997 2012 1994 1993 2000 2014 2011 2011 2014

HR HR HR HR HR HR LR HR UC UC LR UC UC UC LR HR LR

UC UC LR LR LR UC UC LR LR LR UC LR LR LR LR LR LR LR LR UC LR LR LR LR LR LR LR

LR HR LR LR LR LR LR HR LR LR LR LR LR LR LR LR LR LR LR HR LR LR LR LR LR LR LR

LR LR LR LR UC LR LR LR UC UC LR UC UC UC LR LR LR LR LR LR LR LR LR UC LR HR LR

LR HR HR HR HR HR LR HR LR LR LR UC UC UC LR HR LR LR HR LR UC HR UC LR HR UC HR

LR UC LR LR LR UC LR LR LR LR LR LR LR LR LR LR LR LR LR UC LR LR LR LR LR LR LR

LR HR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR LR HR LR LR LR LR LR LR LR

2010_2 LR 2015 2014 1995 HR LR UC

1995_2 UC 2011 2013 2006 2015 2001 UC LR HR UC HR

LR: low risk; HR: high risk; UC: unclear risk

Diagnostic accuracy Overall clinical performance of assays The 52 included studies contributed 127 data points from 52 273 unique test measurements. Some studies contributed additional data points by comparing the accuracy of two or more tests, reporting data from multiple study sites, or reporting the accuracy of a test in more than one type of specimen. The sample sizes of the included studies ranged from 37 to 17 894. Sensitivities of included studies ranged from 0.22 to 1.00, and specificities ranged from 0.77 to 1.00. The overall pooled sensitivity and specificity for all tests were 0.97 (95% CI: 0.97 –0.98) and 0.99 (95% CI: 0.98– Page | 279

0.99), respectively. Figure 2 show estimates of sensitivity and specificity from each study. Fig. 2. Sensitivity and specificity of HCV Ab tests included in the review (n=52)

Manufacturers and accuracy of RDTs among included studies Overall, 32 studies evaluated the accuracy of 30 different RDTs (Table 3). The most commonly evaluated test kit was the OraQuick ADVANCE® from OraSure Technologies.

Page | 280

An Ag–Ab test by BioRad and a combo HIV-HCV test by MedMira were among the tests evaluated.

Page | 281

Table 3. Manufacturers and accuracy of RDTs among included studies First author Manufacturer Sample size 100 674 168 168 194 194 194 407 400 476 385 1081 168 168 674 300 168 168 674 TP FP TN FN SE SP

Montbugnoil et al. O'Connell. et al. O'Connell et al. O'Connell et al. Scalioni Lde, et al. Scalioni Lde et al. Scalioni Lde et al. Jewett et al. Jewett et al. Smith et al. Smith et al. Smith et al. O'Connell et al. O'Connell, et al. O'Connell et al. Maity et al. O'Connell et al. O'Connell et al. O'Connell et al.

Anti-HCV Ab rapid test (1st IRP 75/537 by Thema Ricerca, WHO Geneva) Axiom (Axiom Diagnostics, Burstadt, Germany ) Axiom (Axiom Diagnostics, Burstadt, Germany ) Axiom (Axiom Diagnostics, Burstadt, Germany ) Bioeasy HCV Rapid Test, (Bioeasy Diagnóstica Ltd, Brazil) Bioeasy HCV Rapid Test (Bioeasy Diagnóstica Ltd, Brazil) Bioeasy HCV Rapid Test (Bioeasy Diagnóstica Ltd, Brazil) Chembio DPP HCV Test (Chembio Diagnostic Systems, USA) Chembio DPP HCV test (Chembio Diagnostic Systems,USA) Chembio DPP HCV test (Chembio Diagnostic Systems, USA) Chembio DPP HCV test (Chembio Diagnostic Systems, USA) Chembio DPP HCV test (Chembio Diagnostic Systems, USA) CORE (CORE Diagnostics, United Kingdom) CORE (CORE Diagnostics, United Kingdom) CORE (CORE Diagnostics, United Kingdom) Diagnostics Ltd (other information is not available) FirstVue (AT First Diagnostic, Woodbury, NY, USA) FirstVue (AT First Diagnostic, Woodbury, NY, USA) FirstVue (AT First Diagnostic, Woodbury, NY, USA)

50 326 77 82 137 111 136 101 88 308 264 525 29 24 323 132 66 54 312

1 10 2 5 0 0 0 3 3 12 3 1 1 2 7 0 0 1 3

49 329 82 79 48 48 48 290 294 125 101 543 83 82 332 168 84 83 336

0 9 7 2 9 35 10 8 15 32 17 12 55 60 12 0 18 30 23

1.00 0.97 0.92 0.98 0.94 0.76 0.93 0.93 0.85 0.91 0.94 0.98 0.35 0.29 0.96 1.00 0.79 0.64 0.93

0.98 0.97 0.98 0.94 1.00 1.00 1.00 0.99 0.99 0.91 0.97 1.00 0.99 0.98 0.98 1.00 1.00 0.99 0.99

Page | 282

Al-Tahish et al. Daniel et al. Kim, M. H. et al. Kaur et al. Al-Tahish et al. Ivantes et al. da Rosa et al. Poovoran et al. Mvere et al. Njouom et al. Al-Tahish et al. Yarri et al. Yarri et al. Njouom et al.

Fourth-generation HCV TRI_DOT (J. Mitra Co, India) Fourth-generation HCV TRI_DOT (J. Mitra Co, India) GENEDIA HCV Rapid LF (Green Cross medical science corp., Korea) HCV Bidot (J. Mitra Co., India) HCV one step test device (ACON Laboratories, USA) HCV Rapid Test Bioeasy (Bioeasy Diagnostica Ltd, Brazil) HCV Rapid Test Bioeasy (Standard Diagnostics, South Korea) HCV-SPOT assay (Genelabs Diagnostics Pty Ltd, Singapore) HCV-SPOT assay (Genelabs Diagnostics Pty Ltd, Singapore) Hexagon HCV (Not reported manufacturer located country) ImmunoComb II HCV (Inverness Medical Innovations, USA) ImmunoComb II HCV (Inverness Medical Innovations, USA) ImmunoComb II HCV (Inverness Medical Innovations, USA) ImmunoComb II HCV assay (Orgenics Ltd, Not reported manufacturer located country) Imuno-Rapido HCV (Wama Diagnostica, Brazil). Instant View Cassette (Alfa Scientific Designs, Poway, CA, USA) ® ® ® ® ®

100 2590 100 2754 100 71 307 192 206 329 100 37 37 329

34 138 52 28 34 30 100 41 10 160 34 18 18 103

15 24 0 0 15 3 0 11 4 17 14 4 1 0

50 2427 34 2722 50 38 204 139 191 151 51 15 18 168

1 1 14 4 1 0 3 1 1 1 1 0 0 58

0.97 0.99 0.79 0.88 0.97 1.00 0.97 0.98 0.91 0.99 0.97 1.00 1.00 0.64

0.77 0.99 1.00 1.00 0.77 0.93 1.00 0.93 0.98 0.90 0.78 0.79 0.95 1.00

da Rosa et al. O'Connell et al.

307 674

100 321

0 3

204 336

3 14

0.97 0.96

1.00 0.99

O'Connell et al.

Instant View Cassette (Alfa Scientific Designs, Poway, CA, USA)

168

0.81

0.96

O'Connell et al.

Instant View Cassette (Alfa Scientific Designs, Poway, CA, USA)

168

0.55

0.99

Page | 283

Maity et al. Jewett et al. Nyirenda et al. Tagny et al. Smith et al. Smith et al. Cha et al. Lee et al. Lee et al. Lee et al. Lee et al. Lee et al. O'Connell et al. O'Connell et al. O'Connell et al. Smith et al. Smith et al. Lee et al. Lee et al. Lee et al. Lee et al. Lee et al.

J Mitra Co. India (other information is not available) Rapid HIV/HCV antibody test (Medmira Laboratories, Canada) Monoelisa HCV Ag/Ab Ultra microplate EIA (Bio-Rad, France) Monolisa HCV Ag-Ab Ultra, (BioRad, France) Multiplo Rapid HIV/HCV Antibody Test (MedMira, Canada) Multiplo Rapid HIV/HCV Antibody Test (MedMira, Canada) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA)

300 374 202 1998 1081 432 437 2183 2183 2183 2183 2183 674 168 168 549 266 572 572 572 572 572

120 80 2 26 474 303 134 756 755 753 752 739 333 83 82 375 188 122 123 123 123 123

0 0 7 28 1 8 0 1 2 2 1 5 1 1 0 8 1 0 0 0 1 1

174 274 186 1929 543 40 300 1422 1420 1421 1421 1418 338 83 84 140 72 449 449 449 448 448

6 20 7 15 63 81 3 1 1 2 2 14 2 1 2 26 5 1 0 0 0 0

0.95 0.80 0.22 0.63 0.88 0.79 0.98 1.00 1.00 1.00 1.00 0.98 0.99 0.99 0.98 0.94 0.97 0.99 1.00 1.00 1.00 1.00

1.00 1.00 0.96 0.99 1.00 0.83 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.99 1.00 0.95 0.99 1.00 1.00 1.00 1.00 1.00

Page | 284

Smith et al. Drobnik et al. Lee et al. Lee et al. Lee et al. Lee et al. Lee et al. Gao et al. Ibrahim Scalioni Lde_2014 Hess et al. Buti et al. Yuen et al. Maity et al. Kant et al. Kosack et al. Scalioni Lde et al. Scalioni Lde et al. Hui et al.

OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) OraQuick (OraSure Technologies, PA USA) DPP HIV-HCV-Syphilis Assay (Chembio Diagnostic Systems, Inc. , Medford, NY). Not available SM-HCV Rapid Test (SERO-Med Laborspezialita¨ten GmbH, Eichsta ¨tt, Germany) SPAN Diagnostics, Indi, other information is not available Toyo anti-HCV test (Turklab, Izmir, Turkey) The ImmunoFlow HCV test (Core Diagnostics,United Kingdom) WAMA Imuno-Rápido HCV Kit (WAMA Diagnóstica, Brazil) WAMA Imuno-Rápido HCV Kit (WAMA Diagnóstica, Brazil) Not reported

1081 484 2180 2178 2178 2176 2176 1156 160 172 948 188 290 300 185 82 194 194 197

533 92 756 755 753 752 739 16 53 108 152 135 98 132 82 55 119 134 91

3 3 1 2 2 1 5 6 0 0 6 0 0 0 12 0 3 3 0

541 382 1422 1420 1421 1421 1418 1133 100 50 776 50 189 168 90 26 45 45 88

4 7 1 1 2 2 14 1 7 14 14 3 3 0 1 0 27 12 18

0.99 0.93 1.00 1.00 1.00 1.00 0.98 0.94 0.88 0.89 0.92 0.98 0.97 1.00 0.99 1.00 0.82 0.92 0.83

0.99 0.99 1.00 1.00 1.00 1.00 1.00 0.99 1.00 1.00 0.99 1.00 1.00 1.00 0.88 1.00 0.94 0.94 1.00

Page | 285

Pooled test accuracy for RDT versus EIA alone Overall, five studies evaluated RDT compared to the EIA alone, with a total sample of 15 943. Of the five studies, sample sizes ranged from 197 to 2754, sensitivities ranged from 0.83 to 1.00, and specificities ranged from 0.99 to 1.00. The pooled sensitivity and specificity were 0.98 (95% CI 0.98– 1.00) and 1.00 (95% CI 1.00–1.00), respectively, while heterogeneity was observed between the included studies (P<0.001) (Fig. 3, Table 4). For the three studies that were conducted within past 10 years,8, 31, 42 the total sample size was 12 992, with pooled sensitivity and specificity of 0.99 (95% CI 0.99–1.00) and 1.00 (95% CI 1.00– 1.00), respectively. Fig. 3. Pooled test accuracy of HCV Ab RDTs compared to an EIA reference (5 studies)

RDT accuracy compared to NAT or immunoblot Overall, 13 studies evaluated RDTs compared to NAT or immunoblot, with a total sample of 6 683. Among these studies, sample sizes ranged from 36 to 549, sensitivities ranged from 0.76 to 1.00, and specificities ranged from 0.77 to 1.00. The pooled sensitivity and specificity were 0.93 (95% CI 0.91– 0.95) and 0.98 (95% CI 0.97–0.99), respectively, while heterogeneity was observed between the included studies (P<0.001) (Fig. 5, Table 4).

Page | 286

Fig. 5. Pooled test accuracy of HCV Ab RDTs compared to a NAT or immunoblot reference (n=13 studies)

RDT test accuracy compared to EIA, NAT or Immunoblot Overall, 14 studies evaluated RDT by referencing to EIA with NAT and/or immunoblot, with a total sample of 42 212. Of the 14 studies, sample sizes ranged from 168 to 2754, sensitivities ranged from 0.29 to 1.00, and specificities ranged from 0.90 to 1.00. The pooled sensitivity and specificity were 0.97 (95% CI 0.96–0.98) and 1.00 (95% CI 1.00–1.00), respectively, while heterogeneity was observed between studies (P<0.001) (Fig. 4, Table 4).

Page | 287

Fig. 4. Pooled test accuracy of HCV Ab RDTs compared to EIA, NAT or immunoblot reference standards (n=14 studies)

Pooled test accuracy for oral versus blood samples EIAs using oral fluid samples Overall, 12 studies compared the accuracy of EIAs using oral fluid samples to a blood sample as a reference, with a total sample size of 14 546. Of the 12 studies, sample sizes ranged from 37 to 2176, sensitivities ranged from 0.72 to 1.00, and specificities ranged from 0.91 to 1.00. The pooled sensitivity and specificity were 0.94 (95% CI 0.93–0.96) and 1.00 (95% CI 1.00–1.00), respectively. Heterogeneity was observed between the included studies (P<0.001) (Fig. 6, Table 4).

Page | 288

Fig. 6. Pooled test accuracy for oral HCV Ab RDTs compared to blood as a reference (n=12 studies)

Blood samples Overall, 45 studies used blood samples for evaluations, with a total sample of 89 608. Sample sizes ranged from 37 to 17 894, sensitivities ranged from 0.29 to 1.00, and specificities ranged from 0.18 to 1.00. The pooled sensitivity and specificity were 0.98 (95% CI 0.97–0.98) and 0.98 (95% CI 0.98– 0.98), respectively. Heterogeneity was observed between the included studies (P<0.001) (Fig. 7, Table 4).

Page | 289

Fig. 7. Pooled HCV Ab test accuracy for blood samples (n = 45 studies)

Page | 290

Pooled test accuracy for OraQuick versus other bands on Oral kits OraQuick Overall, eight studies reported sensitivity and specificity of OraQuick (OraSure Technologies, PA, USA), with a total sample of 9024. The sample size of these studies ranged from 172 to 2183, sensitivities ranged from 0.90 to 1.00, and specificities ranged from 0.95 to 1.00. The pooled sensitivity and specificity were 0.98 (95% CI 0.97–0.99) and 1.00 (95% CI 0.90–1.00), respectively. Heterogeneity was observed between the included studies (P<0.001) (Fig. 7, Table 4). Fig. 8. Pooled test accuracy for HCV Ab OraQuick kits (n = 8 studies)

Overall, six studies reported sensitivity and specificity for other three brands of oral kits, with a total sample of 6652. The sample size of these studies ranged from 37 to 1081, sensitivities ranged from 0.72 to 1.00, and specificities ranged from 0.91 to 1.00. The pooled sensitivity and specificity were 0.88 (95% CI 0.84–0.92) and 0.99 (95% CI 0.99–1.00), respectively, while heterogeneity was observed between the included studies (P<0.001) (Fig. 8, Table 4). Figure 9. Pooled test accuracy for other brands of oral HCV Ab test kits (n = 6 studies)

Other findings Our study further found that the overall sensitivity and specificity of studies conducted among general populations were 0.95 (95% CI 0.94–0.96) and 0.99 (95% CI 0.98–0.99), among key populations were 0.97 (95% CI 0.96–0.98) and 0.94 (95% CI 0.94–0.95), and among hospital patients were 0.97 (95% CI 0.96–0.98) and 1.00 (95% CI 1.00–1.00), respectively. Page | 291

Table 4. Pooled test accuracy for different testing strategies (n = 52 studies) Comparison Pooled SE 95%CI Tau-square P-value for heterogeneity 1.00 0.95 <0.001 <0.001 Pooled SP 95% CI Tau-square P-value for heterogeneity

RDT versus EIA only (n = 5) RDT versus NAT or Immunoblot (n = 13) RDT versus EIA, NAT or Immunoblot (n = 14) Oral RDT versus blood reference (n = 12) Sample type Blood samples (n = 45) Oral samples (n = 12) Source population General screening (n = 17) Key population (n = 19) Hospital patients (n = 16) Antibody and antigen combo testing (n = 6) Oral kits brand OraQuick (n = 8) Other brands (n = 6)

0.99 0.93

0.98 0.91

1.00 0.98

1.00 0.97

1.00 0.99

<0.001 <0.001

0.97

0.96

0.98

<0.001

1.00

<0.001

0.94

0.93

0.96

<0.001

1.00

<0.001

0.98 0.94

0.97 0.93

0.98 0.96

<0.001 <0.001

0.98 1.00

0.99 1.00 <0.001

0.95 0.97 0.97 0.86

0.94 0.96 0.96 0.79

0.96 0.98 0.98 0.94

<0.001 <0.001 <0.001 <0.001

0.99 0.94 1.00 0.99

0.98 0.94 1.00 0.98

0.99 0.95 1.00 1.00

<0.001 <0.001 <0.001 <0.001

0.98 0.88

0.97 0.84

0.99 0.92

<0.001 <0.001

1.00 0.99

1.00 1.00

<0.001 <0.001

* Studies conducted in both LMICs and high-income countries were not included here #

Studies conducted across these regions were not included here.

SE: sensitivity; SP: specificity

GRADE GRADE for RDT versus EIA HCV Ab RDTs showed comparable sensitivity and specificity compared to that of EIAs. For the 5 studies evaluated RDT versus EIA, 15 943 of samples were evaluated, and moderate risk of bias was observed, and precision was present, while inconsistency was present for sensitivity, as the sensitivities of the included tests varied. But the consistency of specificity is observed. Since the unit of the analysis varied among studies (Table 4), indirectness was observed. In addition, the overall Page | 292

strength of the pooled evaluation was moderate, with pooled sensitivity and specificity of 0.99 (95% CI 0.98–1.00) and 1.00 (95% CI 1.00–1.00), respectively. Under the pre-test probability of 5%, the post-test probability after a positive test result is 97%, and the post-test probability after a negative test result is 100%. GRADE for oral RDT versus blood reference The use of oral RDT HCV Ab had comparable sensitivity and specificity compared to blood reference standards (Table 4). For the 12 studies evaluated oral RDT versus blood reference, 14 547 samples were evaluated. A moderate risk of bias was observed. Inconsistency present for sensitivity, as the sensitivities of the included studies varied. But the consistency of specificity was observed. Since the unit of the analysis varied with each other among the included studies (Table 4), indirectness was observed for included studies. In addition, the overall strength of the pooled evaluation was moderate, with pooled sensitivity and specificity of 0.94 (95% CI 0.93–0.96) and 1.00 (95% CI 1.00– 1.00), respectively. Assuming a pre-test probability of 5%, the post-test probability after a positive test result was 94%, and the post-test probability after a negative test result was 100%. Table 5. GRADE table RDT versus EIA Unit of analysis Hospital patients, blood donors, injection drug users and other high-risk populations Oral RDT versus blood General population, hospital patients, blood donors, injection drug users and other highrisk population

Sample type Studies, n Risk of bias Consistency Directness/Precision # of samples Strength of evidence Sensitivity (95% CI) Specificity (95%CI) Pretest probability (%) Positive LR (95% CI) PPV Negative LR (95% CI)

Oral fluid serum or plasma 5 Moderate Se: Inconsistent Sp: Inconsistent Indirect/ Precise 15 943 Se: Moderate Sp: Moderate 0.99(0.98–1.00) 1.00 (1.00–1.00) 0.05 618.5 (350.6–2493.2) 0.97 0.01 (0.002–0.02) 314.5 (202.0–684.1) 0.94 0.06 (0.04–0.07) 0.94 (0.93–0.96) 14 547 12

Page | 293

NPV

Page | 294

6. Discussion Acute HCV infection is usually asymptomatic, therefore there is an urgent need to increase screening for individuals who may be at increased risk. In this meta-analysis, we found HCV Ab RDTs, including those using oral fluid, showed a high overall sensitivity and specificity compared to laboratory-based EIAs. This extends the literature by including several new studies that were not included in prior reviews, including a subanalysis focused on RDTs that used oral fluid. Our data suggest that RDTs can be used for HCV Ab detection in a wide range of clinical settings. High HCV Ab RDT sensitivity and specificity were observed across several different populations (general population, key populations, hospital patients). The use of an EIA to detect HCV Ab followed by NAT to confirm active infection is standard practice for diagnosis of HCV infection and recommended by the US Centers for Disease Control and Prevention (CDC) and the WHO.70, 71 However, despite these recommendations, HCV Ab assays have not been widely used because of the complexity of laboratory-based assays, long turnaround time, high cost and requirements for specialized apparatus and trained technicians. To overcome this barrier, companies have developed RDTs for HCV Ab screening.72 They obviate the need for multiple follow-up appointments, shorten wait times, and allow for the simplification and decentralization of testing. However, it is essential for policy-makers, government officials, and health care practitioners engaged in HCV screening, care and treatment to know that the performance of individual RDTs for detection of HCV Ab vary widely. Individual diagnostic accuracy for specific brands should be examined to ensure acceptable performance. In recent years, RDTs that used oral fluid has been developed. Tests that can be used with non-invasive samples allow testing to be decentralised further and can be used in outreach settings.73 Our data suggest that oral tests have slightly lower pooled sensitivity (0.94, 95% CI: 0.93–0.96) compared to blood-based tests (0.98, 95% CI: 0.97–0.98) but comparable specificity. Oral HCV Ab RDTs tests may be particularly useful in contexts where venepuncture may be difficult, such as subsets of people who inject drugs who have difficult veins to access. With the increasing availability of DAAs, countries are seeking information on diagnostic accuracy of different tests, their operational characteristics and cost, to allow them to scale up HCV Ab screening, especially of at risk populations. Deploying which tests at which level of the health care system and for what settings require policy-makers to consider the different attributes of laboratory-based EIA versus blood-based or oral RDTs. Advantages and disadvantages of EIAs and RDTs are listed in Table 6. Performance, cost and accessibility need to be considered. Each country needs to decide on what cannot be compromised and what trade-offs are acceptable, based not only on disease prevalence and the health-care infrastructure, but also on technical, socioeconomic, cultural, behavioural considerations. For example, is it acceptable to buy Test X, which is 10% less accurate than Test Y but is considerably cheaper so that many more people can be screened? Although oral RDTs are less accurate than blood-based RDTs, would their use be more acceptable for outreach to at-risk Page | 295

populations and allow the control programme to identify more HCV cases? In a lowprevalence setting, even a test with 98% specificity can yield more false-positive than truepositive results. All these trade-offs can be modelled to give an estimate of the cost– effectiveness and potential impact of different strategies for HCV Ab screening. Table 6. Advantages and disadvantages of laboratory-based EIAs vs RDTs Laboratory-based EIA Advantages     Accurate High throughput Objective, automated reading of results Within-assay quality control   Disadvantages     Requires laboratory facility, equipment and highly trained staff Reagents need refrigeration Need venepuncture to obtain sera Time to result = 3–4 hours so patients need to return for results       RDTs Accessible to lowest level of the health care system (including outreach) Can be used with non-invasive specimens, and facilitate self-testing Rapid result to enable treatment initiation at the same clinic visit Can be stored at ambient temperature Lower accuracy than EIAs Subjective reading and interpretation of results No built-in quality control Higher cost/test

Our review also underlines some of the common methodological problems encountered in evaluating diagnostic accuracy. Cross-sectional method or case–control methods were used by all 52 included studies, introducing a potential risk of bias. These studies used a broad range of reference standards, which makes the pooled performance data less meaningful. Within the evaluation of diagnostic accuracy, even cross-sectional studies in patients with diagnostic uncertainty and direct comparison of test results with an appropriate reference standard can be considered high quality.74 The majority of the included studies used convenience sampling. In this review, we excluded panel studies because there are not based on clinical settings and our purpose was to produce data that would be relevant in clinical settings as part of detection of HCV Ab. Most studies that reported HIV or HBV coinfection only reported the test performance of the kits among all samples, instead of disaggregated diagnostic accuracy. It may be useful for policy-makers and hospital administrators to know that the diagnostic accuracy of different kits among individuals with coinfections, particularly HIV coinfection.69 Further research is needed to understand HCV Ab test characteristics among coinfected individuals. Our study is subject to a number of limitations. First, we included studies conducted among the general population, hospital patients and key populations. HCV prevalence is variable among these different populations.75 Diagnostic performance can be influenced by disease prevalence.76 Second, we detected substantial heterogeneity that could influence our confidence in the review findings.77 To deal with this problem, we undertook a number of subanalyses. Third, about 20 brands of RDT kits were used in the included studies, and the performance of these RDT kits vary. This limited our ability to summarize the accuracy of Page | 296

different brands, with the exception of comparing OraQuick to other brands. Another concern is publication bias, as studies with poor test performance may be less likely to be published, leading to distorted estimates of accuracy.78

Page | 297

References 1. Hepatitis C (No 164) [Fact sheet]. Geneva: WHO; 2014. 2. Lavanchy D. The global burden of hepatitis C. Liver Int. 2009;29(s1):74–81. 3. Mohd Hanafiah K, Groeger J, Flaxman AD, Wiersma ST. Global epidemiology of hepatitis C virus infection: new estimates of age-specific antibody to HCV seroprevalence. Hepatology. 2013;57(4):1333–42. 4. Papatheodoridis G, Tsochatzis E, Hardke S, Wedemeyer H. Barriers to care and treatment for patients with chronic viral hepatitis in europe: a systematic review. Liver Int. 2014;34(10):1452‒63. 5. Feeney ER, Chung RT. Antiviral treatment of hepatitis C. BMJ. 2014; 348: g3308. 6. Pawlotsky JM. New hepatitis C therapies: the toolbox, strategies, and challenges. Gastroenterology. 2014;146(5):1176–92. 7. Soriano V, Labarga P, Fernández-Montero JV, Vispo E, Poveda E, Martin-Carbonero L,et al. The changing face of hepatitis C in the new era of direct-acting antivirals. Antivir Res. 2013;97(1):36–40. 8. Gao F, Talbot EA, Loring CH, Power JJ, Dionne-Odom J, Alroy-Preis S, et al. Performance of the OraQuick HCV rapid antibody test for screening exposed patients in a hepatitis C outbreak investigation. J Clin Microbiol. 2014;52(7):2650–2. 9. Zheng H, Li M, Chi B, Wu XX, Wang J, Liu DW. IL28B rs12980275 variant as a predictor of sustained virologic response to pegylated-interferon and immunoblotvirin in chronic hepatitis C patients: a systematic review and meta-analysis. Clin Res Hepatol Gastroenterol. 2015;39(5):576–83. 10. Manzano-Robleda Mdel C, Ornelas-Arroyo V, Barrientos-Gutierrez T, Mendez-Sanchez N, Uribe M, Chavez-Tapia NC. Boceprevir and telaprevir for chronic genotype 1 hepatitis C virus infection. A systematic review and meta-analysis. Ann Hepatol. 2015;14(1):46–57. 11. Ambrosino P, Lupoli R, Tarantino P, Di Minno A, Tarantino L, Di Minno MN. Viral hepatitis and anti-phospholipid antibodies positivity: A systematic review and meta-analysis. Dig Liver Dis. 2015;47(6):478–87. 12. Khodabandehloo M, Roshani D. Prevalence of hepatitis C virus genotypes in Iranian patients: a systematic review and meta-analysis. Hepat Mon. 2014;14(12):e22915. 13. Khodabandehloo M, Roshani D, Sayehmiri K. Prevalence and trend of hepatitis C virus infection among blood donors in Iran: a systematic review and meta-analysis. J Res Med Sci. 2013;18(8):674–82. 14. Khuroo MS, Khuroo NS, Khuroo MS. Diagnostic accuracy of point-of-care tests for hepatitis C virus infection: a systematic review and meta-analysis. PloS One. 2015;10(3):e0121450. 15. Shivkumar S, Peeling R, Jafari Y, Joseph L, Pant Pai N. Accuracy of rapid and point-of-care screening tests for hepatitis C: a systematic review and meta-analysis. Ann Intern Med. 2012;157(8):558–66. 16. Whiting PF, Rutjes AW, Westwood ME, Mallet S, Deeks JJ, Reitsma JB, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155(8):529–36. Page | 298

17. Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig LM, et al. The STARD statement for reporting studies of diagnostic accuracy: explanation and elaboration. Ann Intern Med. 2003;138(1):W1–12. 18. Al-Tahish G, El-Barrawy MA, Hashish MH, Heddaya Z. Effectiveness of three types of rapid tests for the detection of hepatitis C virus antibodies among blood donors in Alexandria, Egypt. J Virol Methods. 2013;189(2):370–4. 19. Bonacini M, Lin HJ, Hollinger FB. Effect of coexisting HIV-1 infection on the diagnosis and evaluation of hepatitis C virus. J Acquir Immune Defic Syndr. 2001;26(4):340–4. 20. Buti M, Cotrina M, Chan H, Jardi R, Rodriuez F, Costa X, et al. Rapid method for the detection of anti-HCV antibodies in patients with chronic hepatitis C. Rev Esp Enferm Dig. 2000;92(3):140–6. 21. Caudai C, Padula MG, Bastianoni I, Valensin PE, Shyamala V, Han J, et al. Antibody testing and RT-PCR results in hepatitis C virus (HCV) infection: HCV-RNA detection in PBMC of plasma viremia-negative HCV-Sseropositive persons. Infection. 1998;26(3):151–4. 22. Cha YJ, Park Q, Kang ES, Yoo BC, Park KU, Kim JW, et al. Performance evaluation of the OraQuick hepatitis C virus rapid antibody test. Ann Lab Med. 2013;33(3):184–9. 23. Croom HA, Richards KM, Best SJ, Francis BH, Johnson EIM, Dax EM, et al. Commercial enzyme immunoassay adapted for the detection of antibodies to hepatitis C virus in dried blood spots. J Clin Virol. 2006;36(1):68–71. 24. da Rosa L, Dantas-Correa EB, Narciso-Schiavon JL, Schiavon LL. Diagnostic performance of two point-of-care tests for anti-HCV detection. Hepat Mon. 2013;13(9):e12274. 25. Daniel HD, Abraham P, Raghuraman S, Vivekanandan P, Subramaniam T, Sridharan G. Evaluation of a rapid assay as an alternative to conventional enzyme immunoassays for detection of hepatitis C virus-specific antibodies. J Clin Microbiol. 2005;43(4):1977–8. 26. Denoyel G, van Helden J, Bauer R, Preisel-Simmons B. Performance of a new hepatitis C assay on the Bayer ADVIA Centaur Immunoassay System. Clin Lab. 2004;50(1–2): 75–82. 27. Dokubo EK, Evans J, Winkelman V, Cyrus S, Tobler LH, Asher A, et al. Comparison of hepatitis C virus RNA and antibody detection in dried blood spots and plasma specimens. J Clin Virol. 2014;59(4):223–7. 28. Drobnik A, Judd C, Banach D, Egger J, Konty K, Rude E. Public health implications of rapid hepatitis C screening with an oral swab for community-based organizations serving highrisk populations. Am J Public Health. 2011;101(11): 2151–5. 29. Eroglu C, Yildiz E, Ozturk M, Pinarbasi E. A highly sensitive and specific enzyme-linked immunosorbent assay of antibodies to hepatitis C virus. Acta Virol. 2000;44(1):29–33. 30. Feucht HH, Zollner B, Polywka S, Laufs R. Study on reliability of commercially available hepatitis C virus antibody tests. J Clin Microbiol. 1995;33(3):620–4. 31. Hess KL, Fisher DG, Reynolds GL. Sensitivity and specificity of point-of-care rapid combination syphilis-HIV-HCV tests. PloS One. 2014;9(11):e112190. 32. Hui AY, Chan FK, Chan PK, Tam JS, Sung JJ. Evaluation of a new rapid whole-blood serological test for hepatitis c virus. Acta Virol. 2002;46(1):47–8.

Page | 299

33. Ibrahim S, Al Attas SA, Mansour GA, Ouda S, Fallatah H. Accuracy of rapid oral HCV diagnostic test among a Saudi sample. Clin Oral Investig. 2015;19(2):475–80. 34. Ivantes CA, Silva D, Messias-Reason I. High prevalence of hepatitis C associated with familial history of hepatitis in a small town of south Brazil: efficiency of the rapid test for epidemiological survey. Braz J Infect Dis. 2010;14(5):483–8. 35. Jewett A, Smith BD, Garfein RS, Cuevas-Mota J, Teshale EH, Weinbaum CM. Field-based performance of three pre-market rapid hepatitis C virus antibody assays in STAHR (Study to Assess Hepatitis C Risk) among young adults who inject drugs in San Diego, CA. J Clin Virol. 2012;54(3):213–7. 36. Kant J, Moller B, Heyne R, Herber A, Bohm S, Maier M, et al. Evaluation of a rapid on-site anti-HCV test as a screening tool for hepatitis C virus infection. Eur J Gastroenterol Hepatol. 2013;25(4):416–20. 37. Kaur H, Dhanao J, Oberoi A. Evaluation of rapid kits for detection of HIV, HBsAg and HCV infections. Indian J Med Sci. 2000;54(10):432–4. 38. Kim MH, Kang SY, Lee WI. Evaluation of a new rapid test kit to detect hepatitis C virus infection. J Virol Methods. 2013;193(2):379–82. 39. Kosack CS, Nick S, Shanks L. Diagnostic accuracy evaluation of the ImmunoFlow HCV rapid immunochromatographic test for the detection of hepatitis C antibodies. J Virol Methods. 2014;204:6–10. 40. Lakshmi V, Reddy AK, Dakshinamurty KV. Evaluation of commercially available thirdgeneration anti-hepatitis C virus enzyme-linked immunosorbent assay in patients on haemodialysis. Indian J Med Microbiol. 2007;25(2):140–2. 41. Larrat S, Bourdon C, Baccard M, Garnaud C, Mathieu S, Quesada JL, et al. Performance of an antigen-antibody combined assay for hepatitis C virus testing without venipuncture. J Clin Virol. 2012;55(3):220–5. 42. Lee SR, Kardos K, Yearwood G, Kurtz L, Roehler M, Feiss G. Results of a multi-center evaluation of a new rapid test for detection of HCV infection using whole blood, serum, plasma and oral fluid. J Hepatol. 2010;52:S271. 43. Lee SR, Kardos KW, Schiff E, Berne CA, Mounzer K, Banks AT, et al. Evaluation of a new, rapid test for detecting HCV infection, suitable for use with blood or oral fluid. J Virol Methods. 2011;172(1–2):27–31. 44. Lee SR, Yearwood GD, Guillon GB, Kurtz LA, Fischl M, Friel T, et al. Evaluation of a rapid, point-of-care test device for the diagnosis of hepatitis C infection. J Clin Virol. 2010;48(1):15–17. 45. Maity S, Nandi S, Biswas S, Sadhukhan SK, Saha MK. Performance and diagnostic usefulness of commercially available enzyme linked immunosorbent assay and rapid kits for detection of HIV, HBV and HCV in India. Virol J. 2012;9:290. 46. Montebugnoli L, Borea G, Miniero R, Sprovieri G. A rapid test for the visual detection of anti-hepatitis C virus antibodies in whole blood. Clin Chim Acta. 1999;288(1–2):91-6.

Page | 300

47. Mvere D, Constantine NT, Katsawde E, Tobaiwa O, Dambire S, Corcoran P. Rapid and simple hepatitis assays: encouraging results from a blood donor population in Zimbabwe. Bull World Health Organ. 1996;74(1):19–24. 48. Nalpas B, Thiers V, Pol S, Driss F, Thepot V, Berthelot P, et al. Hepatitis-C viremia and antiHCV antibodies in alcoholics. J Hepatol. 1992;14(2–3):381–4. 49. Njouom R, Tejiokem MC, Zanga MC, Pouillot R, Ayouba A, Pasquier C, et al. A costeffective algorithm for the diagnosis of Hepatitis C virus infection and prediction of HCV viremia in Cameroon. J Virol Methods. 2006;133(2):223–6. 50. Nyirenda M, Beadsworth MB, Stephany P, Hart CA, Hart IJ, Munthali C, et al. Prevalence of infection with hepatitis B and C virus and coinfection with HIV in medical inpatients in Malawi. J Infect. 2008;57(1):72–7. 51. O'Connell RJ, Gates RG, Bautista CT, Imbach M, Eggleston JC, Beardsley SG, et al. Laboratory evaluation of rapid test kits to detect hepatitis C antibody for use in predonation screening in emergency settings. Transfusion. 2013;53(3):505–17. 52. O'Flynn N, Jilg W, McQuillan T, et al. New HCV assay on the Abbott AxSYM random access analyzer. Clin Lab. 1997;43(5):403–9. 53. Park Y, Seok Y, Choi J, Kim HS. Performance evaluation of the Vitros anti-hepatitis C virus antibody assay for use in clinical laboratories. Clin Biochem. 2012;45(1–2):175–7. 54. Poovorawan Y, Theamboonlers A, Chumdermpadetsuk S, Thong CP. Comparative results in detection of HCV antibodies by using a rapid HCV test, ELISA and immunoblot. Southeast Asian J Trop Med Public Health. 1994;25(4):647–9. 55. Prayson RA, Proffitt MR, Sharp DE, Carey WD. Application of a 2nd-generation recombinant immunoblot assay to assess hepatitis-C infection. Lab Med. 1993;24(11):732–8. 56. Rihn B, Hussenet F, Detry MB, Catelle A, Faou AL. Evaluation of a supplemental assay for the diagnosis of hepatitis C virus infections. Int J Infect Dis. 2000;4(1):42–5. 57. Scalioni Lde P, Cruz HM, de Paula VS, Miguel JC, Margues VA, Villela-Nogueria CA, et al. Performance of rapid hepatitis C virus antibody assays among high- and low-risk populations. J Clin Virol. 2014;60(3):200–5. 58. Smith BD, Drobeniuc J, Jewett A, Branson BM, Garfein RS, Teshale E, et al. Evaluation of three rapid screening assays for detection of antibodies to hepatitis C virus. J Infect Dis. 2011;204(6):825–31. 59. Smith BD, Teshale E, Jewett A, Weinbaum CM, Neaigus A, Hagan H, et al. Performance of premarket rapid hepatitis C virus antibody assays in 4 national human immunodeficiency virus behavioral surveillance system sites. Clin Infect Dis. 2011;53(8):780–6. 60. Sommese L, Iannone C, Cacciatore F, De Iorio G, Napoli C. Comparison between screening and confirmatory serological assays in blood donors in a region of South Italy. J Clin Lab Anal. 2014;28(3):198–203. 61. Tagny CT, Mbanya D, Murphy EL, Lefrere JJ, Laperche S. Screening for hepatitis C virus infection in a high prevalence country by an antigen/antibody combination assay versus a rapid test. J Virol Methods. 2014;199:119–23. Page | 301

62. Vrielink H, Zaaijer HL, Reesink HW, Borst-Loef J, Cuypers HT, Lelie PN. Performance of a new third-generation anti-hepatitis C virus ELISA (Monolisa anti-HCV new antigens) in various serum panels. Vox Sang. 1995;69(3):257–8. 63. Vrielink H, Zaaijer HL, Reesink HW, van der Poel CL, Cuypers HT, Lelie PN. Sensitivity and specificity of three third-generation anti-hepatitis C virus ELISAs. Vox Sang. 1995;69(1):14– 7. 64. Yaari A, Tovbin D, Zlotnick M, Mostoslavsky M, Shemer-Avni Y, Hanuka N, et al. Detection of HCV salivary antibodies by a simple and rapid test. J Virol Methods. 2006;133(1):1–5. 65. Yang JF, Lin YY, Hsieh MH, Tsai CH, Liu SF, Yu ML, et al. Performance characteristics of a combined hepatitis C virus core antigen and anti-hepatitis C virus antibody test in different patient groups. Kaohsiung J Med Sci. 2011;27(7):258–63. 66. Yang R, Guan W, Wang Q, Liu Y, Wei L. Performance evaluation and comparison of the newly developed Elecsys anti-HCV II assay with other widely used assays. Clin Chim Acta. 2013;426:95–101. 67. Yoo SJ, Wang LL, Ning HC, Tao CH, Hirankarn N, Kuakarn S, et al. Evaluation of the Elecsys Anti-HCV II assay for routine hepatitis C virus screening of different Asian Pacific populations and detection of early infection. J Clin Virol. 2015;64:20–7. 68. Yuen MF, Hui CK, Yuen JC, Young JL, Lai CL. The accuracy of SM-HCV rapid test for the detection of antibody to hepatitis C virus. Am J Gastroenterol. 2001;96(3):838–41. 69. Khuroo MS, Khuroo NS, Khuroo MS. Diagnostic accuracy of point-of-care tests for hepatitis C virus infection: a systematic review and meta-analysis. PloS One. 2015; 10(3). 70. Alter MJ, Kuhnert WL, Finelli L. Guidelines for laboratory testing and result reporting of antibody to hepatitis C virus. MMWR Recomm Rep. 2003;52(RR–3):1–13. 71. Guidelines for the screening, care and treatment of persons with hepatitis C infection. Geneva: World Health Organization; 2014 (http://apps.who.int/iris/bitstream/10665/111747/1/9789241548755_eng.pdf, accessed 06 June 2016). 72. St John A, Price CP. Economic evidence and point-of-care testing. Clin Biochem Rev. 2013;34(2):61–74. 73. Parisi MR, Soldini L, Vidoni G, Mabellini C, Belloni T, Brignolo L, et al. Point-of-care testing for HCV infection: recent advances and implications for alternative screening. New Microbiol. 2014;37(4):449–57. 74. Schünemann HJ, Oxman AD, Brozek J, Glasziou P, Jaeschke R, Vist GE, et al. Grading quality of evidence and strength of recommendations for diagnostic tests and strategies. BMJ. 2008;336(7653):1106–10. 75. Shepard CW, Finelli L, Alter MJ. Global epidemiology of hepatitis C virus infection. Lancet Infect Dis. 2005;5(9):558–67. 76. Leeflang MM, Moons KG, Reitsma JB, Zwinderman AH. Bias in sensitivity and specificity caused by data-driven selection of optimal cutoff values: mechanisms, magnitude, and solutions. Clin Chem. 2008;54(4):729–37.

Page | 302

77. Irwig L, Macaskill P, Glasziou P, Fahey M. Meta-analytic methods for diagnostic test accuracy. J Clin Epidemiol. 1995;48(1):119–30. 78. Rothstein HR, Sutton AJ, Borenstein M, editors. Publication bias in meta-analysis: prevention, assessment and adjustments. Hoboken, NJ: John Wiley & Sons; 2005.

Page | 303

Annex 5.5 PICO 3 - Testing strategies (HBV) Diagnostic strategies for hepatitis B surface antigen detection: a meta-analysis and review of the literature

London School of Hygiene and Tropical Medicine Team Debi Boeras*, Ali Amini*, Jane Falconer, Helen Kelly, Weiming Tang, Wen Chen, Olivia Varsaneux, Joseph Tucker, Rosanna Peeling (Team lead) London School of Hygiene and Tropical Medicine team *Co-leaders of this review

September 2015

Page | 304

1. Executive study Background: Most individuals with chronic HBV infection are not aware of their serostatus, contributing to delayed diagnosis and complications from advanced disease. Chronic HBV infection, defined as persistence of hepatitis B surface antigen (HBsAg) for at least six months, is a major cause of preventable morbidity and mortality worldwide. Advances in hepatitis B virus (HBV) detection technology create new opportunities for enhancing screening, referral, and treatment. This review will look into what is the best testing strategy (diagnostic accuracy, cost, cost–effectiveness, and other resource utilization) for detection of HBsAg. Methods: A comprehensive literature search algorithm, including Internet searches, using the components hepatitis B, screening, and testing strategies were applied. We reviewed observational and RCT studies that provided original data from patient specimens. Our goal was to compare two broad strategies for HBsAg detection – one-test strategies and two-test strategies. Results: Our search resulted in 3655 literature review references and 7 additional Internet references for PICO 3. Screening of titles/abstracts resulted in 7 selected articles for possible data extraction. None of these 7 articles met all of the data extraction inclusion criteria so no articles were identified as final selection for PICO 3; comparing the diagnostic accuracy, cost, or effectiveness of two different testing algorithms, where possible. These 7 articles are discussed in more detail – 4 of the articles provided 3 national HBV algorithms (Australia, UK, US); 1 discussed testing strategies for select populations; 2 provided a look at cost– effectiveness of given testing strategies. Conclusions: No study compared the diagnostic accuracy, cost, or cost–effectiveness of oneversus two-step HBsAg testing strategies. Studies that may provide contextual information about testing strategies were briefly summarized.

2. Background Hepatitis B virus An estimated 240 million individuals worldwide1 are chronically infected with hepatitis B virus (HBV) and there are an estimated 4 million acute HBV infections each year. Twenty per cent to 30% of those with chronic hepatitis B infection will develop cirrhosis2 or hepatocellular carcinoma,3 leading to approximately 650 000 deaths each year.4 However, most individuals with chronic HBV infection are not aware of their serostatus, contributing to delayed diagnosis and complications from advanced disease.5 HBV testing is critically important in order to refer infected individuals to HBV treatment and care, to refer uninfected individuals to vaccination, and to mobilize prevention and control efforts. In March 2015, the World Health Organization published the first guidelines for the prevention, care, and treatment of individuals with chronic HBV infection.5 These guidelines focused on assessment for treatment eligibility, initiation of first-line therapies, switching, and Page | 305

monitoring. These initial guidelines did not include recommendations on testing strategies that included what test to use and how to test. Given the large burden of HBV in low- and middle-income settings where there are limited or no existing HBV testing guidelines, there is a substantial need for HBV testing guidelines. Description of HBV Ag detection Chronic HBV infection is defined as persistence of hepatitis B surface antigen (HBSAg) for at least six months. However, interpretation of HBV serologies is complex (Table 1). The serological markers most frequently used for HBV testing include HBsAg, total anti-HBc, and anti-HBs (Table 1). Table 1. Hepatitis B serological marker interpretation Serological marker HBsAg (hepatitis B surface antigen) Test results – Total anti-HBc (antibody to hepatitis B core antigen) – IgM anti-HBc (immunoglobulin M to anti-HBc) Anti-HBs (antibody to HBsAg) Interpretation

–

Never infected and susceptible to infection Chronic infection Recovered from past infection and immune Acute infection Immune by natural infection Immune by hepatitis B vaccination Immune by natural infection or possible false positive

+ –

+ +

– –

– +

+ – –

+ + –

+ – –

– + +

–

Source: US Centers for Disease Control and Prevention. (Available at: http://www.cdc.gov/immigrantrefugeehealth/ guidelines/ domestic/hepatitis-screening-guidelines.html)

One test vs two test serological testing strategy The most important marker for the diagnosis of chronic hepatitis B infection requiring further assessment or treatment remains HBsAg. The case definition of chronic hepatitis B is the detection of HBsAg twice six months apart. After an initial positive result for HBsAg, supplementary testing can be undertaken in order to facilitate entry into a care pathway. The detection of HBsAg in blood can include rapid diagnostic tests, or enzyme immunoassays. Confirmation of the specificity of a reactive HBsAg first-line test result is usually carried out by either: Page | 306

repeating the HBsAg testing in a different assay of similar sensitivity, or

ii) performing a neutralization test using a specific anti-HBs-containing reagent in the same first-line assay after appropriate dilution of the specimen under test. Specificity is confirmed when this reagent abolishes reactivity in the assay. WHO recommends standardized testing strategies to maximize the accuracy of hepatitis B and C testing while minimizing cost and increasing simplicity. This PICO question addresses the issue of whether a positive result from a single HBsAg assay has sufficient specificity in order to proceed to supplementary testing and/or entry into a care pathway, or whether confirmatory testing on the same specimen with a different HBsAg assay (or neutralization), performed sequentially after the first assay is required. This is particularly relevant in low prevalence settings where more than one assay may be required to confirm specificity. Two previous reviews6,7 on hepatitis testing focused on the test performance but did not compare testing strategies. Fig. 1. Options for HBV screening, which may include HBsAg in a one-test strategy (e.g. a single HBsAg using a rapid diagnostic test [RDT] or enzyme immunoassay [EIA]) and two-test strategies (second RDT or EIA or neutralization with EIA)

HBsAg one-assay strategy HBsAg (RDT/EIA) (A1)

NonReac ve

Reac ve

Interpreta on: No evidence of HBV infec on Advise retes ng +\- immunisa on if ongoing risk or known exposure

Interpreta on: Compa ble with HBV infec on Proceed to supplementary tes ng

Page | 307

HBsAg (RDT/EIA) (A1)

HBSsAg Two-assay strategy

NonReac ve

Reac ve

Interpreta on: No evidence of HBV infec on Advise retes ng +\- immunisa on if ongoing risk or known exposure

HBsAg (RDT/EIA) (A2) Nonreac ve Interpreta on: Inconclusive result Further tes ng as appropriate

Reac ve

Interpreta on: Compa ble with HBV infec on Proceed to supplementary tes ng

HBsAg (RDT/EIA) (A1)

Two-assay strategy or neutralisa on

Reac on not neutralised

NonReac ve

Reac ve

HBsAg neutralisa on (EIA) (A1+)

Interpreta on: Inconclusive result Further tes ng as appropriate

Interpreta on: No evidence of HBV infec on Advise retes ng +\- immunisa on if ongoing risk or known exposure

HBsAg (RDT/EIA) (A2) Nonreac ve Interpreta on: Inconclusive result Further tes ng as appropriate

Reac on neutralised

Reac ve

Interpreta on: Compa ble with HBV infec on Proceed to supplementary tes ng

PICO 3

HBsAg testing strategy: Among persons identified for hepatitis B testing, what is the best testing strategy (diagnostic accuracy and other outcomes) for detection of HBsAg? (One-test versus two-test strategy) (Figs 1A, 1B) Persons identified for HBV testing One-test strategy; one HBsAg test (Fig. 1A) Two-test strategy; two different HBsAg tests (Fig. 1B) Diagnostic accuracy True negatives (TN), who are screen negative, and do not have HBV infection

P I C O

Page | 308

False negatives (FN), who are screen negative but have HBV infection. These are incorrectly misclassified, and this may result in missed opportunities to recognize and present progression of liver disease. True positives (TP), who are screen positive and have HBV infection. False positives (FP), who are screen positive, but do not truly have HBV infection. These will have additional unnecessary tests and evaluation. Costs (cost of testing strategy, including lab reagents and running costs, cost of further evaluation of a false positive) Cost–effectiveness Acceptability to health-care worker and patients Other outcomes (missed cases of liver disease because of false negative results, unnecessary referral, investigations and/or treatment in false positives)

3. Objectives   To identify quantitative evidence on the sensitivity and specificity of one-test compared to two-test algorithms for detection of HBsAg To evaluate the cost–effectiveness, acceptability, and other outcomes (missed liver disease because of false-negative results, unnecessary referral, investigations) associated with these two types of testing strategies To inform models to optimize hepatitis B screening algorithms.



4. Methods We reviewed observational studies and randomized controlled trials (RCTs) that provided original data from patient specimens. Our goal was to compare two broad strategies for HBsAg detection – one-test strategies and two-test strategies. Search algorithm Literature search strategies were developed by a medical librarian with expertise in systematic review searching. Our search algorithm consisted of the following components: hepatitis B, screening, and testing strategies (Annex 1). We searched MEDLINE (OVID interface, 1946 onwards), EMBASE (OVID interface, 1947 onwards), the Cochrane Central Register of Controlled Trials (Wiley interface, current issue), Science Citation Index Expanded (Web of Science interface, 1970 onwards), Conference Proceedings Citation Index-Science (Web of Science interface, 1990 onwards), SCOPUS (1960 onwards), Literatura Latino-Americana e do Caribe em Ciências da Saúde (LILACS) (BIREME interface) and WHO Global Index Medicus. The search was supplemented by searching for ongoing studies in WHO’s International Clinical Trials Registry. The literature search was limited to the English language and human subjects.

Page | 309

We formulated a comprehensive and exhaustive search strategy in an attempt to identify all relevant studies. After the MEDLINE strategy was finalized, it was adapted to the syntax and subject headings of the other databases. In addition to searching databases, we also searched the Internet for any peerreviewed articles and conference abstracts that might have been missed through our librarian search and also expanded our search to national guidance documents.

5. Results Study selection The librarian search resulted in 3655 references for PICO 3. Because of overlap with objectives and search strategies between PICOs 3 and 4, and to expedite the initial screening, PICO 3 references were combined with the 3060 references identified through the librarian search for PICO 4 (HCV); 2388 searches were immediately excluded. The librarian excluded 835 for not being relevant and there were 1553 duplicates. Thus, 4327 remained for screening. Titles/abstracts were screened according to protocol inclusion and exclusion criteria, for both PICOs 3 and 4. Reasons for excluding 4307 reports were noted (Fig. 2).

Page | 310

Fig. 2. PRISMA for PIC0 3 HBV (Diagnostic strategies for hepatitis B surface antigen and hepatitis C antibody detection)

From the librarian search, no reports were identified for possible data extraction. The Internet searches resulted in 7 additional reports for possible data extraction. Full documents (manuscripts, abstracts, guidelines, etc.) were obtained and assessed against inclusion criteria. Papers were either accepted or rejected and reasons for rejection were explained. The following inclusion criteria were used to evaluate the final selection: evaluations of HBV testing strategies; evaluations based on human clinical materials. The following exclusion criteria were used: studies focused only on evaluation of single-test assays without a two-test comparator group; studies focused on two-test strategies that include other types of test (e.g. anti-HBsAg) studies with primary aims other than evaluation of testing strategies; studies related to disease prevalence, drug resistance, genotyping, sequencing, or nondiagnostic purposes; articles in languages other than English, conference abstracts. Data abstraction and data synthesis Of the 7 selected for possible data extraction, the following variables were collected, when available: first author, title, year, objective, and exclusion criteria (Table 2).

Page | 311

Table 2. Seven reports assessed for eligibility Author or source year 1. Fan et al. 2014 Cost-effectiveness of testing hepatitis B-positive pregnant women for hepatitis B e antigen or viral load To estimate cost-effectiveness of testing with hepatitis B (hepatitis B surface antigen [HBsAg]positive) for hepatitis B e antigen (HBeAg) or hepatitis B virus (HBV) DNA To recommend screening for hepatitis B virus (HBV) infection in persons at high risk for infection Decision tree model to estimate the costs and effects of two sequential testing strategies Either sequential HBeAg testing or sequential HBV load testing was cost-effective. Sequential HBeAg testing dominated sequential HBV load testing with 1000 QALYs and $6.6 million saved Title Objective Exclusion criteria Conclusions

US Preventive Services Task Force, 2014

Hepatitis B, non-pregnant adolescents and adults: screening, May 2014

No data/not a study. USPSTF makes recommendations about the effectiveness of specific clinical preventive services for patients without related signs or symptoms. No data/not a study

Document makes screening recommendation; not relevant to data synthesis for this report

Prepared for the US Preventive Services Task Force by Peter W. Pendergrass and Carolyn DiGuiseppi (Texas Dept State Health and Univ. Colorado), 2014 Chen et al. 2015

Screening for hepatitis B virus infection

To develop recommendations for USPSTF

Document makes screening recommendation; not relevant to data synthesis for this report

Cost-effectiveness of augmenting universal hepatitis B vaccination with immunoglobin treatment

To compare the costeffectiveness of hepatitis B virus (HBV) control strategies combining universal vaccination

Not testing strategies – vaccination strategies

Universal vaccination plus screening for hepatitis B surface antigen (HBsAg) and HBIG treatment for HBsAg-positive mothers’ neonates averted the most infections

Page | 312

with hepatitis B immunoglobulin (HBIG) treatment for neonates of carrier mothers

Public Health England (PHE), National Health Service (NHS), 2014

UK Standards for Microbiology Investigations

To develop a set of standards for hepatitis B diagnostic serology in the immunocompetent (including hepatitis B in pregnancy) To provide diagnostic strategies for HBV To thoroughly assess the performance of the HBsAg assays and testing algorithm currently used in clinical settings

No data/not a study

No conclusions

Australian Government, 2012 Peng et al. 2011

National HBV Testing Strategy Development of an economic and efficient strategy to detect HBsAg: Application of “grey-zones” in ELISA and combined use of several detection assays

No data/not a study

No conclusions

Does not compare strategies

Combined use of “grey-zones” in ELISA and several different detection assays can significantly increase the efficiency of HBsAg detection

References found in Annex 2.

Page | 313

None of the reports compared the cost or effectiveness of two different testing algorithms. Of the 13 documents selected, 4 referenced algorithms and are therefore shown in Table 3. Types of tests performed and exclusion criteria are also included. Table 3. Four reports that examined testing strategies References 1 Fan et al. 2014 Sequential HBV load 2* Public Health England (PHE), National Health Service (NHS), 2014 HBsAg Repeat HBsAg Confirm by neutralization Anti-HBc HBV DNA or IgM Test 1 Sequential HBeAg Test 2 Test 3 Test 4 Exclusion criteria Decision tree model to estimate the costs and effects of two sequential testing strategies

HBsAg

Repeat HBsAg

No data/not a study

3* Australian Government, 2012 4* Peng et al. 2011

HBsAg

Anti-HBs

Anti-HBc No data/not a study

ELISA

CMIA

Confirm by HBsAg

Does not compare strategies

* Algorithm schematics attached (Annex 3)

Extensive review of the literature found no articles, reports, etc. that met all of the eligibility criteria for data extraction to be used to address this question. Most of the literature focused on screening blood donations. Seven reports were identified that might be useful for modelling exercises to address this PICO question. This short narrative will provide an overview of these 7 articles, also drawing on other informative reviews and personal communications. Cost Fan et al. (2014) examined the cost–effectiveness of testing hepatitis B-positive pregnant women for hepatitis B e antigen or viral load. In this select population of mothers of a neonate birth cohort, either sequential HBeAg testing or sequential HBV load testing was found cost– effective. Sequential HBeAg testing dominated sequential HBV load testing with 1000 QALYs Page | 314

and $6.6 million saved. It is important to note that this study used a decision tree model to estimate the costs and effects of two sequential testing strategies. Chen et al. (2015) also examined cost–effectiveness of three strategies using a cohort of hospital patients in China. In this case, costing was not related to testing strategies but vaccination strategies, specifically universal HBV vaccination with immunoglobulin treatment. Their study found that while screening tests may be cost–effective, they require more infrastructure than is needed for vaccination, including laboratory services and adequate numbers of medical professionals to interpret test results and administer HBIG. As previously mentioned, it appears that vaccination should strongly be taken into account with considering HBV testing strategies. Quality assessment Study quality was not evaluated using the QUADAS-2 tool8 and the STARD checklist, as it was not applicable since none of the studies met inclusion criteria.9

6. Discussion Testing strategies Although four of the seven reports identified provided national testing algorithms for Australia and the UK, none published data on supporting evidence. This has been confirmed by personal communications. In general, HBV screening typically includes HBsAg in both the one-test (e.g. HBsAg) and two-test strategies. Beyond this it is unclear how to select other tests. This may depend on findings from PICO 1 to better understand the performance characteristics of HBV tests, and from this to model various testing strategies for feasibility and utility. The “simplest” testing strategies seemed to include all 3 tests below (as seen in the US and Australia algorithms).    hepatitis B surface antigen (HBsAg) hepatitis B core antigen (anti-HBc) hepatitis B surface antibody (anti-HBs)

Public Health England (PHE) provides testing strategies to confirm HBsAg by an alternative assay or neutralization. Without much other evidence to support this, it seems as though confirming with an alternative assay might be a simple, cost–effective approach. It is also important to note that almost all reports reviewed also discussed comparing cost of screening and a testing strategy to cost of vaccination. Testing recommended for select populations During domestic medical examination of refugees, CDC screens using the above tests for chronic HBV (HBsAg) for all persons from countries with intermediate (≥2%–7%) or high (≥8%)

Page | 315

prevalence of chronic HBV infection. The only exception would be if a negative HBsAg test result is documented on their medical form. Peng et al. (2011) examined novel strategies to detect HBsAg using ELISA “greyzones”, in combination with other detection assays. As noted, clinical HBV detection methods differ between countries with high and low levels of endemic HBV infection. This study focused on a select population to test the algorithm currently used in clinical settings in China, specifically assessing the performance of the KHB (Kehua Bio-engineering Co. Ltd., Shanghai, China) and CMIA (Chemiluminescent Micropartical Immunoassay) HBsAg tests. While KHB is one of the most commonly used kits in China, this was a major limitation for the purposes of this systematic review. Yet, they presented a novel approach of combining strategies using ELISA “grey-zones”, which was found to significantly increase the efficiency of HBsAg detection. Although this approach “broadened the range” to allow for increased sensitivity, it may prove to be rather complicated requiring the establishment of numerous populationspecific “grey-zones” and complex interpretations. Again, almost all reports discussed testing based on select populations but this study did not focus on identifying algorithms to be used on select populations. The choice between a one-test versus two-test strategy depends on the diagnostic accuracy of HBsAg tests. Results from a systematic review of the diagnostic accuracy of HBsAg tests (PICO 1) across 21 studies that evaluated 25 brands of RDTs using 15 EIA reference assays, with 36 919 total samples, including serum, plasma, venous and capillary whole blood, showed that the overall pooled clinical sensitivity and specificity of rapid HBsAg tests were 90.0% (95% CI: 89.1, 90.8) and 99.5% (95% CI: 99.4, 99.5), respectively, compared to laboratory-based immunoassay reference standards. Sensitivities ranged from 50% to 100% with overall pooled sensitivity of 90.0% (95% CI: 89.1, 90.8). Specificities ranged from 69% to 100%, with overall pooled specificity of 99.5% (95% CI: 99.4, 99.5). Pooled positive likelihood ration (PLR) and negative likelihood ratio (NLR) were 117.5 (95% CI: 67.7, 204.1) and 0.095 (95% CI 0.067, 0.136), respectively, with tau-square 3.89, 1.72, respectively, suggestive of significant heterogeneity between studies. Pooled sensitivity in studies of HIV-positive persons was lower than in known HIVnegative patients; 72.3% (95% CI: 67.9, 76.4) compared to 92.6% (95% CI: 89.8, 94.8), respectively. Pooled sensitivity and specificity in blood donors were 91.6% (95% CI: 90.1, 92.9) and 99.5% (95% CI: 99.5, 99.9), respectively. Samples using whole blood specimens (venous or capillary) were 91.7% (95% CI: 89.1, 93.9) and 99.9% (95% CI: 99.8, 99.9) sensitive and specific compared to serum. The overall pooled clinical sensitivity and specificity of laboratory-based HBsAg tests were 88.9% (95% CI: 87.0, 90.6) and 98.4% (95% CI: 97.8, 98.8) sensitivity and specificity, respectively, compared to state-of-the-art chemiluminescent microparticle enzyme immunoassays. Although these tests appeared to have excellent specificity and would have required a 2-test strategy, in a low-prevalence setting, even tests that have excellent specificities may produce false-positive results. This would then require the use of a second test to reduce the number of false-positive results. This can be illustrated as follows:

Page | 316

In a hypothetical population of 1000 people where the HBV prevalence is 2%, a test with a sensitivity of 100% and a specificity of 99% may lead to 10 false-positive and 20 truepositive results. This means that 1 in 3 positive results may be a false-positive result. Table 4. Diagnostic accuracy in a low-prevalence setting example Reference test + Index test + Index test – Total 20 0 20 – 10 970 980 30 970 1000

Sensitivity = 20/20 = 100%; Specificity = 970/980 = 99%; PPV = 20/30 = 67%; NPV = 970/970 = 100%.

The systematic review also showed that HBsAg tests have a lower sensitivity in HIV-positive individuals. A second test may be useful for increasing the performance of testing overall. Worked example to illustrate the effect of prevalence on predictive values for the two different testing strategies Assuming the following assay performance characteristics: If Assay 1 has sensitivity of 99% and specificity of 98% If Assay 2 has sensitivity of 99.4 and specificity of 99.5% Table 5. Effect of prevalence on predictive values for the two different testing strategies Prevalence of analyte 0.1% Positive predictive values Assay 1 Assay 1 + Assay 2 (serial) Negative predictive values Assay 1 99.9% 99.99% 99.99% 4.7% 90.7% 33.3% 99% 84.6% 99.9% 1% 10%

Using the following equation for PPV and NPV that incorporates prevalence more correctly,

Page | 317

Reference: Altman DG, Bland JM. Diagnostic tests 2: predictive values. BMJ. 1994 Jul 9; 309(6947):102. Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/ PMC2540558 /pdf/bmj00448-0038 za.pdf 7. Conclusions No study compared diagnostic accuracy, cost, cost–effectiveness of one- vs two-step testing strategies for the detection of HBsAg. Diagnosis of HBV is very complex and there may not be simple algorithms that will cover all settings. All PICOs related to HBV will need to be looked at together to address PICO 3. The decision tree model described by Fan et al. may prove to be useful for modelling costs of testing strategies. Vaccination strategies should also be taken into account.

References 1. 2. 3. 4. Ott JJ, Stevens GA, Groeger J, Wiersma ST. Global epidemiology of hepatitis B virus infection: new estimates of age-specific HBsAg seroprevalence and endemicity. Vaccine. 2012;30(12): 2212–9. Ganem D, Prince AM. Hepatitis B virus infection – natural history and clinical consequences. N Engl J Med. 2004;350(11):1118–29. Fattovich G, Stroffolini T, Zagni I, Donato F. Hepatocellular carcinoma in cirrhosis: incidence and risk factors. Gastroenterology. 2004;127(5 Suppl 1): S35–S50. Lozano R, Naghavi M, Foreman K, Lim S, Shibuya K, ABoyans V et al. Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380(9859):2095–128. Guidelines for the prevention, care, and treatment of persons with chronic hepatitis B infection. Geneva: WHO; 2015. Hwang SH, Oh HB, Choi SE, et al. [Meta-analysis for the pooled sensitivity and specificity of hepatitis B surface antigen rapid tests]. Korean J Lab Med. 2008; 28(2): 160–8. Shivkumar S, Peeling R, Jafari Y, Joseph L, Pai NP. Rapid point-of-care first-line screening tests for hepatitis B infection: a meta-analysis of diagnostic accuracy (1980–2010). Am J Gastroenterol. 2012;107(9):1306–13. Whiting PF, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155(8):529– 36. Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig LM et al. The STARD statement for reporting studies of diagnostic accuracy: explanation and elaboration. Ann Intern Med. 2003;138(1):W1–W12.

5. 6. 7.

Page | 318

Appendices Appendix 1. Librarian search 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. Hepatitis, Viral, Human/ Hepatitis Viruses/ Hepatitis Antibodies/ exp Hepadnaviridae Infections/ Hepatitis B Antibodies/ Hepatitis B virus/ Hepadnaviridae/ Hepatitis B Surface Antigens/ (heptatitis-b or hep-b or (hepatitis adj5 b) or (hep adj5 b) or hbv).ti,ab. hbsag.ti,ab. exp Hepatitis C/ Hepacivirus/ Hepatitis C Antibodies/ (heptatitis-c or hep-c or (hepatitis adj5 c) or (hep adj5 c) or hcv or aghcv or hepacivirus*).ti,ab. hcvab.ti,ab. or/1-15 [HEP B or HEP C] exp Mass Screening/ screen*.ti,ab. 17 or 18 [MASS SCREENING] (one-test* or two-test*).ti,ab. ("1-test*" or "2-test*").ti,ab. ((one or two or "1" or "2" or strateg* or algorithm* or approach or procedure* or system*) adj5 (test or tests or testing or detect* or diagnos* or kit or kits or assay* or device*)).ti,ab. or/20-22 [TESTING STRATEGIES] 16 and 19 and 23 Humans/ Animals/ 25 and 26 26 not 27 24 not 28 Limit 29 to English language

23. 24. 25. 26. 27. 28. 29. 30.

Page | 319

Appendix 2. Seven full text articles assessed for eligibility (comparing algorithms, including costing) 1. Final recommendation statement: screening for hepatitis B virus infection in non-pregnant adolescents and adults. In: US Preventive Services Task Force [website]. May 2014. (http://www.uspreventiveservicestaskforce.org/Announcements/News/ Item/finalrecommendation-statement-screening-for-hepatitis-b-virus-infection-in-nonpregnantadolescents-and-adults, accessed 08 June 2016). UK Standards for Microbiology investigations: hepatitis B diagnosis in the immunocompetent (including hepatitis B in pregnancy). Standards Unit, Microbiology Services, PHE Virology. V 4, Issue no: 5.3, Issue date: 31.03.14. (https://www.gov.uk/government/ uploads/system/uploads/attachment_data/file/344145/V_4i5.3.pdf, accessed 8 June 2016). Internet Citation: National Hepatitis B Testing Policy. 2012 National Hepatitis B Testing Policy v1.1. Commonwealth of Australia; 2012. (http://testingportal.ashm.org.au/images/HepB_TESTING_POLICY_MARCH2014_V1.1_FOR%20 PRINT.pdf, accessed 11 June 2016). Chen SC, Toy M, Yeh JM, Wang JD, Resch S. Cost-effectiveness of augmenting universal hepatitis B vaccination with immunoglobin treatment. Pediatrics. 2013;131(4):e1135–e1143. Fan L, Owusu-Edusei K, Jr., Schillie SF, Murphy TV. Cost-effectiveness of testing hepatitis Bpositive pregnant women for hepatitis B e antigen or viral load. Obstet Gynecol. 2014;123(5):929–37. Peng J, Cheng L, Yin B, Guan Q, Liu Y, Wu S et al. Development of an economic and efficient strategy to detect HBsAg: application of "gray-zones" in ELISA and combined use of several detection assays. Clin Chim Acta. 2011 412(23–24):2046–51. Summaries for patients. Screening for hepatitis B virus infection: US Preventive Services Task Force recommendation statement. Ann Intern Med. 2014;161(1):I–28.

4. 5.

Page | 320

Appendix 3. Testing schematics

Public Health England, 2014: hepatitis B virus serology – HBsAg confirmation by alternative assay

Public Health England 2014: hepatitis B surface antigen (HBsAg) confirmation by neutralization

Page | 321

Public Health England, 2014: hepatitis B surface antigen confirmed reactives

Australia Government, 2012: suspected acute HBV

Page | 322

Australia Government 2012: suspected chronic HBV

Peng et al. 2011

Page | 323

Annex 5.6 PICO 4 - How to test (HCV) Diagnostic strategies for hepatitis C antibody detection: a meta-analysis and review of the literature

London School of Hygiene and Tropical Medicine team Debi Boeras*, Ali Amini*, Jane Falconer, Helen Kelly, Rosanna Peeling (Team lead), Weiming Tang, Joseph Tucker London School of Hygiene and Tropical Medicine team *Co-leaders of this review

September 2015

Page | 324

1. Executive summary Background: An estimated 130–150 million people have chronic hepatitis C infection worldwide, leading to 350 000–500 000 deaths per year. Although HCV treatment is successful in a majority of people, most HCV-infected individuals remain undiagnosed and untreated. Advances in HCV detection technology create new opportunities for enhancing screening, referral, and treatment. Methods: A comprehensive literature search algorithm, including Internet searches, using the components hepatitis C, screening, and testing strategies were applied. We reviewed observational studies and randomized controlled trials (RCTs) that provided original data from patient specimens. Our goal was to compare the effects of two broad strategies for hepatitis C antibody detection – one-test strategies and two-test strategies on diagnostic accuracy, costs, and resource utilization. Results: Our search resulted in 3060 literature review references and 3 additional Internet references for PICO 4. Screening of titles/abstracts resulted in the selection of 8 articles for possible data extraction. Two of these 8 articles met all of the data extraction inclusion criteria so no articles were identified as final selection for PICO 4; comparing the diagnostic accuracy, cost or effectiveness of two different testing algorithms. These 8 articles are discussed in more detail – 1 of the articles provided a comprehensive overview of antibody/antigen testing; 2 articles delved into core antigen testing; 3 articles exemplified other testing such as recombinant immunoblot (IB) tests, signal-to-cut-off ratios, point-of-care tests (POCT), and antibody-based rapid diagnostic tests (RDT); 1 discussed testing strategies; 1 provided a look at comparison and cost–effectiveness of given testing strategies. Conclusions: Two studies compared the diagnostic accuracy, cost, cost–effectiveness of a 1test versus 2-test strategy for detection of HCV antibody. One study found that in individuals who are HCV antibody positive, the use of an IB assay with defined signal-to-cut-off ratios can be used to distinguish between those who are viraemic and those who are not. This reduces the number of nucleic acid tests (NATs) required to confirm active infection is a cost–effective strategy. Another study found that screening with a highly sensitive EIA followed by another EIA as confirmation assay in a routine clinical laboratory can be effective in nonimmunocompromised populations. In immunocompromised patients, IB may be more effective as these patients tend to have low antibody levels.

2. Background Hepatitis C virus Hepatitis C virus (HCV) causes acute infection which can progress to chronic infection and liver disease.1,2 An estimated 130–150 million people have chronic hepatitis C infection worldwide, leading to 350 000–500 000 deaths per year. 1–3 Approximately 15–45% of individuals who have acute HCV infection will spontaneously clear it without any treatment. Most individuals will go on to develop chronic active HCV infection which is defined by the presence of HCV Page | 325

RNA.1–3 Although HCV treatment is successful in a majority of people, most HCV-infected individuals remain undiagnosed and untreated.4 As a result, approximately 15–30% of individuals with chronic HCV infection progress to cirrhosis, leading to end-stage liver disease and hepatocellular carcinoma.1,2 The recent introduction of direct-acting antivirals (DAAs) have led to sustained virological response (SVR) in greater than 90% of all individuals5,6 and are recommended by the WHO.7 DAAs will not only improve SVR rates, but also may simplify HCV management algorithms and allow smaller health facilities to manage HCV-infected individuals.8 In April 2014, the World Health Organization published guidelines for the screening, care, and treatment of individuals with HCV infection.9 These guidelines included recommendations on who to screen for HCV and how to confirm HCV infection, but not which tests are optimal for initial screening. A test for HCV antibody (Ab) is an important first step in the diagnosis of hepatitis C infection as the presence of Ab is a marker of exposure to HCV. After an initial positive result for HCV Ab, supplementary testing can be undertaken in order to confirm active infection and facilitate entry into a care pathway. The detection of HCV Ab in blood can include rapid diagnostic tests, or enzyme immunoassays (EIA). Confirmation of the specificity of a reactive HCV Ab first-line test result can be carried out by repeating the HCV Ab testing in a different assay of similar sensitivity. Specificity is confirmed when this reagent abolishes reactivity in the assay. WHO recommends standardized testing strategies to maximize the accuracy of hepatitis B and C testing while minimizing cost and increasing simplicity. This PICO question addresses the issue of whether a positive result from a single HCV Ab assay has sufficient specificity in order to proceed to supplementary testing and/or entry into a care pathway, or whether confirmatory testing on the same specimen with a different HCV Ab assay performed sequentially after the first assay is required. This is particularly relevant in low-prevalence settings where more than one assay may be required to confirm specificity. Description of HCV antibody testing Antibodies to HCV infection begin during early infection and persist throughout life in most individuals. Hence, an HCV Ab test is the best marker of exposure to HCV but cannot be used to distinguish between active and treated or resolved past infection. Screening for exposure to HCV is dependent on assays that detect antibodies to HCV (anti-HCV). Once antibody status is confirmed, the patient can undergo supplementary testing to determine the presence of HCV RNA or core antigen (HCV cAg) as markers of active infection. It is important to note that the latest generation of assays designed to detect antiHCV are combined with cAg to increase the sensitivity of the assay in detecting active infection. The question this PICO aims to address is whether one or two serological assays (antiHCV or HCV Ag/Ab combo assays) performed sequentially are required, in terms of specificity and positive predictive value, in order to proceed to supplementary testing.

Page | 326

PICO 4

Among persons identified for hepatitis C testing, what is the best testing strategy (diagnostic accuracy and other outcomes) for detection of HCV antibodies? (One-test versus two-test strategy) (Fig. 1A,1B) Persons identified for HCV testing One-test strategy; One HCV Ab test (Fig. 1A) Two-test strategy; Two different HCV Ab tests (Fig. 1B) Diagnostic accuracy True negatives (TN), who are screen negative, and do not have HCV infection. False negatives (FN), who are screen negative but have HCV infection. These are incorrectly misclassified, and this may results in missed opportunity to recognize and prevent progression of liver disease. True positives (TP), who are screen positive and have HCV infection. False positives (FP), who are screen positive, but do not truly have HBV infection. These will have additional unnecessary tests and evaluation. Costs (cost of testing strategy including lab reagents and running costs, cost of further evaluation of a false positive) Cost–effectiveness Acceptability to health-care worker and patients Other outcomes (missed cases of liver disease because of false-negative results, Unnecessary referral, investigations and/or treatment in false positives)

P I C O

Two systematic reviews on diagnostic performance of different hepatitis C serological assays focused on evaluating point-of-care tests compared to EIAs and other reference tests.10,11 None of the existing reviews compared one-test and two-test strategies for detection of hepatitis C Ab.

Page | 327

Fig. 1: What is the best testing strategy for detection of HCVAb? (A. One test, B. Two-test strategy)

An -HCV one-assay strategy (HCV-exposure) An -HCV (RDT/EIA) (A1)

NonReac ve

Reac ve

Interpreta on: No serological evidence of exposure to HCV

Interpreta on: Compa ble with exposure to HCV Proceed to supplementary tes ng

Fig. 1a. One-assay testing strategy for exposure to HCV (detection of anti-HCV) An -HCV (RDT/EIA) (A1)

NonReac ve

Reac ve

An -HCV one-assay strategy for HCV exposure, with addi onal step for diagnosis of ac ve HCV infec on

HCV RNA or HCV cAg (A2)

Detected

Not detected

Interpreta on: No exposure to HCV

Interpreta on: Ac ve HCV infec on

Interpreta on: No ac ve HCV infec on

Link to care

Addi onal tes ng as appropriate

Fig. 1b. Two-assay testing strategy for diagnosis of HCV (detection of anti-HCV, followed by HCV RNA/core Ag)

3. Objectives  To identify quantitative evidence on the sensitivity and specificity of one-test compared to two-test algorithms for detection of hepatitis C antibody

Page | 328



To evaluate the cost–effectiveness, acceptability, and other outcomes (missed liver disease because of false-negative results, unnecessary referral, investigations) associated with these two types of testing strategies To inform models to optimize hepatitis C screening algorithms.



4. Methods We reviewed observational studies and RCTs that provided original data from patient specimens. Our goal was to compare two broad strategies for hepatitis C antibody detection – one-test strategies and two-test strategies. Search algorithm Literature search strategies were developed by a medical librarian with expertise in systematic review searching. Our search algorithm consisted of the following components: hepatitis C, screening, and testing strategies (Annex 1). We searched MEDLINE (OVID interface, 1946 onwards), EMBASE (OVID interface, 1947 onwards), the Cochrane Central Register of Controlled Trials (Wiley interface, current issue), Science Citation Index Expanded (Web of Science interface, 1970 onwards), Conference Proceedings Citation Index-Science (Web of Science interface, 1990 onwards), SCOPUS (1960 onwards), Literatura Latino-Americana e do Caribe em Ciências da Saúde (LILACS) (BIREME interface) and WHO Global Index Medicus. The search was supplemented by searching for ongoing studies in WHO’s International Clinical Trials Registry. The literature search was limited to the English language and human subjects. We formulated a comprehensive and exhaustive search strategy in an attempt to identify all relevant studies. After the MEDLINE strategy was finalized, it was adapted to the syntax and subject headings of the other databases. In addition to searching databases, we also searched the Internet for any peerreviewed articles and conference abstracts that might have been missed through our librarian search and also expanded our search to national guidance documents.

5. Results Study selection The librarian search resulted in 3060 references for PICO 4. Because of overlap with objectives and search strategies between PICOs 3 and 4, and to expedite the initial screening, PICO 4 references were combined with the 3655 references identified through the librarian search for PICO 3 (HBV) for a total of 6715 references. 2388 searches were immediately excluded: the librarian excluded 835 as not relevant and there were 1553 duplicates; 4327 remained for screening. Titles/abstracts were screened according to protocol inclusion and exclusion criteria, for both PICOs 3 and 4; 4307 reports were excluded. Reasons for excluding them were noted (Fig. 2).

Page | 329

From the librarian search, 5 reports were identified for possible data extraction. The Internet searches resulted in 3 additional reports for possible data extraction. Full documents (manuscripts, abstracts, guidelines, etc.) were obtained and assessed against inclusion criteria. Papers were either accepted or rejected and reasons for rejection were explained. Fig. 2. PRISMA for PIC0 4 HCV (diagnostic strategies for hepatitis C antibody detection)

Page | 330

The following inclusion criteria were used to evaluate the final selection: evaluations of HCV testing strategies; evaluations based on human clinical materials. The following exclusion criteria were used: studies only focused on evaluation of single-test assays without a two-test comparator group; studies focused on two-test strategies that include other types of test (e.g. HCV RNA) studies with primary aims other than evaluation of testing strategies; studies related to disease prevalence, drug resistance, genotyping, sequencing, or nondiagnostic purposes; articles in languages other than English, conference abstracts. Study characteristics Of the 8 selected for possible data extraction, the following variables were collected, when available: first author, title, year, objective, and exclusion criteria (Table 2). Table 2. Eight reports assessed for eligibility Author or source, year Title Objective Exclusion criteria No comparison of testing strategies Conclusions

Cresswell. et al. Hepatitis C core antigen testing: a reliable, quick 2014 and potentially costeffective alternative to hepatitis C polymerase chain reaction in diagnosing acute hepatitis C virus infection Krajden 2000 Hepatitis C virus diagnosis and testing

To compare the utility of HCV core-antigen compared to qRT-PCR in the diagnosis on acute HCV in an HIVpositive cohort

HCV core-antigen detection compared to HCV PCR is a quick, simple, cost–effective test in screening for acute HCV

To identify how antiHCV serology and NAT can be combined to provide a definitive answer as to whether or not an individual has been or is actively infected To describe the accuracy of an algorithm that combines two HCV rapid tests to diagnose and predict viraemia of HCV in Cameroon

No data/not a Report describes how antiHCV serology and NAT can study be combined to provide a definitive answer as to whether or not an individual has been or is actively infected No comparison of testing strategies A comparison of 2 HCV rapid tests suggests an algorithm using the more sensitive test first to screen followed by the 2nd test to discriminate between viraemic and non-viraemic HCV seropositive subjects. Not relevant for this review as the second test is for HCV RNA

Njouom 2006

A cost-effective algorithm for the diagnosis of hepatitis C virus infection and prediction of HCV viraemia in Cameroon

Page | 331

Shivkumar 2012

Accuracy of rapid and point-of-care screening tests for hepatitis C: a systematic review and meta-analysis

To review evidence on the diagnostic performance of globally available RDTs and POCTs to screen for hepatitis C To review the current knowledge on 4 newer assays with decreased sensitivity, in different scenarios and reflect on their utility

No comparison of testing strategies

POCTs (blood) have highest accuracy, followed by RDTs (serum, plasma) and POCTs (oral fluids). RDTs and POCTs may be useful in expanding first-line screening for hepatitis C HCV core antigen has relative strong role in a diagnostic algorithm for HCV infection, while it is too insensitive in its present form to substitute for HCV RNA testing in the blood bank setting Study evaluated and costed 3 algorithms (2 CDC algorithms and Brazilian). The more practical and economical algorithm requires the establishment of a specific level of signal-to-noise ratio to determine the need for reflex supplemental testing (i.e. immunoblot anti-HCV) Monolisa Plus can be used as an alternative to immunoblot for the confirmation of AxSYMpositive sera

Tillmann 2014

Hepatitis C virus core antigen testing: role in diagnosis, disease monitoring and treatment

No comparison of testing strategies

Barreto 2008

Cost–effective analysis of different algorithms for the diagnosis of hepatitis C virus infection

To compare diagnostic performance and cost– benefit of two new algorithms with the conventional one in Brazilian blood donors who showed positive or inconclusive anti-HCV results in screening tests

Study was performed using blood donors

Vermeersch 2008

Validation of a strategy for HCV antibody testing with two enzyme immunoassays in a routine clinical laboratory

To compare the Comparison performance of a of testing strategy using AxSYM strategies HCV 3.0 as screening test and Monolisa Plus anti-HCV version 2 as confirmation to AxSYMpos sera with PCR and immunoblot

CDC MMWR 2013

Testing for HCV infection: To provide guidance to an update of guidance for for clinicians and clinicians and laboratorians laboratorians on testing for HCV infection

No data/not a Update to CDC guidance for diagnosis of acute study hepatitis C: rapid or a laboratory-conducted assay for HCV antibody, reactive followed by NAT for HCV RNA

References listed in Annex 2.

Of the 8 included reports, 6 described algorithms with the types of tests used (Table 3). Reports 1–4 were excluded from the systematic review as they did not compare testing Page | 332

strategies. Two studies, Boretto et al. (2008) and Vermeersch et al. (2008), determined costs and effectiveness. Table 3. Six reports of HCV testing algorithms Report 1 Cresswell 2014 2* Njouom 2006 Anti-HCV EIA Anti-HCV EIA Anti-HCV testing HCV RNA PCR HCV RNA PCR RIBA RT RNA PCR RT No comparison of testing strategies No comparison of testing strategies Test 1 HCV coreantigen Test 2 HCV RNA Test 3 HCV Ab Test 4 Exclusion Criteria No comparison of testing strategies

Tillmann 2014

CDC 2013

HCV antibody

HCV RNA

No data/not a study

Barreto 2008

See schematic below (algorithm depends on a specific level of signal-to-cut-off ratio)

Study was performed using blood donors

Vermeersch

MEIA MEIA

Confirm by EIA Confirm by PCR Confirm by immunoblot

* Algorithm schematics shown in Annex 3.

Although the study of Barreto et al. was conducted in a blood donor setting, the study did compare 3 testing strategies and determined cost–effectiveness. In this study the authors recognized that new anti-HCV tests have increased sensitivity but it means that there may be more false-positive results. These tests would be falsely negative in individuals who are newly infected as antibodies are absent or at low levels during this immunological window period. The use of a confirmatory diagnostic assay that targets different antigens can lower the risk of detecting false reactive results. Supplemental testing can be used to ensure a reliable diagnosis but this also means increased costs. The authors compared 2 CDC algorithms to the national Brazilian algorithm to determine effectiveness and cost–benefit. The figure below depicts the testing of 517 individuals identified as ELISA-positive or inconclusive by anti-HCV test using 3 different algorithms. Algorithms A and B are the CDC recommended algorithms while Algorithm C is the national Brazilian algorithm. Page | 333

The authors found that all three algorithms had similar diagnostic performance, revealing a remarkable agreement in the results obtained by the algorithms. As shown above, PCR was performed to resolve indeterminate results from immunoblots (139 samples from algorithm A and 141 samples from conventional algorithm C). Algorithm A (CDC) was recommended for populations with a high prevalence of HCV infection. The algorithm showed high concordance with true-positive results. IB testing was required only for weakly reactive samples. Algorithm B (CDC) used PCR to speed up clinical decision and was found more suitable for the immunosuppressed patient population for whom the IB test could represent a problem because of its low antibody level, leading to occasional false-negative results. Algorithm C (Brazil) was found to be useful for determining the immune status of the patients against HCV infection and also for confirming the specificity of positive enzyme-linked immunoassay (ELISA) results. It is recommended for low prevalence populations for which false-positive antibody results are usually high. However, in the present study, this algorithm yielded a high frequency of IB-indeterminate results, producing no conclusive diagnosis. This algorithm also did not differentiate between active and past infections. While algorithms A and B were found to be highly sensitive, the choice of an algorithm must take into account its purpose, the population and the prevalence of HCV infection, as well as the financial and infrastructure conditions of the laboratory. In the end they concluded that algorithm A is the best in terms of cost and feasibility, and particularly suitable for laboratories in resource-limited settings as it minimizes the number of samples requiring supplemental testing. Supplemental PCR tests were still required to detect active infection. The Vermeersch study also investigated the CDC guidelines, specifically the required confirmation of HCV screening-test-positive sera with a low signal/cut-off (S/CO) ratio by Page | 334

recombinant immunoblot or PCR. The UK Health Protection Agency suggested that a second EIA could be used as an alternative for confirmation in non-immunocompromised patients. A total of 17 936 consecutive in-house sera were evaluated in this study; AxSYM-positive sera were tested by Monolisa Plus and confirmed with IB (per CDC guidelines) or PCR. This study specifically determined the performance of a strategy using AxSYM as screening test and Monolisa Plus as confirmation assay in a routine clinical laboratory and found that Monolisa Plus can be used as an alternative to immunoblot for the confirmation of AxSYM-positive sera in nonimmunocompromised. Although the study of Barreto et al. was conducted in a blood donor setting, the study did compare 3 testing strategies and determined cost–effectiveness. Cost Barreto et al. performed a cost–effective analysis of the two CDC-recommended algorithms compared to the current Brazilian national algorithm for the diagnosis of HCV infection. The cost of each algorithm depended on the number of supplemental tests required.

Algorithm A (CDC) Based on signal-to-cut-off (s/co) ratio of ELISA anti-HCV samples that show s/co ratio ≥95% concordance with immunoblot (IB) positivity. US$ 21 299.39 This was determined to be the more practical and economical one since it requires supplemental tests for only 54% of the samples

Algorithm B (CDC) Reflex nucleic acid amplification testing by PCR was required for ELISApositive or -inconclusive samples and IB for PCR-negative samples US$ 32 397.40 This one provided early information about the presence of viraemia

Algorithm C (Brazil) All positive or inconclusive ELISA samples were submitted to immunoblot US$ 37 673.79

Quality assessment Study quality was not evaluated using the QUADAS-2 tool12 and the STARD checklist, as these do not apply to the two studies.13

6. Discussion Although none of the studies met inclusion criteria, eight references were identified that might be useful for modelling exercises to address this PICO question. This short narrative will provide an overview of these 8 articles, also drawing on other informative reviews and personal communications.

Page | 335

Antibody and antigen tests In 2000, Krajden described strengths and weaknesses of serological and molecular tests for diagnosing hepatitis C. In general, serological tests detect antibodies to hepatitis C while molecular tests detect or quantify HCV RNA. This combination of an antibody test followed by a confirmatory NAT (RNA) has generally been accepted. HCV antibody detection by enzyme immunotests (EIA) are simple, inexpensive, and often less time consuming, although they cannot distinguish between acute, active or chronic, non-viraemic HCV infection. In chronically infected persons, EIA sensitivity approaches 97– 99% while in acutely infected individuals, EIA sensitivity is as low as 50–70%. The rapid antibody tests are typically more expensive and not designed for testing large batches of specimens. However, in non-clinical (field) settings and laboratories that conduct low-volume testing, adoption of rapid testing can be cost-effective. Nucleic acid testing (NAT) remains the gold standard for identifying active infection (HCV RNA is detectable in serum or plasma as early as 1 week after exposure) but is costly, requires skilled technicians, extensive equipment and reagents, and a robust transport system to ensure sample integrity. The various forms of NAT testing include polymerase chain reaction (PCR), branched DNA signal amplification, and transcription-mediated amplification. NATs exhibit high specificities of up to 99% across all 6 genotypes of HCV. Recently, HCV core antigen testing has become widely commercially available. Two of the eight papers selected for this narrative discussed HCV core antigen testing (Cresswell 2014; Tillmann 2014). Tillmann describes the use of core testing as a serological test capable of identifying active infection, and as a possible replacement for NAT as a confirmatory test. Overall, the core test is less sensitive than HCV RNA tests, but as Tillmann reports, more than 50% of anti-HCV positive persons will be HCV core antigen positive making core antigen testing a cost–effective reflex test to confirm infection, and can easily be applied on the same platform. (Current HCV RNA assays have a lower level of detection between about 5–15 IU/mL. The sensitivity for the currently available HCV core antigen assay by Abbott was improved to about 3.00 fmol/L [0.0 6 pg/mL].) Cresswell examined the efficacy and cost of HCV core antigen in diagnosing acute HCV in a high-risk, high-prevalence population (HIV-positive cohort of MSM). Compared to HCV NAT PCR, core antigen proved sensitive (100%), specific (97.9%), and cost–effective. In their cohort, they calculated cost per individual tests to be $108 for PCR versus $23.4 for HCV cAg. Their conclusion was that in high-risk, high-prevalence populations, the core test can be used as a quick, simple and cost–effective test in screening for acute HCV. Other possible tests Three other possibilities for testing were discussed in the literature and briefly mentioned here; recombinant immunoblot tests, signal-to-cut-off ratios and point-of-care tests (POCT) and antibody-based rapid diagnostic tests (RDT).

Page | 336

IBs are highly specific serological tests. They can be performed on the same sample used in the screening test; however, they are not amenable to routine use, as they do not have high sensitivity, are costly, with a testing procedure that is technically complex, and lengthy. Confirmation of active infection still requires testing for HCV RNA. The CDC guidelines now include an option to use signal-to-cut-off ratios to limit the number of samples needing supplemental testing. Signal-to-cut-off ratios are test specific and slightly complicated to put in use and interpret. This approach might be better suited in a clinical laboratory setting (reference laboratory) that would use only one test, employ skilled technicians, and have a high volume throughput. Shivkumar (2012) published a meta-analysis specifically on diagnostic accuracy of POCTs and RDTs to screen for hepatitis C. This analysis showed POCTs of blood (serum, plasma, or whole blood) have the highest accuracy, followed by RDTs of serum or plasma and then by POCTs of oral fluids. More evidence is needed to consider using these newer tests in a diagnostic algorithm. Testing recommended for select populations Many of the articles identified through the librarian search did not meet the inclusion criteria because they recommended HCV testing in select populations based on demography, prior exposures, high-risk behaviours, and medical conditions. For example, one-time HCV testing is recommended for persons born between 1945 and 1965, without prior ascertainment of risk. Smith et al. (2012) note that the cost– effectiveness of one-time birth cohort testing is comparable to that of current risk-based screening strategies. Other major groups discussed under “risk behaviours” or “risk exposures” include injection drug use, children born to HCV-infected women, HIV infection. Because these studies were so specific to populations, they were not included as applicable to PICO 4. Testing strategies CDC MMWR (2013) describes CDC guidelines for HCV diagnostic testing: an anti-HCV test, and if the result is positive, active infection should be confirmed by a sensitive HCV RNA test. CDC recommends using US Food and Drug Administration (FDA)-approved tests (laboratory-based tests and POCT) such as OraQuick HCV rapid antibody test which has sensitivity and specificity similar to those of FDA-approved laboratory-based HCV antibody tests). An FDA-approved quantitative or qualitative NAT with a detection level of 25 IU/mL or lower should be used to detect HCV RNA. Persons positive for an anti-HCV test and negative for HCV PCR RNA are informed that they do not have current (active) HCV infection, with no further testing necessary, unless there are ongoing risk factors for and suspicion of recent infection. In this case, repeat HCV RNA test is recommended. To determine if the HCV antibody test represents a remote HCV infection that has resolved (true positivity) or a false-positive result (biological false positivity), CDC recommends Page | 337

a second FDA-approved HCV antibody test that is different from the test used for initial antibody testing. A biological false result is not likely to occur with 2 different tests (Vermeersch 2008). For patients with no apparent risk for HCV infection, the likelihood of a false-positive HCV antibody test is directly related to the HCV prevalence in the tested population; falsepositive test results for anti-HCV are most common for populations with a low prevalence of HCV infection. Njoum et al. (2006) performed a study in Cameroon comparing HCV rapid tests. In this study, using the more sensitive test first to screen followed by the second test to discriminate between viraemic and non-viraemic HCV seropositive proved to be a cost–effective algorithm for the diagnosis of HCV infection and prediction of HCV viraemia in Cameroon. The two rapid tests evaluated were the ImmunoComb® II HCV assay and Hexagon® HCV assay. The ImmunoComb® II HCV test had a higher sensitivity than the Hexagon® HCV assay for detecting anti-HCV.

ImmunoComb II Sensitivity Specificity Sensitivity Specificity 99.4 89.9 100.0 2.1

Hexagon 64.0 100.0 87.7 93.6

Reference assay HCV antibody detection

HCV RNA detection

Their study did not actually report on cost but mentioned that EIAs are less expensive than PCR technology and in this case a second EIA can be substituted in the algorithm for the confirmatory PCR test.

7. Conclusions and recommendations for research  Two studies compared the diagnostic accuracy, cost, cost–effectiveness of a one-test versus two-test strategies for detection of HCV antibody. One study found that in individuals who are HCV antibody positive, the use of an immunoblot assay with a defined signal-to-cut-off ratio can be used to distinguish between those who are viraemic and those who are not. This reduces the number of NATs required to confirm active infection is a cost–effective strategy. The challenge of using immunoblot assays is that they are lengthy and technically complex laboratory procedures, often leading to indeterminate results. Another study found that screening with a highly sensitive EIA followed by another EIA as confirmation assay in a routine clinical laboratory can be effective in nonimmunocompromised populations. In immunocompromised patients, immunoblot is more effective as these patients tend to have low antibody levels.

 

Page | 338

References 1. Hepatitis C (No. 164) [Fact sheet]. Geneva: WHO; 2014. 2. Lavanchy D. The global burden of hepatitis C. Liver Int. 2009;29(s1):74–81. 3. Mohd Hanafiah K, Groeger J, Flaxman AD, Wiersma ST. Global epidemiology of hepatitis C virus infection: new estimates of age-specific antibody to HCV seroprevalence. Hepatology. 2013;57(4):1333–42. 4. Papatheodoridis G, Tsochatzis E, Hardke S, Wedemeyer H. Barriers to care and treatment for patients with chronic viral hepatitis in europe: a systematic review. Liver Int. 2014;34(10):1452–63. 5. Feeney ER, Chung RT. Antiviral treatment of hepatitis C. BMJ. 2014;348 g3308. 6. Pawlotsky JM. New hepatitis C therapies: the toolbox, strategies, and challenges. Gastroenterology. 2014;146(5):1176–92. 7. Guidelines for the screening, care and treatment of persons with hepatitis C infection. Geneva: WHO; 2014. 8. Soriano V, Labarga P, Fernández-Montero JV, Benito JM, Poveda E, Rallon N et al. The changing face of hepatitis C in the new era of direct-acting antivirals. Antivir Res. 2013;97(1):36–40. 9. Duskova D, Darebnicek L. Nucleic acid testing of hepatitis B virus, hepatitis C virus, and human immunodeficiency virus 1, 2 in blood donors in the General University Hospital, Prague. Acta Virol. 2014;58(2):146–51. 10. Khuroo MS, Khuroo NS, Khuroo MS. Diagnostic accuracy of point-of-care tests for hepatitis C virus infection: a systematic review and meta-analysis. PloS One. 2015;10(3):e0121450. 11. Shivkumar S, Peeling R, Jafari Y, Joseph L, Pant Pai N. Accuracy of rapid and point-of-care screening tests for hepatitis C: a systematic review and meta-analysis. Ann Intern Med. 2012;157(8):558–66. 12. Whiting PF, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155(8):529–36. 13. Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig LM et al. The STARD statement for reporting studies of diagnostic accuracy: explanation and elaboration. Ann Intern Med. 2003;138(1):W1–W12.

Page | 339

Appendices Appendix 1. Librarian search 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. Hepatitis, Viral, Human/ Hepatitis Viruses/ Hepatitis Antibodies/ exp Hepadnaviridae Infections/ Hepatitis C Antibodies/ Hepatitis B virus/ Hepadnaviridae/ Hepatitis B Surface Antigens/ (heptatitis-b or hep-b or (hepatitis adj5 b) or (hep adj5 b) or hbv).ti,ab. hbsag.ti,ab. exp Hepatitis C/ Hepacivirus/ Hepatitis C Antibodies/ (heptatitis-c or hep-c or (hepatitis adj5 c) or (hep adj5 c) or hcv or aghcv or hepacivirus*).ti,ab. hcvab.ti,ab. or/1-15 [HEP B or HEP C] exp Mass Screening/ screen*.ti,ab. 17 or 18 [MASS SCREENING] (one-test* or two-test*).ti,ab. ("1-test*" or "2-test*").ti,ab. ((one or two or "1" or "2" or strateg* or algorithm* or approach or procedure* or system*) adj5 (test or tests or testing or detect* or diagnos* or kit or kits or assay* or device*)).ti,ab. or/20-22 [TESTING STRATEGIES] 16 and 19 and 23 Humans/ Animals/ 25 and 26 26 not 27 24 not 28 limit 29 to english language

53. 54. 55. 56. 57. 58. 59. 60.

Page | 340

Appendix 2. Eight full-text articles assessed for eligibility (comparing algorithms, including costing). 1. Barreto AM, Takei K, E CS, Bellesa MS, Salles NA, Barreto CC et al. Cost-effective analysis of different algorithms for the diagnosis of hepatitis C virus infection. Braz J Med Biol Res. 2008;41(2):126–34. 2. Centers for Disease C, Prevention. Testing for HCV infection: an update of guidance for clinicians and laboratorians. MMWR Morb Mortal Wkly Rep. 2013;62(18):362–5. 3. Cresswell FV, Fisher M, Hughes DJ, Shaw SG, Homer G, Hassan-Ibrahim MO. Hepatitis C core antigen testing: a reliable, quick, and potentially cost-effective alternative to hepatitis C polymerase chain reaction in diagnosing acute hepatitis C virus infection. Clin Infect Dis. 2015;60(2):263–6. 4. Krajden M. Hepatitis C virus diagnosis and testing. Can J Public Health. 2000;91 (Suppl 1):S34–S39, S6–S42. 5. Njouom R, Tejiokem MC, Zanga MC, Pouillot R, Ayouba A, Pasquier C et al. A costeffective algorithm for the diagnosis of hepatitis C virus infection and prediction of HCV viremia in Cameroon. J Virol Methods. 2006;133(2):223–6. 6. Shivkumar S, Peeling R, Jafari Y, Joseph L, Pant Pai N. Accuracy of rapid and point-ofcare screening tests for hepatitis C: a systematic review and meta-analysis. Ann Intern Med. 2012;157(8):558–66. 7. Tillmann HL. Hepatitis C virus core antigen testing: role in diagnosis, disease monitoring and treatment. World J Gastroenterol. 2014;20(22):6701–6. 8. Vermeersch P, Van Ranst M, Lagrou K. Validation of a strategy for HCV antibody testing with two enzyme immunoassays in a routine clinical laboratory. J Clin Virol. 2008;42(4):394–8.

Page | 341

Appendix 3. Testing schematics

Vermeersch et al. 2008

Tillman et al 2014

Page | 342

Njoum et al. 2006

Baretto et al. 2008

Page | 343

CDC MMWR 2013

Page | 344

Annex 5.7 PICO 6 - How to test (confirmation of HCV viraemia) Diagnostic accuracy of HCV RNA tests to detect active HCV infection: a meta-analysis and review of the literature

London School of Hygiene and Tropical Medicine team Olivia Varsaneux*, Ali Amini*, Weiming Tang, Wen Chen, Debi Boeras, Jane Falconer, Helen Kelly, Joseph Tucker, Rosanna Peeling (Team lead) London School of Hygiene and Tropical Medicine team *Co-leaders of this review

September 2015

Page | 345

1. Executive summary Background: Advances in hepatitis C virus detection technology create new opportunities for enhancing screening, referral and treatment. The purpose of this review was to determine the accuracy of qualitative NAT methods versus quantitative NAT methods for HCV RNA for detection and/or quantification to confirm active HCV infection. Method: A literature search was conducted focused on hepatitis C, diagnostic tests and diagnostic accuracy. Studies were included if they evaluated an assay to determine the sensitivity and specificity of a single qualitative hepatitis C RNA test compared to a quantitative HCV RNA reference among humans. Two reviewers performed a quality assessment of the studies and extracted data for estimating test accuracy. Results: Traditionally, qualitative nucleic acid amplification (NAT) assays are at least 10 times more sensitive than quantitative assays. This systematic review shows that for HCV, the lower limit of detection of most commercial qualitative assays was in the 10–15 IU/mL range measured against a WHO standard, whereas the lower limit of detection for quantitative assays is at 600–1100 IU/mL. This systematic review shows that the sensitivity of HCV viral quantitative assays range from 87% to 100% compared to qualitative assays. Conclusions: Although HCV qualitative assays have a lower limit of detection than quantitative assays, the range of sensitivity found in this systematic review demonstrate that HCV viral loads are rarely in the lower range of the limit of detection of these quantitative assays. New technology platforms are now available which have linear range of quantitation between 12 and 108 IU/mL, with the result that there is no longer any difference between the lower limit of detection of a qualitative assay compared to a quantitative assay. New point-of-care (POC) devices for quantitation of HCV viral load will soon be available. These devices are more affordable than the laboratory-based assays and can potentially be used to improve access to HCV detection and treatment monitoring.

screening, care and treatment of individuals with HCV infection.6 These guidelines included recommendations on who to screen for HCV and how to confirm HCV infection, but not which tests are optimal for initial screening. The World Health Assembly has passed several resolutions highlighting the importance of viral hepatitis for global health. Advances in HCV detection technology create new opportunities for enhancing screening, referral and treatment. Previous systematic reviews on hepatitis C infection have focused on treatment response,7,8 clinical complications,9 and epidemiology.10,11 Two systematic reviews on hepatitis C testing focused on evaluating point-of-care tests compared to EIAs and other reference tests.12,13 This review instead focuses on individuals with detectable HCV antibodies to evaluate qualitative versus quantitative detection methods to confirm active HCV infection. The purpose of this review was to identify evidence on the sensitivity and specificity of qualitative HCV RNA tests compared to quantitative HCV RNA tests for the detection of active HCV infection, to summarize the key test characteristics associated with detection of active HCV infection.

PICO 6

Among HCV Ab positive patients, what is diagnostic test accuracy of qualitative NAT methods versus quantitative NAT methods for HCV RNA for detection and/or quantification to confirm active HCV infection? Persons with detectable HCV antibodies Qualitative NAT methods Quantitative NAT methods Diagnostic accuracy: True negatives (TN), who are screen negative and do not have HCV infection. False negatives (FN), who are screen negative but have HCV infection, These are incorrectly misclassified and this may results in missed opportunity to recognise and present progression of liver disease. True positives (TP), who are screen positive and have HCV infection. False positives (FP), who are screen positive, but do not truly have HCV infection. These will have additional unnecessary tests and evaluation. Costs (Cost of testing strategy including lab reagents and running costs, cost of further evaluation of a false positive) Cost–effectiveness Acceptability to health-care worker and patients

P I C O

3. Objectives The purpose of this review was to identify evidence on the sensitivity and specificity of qualitative HCV RNA tests compared to quantitative HCV RNA tests for the detection of active HCV infection and to summarize the key test characteristics associated with detection of active HCV infection.

Page | 347

4. Methodology We followed standard guidelines and methods for systematic review and meta-analyses of diagnostic tests.14,15 We prepared a protocol for the literature search, article selection, data extraction and assessment of methodological quality. Selection criteria Types of studies We included observational and RCT studies that provide original data from patient specimens, including cross-sectional and case–control studies and studied qualitative NAT tests used to detect HCV RNA compared to a reference standard of quantitative HCV RNA tests. Participants Little information on participants was provided in the selection of papers included in the systematic review; therefore, we set a wide inclusion criterion. We included patients of all age groups from all settings and countries as well as all types of specimens. Index tests Studies that utilized a commercially available HCV NAT test were eligible for inclusion. The following seven are the index tests included:        AMPLICOR HCV test, version 2.0, Roche CAP/CTM, Roche COBAS AMPLICORTM HCV Test v1.0 assay, Roche COBAS AMPLICORTM HCV Test v2.0 assay, Roche COBAS HCM-2, Roche Real-Time Assay, Abbott Versant HCV genotype assay, Bayer.

Reference standard The reference standards accepted for a definitive diagnosis included tests for detection of HCV RNA by the following quantitative NAT techniques: polymerase chain reaction (PCR), branched-chain DNA (bDNA), or transcription mediated amplification (TMA). The performance characteristics of NATs are very similar above 50 IU/mL; thus all NATs were considered as one reference standard. Outcome measures Sensitivity refers to the proportion of samples with true HCV infection diagnosed with positive qualitative NAT tests confirmed with a positive quantitative NAT tests.

Page | 348

Specificity refers to the proportion of samples with negative qualitative NAT tests confirmed with a negative quantitative NAT tests. Search methods A database search of LILACS, MEDLINE, EMBASE, PubMed, Scopus, Web of Science, Cochrane and WHO Global Index Medicus was performed through April 2015. No language restriction was applied. The references of published articles found in the above databases were searched for additional pertinent materials. Study selection proceeded in three stages. First, titles/abstracts were screened by a single reviewer according to standard inclusion and exclusion criteria. Second, full manuscripts were obtained and assessed against inclusion criteria. Papers were accepted or rejected and reasons for rejection were specified. Third, two independent reviewers assessed each manuscript and differences were resolved by a third independent reviewer. Data extraction Information on the following variables were extracted by a reviewer if the study met the exclusion and inclusion criteria: first author, total sample size, country (and city) of sampling, sample type (oral fluid, finger-prick, venous blood, etc.), point-of-care (Y/N), eligibility criteria, reference standard, manufacturer, raw cell numbers (true positives, false negatives, false positives, true negatives), sources of funding and reported conflict of interest. We define point-of-care as being able to give a result within 60 min and having the results guide clinical management at the same encounter. Assessment of methodological quality Study quality was evaluated using the QUADAS-2 tool,14 the STARD checklist15 and the GRADE method.16 QUADAS includes domains to evaluate bias in the following categories: risk of bias (patient selection, index test, reference standard, flow and timing); applicability concerns (patient selection, index test, reference standard). The GRADE method evaluates the strength of evidence by assessing the risk and probability of bias, imprecision and inconsistency as well as dose-respondent gradient and residual confounding.16

5. Results PRISMA flowchart A total of 17 109 citations were identified and 9623 duplicates were removed. Each of the 7486 titles was examined according to pre-specified inclusion and exclusion criteria. A total of 4 research studies were included in the final analysis (Fig. 1 below).

Page | 349

Fig. 1. PRISMA flow diagram outlining study selection examining diagnostic accuracy HCV RNA tests to detect active HCV infection

Characteristics of included studies A total of four studies met the PICO criteria and data was extracted from each of these studies. Two of the four studies took place in the United States of America, with the remaining two in Taiwan and Germany. Of these studies only one included a population of patients at risk of HCV infection, while the others were either patients who have an acute or chronic HCV infection. The assays evaluated in these analyses were Abbott Real-Time Assay, AMPLICOR

Page | 350

HCV test, v2.0 assay, COBAS AMPLICORTM HCV Test v2.0 assay and Versant HCV genotype assay. This systematic review shows that for HCV, the lower limit of detection of most commercial qualitative assays was in the 10–15 IU/mL range measured against a WHO standard, whereas the lower limit of detection for quantitative assays is at 600–1100 IU/mL. The sensitivities of qualitative NAT methods reported in the selected articles showed a relatively wide range (87–100%), contrary to the narrow range reported for specificity (97– 100%). A large limitation in the quality of the studies was a lack of information on the populations studied, randomization and sample collection as well as poor standardization in the evaluation of diagnostic test accuracy. Table 1. Description of study design, study population and setting of all studies (n=4) No First author, country Sample type and number Serum N = Not stated Study population Diagnostic test (quantitative) Reference test (qualitative) Sensitivity Specificity

Lee, 2000, United States of America

Patients at risk of HCV infection

AMPLICOR HCV test, version 2.0 Roche

COBAS AMPLICORTM HCV Test v2.0 assay. Roche

94%

97%

Yu 2000, Taiwan

Serum N = 215

Patients with chronic hepatitis C Patients with HCV infection

COBAS HCM-2 Roche

COBAS AMPLICORTM HCV Test v2.0 assay. Roche COBAS AMPLICORTM HCV Test v1.0 assay Roche

95%

–

Ferreira-Gonzalez, 2007, United States of America

Plasma N = 76

Versant HCV genotype assay, Bayer

100%

Sarrazin, 2008, Germany

Serum N = 65

Patients with HCV infection

Versant HCV genotype assay. Bayer CAP/CTM* Roche

Abbott Real-time Assay

87%

–

*CAP/CTM= Roche Cobas Ampliprep/Cobas Taqman HCV assay

Lee et al. (2000) investigated the performance characteristics of AMPLICOR HCV test, version 2.0 Roche (quantitative assay) and COBAS AMPLICORTM HCV Test v2.0 assay Roche (qualitative assay). This was done by measuring the limit of detection, sensitivity, specificity, linear range, agreement between test formats and genotypic reactivity for both tests. The genotypic reactivity for both tests showed that samples with 10 copies per reaction yielded positive results at least 95% of the time for all genotypes with the exception of genotype 5. When testing for the agreement between test formats, it was shown that the COBAS AMPLICOR v.2 assay format produced values with a range of 0.02–0.13 log10 higher than those obtained for the AMPLICOR assay with samples containing titres within the linear range of the assays. Page | 351

Yu et al. (2000) investigated the performance characteristics of COBAS HCM-2 (quantitative assay) compared to the COBAS AMPLICORTM HCV Test v2.0 assay (qualitative) and evaluated the clinical utility of COBAS HCM-2. This study looked at quantitative range, reproducibility of COBAS HCM-2 and linearity of HCV RNA quantifications. The quantitative range for the COBAS HCM-2 assay from 1.0 X 103 to 3.88 X 106 copies/mL and the within-run reproducibility showed serum HCV RNA levels with standard deviations of 0.03, 0.09 and 0.12. The linearity of HCV RNA quantifications ranged from 6.11, 6.44, 6.46 to 6.49 logs for genotypes 1b, 2a, 2b and 1b. Ferreira-Gonzalez et al. (2007) evaluated the qualitative (COBAS Taqman HCV) against the quantitative (Versant HCV genotype assay, Bayer) using the same clinical specimens. Analytical sensitivity was measured by the ability of a system to detect replicates, both tests were capable of detecting all six replicates with 10 HCV RNA IU/mL and 100% of all replicates with 1.0 Log10 HCV RNA IU/mL. Sarrazin et al. (2008) focuses on the evaluation and comparison of performance characteristics of HCV qualitative (CAP/CTM Roche Cobas Ampliprep/Cobas Taqman HCV assay, Roche) and quantification (Versant HCV genotype assay, Bayer) methods. The study compared intra-assay variability, analytic sensitivity, limit of detection, HCV WHO standard RNA unitage, genotypic specific assay linearity. The intra-assay variability of the two assays varied from 0.72% to 1.3% for CAP/CTM assay and 1.4–3.02% for Versant HCV genotype assay. For sensitivity, the positive hit rates for WHO HCV RNA standards for the Abbott RealTime HCV assay were 87% at 15 IU/mL, where as for the CAP/CTM they were 100% at 15 IU/mL. The limit of detection was higher for RealTime HCV assay at 16.8 (95% CI; 13.1 to 27.9) than for CAP/CTM 10.3 (95% CI; 8.4 to 15.1). HCV WHO standard RNA unitage deviated between – 0.2 log10 IU/mL at 3.2 Log10 for RealTime HCV assay and –0.3 log10 IU/mL at 3.2 Log10. For RealTime HCV assay and CAP/CTM assay, the quantification of HCV RNA of five different HCV genotypes was mostly linear between concentrations of 4.0 X 103 and 1.0 X 106 IU/mL though the sample harbouring genotype 4 showed lower results than expected at concentrations above 1.0 X 104 UI/mL. Narrative summary of each systematic review’s findings Traditionally, qualitative nucleic acid amplification (NAT) assays are at least 10 times more sensitive than quantitative assays. This systematic review shows that for HCV, the lower limit of detection of most commercial qualitative assays was in the 10–15 IU/mL range measured against a WHO standard, whereas the lower limit of detection for quantitative assays is at 600‒1100 IU/mL. Though very sensitive for determining the presence of virus, qualitative assays do not allow for determination of viral load. The sensitivity of qualitative assays makes them essential during screening blood donors and monitoring treatment progression as they are used to show the presence of virus as a marker of an on-going HCV infection. 17‒20 This systematic review showed sensitivities as low as 5 IU/mL for HCV NAATs. Verification of the presence of RNA is complicated due to the lack of a standardized commercial HCV assay with sufficient sensitivity that is capable of testing discrepant specimens at such low concentrations.21 This is Page | 352

particularly important when qualitative assays are used prove the absence of HCV-RNA at the end of treatment (ETR) or at the end of follow up (EFU).22 Another use of qualitative assays is to discriminates sustained responders (SRs) from relapsers (RELs).23‒25 Despite the lower sensitivity, quantitative assays have been found to be a reproducible method to detect and quantify HCV RNA in plasma or serum.25 This systematic review also highlighted potential issues with accuracy as the level of imprecision in a number of quantitative tests (COBAS) AMPLICOR HCV Monitor assay) was seen to be between a factor of 3–5 (0.5–0.7 log10) difference from the actual titre.26 These assays have been proven to be crucial in the measurement of the viral load at the start of therapy and after 12 weeks of treatment to decide about the usefulness of further treatment (stopping rule). These assays have also been proven to have a broad dynamic range of 615–7 700 000 IU/mL ([COBAS] AMPLICOR HCV Monitor assay, Roche). The ability of a quantitative tests to detect viral loads as low as 650 IU/mL allow them to measure early response to treatment as often viral loads drop rapidly at the start of interferon treatment. Viral loads seldom exceed the upper limit of this assay so that retesting of diluted samples is often unnecessary.27–28 This explains the reasoning behind laboratories in high-income countries employ HCV RNA quantitative assays in serum or plasma, notwithstanding its higher cost, as it can be used to monitor treatment efficacy and chronic HCV disease progression.29–32 Since these studies were published, companies have been steadily improving the sensitivity of quantitative assays. Table 1 shows that currently 5 HCV quantitative assays are commercially available with another two in the pipeline (UNITAID Hepatitis C diagnostic technology landscape report, 2015).39 These assays have linear range of quantitation of 12– 108 IU/mL using plasma or serum and the time to result ranges from 70 min to 5‒6 h. With these new technology platforms, there is no longer any difference between the lower limit of detection of a qualitative assay compared to a quantitative assay. From this systematic review, the finding of sensitivities of 87‒100% for a quantitative assay compared to qualitative assays demonstrates that viral loads are rarely in the lower range of the limit of detection of these older quantitative assays. These new quantitative assays and the equipment are costly and can be prohibitive to control programmes. However, since other assays such as HIV viral load can be performed on these platforms, HCV control programmes can leverage investments made by HIV programmes for the procurement of these technologies. In resource-limited settings where these assays are not affordable, HCV antigen detection can be considered as a surrogate marker of ongoing virus replication. Investments in the development of point-of-care (POC) devices that can be used to measure HIV viral load have yielded several technological platforms that can be used to quantitate HCV viral load. Table 2 shows that four companies have developed HCV quantitative assays that would soon be available with another three in the pipeline. These devices cost much less than the laboratory based instruments and can be used outside of laboratory settings as some of them can run on batteries. They are sample in-answer out type of technologies that will require minimal training and all have connectivity capacity so that surveillance for hepatitis C can be automated. The evaluation of these POC platforms will be

Page | 353

important to inform countries of whether these devices can be used to improve access to HCV detection and treatment monitoring.

Page | 354

Table 2. Qualitative/quantitative HCV RNA platforms currently available (1–5) and soon to be available (6–7)

Roche Molecular Systems (1) Qualitative assays COBAS AmpliPrep /COBAS TaqMan HCV Qualitative Test v.2 (LLOD: 15 IU/mL) COBAS AmpliPrep /COBAS TaqMan HCV Quantitative Test v.2 Linear range of quantitation, IU/mL Sample type (mL) Cost/test (US$) Price of instrument (US$) 0.650 mL plasma/ serum 36‒38/43‒51 COBAS AmpliPrep: 80 000–100 000 COBAS Taqman 48: 15‒108 –

Abbott Diagnostics (2)

Siemens Healthcare Diagnostics (3) –

Sacace Biotechnologies (4) – –

QIAGEN (5) –

Beckman Coulter (6)

Hologic Inc (7)

–

Quantitative assays

Abbott RealTime HCV Assay

VERSANT kPCR HCV RNA Assay

HCV Real-TM Quant Dx Assay

artus HCV QS-RCQ Kit

VERIS MDx

RT-TMA Technology for the Panther® System

12‒108

15‒108

13‒108

35–1.77x106

–

0.5 mL plasma/serum 13‒35 248 000 (45 000 + 162 000 + 80 000)

0.5 mL plasma or serum 72‒100 Pricing for the assay and instrument is available from Siemens

1 mL plasma >20 113000 (95 000+18 645)

1 mL plasma 16‒45 Available from company

plasma, serum – –

0.24 mL plasma 10‒15 –

Page | 355

40–50 000 # Specimen/run and Time to result 24 specimens in 2 h, can process up to 72 samples at one time 96 samples at a time in about 3 h 89 samples per run with a total time to result of <6 h 24 samples/run in 5–6 h Continuous loading in batches of up to 24 samples plus internal controls 48 samples can be lined up on 12 racks; DNA tests takes ~70 min and RNA tests ~110 min First results available 3 h after loading samples and five results after every 5 min thereafter. Samples can be continuously loaded, with up to 120 samples on the Panther® System

Alere Quantitative Assays Alere q HCV VL

Molbio Diagnostics Pvt Ltd Truelab Real Time micro PCR System

Cepheid GeneXpert HCV quantitative assay

Ustar Biotechnologies RT-CPA HCV Viral Load Test

Linear range of quantitation, IU/mL Sample type (mL) Cost/test (US$)

– 0.5‒1 plasma 15–25

– 0.1 plasma 14 per chip; 2 per extraction

10 –10

1 plasma/serum <US$ 20

0.1 blood –

– Price of instrument (US$)

–

8000

17 000

–

Time to result (min) Other tests on platform

<60 HIV, Ebola

60 MTB, HBV, dengue, chickungunya, HINI, malaria, HIV VL, HCV VL

105 HIV

20‒45 (<500 IU/mL) –

Table 3. Point-of-care HCV RNA platforms in the pipeline

Page | 356

Page | 357

References A. Reference list of studies that met criteria for inclusion in the analysis 1. Lee SC, Antony A, Lee N, Leibow J, Yang JQ, Soviero S, et al. Improved version 2.0 qualitative and quantitative AMPLICOR reverse transcription-PCR tests for hepatitis C virus RNA: Calibration to International Units, Enhanced Genotype Reactivity, and Performance Characteristics. J Clin Microbiol. 2000;38(11):4171‒9. Sábato MF, Shiffman ML, Langley MR, Wilkinson DS, Ferreira-Gonzalez A. Comparison of performance characteristics of three Real-Time reverse transcription-PCR test systems for detection and quantification of hepatitis C virus. J Clin Microbiol. 2007;45(8):2529‒36. Vermehren J,Kau A,Gärtner BC,Göbel R, Zeuzem S, Sarrazin C. Differences between two Real-Time PCRbased hepatitis C virus (HCV) assays (RealTime HCV and Cobas AmpliPrep/Cobas TaqMan) and one signal amplification assay (Versant HCV RNA 3.0) for RNA Detection and Quantification. J Clin Microbiol. 2008;46(12):3880‒91. Yu ML, Chuang WL, Dai CY, Chen SC, Lin ZY, Hsieh MY, et al. Clinical evaluation of the automated Cobas Amplicor HCV monitor test version 2.0 for quantifying serum hepatitis C virus RNA and comparison to the Quantiplex HCV Version 2.0 Test. J Clin Microbiol. 2000;38(8):2933‒9.

B. Reference list from background 1. 2. 3. WHO. Hepatitis C (No 164) [Fact sheet]. 2014. Lavanchy D. The global burden of hepatitis C. Liver Int. 2009;29(s1):74‒81. Mohd Hanafiah K, Groeger J, Flaxman AD, Wiersma ST. Global epidemiology of hepatitis C virus infection: new estimates of age-specific antibody to HCV seroprevalence. Hepatology. 2013;57(4):1333‒42. Papatheodoridis G, Tsochatzis E, Hardke S, Wedemeyer H. Barriers to care and treatment for patients with chronic viral hepatitis in europe: a systematic review. Liver Int. 2014;34(10):1452‒63. Guidelines for the screening, care and treatment of persons with hepatitis C infection. Geneva: WHO, 2014. Zeuzem S, Berg T, Moeller B, Hinrichsen H, Mauss S, Wedemeyer H, Sarrazin C, et al. Expert opinion on the treatment of patients with chronic hepatitis. J Viral Hepat. 2009;16(2):75‒90. Zheng H, Li M, Chi B, Wu XX, Wang J, Liu DW. IL28B rs12980275 variant as a predictor of sustained virologic response to pegylated-interferon and ribavirin in chronic hepatitis C patients: a systematic review and meta-analysis. Clin Res Hepatol Gastroenterol. 2015;39(5):576‒83. Manzano-Robleda Mdel C, Ornelas-Arroyo V, Barrientos-Gutierrez T, Mendez-Sanchez N, Uribe M, Chavez-Tapia NC. Boceprevir and telaprevir for chronic genotype 1 hepatitis C virus infection. A systematic review and meta-analysis. Ann Hepatol. 2015;14(1):46‒57. Ambrosino P, Lupoli R, Tarantino P, Di Minno A, Tarantino L, Di Minno MN. Viral hepatitis and antiphospholipid antibodies positivity: a systematic review and meta-analysis. Dig Liver Dis. 2015;47(6):478‒87.

4. 5. 6. 7.

10. Khodabandehloo M, Roshani D. Prevalence of hepatitis C virus genotypes in Iranian patients: a systematic review and meta-analysis. Hepat Mon. 2014;14(12):e22915. 11. Khodabandehloo M, Roshani D, Sayehmiri K. Prevalence and trend of hepatitis C virus infection among blood donors in Iran: a systematic review and meta-analysis. J Res Med Sci. 2013;18(8):674‒82.

Page | 358

12. Khuroo MS, Khuroo NS, Khuroo MS. Diagnostic Accuracy of point-of-care tests for hepatitis C virus infection: a systematic review and meta-analysis. PLoS One. 2015;10(3):e0121450. 13. CRD’s guidance for undertaking reviews in health care. York: Centre for Reviews and Dissemination, 2008. 14. Pai M, Mcculloch M, Gorman JD, PaiNEnanoria W, Kennedy G, et al. Systematic reviews and metaanalyses: an illustrated, step-by-step guide. Natl Med J India. 2004;17(2):86‒95. 15. Shivkumar S, Peeling R, Jafari Y, Joseph L, Pant Pai N. Accuracy of rapid and point-of-care screening tests for hepatitis C: a systematic review and meta-analysis. Ann Intern Med. 2012;157(8):558‒66. 16. Whiting PF, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155(8):529‒36. 17. Malmivaara A. Methodological considerations of the GRADE method. Ann Med 2015;47(1):1‒5. 18. National Institutes of Health. 2002. Management of hepatitis C: 2002. NIH Consensus and State-of-theScience Statements 19(3):1‒46. 19. National Institutes of Health. 2002. Management of hepatitis C: 2002. NIH Consensus and State-of-theScience Statements 19(3):1‒46. 20. Krajden M, Ziermann R, Khan A, Mak A, Leung K, Hendricks D, et al. Qualitative detection of hepatitis C virus RNA: comparison of analytical sensitivity, clinical performance, and workflow of the Cobas Amplicor HCV Test Version 2.0 and the HCV RNA transcription-mediated amplification qualitative assay. J Clin Microbiol. 2002;40(8):2903‒07. 21. Martinot-Peignoux M, Marcellin P, Pouteau M, Castelnau C, Boyer N, Poliquin M, et al. Pretreatment serum hepatitis C virus RNA levels and hepatitis C virus genotype are the main and independent prognostic factors of sustained response to interferon alfa therapy in chronic hepatitis C. Hepatology. 1995;22(4 Pt 1):1050‒6. 22. McHutchison JG, Poynard T. Combination therapy with interferon plus ribavirin for the initial treatment of chronic hepatitis C. Semin Liver Dis. 1999;19(Suppl. 1):57‒65. 23. Morishima C, Gretch DR. Clinical use of hepatitis C virus tests for diagnosis and monitoring during therapy. Clin Liver Dis. 1999;3:717‒40. 24. Lee SC, Antony A, Lee N, Leibow J, Yang JQ, Soviero S, et al. Improved version 2.0 qualitative and quantitative AMPLICOR reverse transcription-PCR tests for hepatitis C virus RNA: calibration to international units, enhanced genotype reactivity, and performance characteristics. J Clin Microbiol. 2000;38(11):4171‒79. 25. OritoE, Mizokami M, Suzuki K, Ohba K, Ohno T, Mori M, et al. Loss of serum HCV RNA at week 4 of interferon-alpha therapy is associated with more favorable long-term response in patients with chronic hepatitis C. J Med Virol. 1995;46:109‒15. 26. Reichard O, Norkrans G, Frydén A, Braconier JH, Sonnerborg A, Weiland O. Comparison of 3 quantitative HCV RNA assays – accuracy of baseline viral load to predict treatment outcome in chronic hepatitis C. Scand J Infect Dis. 1998;30:441‒6. 27. Van Vlierberghe H, Leroux-Roels G, Adler M, Bourgeois N, Nevens F, Horsmans Y, et al. Daily induction combination treatment with alpha 2b interferon and ribavirin or standard combination treatment in naive chronic hepatitis C patients. A multicentre randomized controlled trial. J Viral Hepat. 2003;10:460‒6. 28. Zeuzem S, Lee JH, Franke A, Rüster B, Prümmer O, Herrmann G, et al. Quantification of the initial decline of serum hepatitis C virus RNA and response to interferon alfa. Hepatology. 1998;27:1149‒56.

Page | 359

29. Mancini C, Pisani G, Azzi A, Luisa Zerbini M, Gentili G, Mario Bisso G, et al. Inter-laboratory comparison of qualitative and quantitative detection of hepatitis C (HCV) virus RNA in diagnostic virology: a multicentre study (MS) in Italy. J Clin Virol. 2004;30:313‒9. 30. PawlotskyJM. Molecular diagnosis of viral hepatitis. Gastroenterology. 2002;122:1554‒68. 31. Trimoulet P, Halfon P, Pohier E, Khiri H, Chêne G, Fleury H. Evaluation of the VERSANT HCV RNA 3.0 assay for quantification of hepatitis C virus RNA in serum. J Clin Microbiol. 2002;40:2031‒6. 32. Anderson JC, Simonetti J, Fischer DG, Williams J, Yamamura Y, Rodriguez N, et al. Comparison of different HCV viral load and genotyping assays. J Clin Virol. 2003;28(1):27–37. 33. European association for the study of the liver (EASL), 1999. EASL International consensus conference on hepatitis C. Consensus Statement J Hepatol. 30:956–61. 34. Gretch D. Standardization of hepatitis C virus – RNA quantification advances and unfinished business. Hepatology. 2000;31:788–9. 35. Martinot-Peignoux M, Boyer N, Le Breton V, Le Guludec G, Castelau C, Akremi R, et al. A new step toward standardisation of serum hepatitis C virus RNA quantification in patients with chronic hepatitis C. Hepatology. 2000;31:726–9. 36. Lo Re V 3rd, Amorosa VK, Localio AR, O'Flynn R, Teal V, Dorey-Stein Z, et al. Adherence to hepatitis C virus therapy and early virologic outcomes. Clin Infect Dis. 2009;48(2):186–93. 37. McHutchinson JG, Manns M, Patel K, Poynard T, Lindsay KL, Trepo C,et al. Adherence to combination therapy enhances sustained. Response in genotype-1-infected patients with chronic hepatitis C. Gastroenterology. 2002;123(4):1061–9. 38. UNITAID. Hepatitis C diagnostic technology landscape. 1st edn. Geneva: WHO; 2015. (http://unitaid.org/images/marketdynamics/publications/UNITAID-HCV_Diagnostic_ Landscape1st_edition.pdf, accessed 06 June 2016).

Page | 360

Annex 5.8 PICO 5a and PICO 9 - How to test (confirmation of HCV viraemia) HCV core antigen testing for presence of active HCV infection and monitoring for treatment response and cure: a systematic review

J. Morgan Freiman1; Trang M. Tran2; Samuel G Schumacher3; Laura F. White4; Stefano Ongarello3; Jennifer Cohn2,5; Benjamin P. Linas1,6; Claudia M. Denkinger3,7 (Team lead) 1. 2. 3. 4. 5. 6. 7. HIV Epidemiology and Outcomes Research Unit, Section of Infectious Diseases, Boston Medical Center, Boston, USA Médecins Sans Frontières – Access Campaign, Geneva, Switzerland FIND, Geneva, Switzerland Department of Biostatistics, Boston University School of Public Health, Boston, USA Infectious Diseases Division, University of Pennsylvania, Philadelphia, USA Department of Epidemiology, Boston University School of Public Health, Boston, USA Division of Infectious Diseases, Beth Israel Deaconess Medical Center, Boston, USA

Corresponding author Claudia M. Denkinger Head of Tuberculosis and Hepatitis Programme FIND, Geneva, Switzerland Email: claudia.denkinger@finddx.org Tel: Office: +41-22-749-2931

2015

Page | 361

Abstract Background: Chronic hepatitis C virus (HCV) infection with viraemia is prevalent in approximately 1.1% of the world population.1 Current diagnosis of active infection requires a positive HCV antibody (Ab) as well as nucleic acid testing (NAT) to detect HCV ribonucleic acid (RNA) indicative of active replication. HCV core antigen (HCVcAg) testing was developed as an alternative to NAT. This systematic review aims to summarize (1) the diagnostic accuracy of HCVcAg testing in those with and without positive HCV Ab (PICO 5a), (2) inform the best testing strategy for identification of active HCV infection (PICO 5b), and (3) examine the utility of HCVcAg monitoring for those on HCV treatment (PICO 9). Methods: We performed a literature search in multiple databases for all published and peer reviewed literature without language restriction through March 2015. Studies were included if a commercially available HCV Core Ag test result was compared with NAT in at least 10 independent clinically collected samples. We contacted authors for missing data to complete extraction. We assessed the quality of studies using an adapted QUADAS-2 tool. Data were classified by HCV Core Ag test manufacturer. For PICO 5a, bivariate meta-analyses were performed for the Abbott ARCHITECT, Hunan Jynda, and Ortho ELISA to obtain pooled sensitivity (Se) and specificities (Sp) with 95% confidence intervals (CI). Due to limited number of studies and specificity data descriptive statistics were derived for the Murex EIA, Bio-RAD Monolisa, EIKEN Lumispot and Fujirebio Lumipulse. We assessed non-parametric regression of quantitative data and identified outliers. Due to the absence of published studies to inform PICO 5b, a decision analysis was performed and is reported separately. Only a descriptive analysis was possible on the use of HCV core Ag in treatment monitoring and assessment of SVR (PICO 9). Results: We identified 50 published studies for inclusion in the analysis of PICO 5a, 1 study relevant to PICO 5b, and 5 studies relevant to PICO 9. For PICO 5a, 7 index tests were included with 30 studies utilizing Abbott ARCHITECT, 5 studies for Bio-RAD Monolisa, 4 for Murex Ag/Ab EIA, 6 for Ortho ELISA-Ag, 2 for EIKEN Lumispot HCV Ag, 1 for Ortho Lumipulse-Ag, and 4 for Hunan Jynda Bioengineering Group HCV Core Ag ELISA. Among these, 1 directly compared the ARCHITECT with the Lumipulse and Lumispot, and 1 compared the Monolisa with the Murex. From bivariate analyses, the pooled sensitivity and specificity with 95% CI were: ARCHITECT 93.4% (88.7, 96.2) and 98.7% (96.9, 99.4), Ortho ELISA 93.2% (81.6, 97.7) and 99.2% (87.9, 100), and Hunan Jynda 59.5% (46% 71.7) and 82.9% (58.6, 94.3). The sensitivity for the Lumipulse was 95% (90.2, 99.8) in one study; specificities could not be calculated. Three studies using the ARCHITECT provided quantitative data. The few points with negative HCVcAg were shown to occur at RNA levels below 3000 IU/mL where loss of linearity was also noted in pooled non-parametric regression. Accuracy of HCVcAg for treatment monitoring and as a test of cure was assessed by descriptive analysis in 5 studies (PICO 9). The sensitivity of ARCHITECT in EVR ranged from 74–100% with specificity from 70% to 100%. SVR was only assessed in 2 studies with 100% sensitivity and specificity ranging from 94% to 100%. Data on accuracy in prediction of SVR were limited and assessed in only 3 small studies.

Page | 362

Conclusions: HCV core antigen assays can have high sensitivity (up to 93.4% for Abbott ARCHITECT HCVcAg test), high specificity, and good correlation with HCV RNA to a detection limit of roughly 3000 IU/mL. The data on core antigen for treatment monitoring and as a test of cure is too limited to reach reliable conclusions.

GRADE summary tables I. PICO 5a: What is the best strategy (diagnostic accuracy and other outcomes); comparing HCV core Ag test versus NAT for HCV RNA for detection (and/or) quantification to confirm active HCV infection? SR Outcome: Diagnostic accuracy, sensitivity and specificity Patients/population: Persons with detectable HCV RNA with or without positive HCV antibody Setting: Any Index tests: HCV core antigen assay Importance: Inform best strategy for HCV diagnosis in a variety of clinical settings and economies Reference standard: HCV RNA testing Studies: Cohort, cross-sectional, or randomized controlled trials that use HCV NAT as gold standard reference test compared with a commercially available HCV core Ag index test A) Strength of evidence SR outcome: diagnostic accuracy Index test Outcome Measure Sensitivity # Studies Design (# samples) 30 (12,788) Quality Strength of Evidence Inconsistency Indirectness Imprecision

Risk of Bias

Abbott ARCHITECT HCV Ag Assay

Cohort and cross- Low1 sectional Cohort and cross- Low1 sectional Cohort and cross- High1 sectional (–2) Cohort and cross- High1 sectional (–2) Cohort and cross- Low1 sectional Cross-sectional

Low2

Moderate3 (–1)

Low4

Moderate

⨁⨁⨁◯ Low4 Moderate

Specificity

20 (11,820)

Low2

Moderate3 (–1)

⨁⨁⨁◯ Moderate4 (–1) Moderate4 (–1) Low4 Very low

Ortho ELISA-Ag

Sensitivity

6 (1,423)

Moderate2 (–1) Moderate2 (–1) High2 (–2)

Moderate3 (–1) Moderate3 (–1) Moderate (–1) Moderate3 (–1) Moderate3 (–1)

⨁◯◯◯ Very low

Specificity

5 (1,177)

⨁◯◯◯ Very low

Bio-RAD Monolisa HCV Sensitivity Ag-Ab ULTRA Specificity

5 (525)

⨁◯◯◯ NA4 Very low

1 (337)

Moderate1 NA2 (–1) (–1)

⨁◯◯◯ Moderate4 (–1) Very low

EIKEN Lumispot HCV Ag

Sensitivity

2 (235)

Cross-sectional

Moderate1 Low2 (–1)

⨁◯◯◯

Page | 363

Specificity Fujirebio Lumipulse Ortho HCV Ag Sensitivity

0 1 (80)

NA Cross-sectional

NA Moderate3 (–1) NA Moderate3 (–1) Moderate3 (–) Moderate3 (–1) Moderate3 (–1)

NA NA4

NA Very low

Moderate1 NA2 (–1) (–1) NA

⨁◯◯◯ NA Low4 NA Very low

Specificity Hunan Jynda HCV Core Ag ELISA Sensitivity

0 4 (524)

Cohort and cross- Moderate1 High2 sectional (–1) (–2) Cohort and cross- Moderate1 High2 sectional (–1) (–2) Cohort and cross- Low1 sectional Cohort Low1 High2 (–2) Moderate2 (–1)

⨁◯◯◯ Low4 Very low

Specificity

4 (524)

⨁◯◯◯ Low4 Very low

DiaSorin S.A. Murex Ag/Ab EIA

Sensitivity

4 (770)

⨁◯◯◯ Low4 Low

Specificity

3 (658)

⨁⨁◯◯

NA= not applicable Footnotes: For each index test, quality of evidence started high when there were several high-quality observational studies (prospective cohort studies, cross-sectional studies with direct comparison of index test results with a reference standard). We then downgraded one point when there was moderate concern identified and two points when a there was a high concern identified in any of the four factors that may decrease the quality of evidence: risk of bias, inconsistency, indirectness, and imprecision. 1

We used QUADAS-2 to assess risk of bias.  For ARCHITECT, in half of the studies it was unclear how participants were selected and one study used only healthy blood donors; however, the data from all studies is consistent and unclear selection does not appear to cause bias thus we did not downgrade. For the Ortho ELISA, two studies of five used convenience enrolment for participant selection, and one enrolled only healthy blood donors thus we downgraded 2 points. For the Monolisa, four of five studies had unclear patient selection. For one it was unclear if the index and reference test were performed within 30 days. Given that there were no high-risk concerns for bias we did not downgrade. For specificity, there was only one study with data that had unclear participant selection, thus we downgraded one point, as there were no data from studies with random or consecutive selection to compare to and identify possible selection bias (as was possible with the ARCHITECT). For the Lumispot, both studies had unclear patient selection. As there were no data from studies with random or consecutive selection to compare, we downgraded one point. The Lumipulse only included one study with unclear participant selection and was downgraded 1 point. The Hunan Jynda had one of four studies with unclear participant selection, one in only healthy blood donors, and one for which it was unclear whether the index and reference were performed within 30 days. As the use of only healthy blood donors was considered a high-risk category, in combination with the other unclear factors, we downgraded one point. For the Murex test, three of four studies had unclear participant selection but no other high-risk concerns for bias and thus we did not downgrade.

 

  

 2

Unexplained heterogeneity in remaining studies may be related to covariates that could not be adjusted for in metaregression due to limited data (HIV and HBV coinfections, HCV genotype). Additionally, not all studies identified HCV antibody status or stratified by acute and chronic infection thus variability of HCV replication could contribute to higher false negative HCVcAg.

Page | 364

  



  3

There was little heterogeneity noted in the ARCHITECT studies; thus we did not downgrade. For the Ortho ELISA, there was moderate heterogeneity with largely one outlier study, thus we downgraded 1 point. For the Monolisa sensitivity outcome, heterogeneity between studies precluded meta-analysis and thus we downgraded 2 points. For specificity, there is only 1 study and we cannot assess heterogeneity and downgrade 1 point. For the Murex sensitivity outcome there was too much heterogeneity to pool the data, and thus we downgraded 2 points. For specificity, there were not enough studies to perform meta-analysis and heterogeneity could not be formally assessed, however there is a broad range among results and thus we downgraded one point. The EIKEN Lumispot was only used in 2 studies. Sensitivity was similar in both studies suggesting little heterogeneity, thus we did not downgrade. For the Fujirebio Lumipulse, there is only 1 study and we cannot assess heterogeneity and downgrade 1 point.

All studies were performed in reference laboratories, and the majorities were in high and middle-income countries. Thus the patient population, the viral population tested (e.g. genotype distribution), and the test users are not representative of the limited-resource settings for which these guidelines are envisioned. All were downgraded 1 point. We considered imprecision as present when the pooled confidence intervals were >10% and when there were fewer than 250 samples in the analysis. As such, we downgrade the Ortho ELISA, and Hunan Jynda one point for wide confidence intervals, and downgraded the Lumispot one point for small sample size. Additionally, imprecision could not be graded for the Monolisa specificity outcome, and the Lumipulse test as these only included one study.

B) Summary of findings, PICO 5a SR outcome: diagnostic accuracy Index test # Studies (# samples) 20 (11,820) Unit of analysis Sample Effect accuracy (95% confidence interval) Sensitivity Specificity Positive LR Negative LR Effect likelihood ratio (LR)

Abbott ARCHITECT HCV Ag Assay Ortho ELISA-Ag Bio-RAD Monolisa HCV AgAb ULTRA EIKEN Lumispot HCV Ag Fujirebio Lumipulse Ortho HCV Ag Hunan Jynda HCV Core Ag ELISA DiaSorin S.A. Murex Ag/Ab EIA

93.4% (88.7, 96.2)

98.7% (96.9, 99.4)

71.8 (28.6, 160.3)

0.07 (0.04, 0.12)

5 (1,177) 5 (525)

Sample Sample

93.2% (81.6, 97.7) 28.6–95%*

99.2% (87.9, 99.9) 94.9% (89.9, 99.8)**

116.5 (6.7, 977) NA

0.06 (0.02, 0.07) NA

2 (235) 1 (80)

Sample Sample

97.5–98.1%*

NA NA

95% (90.2, 99.8)** ND

4 (524)

Sample

59.5% (46, 71.7)

82.9% (58.6, 94.3)

3.5 (1.1, 12.6)

0.28 (0.2, 0.3)

4 (730)

Sample

50–100%*

83.8–100%*

ND: no data, NA = not applicable – if sensitivity and specificity results were not available from meta-analysis, likelihood ratios were not calculated. * Meta-analysis not possible. Range of results seen across studies reported. **Result from one study only.

Page | 365

C) Impact of findings in different prevalence settings Outcome Effect per 1000 patients with presumed HCV for varying prevalence settings comparing HCV core Ag against HCV RNA Prevalence 2%* Abbott ARCHITECT HCV Ag Assay True positives (patients with HCV) True negatives (patients without HCV) False positives (patients incorrectly classified as having HCV) False negatives (patients incorrectly classified as not having HCV) Ortho ELISA-Ag True positives (patients with HCV) True negatives (patients without HCV) False positives (patients incorrectly classified as having HCV) False negatives (patients incorrectly classified as not having HCV) Hunan Jynda HCV Core Ag ELISA True positives (patients with HCV) True negatives (patients without HCV) False positives (patients incorrectly classified as having HCV) False negatives (patients incorrectly classified as not having HCV) 12 (9, 14) 813 (578, 921) 167 (59, 402) 60 (46, 72) 747 (531, 846) 153 (54, 369) 179 (138, 216) 581 (413, 658) 119 (42, 287) 19 (16, 20) 970 (862,980) 10 (0, 118) 93 (82, 98) 891 (792, 900) 9 (0, 108) 279 (246, 294) 693 (616, 700) 7 (0, 84) 19 (18, 19) 967 (951, 974) 13 (6, 29) 93 (89, 96) 888 (873, 895) 12 (5, 27) 279 (267,288) 691 (697, 696) 9 (4, 21) Prevalence 10%* Prevalence 30%*

1 (1, 2)

7 (4, 11)

21 (12, 33)

1 (0,4)

7 (2, 18)

21 (6, 54)

8 (6, 11)

41 (28, 54)

122 (84, 162)

*Numbers in parentheses consider 95% confidence intervals of accuracy estimate

II. PICO 9: Among patients receiving treatment for HCV, what is the diagnostic accuracy of HCV core Ag test versus NAT for HCV RNA detection (and/or) quantification to confirm successful treatment response with viral clearance? SR Outcome 1: Diagnostic accuracy, sensitivity and specificity of HCVcAg at SVR SR Outcome 2: Timing and predictive accuracy of HCVcAg for SVR Patients/population: Persons with detectable HCV RNA with or without positive HCV antibody Setting: Any Index tests: HCV core antigen assay

Page | 366

Importance: Inform best strategy for treatment monitoring and test of cure in a variety of clinical settings and economies Reference standard: HCV RNA Testing Studies: Longitudinal cohort or randomized controlled trials that use HCV NAT as gold standard reference test compared with a commercially available HCV core Ag index test SR outcome 1: Diagnostic Accuracy at SVR Index test Outcome measure # Studies Design (# samples) 2 (67) RCT, cohort Quality Effect* Strength of evidence

Risk of Inconsistency Indirectness Imprecision bias Low1 Low2 Moderate3 (–1) Low4 100%* Moderate

Abbott ARCHITECT Sensitivity HCV Ag Assay Specificity

⨁⨁⨁◯ Low4 94–100%* Low

2 (67)

RCT, cohort

Low1

Moderate2 (–1)

Moderate3 (–1)

⨁⨁◯◯

SR outcome 2: Predictive accuracy of SVR

Quality

Effect*

Strength of evidence

Index test

Outcome # Studies Design Risk of bias Inconsistency Indirectness Imprecision measure (# individuals) 1 (23) Cohort Low1 NA2 (–1) 1 (23) Cohort Low1 NA2 (–1) Moderate3 (–1) Moderate3 (–1) Moderate3 (–1) Moderate3 (–1) Moderate4 (–1) Moderate4 (–1) 88.5–99.3%* 57.1–79.4%* NA4 70%** NA4 95.2%** Low

Abbott Sensitivity ARCHITECT HCV Ag Assay Specificity

⨁⨁◯◯ Low

⨁⨁◯◯ Very low

Fujirebio Sensitivity Lumipulse Ortho HCV Ag Specificity

2 (134)

Cohort Moderate1 (–1)

Moderate2 (–1) Moderate2 (–1)

⨁◯◯◯ Very low

2 (134)

Cohort Moderate1 (–1)

⨁◯◯◯

* Results reported are range across studies or **individual result, NA= not applicable Footnotes: For each index test, quality of evidence started high when there were several high quality observational studies (prospective cohort studies, cross-sectional studies with direct comparison of index test results with a reference standard). We then downgraded one point when a serious issue was identified and two points when a very serious issue was identified in any of the four factors that may decrease the quality of evidence: risk of bias, inconsistency, indirectness, and imprecision. SR outcome 1: diagnostic accuracy at SVR 1.

We used QUADAS-2 to assess risk of bias. There were no concerns raised for the two studies that utilized the ARCHITECT assay.

Page | 367

The limited number of studies precluded a meta-analysis and formal assessment of heterogeneity. However, the data between studies for sensitivity is consistent thus we did not downgrade. There is some variability seen in the data for specificity, thus we downgrade 1 point. Both studies were performed in reference laboratories in high and middle-income countries, which is not representative of broad use throughout the world thus we downgraded 1 point. The range in specificity results was attributed to possible unexplained heterogeneity and is less likely from verification bias given the excellent reference standard. Given that we already downgraded for heterogeneity, we did not downgrade for imprecision.

SR outcome 2: predictive accuracy of early HCV Ag on SVR 1.

We used QUADAS-2 to assess risk of bias. For the ARCHITECT, there were no concerns raised so we did not downgrade. In the 2 Fujirebio Lumipulse studies, participant selection was unclear in one, and one did not include all patients initially enrolled in the analysis so we downgraded 1 point. For the ARCHITECT, there was only one study thus we could not assess heterogeneity and downgrade one point. For the Lumipulse assay, there were only 2 small studies and no formal heterogeneity could be assessed. However, neither study included covariate information aside from genotype and the results between studies are broad, thus we downgraded one point. All studies were performed in reference laboratories in high- and middle-income countries, which is not representative of broad use throughout the world. All were downgraded 1 point. For the ARCHITECT, imprecision could not be graded as there was only one study. There is a broad range of effect between the Fujino studies, which may in part be from unexplained heterogeneity already discussed, but may also be from imprecision as only absolute values of decline in HCVcAg were examined instead of log decline thus we downgraded 1 point.

2. Background Chronic hepatitis C virus (HCV) infection with viraemia is prevalent in approximately 1.1% of the world population, or 64–103 million people, with an estimated 75% of all cases occurring in low- to middle-income countries (LMICs).1 HCV is a small, enveloped, single stranded ribonucleic acid (RNA) virus belonging to the Flaviviridae family with seven genotypes and more than sixty-seven subtypes.2 The genome is contained in an internal capsid formed by three domains of the HCV core protein, which is highly conserved and antigenic.3,4 During viral assembly, nucleocapsid peptides 22 (p22) are released into plasma5 and can be detected early in the course of infection. Screening assays to assess for anti-HCV antibodies (HCV Ab) were among the first diagnostic tools developed to identify HCV infection, but can only inform about exposure to the virus and not active replication or ongoing infection. The serological window for conversion to a positive antibody is highly variable with an average of 60 days6 and antibodies may remain persistently negative among patients on haemodialysis and those with poorly controlled HIV infection or other immunocompromised states. Thus, diagnosis of active HCV infection requires antibody testing followed by an assessment for viraemia both for confirmation of true infection in antibody-positive patients and for high-risk antibody negative patients. Confirmatory testing can be based on nucleic acid testing (NAT) to detect HCV-RNA or an antigen testing to detect core antigen. HCV core antigen (HCVcAg) tests largely targeting p22 have been in development as an alternative to NAT since Tanaka et al. first demonstrated detection of circulating antigen in those

Page | 368

with chronic HCV infection in 19957 and the first commercial assay was released in 2000.8 HCVcAg tests have the potential to be less costly and less centralized than NAT. Detection of HCV viraemia is also important during treatment of chronic HCV infection. Current guidelines recommend virological confirmation pretreatment with the measurement of a baseline viral load with NAT. For interferon-based treatments, viral load is assessed at week 4 of therapy for the “rapid viral response” (RVR) to help predict efficacy of therapy, and repeated at week 6 if elevated at week 4 to see further viral response and guide whether treatment should be continued. NAT is performed again at week 12 (early viral response, EVR), at the end of treatment, and 12 and 24 weeks after therapy is completed to test for cure, “sustained viral response” (SVR). With the development of direct-acting antivirals (DAAs), NAT during therapy may no longer be necessary.9 Additionally, DAA has made treatment for HCV possible in LMICs10 making access to an affordable diagnostic and monitoring test even more important. This systematic review of the published literature aims to assess the diagnostic accuracy of HCVcAg testing for HCV detection and inform the best testing strategy for identification of chronic HCV infection. Furthermore, the review looks at the utility of HCVcAg for monitoring on HCV treatment and to test for cure.

Tests included in this systematic review Only commercially available tests were included in the systematic review. The most widely studied is the Abbott ARCHITECT HCV Ag assay, a two-step automated chemiluminescent microparticle immunoassay (CMIA) that allows quantitative determination of HCVcAg in serum or plasma. The assay uses the Abbott ARCHITECT i System (i2000/i2000SR/i1000SR modules), a reference laboratory instrument with ARCHITECT System Software version 5.0 or higher. The Fujirebio Lumipulse Ortho HCV Ag test and EIKEN Lumispot HCV Ag are similar automated chemiluminescent enzyme immunoassays (CLEIA) available in Japan and China. There are two available Ab–Ag combination enzyme immunoassays (EIA), the DiaSorin S.A. Murex HCV Ag–Ab combination and Bio-RAD MonolisaTM HCV Ag-Ab ULTRA. The Monolisa uses a spectrophotometer to read absorbance values that detects presence or absence of Ab and/or HCVcAg with the colour intensity being proportional to quantity of Ab or Ag to HCV bound on the solid phase. Lastly, there are two enzyme-linked immunosorbent assay (ELISA)-based HCVcAg tests, Hunan Jynda Bioengineering Group HCV Core Ag ELISA and Ortho ELISA-Ag.

Page | 369

3. Objectives This systematic review addresses predefined PICO questions 5a and 9. Question 5b will be addressed in a separate report. PICO 5a What is the best strategy (diagnostic accuracy and other outcomes); comparing HCV core Ag test versus NAT for HCV RNA for detection (and/or) quantification to confirm active HCV infection? Persons with detectable HCV RNA with or without positive HCV antibody HCV core antigen assay HCV RNA testing  Diagnostic accuracy 1. 2. True negatives (TN), who are screen negative, and do not have HCV infection. False negatives (FN), who are screen negative but have HCV infection. These are incorrectly misclassified, and this may results in missed opportunity to recognize and present progression of liver disease. True positives (TP), who are screen positive and have HCV infection. False positives (FP), who are screen positive, but do not truly have HCV infection. These will have additional unnecessary tests and evaluation.

P I C O

3. 4.

PICO 5b

What is best testing strategy (diagnostic accuracy and other outcomes); between using sequential testing strategy (HCV core Ag followed by NAT if negative) versus NAT alone for diagnosis of active HCV infection? Persons with detectable HCV RNA with or without positive HCV antibody Sequential testing strategy (HCV core Ag followed by NAT if negative) (Fig. 4A) Standalone NAT test (Fig. 4B)  Diagnostic accuracy 1. 2. True negatives (TN), who are screen negative, and do not have HCV infection. False negatives (FN), who are screen negative but have HCV infection. These are incorrectly misclassified, and this may results in missed opportunity to recognize and present progression of liver disease. True positives (TP), who are screen positive and have HCV infection. False positives (FP), who are screen positive, but do not truly have HCV infection. These will have additional unnecessary tests and evaluation.

P I C O

3. 4.

Page | 370

PICO 9

Among patients receiving treatment for HCV, what is the diagnostic accuracy of HCV core Ag test versus NAT for HCV RNA detection (and/or) quantification to confirm successful treatment response with viral clearance? (Fig. 5A, 5B) Patients receiving treatment for HCV HCV core Ag test (Fig. 5A) NAT for HCV RNA detection (and/or) quantification (Fig. 5B)  Diagnostic accuracy 1. 2. True negatives (TN), who are screen negative, and cleared the HCV infection. False negatives (FN), who are screen negative but have HCV infection. (These will be misclassified, and treatment will be stopped resulting in disease progression leading to Liver related morbidity (fibrosis, cirrhosis, end-stage liver disease, hepatocellular carcinoma), progression of liver disease, and mortality. True positives (TP), who are screen positive and truly have HCV infection, this will increase the number of treated cases and cured rate. False positives (FP), who are screen positive, but do not have HCV infection. (These will continue treatment inappropriately, and will have unnecessary referral).

P I C O

3. 4.

4.Methods We followed standard guidelines and methods for systematic review and meta-analyses of diagnostic tests.11–13 We prepared a protocol for the literature search, article selection, data extraction, and assessment of methodological quality.

Selection criteria Types of studies We included case–control, cross-sectional, cohort studies and randomized trials that used HCV NAT as gold standard reference test compared with a commercially available HCV core Ag index test for the diagnosis of active HCV infection or in the monitoring of HCV infection while on treatment. Participants We included patients of all age groups from all settings and countries. Specimen types were limited to whole blood, plasma or serum, and we only included studies that examined at least 10 independent HCV NAT positive samples. Saliva specimens were also considered, but only one study was identified during the search and it did not use NAT as reference test and was thus excluded from further analysis. Commercially prepared reference panel specimens were excluded. Index tests

Page | 371

Studies that utilized a commercially available HCV Core Ag test were eligible for inclusion. The following seven are the index tests included:        Abbott ARCHITECT HCV Ag Bio-RAD MonolisaTM HCV Ag-Ab ULTRA EIKEN Lumispot HCV Ag Fujirebio Lumipulse Ortho HCV Ag Hunan Jynda Bioengineering Group HCV Core Ag ELISA DiaSorin S.A. Murex Ag/Ab EIA Ortho ELISA-Ag

Target conditions PICO 5a/b  Acute HCV infection: the 6-month time period following acquisition of hepatitis C virus. HCV Ab may be positive or negative; the time period between initial infection to seroconversion of antibody is the “window period”. HCV RNA is detectable. Chronic HCV infection: duration of HCV infection more than 6 months from time of acquisition. HCV Ab and RNA are detectable.



PICO 9:  Monitoring of viral clearance while on treatment with rapid viral response (RVR) at 4 weeks and early viral response (EVR) at 12 weeks and sustained viral response (SVR) at 24 weeks after completion of treatment

Reference standard The reference standards accepted for a definitive diagnosis included tests for detection of HCV RNA by any of the following NAT techniques: polymerase chain reaction (PCR), branched-chain DNA (bDNA), or transcription mediated amplification (TMA). Tests were noted to be either qualitative or quantitative. The performance characteristics of NATs are very similar above 50 IU/mL, thus all NATs were considered as one reference standard.

Outcome measures Sensitivity refers to the proportion of samples with true HCV infection diagnosed with positive HCV Core Ag test confirmed with a positive NAT result. Specificity refers to the proportion of samples with negative HCV Core Ag test and no evidence of active HCV infection confirmed with a negative NAT result. Search methods Page | 372

A database search of EMBASE, PubMed, Scopus, Web of Science, and Cochrane was performed through March 2015. No language restriction was applied. The search terms used for each database are outlined in Appendix A. Two review authors (JMF and TT) independently assessed titles and abstracts identified by the literature search to select potentially eligible studies (screen 1). Any citation identified by either review author during screen 1 was selected for full text-review. Full papers of each potentially eligible article were retrieved. Two review authors (JMF and TT) independently assessed the full text articles for inclusion using the predefined inclusion and exclusion criteria (screen 2). Three articles were excluded because of inability to find appropriate language interpretation. Discrepancies were resolved by discussion between the review authors, and for several studies by the decision of a third review author (CMD). The included studies were divided into those applicable for each PICO question. A list of excluded studies and reasons for exclusion can be found in Appendix D. Data extraction We created a data extraction form, pilot-tested the form with a subset of eligible studies, and then finalized the form (Appendix B). Two review authors (JMF and TT) independently extracted data from the included studies with the standardized form and crosschecked to ensure accuracy. Disagreement between review authors on data extraction was resolved by discussion or by a third reviewer (CMD). For studies without complete extraction information available, authors were contacted to request further data. Studies without extractable sensitivity and specificity data were excluded if no further information was acquired after three attempts to contact the study authors. Assessment of methodological quality We adapted the QUADAS-2 instrument, a validated tool for diagnostic studies,14 to assess study quality. The information needed to answer QUADAS-2 questions was incorporated in the data extraction sheet. A description of the QUADAS-2 items and the interpretation in the study context can be found in Appendix C. Statistical analysis and data synthesis Statistical analyses were performed using STATA (version 14; STATA corporation, College Station, TX). The studies were grouped by type of index test used. QUADAS analysis was performed using Excel (version 14.5.3; Microsoft, Seattle, WA). Approach to indeterminate index test results We excluded indeterminate test results from the analyses for determination of sensitivity and specificity, as it was less than 1% for all index tests. Assessment of publication bias

Page | 373

We did not perform formal assessment of publication bias (tests for funnel plot asymmetry), as these techniques are not recommended for diagnostic test accuracy studies. We reviewed the EASL and AASLD conference abstract books for abstracts of studies that have not been published subsequently and did not find anything between 2010 and 2013. We did not include unpublished data in this review.

Meta-analysis Meta-analysis for each index test type was performed if at least four studies were available with the same index test with at least ten independent samples in each study. Bivariate random effects meta-analyses were performed for index tests with enough studies that included data to calculate sensitivity, specificity, and 95% confidence intervals for each. Several studies did not contribute to both sensitivity (no true positives and false negatives) and specificity (no false positives and true negatives) but only to one of the two. In such cases, we examined the correlation between sensitivity and specificity visually from a scatter plot of the sensitivity versus 1-specificity across studies. If the correlation was limited, we performed a univariate random effects meta-analysis of the sensitivity and/or specificity estimates separately, so as to make complete use of the available data. We then compared the results from the univariate analysis (including all studies) with the results from the bivariate analysis of the subset of studies that contributed to both sensitivity and specificity estimates. For index tests with data that contributed only to sensitivity but had at least 4 studies, we performed univariate random effects meta-analysis only for sensitivity. A descriptive analysis was performed for index tests with less than four studies available or when substantial heterogeneity was evident on forest plots that precluded a meta-analysis. We visually assessed forest plots for heterogeneity among the studies within each index test and in the summary plots we examined the variability in estimates and the width of the prediction region, with a wider prediction region suggesting more heterogeneity. We also report an estimate of τ2 (along with its standard error) corresponding to the variance of the logit-transformed specificity and sensitivity, which can be interpreted as a measure of between-study variability. The initial protocol planned for sensitivity analysis excluding case-control studies, but none were identified amongst the studies included. We anticipated that studies included in the meta-analysis would be heterogeneous in many respects. Therefore, we pre-specified subgroups by antibody status. Furthermore, we planned to examine the effect of specimen condition (fresh vs. frozen), HBV and HIV status and genotype in a meta-regression. Where meta-regression to assess impact of covariates was not possible due to limited data, we showed descriptive statistics for HIV and HBV coinfection and genotype distribution. The impact of specimen condition could not be assessed as all studies either used frozen samples or did not specify condition.

Page | 374

Analysis of quantitative data Where quantitative data were available from the studies, a locally weighted regression smoother was used to visually assess the correlation between quantitative HCV Ag measured in fmol/L to HCV RNA measured in IU/mL.15 We identified outliers and performed descriptive statistics of these points. There was only enough quantitative data to assess the Abbott ARCHITECT assay. 5. Results Results of the search From the literature search, 8146 citations were identified and a total of 313 full-text articles were reviewed: 283 applied to PICO 5a, 11 applied to PICO 5b (reported separately), and 44 applied to PICO 9. For PICO 5a, 50 studies were included. For PICO 9, 4 studies were included. Figure 1a-d shows the PRISMA diagram with the flow of studies for each PICO and reasons for exclusion. Description of studies Core antigen for HCV detection (PICO 5a) – included studies Fifty included studies utilized the 7 different HCV cAg assays described above, with two performing direct comparisons between two or more antigen tests.16,17 Four studies were translated from Mandarin,18 1 from German,19 1 from French,20 and 2 from Japanese.21, 22 Characteristics for each study are presented in Table 1a. The Abbott ARCHITECT HCV Ag assay was assessed in 30 studies.5, 16, 18, 19, 23–49 All study designs were either cross-sectional or cohort, with a broad study population (included patients with HCV disease, and those susceptible to HCV disease) with the exception of one study that evaluated only healthy blood donors.35 Only 20 had enough data to be included in the bivariate analysis.34 Ten did not have enough data to calculate specificity16, 17, 18, 28, 31, 36, 37, 41, 42, 45 and were only included in the univariate pooled sensitivity estimate. All but 3 studies specified positive HCV Ab status of specimens19, 25, 26 and 4 included data for HCV Ab negative samples.18,24,42,45, 48 Demographic data was available in 18 studies, the remainder utilized anonymous specimens and authors were unable to provide further information. HIV status was known in 15 of the studies with 2 including only HIV-coinfected subjects.46, 48 HBV status was known in 13 studies and all but 4 excluded patients with HBV coinfection. The study with highest prevalence included 50.5% with HBV coinfection.40 Only 1 study included children.28 The Bio-RAD MonolisaTM HCV Ag-Ab ULTRA was used in 5 studies;17, 20, 50–52 all were cohort or cross-sectional in design with a broad study population. One study had an unknown amount of participants with at least 25 known subjects and an additional 94 samples from an unknown amount of donors.48 Two included only HIV-coinfected adult subjects,20, 51 the remaining 3 had unknown subject demographic information. The EIKEN Lumispot HCV Ag was performed in one cross-sectional study with a broad study population.53 Further demographic information was unavailable. Page | 375

The EIKEN Lumispot HCV Ag, Fujirebio Lumipulse Ortho HCV Ag, and Abbott ARCHITECT HCV Ag were compared in 1 cross-sectional study,16 with unknown demographic information. Four studies assessed the Hunan Jynda Bioengineering Group HCV Core Ag ELISA.54–57 Two studies had a cohort design, 2 cross-sectional, and 1 assessed a healthy blood donor population56 while the others included broad study populations. HIV and HBV coinfection status was unknown in all studies. One included children,57 and the remaining had unknown age groups included. The DiaSorin S.A. Murex Ag/Ab EIA was used in 4 articles, 3 adult cohort studies 58–60 and 1 cross-sectional study that compared performance with the Bio-RAD MonolisaTM HCV Ag-Ab ULTRA;17 this is the same study as above with an unknown total number of participants. One study included 25% HIV-coinfected patients,59 and one included 6.1% HBV co-infected patients.60 Finally, 6 articles utilized the Ortho ELISA-Ag test.61–64 All were either cross-sectional or cohort designs in broad study populations except for 1 study performed in healthy blood donors.63 All had unknown demographic information. Core antigen in best testing strategy for identification of active HCV infection (Pico 5b) – included studies Only 1 study was found to meet inclusion criteria.43 Given limited data, a decision analysis was performed to address this PICO question and was reported separately. Core sntigen for treatment monitoring and test of cure (PICO 9) – included studies Two studies evaluated the Abbott ARCHITECT compared to NAT at baseline, EVR and SVR, one used patients enrolled in a randomized controlled trial65 and one used a cohort design.66 One study used the ARCHITECT to assess correlation of HCVcAg and NAT at EVR only. Three cohort studies evaluated HCVcAg kinetics during EVR to assess predictive accuracy of SVR, but did not compare HCVcAg to NAT at SVR. One used the ARCHITECT,67 and 2 employed the Fujirebio Lumipulse.68,69 None required translation. Characteristics are presented in Table 1b. All studies included patients with active HCV infection who were initiated on interferon based treatment regimens. Excluded studies A list of excluded studies for each PICO and the reasons for exclusion is presented in appendix D. Methodological quality (QUADAS-2) The overall methodological quality of all included studies for each PICO question organized by QUADAS-2 domain is summarized in Fig. 2 and presented for each individual study in Fig. 3.

Page | 376

Patient selection PICO 5a In the “patient selection” domain, we judged 2 studies to have “high risk of bias” because they used convenience sampling of participants for enrolment. Twenty-three studies were judged to be “low risk of bias”. In 25 studies risk of bias was “unclear” with 12 having both unspecified enrolment and prior exclusion strategies and the remaining 13 with a mix of the two. Applicability in this domain was judged to be “high risk” in 3 studies that included only healthy blood donors, and the remaining 47 were determined to be “low risk”. Setting of testing was not considered for this review as currently available tests can only be operated in specialized laboratories. PICO 9 In the “patient selection” domain, 4 studies were judged to be “low risk of bias” as sampling was consecutive or random. One was judged to be “unclear risk of bias” as patient selection was not specified. Applicability was judged to be “low risk” in all studies as all included patients with active HCV infection. Index test PICO 5a All studies were determined to have “low risk of bias” as all index tests had predefined thresholds of positivity and interpretation does not require judgement thus all were considered blinded with respect to the results of the reference test. Applicability in this domain was assessed by whether or not the index test was performed per recommendations of the manufacturer. In 3 studies, this was unclear and information could not be obtained from the study authors, thus 3 were determined to be “unclear risk” while the remaining 47 were “low risk”. PICO 9 All studies were determined to have “low risk of bias” as all index tests had predefined thresholds of positivity and interpretation does not require judgement. For applicability, it was unclear in 1 study whether the index test was performed per recommendation of the manufacturer and was thus judged “unclear”. The remaining 4 studies were “low risk”. Reference standard PICO 5a and PICO 9 All studies were judged to be “low risk of bias” per our QUADAS-2 rules. Though studies used a variety of NAT techniques, all are considered highly sensitive and results are objective and do not require interpretation. As far as “applicability”, this was also determined to be “low risk” for all studies as circulating virus detected by NAT is by definition associated with active infection and the specificity of the reference standard is high. Page | 377

Flow and timing PICO 5a In the “flow and timing” domain, 42 studies were judged “low risk of bias”. Eight studies were judged to be “unclear risk of bias”. In 7 it could not be determined whether the index and reference tests were performed on the same specimen or within <1 month, and in 1 there were an unknown number of participants so we could not judge if all were included in the final analysis. PICO 9 Four studies were judged to be “low risk” as index test and reference testing were performed on the same specimens at various time points throughout, and all patients were included in the final analyses. One study was judged to be “high risk” as not all patients enrolled were included in the analysis, only those who completed protocol. HCV core Ag for diagnosis of active HCV infection Abbott ARCHITECT There were 20 studies included in the bivariate analysis with 11 820 total samples. Based on studies reporting paired (sensitivity and specificity) data, the pooled sensitivity regardless of HCV Ab status was 93.4% (95% CI 88.7, 96.2), sensitivity was 98.7% (95% CI 96.9, 99.4), positive likelihood ratio (LR) was 71.8 (95% CI 28.6, 160.3), and negative LR 0.07 (0.04, 0.12) (Table 2, Fig. 4a). The pooled sensitivity estimate from a univariate analysis was 94.1% (95% CI 92.4, 95.7) and included 10 additional studies that only contributed data for sensitivity with a total of 12,788 samples (Table 2, Figure 4b). Among 16 studies with known HCV Ab positive samples, the sensitivity was 92.5% (95% CI 86.9, 95.8) and specificity 97.8% (95% CI 94.7, 99.1) (Table 2, Figure 4c). From 4 studies that analysed HCV Ab negative samples, the pooled sensitivity was 74.4% (95% CI 6.2, 99.2) and specificity was 98.8% (97.2, 99.5) (Table 2, Fig. 4d). Figure 5 presents the pooled sensitivity and specificity estimates (a) regardless of HCV Ab status, (b) for HCV Ab-positive samples only, and (c) for HCV Ab-negative samples. In plots (a) and (b) the summary point approached the upper left corner suggesting good accuracy of the ARCHITECT test for diagnosis of HCV infection. Plot (c) demonstrated the broad 95% confidence interval among Ab-negative specimens. Heterogeneity was visually assessed in Figs 4 and 5 and with τ2 (Fig. 2). The studies were relatively homogeneous (Fig. 4a). A meta-regression was not possible given the limited amount of data on predefined covariates. There were three outlier studies in respect to sensitivity: Ergünay, Florea and Gu (72%, 74% and 44% sensitivity, respectively). Antibody status was known for Gu and performance was similar across antibody-positive and antibody-negative samples (44.0% and 41.7%, respectively). Other covariates were examined to assess reasons for low sensitivity. In the Ergünay study, HIV and HBV coinfection status were unknown, 60.2% of participants had HCV genotype 1b infection, 2.2% genotype 1a, 0.8% genotypes 3 and 4, and 35.8% were unknown (Table 3). In the Florea study, there were Page | 378

no HIV or HBV infected patients, but genotype status was unknown. For specificity, the results are even more homogeneous with only 1 outlier, the Medici study. There are no demographic data for this study as it was performed on anonymous samples. Overall, genotype distribution was reported for 15 studies (Table 3a) with genotype 1b being the most prevalent and genotypes 5 and 6 minimally studied. Ortho ELISA-Ag Five studies were included in the bivariate analysis with 1177 total samples. The pooled sensitivity was 93.2% (95% CI 81.6, 97.7), specificity 99.2% (95% CI 87.9, 100), positive LR 116.5 (95% CI 6.7, 977), and negative LR 0.06 (95% CI 0.02, 0.07) (Fig. 6, Table 2). Univariate analysis with one additional study by Agha resulted in a pooled sensitivity of 90.8% (95% CI 83.5, 98.2) (Table 2). Figure 7 demonstrates the bivariate pooled sensitivity and specificity estimates, with the summary point approaching the upper left corner suggesting good accuracy of the Ortho ELISA-Ag test for diagnosis of HCV infection though the data exhibit some heterogeneity demonstrated by the wide 95% CI. Heterogeneity was also visually assessed in the forest plot (Fig. 6) and with τ2 (Fig. 2) with two outlier studies, Nübling and El-Sayed. Both studies reported unknown HIV or HBV coinfection information, and genotype distribution was unknown for El-Sayed. The genotype distribution in the Nübling study was 11.5% genotype 1 not specified, 42.3% genotype 1a, 19.2% genotype 1b, 11.5% genotype 2, and 15.4% genotype 3. This study was performed in 494 total plasma samples from 52 subjects at various time points during HCV infection with varying levels of HCV RNA. The data were not stratified by antibody status, and the raw quantitative information was no longer available. The authors noted that panels later in the course of infection with higher and more consistent HCV RNA levels had improved correlation with HCVcAg detection but no sensitivity or specificity data were calculated. Bio-RAD Monolisa HCV Ag-Ab ULTRA Five studies with 525 total samples were included. Given heterogeneity observed in the forest plot, a pooled analysis was not performed and only descriptive statistics were examined (Fig. 8). The Nastouli and Schnuriger studies have substantially different results – sensitivities of 61.9 (95% CI 38.6, 81.9) and 95% (95% CI 75.1, 99.9), respectively. Each study was performed in participants with 100% HIV coinfection, though the genotype distribution differed with more genotype 1 patients in the Nastouli study, and more genotype 3 and 4 in the Schnuriger study (Table 3). The Tuke study demonstrated the lowest sensitivity of 28.6% (95% CI 20.4, 37.7). This study was performed in pre-seroconversion HCV Ab negative specimens only. Among the HCV genotypes, sensitivity was 33% for genotype 1a, 41% for genotype 1b, 29% for genotype 2, 0% for genotype 3, and 0% for unknown genotype (data not shown, obtained from original article). The authors also noted the sensitivity improved to 71% when limited to specimens with HCV RNA >106IU/mL, though remained negative in 7 genotype 3 samples with viral load >2 million IU/mL. The Laperche study was also performed in HCV Ab negative specimens with a broad distribution Page | 379

of genotypes: 11.4% genotype 1a, 34.3% genotype 1b, 25.7% genotype 2, 14.3% genotype 3, 5.7% genotype 4, and 2.9% unknown. The sensitivity was 40.9% (95% CI 29.3, 53.2). The Vermeersch study was the largest with 337 samples and was the only with data to calculate specificity. The reported sensitivity was 93.83% (95% CI 90.2, 96.4) and specificity 94.9% (95% CI 89.9, 99.8). The study was done on anonymous samples without known HIV or HBV status or genotype. The reference standard was incomplete as RNA testing was done only on 61 random samples and all samples with discordant Ab and Ag result. Seventy-eight samples were antibody negative. Genotype distribution was unknown. EIKEN Lumispot HCV Ag Two studies only utilized the Lumispot assay. The first included 155 samples and the sensitivity reported was 98.1% (95% CI 95.9, 100) (Table 2).53 The majority of samples were genotype 1 (65.2%) with the remaining genotype 2. The second study (Murayama et al.) compared the Lumispot to Fujirebio Lumipulse and Abbott ARCHITECT.16 There were 80 participants, and the reported sensitivity was 97.5% (95% CI 94.1, 100). The Abbott ARCHITECT sensitivity in that study was 100%, suggesting a bias towards better performance. Not enough data were reported to determine specificity in either study. Fujirebio Lumipulse Ortho HCV Ag Only one study was performed using the Lumipulse test with 80 participants comparing against Lumispot and Abbott ARCHITECT.16 Sensitivity for the Lumipulse was reported as 95% (95% CI 90.2, 99.8) (Table 2). The Abbott ARCHITECT sensitivity in that study was 100%, suggesting a bias towards better performance. Not enough data was reported to determine specificity. Hunan Jynda Bioengineering Group HCV Core Ag ELISA There were 4 studies included in the bivariate analysis with 524 total samples. The pooled sensitivity was 59.5% (95% CI 46, 71.7), specificity 82.9% (95% CI 58.6, 94.3), positive LR 3.5 (95% CI 1.1, 12.6), and negative LR 0.28 (95% CI 0.2, 0.3; Table 2). Both the forest plot (Fig. 9) and bivariate analysis (Fig. 10; Table 2) demonstrated heterogeneity among the four studies, which limited confidence in the pooled estimate. No covariate assessment was performed as HIV status, HBV status and genotype distribution were unknown for all studies. DiaSorin S.A. Murex Ag/Ab EIA Four studies with a total sample size of 770 were available; however, given substantial heterogeneity in the forest plot (Fig. 11) a pooled estimate was not calculated. The sensitivity estimates varied between 50% and 100%. Heterogeneity was largely secondary to one outlier study by Tuke where a sensitivity of 50% (95% CI 40.4%, 59.6%) was reported. As reported above, this study was performed in HCV Ab-negative specimens in the “window period” of acute HCV infection. The authors note that when analysis was limited to specimens with viral load HCV RNA >106 IU/mL, there was an increase in sensitivity from 50% to 98% Page | 380

(data not shown). Specificity could not be calculated. The El-Emshaty study is the smallest with 39 participants and reported a sensitivity of 91.3% (95% CI 71.9, 98.9), and specificity of 100% (95% CI 75.9, 100). Genotype distribution was unknown, though the study was performed in Egypt where genotype 4 is most prevalent. The Alzahrani study included 418 samples from 118 female adult participants and reported a sensitivity of 97.4% (95% CI 92.6%, 99.5%) and specificity of 100% (95% CI 98.4, 100). Finally, the Yang study conducted in Taiwan included 201 participants, 25% with HIV coinfection, and unknown genotype distribution. The reported sensitivity was 100% (95% CI 96.5, 100) and specificity 83.8% (95%CI 74.8, 90.2). Quantitative data Three studies provided quantitative data for analysis.33, 41, 44 All used the Abbott ARCHITECT HCV Core Ag Assay in comparison with NAT. Non-parametric regression of these pooled quantitative data was used to visually assess the correlation between HCVcAg and RNA (Fig. 12). The few points with negative HCVcAg were shown to occur at RNA levels < 3000 IU/mL where loss of linearity was also noted. There were two outlier points between 10 000 and 100 000 IU/mL and an additional point on the threshold cut-off for positivity. No further data on genotype or coinfection information was provided to further characterize these points. HCV core Ag for treatment monitoring Two studies evaluated HCVcAg compared to NAT at SVR, both using the Abbott ARCHITECT index test.65, 66 One additional study assessed the accuracy of HCVcAg compared to NAT at EVR only.67 There were not enough studies to perform a meta-analysis. Results for sensitivity and specificity of the index test compared to NAT at baseline, EVR, and SVR were calculated and summarized in Table 4. These data do not evaluate the accuracy of HCVcAg at EVR to predict SVR but rather assess how the tests correlate at each specific time point and thus shed light on differences in the kinetics of core antigen and RNA. Two additional studies assessed timing of HCVcAg in EVR as a predictor of SVR using the Fujiribio Lumipulse test.68,69 Descriptive statistics of each including demographic data, HIV and HBV coinfection, and HCV genotype distribution are presented in Tables 1b and 3b. HCV core Ag performance at different time-points during treatment The Feng study included 32 adults without HIV or HBV coinfection. All participants had genotype 1b chronic HCV infection with viral loads >2000 IU/mL. The sensitivity of HCVcAg at baseline, EVR, and SVR was reported to be 100%, though specificity of EVR was 88.9% (95% CI 68.4, 100). There were 21 patients who achieved SVR, and 11 whose HCVcAg and HCV RNA remained positive 24 weeks after completion of therapy. In all 11 patients, the HCV viral load was >104 IU/mL. The Loggi-study enrolled 35 adult patients without HIV or HBV coinfection; 20% had genotype 1a, 80% had genotype 1b HCV infection. Seventeen patients achieved sustained Page | 381

virological response. The baseline sensitivity of the HCVcAg was 100% without enough data to calculate specificity. Sensitivity at EVR was 73.5% (95% CI 58.7, 88.4) with 100% specificity. The false negatives occurred in samples with viral loads between 15 and 10 000 IU/mL. For SVR, the sensitivity was 100% with 94.1% specificity (95% CI 82.9, 100). The Moscato study analysed samples from 23 patients with unknown demographic information and included 4% genotype 1a, 39.1% genotype 1b, 26.1% genotype 2, 21.7% genotype 3, and 8.7% genotype 4. Baseline and SVR data comparing HCVcAg to NAT were not reported. Direct comparison of HCVcAg compared to NAT at 4 weeks for 10 patients had 100% sensitivity with 70% specificity (95% CI 41.6, 98.4). Three false-positive HCVcAg results were obtained: for 1 of which HCVcAg turned negative 1 month later and 2 turned negative 2 months later (12 weeks into therapy). For three patients tested at 12 weeks, correlation between qualitative results of HCVcAg and HCV NAT was complete. In 9 additional patients with unknown demographic data, correlation between quantitative HCV RNA and HCVcAg was assessed in 54 serum specimens collected at various time points during treatment. Among these samples, authors reported 100% sensitivity of cAg, including 9 specimens with low-level viraemia between 100 and 1000 IU/mL. HCV Core Ag as a predictor of SVR Abbott ARCHITECT Included in the Feng study presented above, was an assessment of measurement of HCVcAg for EVR as a predictor of SVR. The study found a sensitivity of 100% with a specificity of 28% of EVR to predict SVR. This translated into a positive predictive value (PPV) 72% and a negative predictive value (NPV) of 100%. At 4 weeks after therapy initiation (RVR), the performance was inversed with a sensitivity of 29% and a specificity of 100%. The best measure was identified to be a log10 reduction in HCVcAg (ΔHCV Ag) at 144 hours with 95% sensitivity and 73% specificity.

Fujirebio Lumipulse The Takahashi study included 60 genotype 1b patients, and 30 genotype 2 patients with unknown HIV and HBV coinfection status. Serum HCV core Ag was measured at baseline and at 3 days, 1 week, 2 weeks, 4 weeks, and 12 weeks of treatment while qualitative NAT was performed at 12 weeks to assess EVR, and 24 weeks after completion of therapy for SVR. SVR was achieved in 50% of genotype 1b patients, and 90% of genotype 2 patients. In genotype 1b patients, HCVcAg was higher at each time point among the non-SVR group compared to the SVR group, while in genotype 2 patients there was no difference seen in HCVcAg quantity over time between the 3 non-SVR patients and those who achieved SVR, and HCVcAg was below detection limit in all genotype 2 patients by day 14. For genotype 1b, HCVcAg level on day 7 was found to be the best predictor for SVR with sensitivity 79.4%, specificity 88.5%, PPV 90%, NPV 76.7%, and accuracy of 83.3% (Table 5).

Page | 382

In the Fujino study, 49 adult genotype 1b patients were initially enrolled, though 44 completed protocol and were included in the analysis. Patients with HBV were excluded and HIV status was not described. SVR was achieved in 10 patients. HCVcAg and RNA were measured on days 1, 7, and 14 of therapy. Four of the SVR group had negative HCVcAg on day 1 of therapy while all had positive NAT. Negative HCVcAg on day 7 of therapy gave sensitivity 57.1%, specificity 93.3%, PPV 80%, NPV 82.4%, and accuracy of 81.8% in prediction of SVR (Table 5), while undetectable HCV RNA on day 7 yielded sensitivity 100%, specificity 87.2%, PPV 50%, NPV 100%, and accuracy of 88.6%.

6. Discussion This systematic review addressed diagnostic accuracy of HCV core antigen tests for identification of active HCV infection among those with and without positive HCV antibody through an analysis of 50 published studies (PICO5a) that utilized 7 different index tests. Additionally, accuracy of HCV core antigen tests for treatment monitoring and as a test of cure was assessed in 5 published studies (PICO 9). The Abbott ARCHITECT HCV Core Antigen test had the highest sensitivity (93.4%), while specificity was similar to that of the Ortho ELISA-Ag (98.7% vs 99.2%). The estimates for both sensitivity and specificity were more precise for the ARCHITECT assay. This was partly because the ARCHITECT was the most extensively studied, with 30 publications included in this review compared to 5 studies included for Ortho, but also partly because of the greater homogeneity among the ARCHITECT studies. The likelihood ratios for both tests were also very favourable with the positive LR >10 indicative of a large increase in probability of disease with a positive result and negative LR <0.1 indicative of a moderate decrease in the probability of disease with a negative result. The EIKEN Lumispot and Fujirebio Lumipulse were designed with the same principle of technology as the ARCHITECT and have similar sensitivity and specificity, though assessment was limited to 1 and 2 studies. Tests such as the Hunan Jynda assay have the lowest sensitivity (59.5%), which supports the notion that signal amplification (as with chemiluminesence) is necessary to achieve adequate detection limits. Quantitative analysis of data available from 3 studies using the ARCHITECT demonstrated close correlation between HCVcAg and RNA, though the linearity declined around an HCV RNA level of 3000 IU/mL, which is consistent with the analytical limit of detection reported by Abbott. All studies included with treatment monitoring and SVR data were in patients on interferon (IFN)-based therapies. Data was limited and a meta-analysis was not possible. Descriptive analysis found the sensitivity of ARCHITECT at EVR in comparison to RNA ranging from 74% to 100% with specificity from 70% to 100% and at SVR (only assessed in 2 studies) sensitivity was 100% and specificity ranged from 94% to 100%. HCVcAg predictive accuracy for SVR was described in only 3 studies, 1 using the ARCHITECT and 2 using the Lumipulse. All three included mostly genotype 1b patients, and results indicated best

Page | 383

predictive accuracy of core antigen for SVR from the decline or reversion to negative early on in therapy at 6–7 days. There were limitations in the data summarized in this review. For several index tests, there were not enough studies to derive pooled estimates and descriptive analyses only could be completed. There was substantial heterogeneity among all index tests aside from the ARCHITECT, and there were not enough data to perform planned sensitivity analyses of covariates and meta-regression of subgroups; thus descriptive statistics were substituted. From the limited data available, it is clear that data on core antigen test performance in genotypes 4, 5 and 6 is largely lacking, which limits the conclusions. Most of the studies were performed in high-resource settings and might not be reflective of the population that will be tested if HCVcAg diagnostics are implemented in LMICs. Furthermore, most studies were performed in reference laboratories and test performance might be decreased if tests are applied in routine laboratories. To assess treatment monitoring, only 2 studies measured HCVcAg in comparison with NAT at SVR and only descriptive analyses could be conducted. Several of the studies found in the search were designed to answer a different question from that of the PICO structure – whether an early decline in HCVcAg could predict SVR and at what time period this was most accurate. These data were also described, though again meta-analysis could not be performed. Additionally, there were no studies using DAA IFN-free treatment regimens, thus the results from this descriptive analysis might not be reflective of the results that are to be expected with DAAs. The timing of this review has also occurred during a rapidly changing landscape; the utility of viral load monitoring while on treatment with these highly effective therapies, and thus utility of HCVcAg as a surrogate of NAT, has been called into question.9 Strengths and limitations of the systematic review Strengths of this review include the development of an a priori protocol for the literature search, article selection, data extraction, and assessment of methodological quality. The search was performed without language restriction, though ultimately 3 articles were excluded for inability to find appropriate translation for Russian, Korean and Polish. Nevertheless, studies may have been missed in the comprehensive search, and subsequent studies published after the search date could not be included. Article selection and standardized data extraction in accordance with the predefined protocol was ensured by independent reviewers. Authors were contacted for missing data and clarifications, though some studies were excluded due to lack of author response or inability to provide original data. In the analysis, bivariate random effects modeling was used when appropriate to derive pooled estimates and univariate analyses were performed in effort to utilize all available data. Further research suggested The data limitations in this review highlight a need for better surveillance data that will inform an understanding of how many patients are missed by assays that have higher limits Page | 384

of detection (e.g. 3000 IU/mL for ARCHITECT). Furthermore, a better understanding is necessary on the outcomes of patients with low viral loads: are these patients more likely to resolve their infection or at least less prone to develop HCV disease, or do they still have notable disease progression that would make them eligible for treatment? Similarly, more information is necessary on patients with high viral loads and negative HCVcAg to inform the optimization of antigen detection. The fluctuation in RNA during the pre-seroconversion phase and correlation of core antigen is also poorly understood. Additionally, research is required to determine how covariates such as HIV or HBV coinfection or genotype may impact the accuracy of HCVcAg for diagnosing active infection as well as for monitoring treatment outcomes. The kinetics of HCVcAg with treatment also need to be evaluated further, particularly in the context of new DAA regimens.

7. Summary In summary, this systematic review showed that there are several HCVcAg assays associated with high sensitivity (>90%) and specificity (>98%) compared with NAT. While even those with the highest performance do not reach the sensitivity of NAT, well-performing HCVcAg tests with an analytical sensitivity reaching into the femtomolar range (~3000 IU/mL), which translates into diagnostic sensitivity of about 95%, could serve as a replacement for NAT for HCV detection. This is the case particularly if HCVcAg are more affordable than NAT, which is conceivable from the cost of goods for the test. Furthermore, HCVcAg tests could be applied for a one-step screening test as they turn positive earlier than antibody tests (1–2 days after HCV RNA appears) and have a high specificity, thus not requiring any further confirmatory testing. For both core antigen tests and NATs to reach a larger population at risk in LMICs, tests with better point-of-care (POC) suitability need to be developed and sample processing and transport mechanisms need to be improved to optimize the use of platforms requiring reference laboratories. HCVcAg tests are possible on a POC platform; however, given the need for signal amplification (as suggested by this review), an instrument-free assay is not conceivable in the near future. Furthermore, sample processing is necessary. The role for HCVcAg as a substitute for NAT in assessment for SVR remains less clear. While the two studies presented show excellent results, and the quantitative data from PICO5a supports close correlation of HCVcAg with RNA above 3000 IU/mL, the kinetics of HCVcAg with treatment are not fully understood. Particularly, data on the early kinetics of HCVcAg, and the appropriate timing of assessment for predicting SVR are limited.

Acknowledgements We would like to thank all of the authors of the studies who provided additional data necessary to complete this review. We also wish to thank the following individuals: Kuniaki Arai at the World Health Organization for translation and extraction of articles in Japanese, Jeanne Chauffor at Médecins Sans Frontières and Audrey Albertini at FIND for translation articles in French, Wei Huang at Boston Medical Center for translation and extraction of Page | 385

articles in Mandarin, Joseph Tucker at University of North Carolina for translation of author communication in Mandarin, David Flynn from Boston University School of Medicine Alumni Medical Library, Genevieve Gore from McGill University and Melanie Cedrone from the Biomedical Library at University of Pennsylvania for their help with search terms, and Ranald Sutherland and Martin Brusdeilins from FIND for their input on the technical aspects and commercial availability of HCV Core Ag tests.

References 1. 2. Gower E, Estes C, Blach S, Razavi-Shearer K, Razavi H. Global epidemiology and genotype distribution of the hepatitis C virus infection. J Hepatol. 2014;61(1 Suppl):S45‒S57. Smith DB, Bukh J, Kuiken C, Muerhoff AS, Rice CM, Stapleton JT, et al. Expanded classification of hepatitis C virus into 7 genotypes and 67 subtypes: updated criteria and genotype assignment web resource. Hepatology. 2014;59(1):318‒27. Reed KE, Rice CM. Overview of hepatitis C virus genome structure, polyprotein processing, and protein properties. Curr Top Microbiol Immunol. 2000;242:55‒84. ubuisson J. Hepatitis C virus proteins. World J Gastroenterol. 2007;13(17):2406‒15. Tedder RS, Tuke P, Wallis N, Wright M, Nicholson L, Grant PR. Therapy-induced clearance of HCV core antigen from plasma predicts an end of treatment viral response. J Viral Hepat. 2013;20(1):65‒71. Sanz C, Tassies D, Costa J, Freire C, Pereira A. The first case of HCV infection detected before seroconversion in blood donors tested by HCV core antigen ELISA. Transfusion. 2002;42(4):505‒6. Tanaka E, Kiyosawa K, Matsumoto A, Kashiwakuma T, Hasegawa A, Mori H, et al. Serum levels of hepatitis C virus core protein in patients with chronic hepatitis C treated with interferon alfa. Hepatology. 1996;23(6):1330‒3. Tillmann HL. Hepatitis C virus core antigen testing: role in diagnosis, disease monitoring and treatment. World J Gastroenterol. 2014;20(22):6701‒6. Sidharthan S, Kohli A, Sims Z, Nelson A, Osinusi A, Masur H, et al. Utility of hepatitis C viral load monitoring on direct-acting antiviral therapy. Clin Infect Dis. 2015;60(12):1743‒51. Ford N, Swan T, Beyer P, Hirnschall G, Easterbrook P, Wiktor S. Simplification of antiviral hepatitis C virus therapy to support expanded access in resource-limited settings. J Hepatol. 2014;61(1 Suppl):S132‒S138. Deville WL, Buntinx F, Bouter LM, Montori VM, de Vet HC, van der Windt DA, et al. Conducting systematic reviews of diagnostic studies: didactic guidelines. BMC Med Res Methodol. 2002;2:9. Pai M, McCulloch M, Gorman JD, Pai N, Enanoria W, Kennedy G, et al. Systematic reviews and metaanalyses: an illustrated, step-by-step guide. Natl Med J India. 2004;17(2):86‒95. Committee on Standards for Systematic Reviews of Comparative Effectiveness Research. In: Eden J, Levit L, Berg A, Morton S, editors. Finding what works in health care: standards for systematic reviews. Washington (DC): National Academy of SCiences; 2011. Whiting PF, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155(8):529‒36. Cleveland W. Robust locally weighted regression and smoothing scatterplots. J Am Stat Assoc. 1979;74(368):829‒36. Murayama A, Sugiyama N, Watashi K, Masaki T, Suzuki R, Aizaki H, et al. Japanese reference panel of blood specimens for evaluation of hepatitis C virus RNA and core antigen quantitative assays. J Clin Microbiol. 2012;50(6):1943‒9.

3. 4. 5. 6. 7.

8. 9. 10.

11. 12. 13.

14. 15. 16.

Page | 386

17. 18. 19. 20.

Tuke PW, Grant PR, Waite J, Kitchen AD, Eglin RP, Tedder RS. Hepatitis C virus window-phase infections: closing the window on hepatitis C virus. Transfusion. 2008;48(4):594‒600. Gu J, Yu T, Liang Z. Performances of HCV Ag or HCV RNA kits for screening of HCV-infected samples. Chinese Journal of Biologicals. 2014;27(9):1181‒84. van Helden J, Weiskirchen R. [Hepatitis C diagnostics: clinical evaluation of the HCV-core antigen determination]. Z Gastroenterol. 2014;52(10):1164‒70. Schnuriger A, Dominguez S, Valantin MA, Tubiana R, Duvivier C, Ghosn J, et al. Early detection of hepatitis C virus infection using a new combined antigen-antibody detection assay: potential use in HIV coinfected individuals. Pathol Biol (Paris). 2006;54(10):578‒86. Ohta H, Takemura M, Furuta N, Akiyama M, Katagiri Y, Ohashi H, et al. Clinical significance and problems in HCV measurement – comparison of CLEIA method with PCR method. Rinsho Byori. 2004;52(10):813‒ 8. Okazaki K, Nishiyama Y, Saitou T, Shibata N, Yamamoto C, Oosaga J, et al. [Fundamental evaluation of HCV core antigen method comparison with Cobas Amplicor HCV monitor v2.0 (high range method)]. Rinsho Byori. 2008;56(2):95‒100. Buket CA, Ayse A, Selcuk K, Suleyman O, Emel SC. Comparison of HCV core antigen and anti-HCV with HCV RNA results. Afr Health Sci. 2014;14(4):816‒20. Chevaliez S, Soulier A, Poiteau L, Bouvier-Alias M, Pawlotsky JM. Clinical utility of hepatitis C virus core antigen quantification in patients with chronic hepatitis C. J Clin Virol. 2014;61(1):145‒8. Descamps V, Op de Beeck A, Plassart C, Brochot E, Francois C, Helle F, et al. Strong correlation between liver and serum levels of hepatitis C virus core antigen and RNA in chronically infected patients. J Clin Microbiol. 2012;50(2):465‒8. Durante-Mangoni E, Vallefuoco L, Sorrentino R, Iossa D, Perna E, Molaro R, et al. Clinico-pathological significance of hepatitis C virus core antigen levels in chronic infection. J Med Virol. 2013;85(11):1913‒8. Thong VD, Akkarathamrongsin S, Avihingsanon A, Theamboonlers A, Poovorawan Y, Tangkijvanich P. The correlation between hepatitis C core antigen and hepatitis C virus RNA levels with respect to human immunodeficiency virus status, hepatitis C virus genotype and interferon-lambda-4 polymorphism. Intervirology. 2015;58(2):73‒9. Ergunay K, Sener B, Alp A, Karakaya J, Hascelik G. Utility of a commercial quantitative hepatitis C virus core antigen assay in a diagnostic laboratory setting. Diagn Microbiol Infect Dis. 2011;70(4):486‒91. Florea D, Neaga E, Nicolae I, Maxim D, Popa M, Otelea D. Clinical usefulness of HCV core antigen assay for the management of patients with chronic hepatitis C. J Gastrointestin Liver Dis. 2014;23(4):393‒6. Garbuglia AR, Monachetti A, Galli C, Sabatini R, Ferreri ML, Capobianchi MR, et al. HCV core antigen and HCV-RNA in HIV/HCV co-infected patients with different HCV genotypes. BMC Infect Dis. 2014;14:222. Hadziyannis E, Minopetrou M, Georgiou A, Spanou F, Koskinas J. Is HCV core antigen a reliable marker of viral load? An evaluation of HCV core antigen automated immunoassay. Ann Gastroenterol. 2013;26(2):146‒9. Heidrich B, Pischke S, Helfritz FA, Mederacke I, Kirschner J, Schneider J, et al. Hepatitis C virus core antigen testing in liver and kidney transplant recipients. J Viral Hepat. 2014;21(11):769‒79. Kadkhoda K, Smart G. HCV antigen testing for the diagnosis of hepatitis C infection: a cost-efficient algorithm. Clin Lab. 2014;60(4):677‒80. Karabay O, Öğütlü A, Ankarali H, Özdemir F, Karabay M, Gozdas H. Comparison of HCV RNA and HCV core antigen in genotype 1 chronic hepatitis C patients. Acta Medica Mediterranea. 2014;30(5):1025‒28. Kesli R, Polat H, Terzi Y, Kurtoglu MG, Uyar Y. Comparison of a newly developed automated and quantitative hepatitis C virus (HCV) core antigen test with the HCV RNA assay for clinical usefulness in confirming anti-HCV results. J Clin Microbiol. 2011;49(12):4089‒93.

21.

22.

23. 24. 25.

26. 27.

28. 29. 30. 31.

32. 33. 34. 35.

Page | 387

36.

Köroğlu M, Ak S, Ak M, Yakupoğulları Y, Özer A. Evaluation of diagnostic performance of new antigen based enzyme immune assay for diagnosis of hepatitis C virus (HCV) infections. Afr J Microbiol Res. 2012;6(4):809‒12. Kuo YH, Chang KC, Wang JH, Tsai PS, Hung SF, Hung CH, et al. Is hepatitis C virus core antigen an adequate marker for community screening? J Clin Microbiol. 2012;50(6):1989‒93. Li Cavoli G, Zagarrigo C, Schillaci O, Servillo F, Tralongo A, Coglitore M, et al. Hepatitis C virus core antigen test in monitoring of dialysis patients. Hepat Res Treat. 2012;2012:832021. Mederacke I, Ciesek S, Raupach R, Wursthorn K, Deterding K, Steinmann E, et al. Kinetics of HCV core antigen during antiviral treatment of acute and chronic hepatitis C as determined by a novel chemiluminescent microparticle immunoassay. In: The International Liver CongressTM 2009. 44th Annual Meeting of the European Associateion for the Study of the Liver (EASL); Copenhagen, Denmark; 2009 [S129]. Mederacke I, Potthoff A, Meyer-Olson D, Meier M, Raupach R, Manns MP, et al. HCV core antigen testing in HIV- and HBV-coinfected patients, and in HCV-infected patients on hemodialysis. J Clin Virol. 2012;53(2):110‒5. Medici MC, Furlini G, Rodella A, Fuertes A, Monachetti A, Calderaro A, et al. Hepatitis C virus core antigen: analytical performances, correlation with viremia and potential applications of a quantitative, automated immunoassay. J Clin Virol. 2011;51(4):264‒9. Miedouge M, Saune K, Kamar N, Rieu M, Rostaing L, Izopet J. Analytical evaluation of HCV core antigen and interest for HCV screening in haemodialysis patients. J Clin Virol. 2010;48(1):18‒21. Ottiger C, Gygli N, Huber AR. Detection limit of architect hepatitis C core antigen assay in correlation with HCV RNA, and renewed confirmation algorithm for reactive anti-HCV samples. J Clin Virol. 2013;58(3):535‒40. Park Y, Lee JH, Kim BS, Kim do Y, Han KH, Kim HS. New automated hepatitis C virus (HCV) core antigen assay as an alternative to real-time PCR for HCV RNA quantification. J Clin Microbiol. 2010;48(6):2253‒6. Reyes-Mendez MA, Juarez-Figueroa L, Iracheta-Hernandez P, Medina-Islas Y, Ruiz-Gonzalez V. Comparison of two diagnostic algorithms for the identification of patients with HCV viremia using a new HCV antigen test. Ann Hepatol. 2014;13(3):337‒42. Rouet F, Deleplancque L, Mboumba BB, Sica J, Mouinga-Ondeme A, Liegeois F, et al. Usefulness of a fourth generation ELISA assay for the reliable identification of HCV infection in HIV-positive adults from Gabon (Central Africa). PLoS One. 2015;10(1):e0116975. Russi S, Sansonno D, Mariggio MA, Vinella A, Pavone F, Lauletta G, et al. Assessment of total hepatitis C virus (HCV) core protein in HCV-related mixed cryoglobulinemia. Arthritis Res Ther. 2014;16(2):R73. Vanhommerig JW, van de Laar TJ, Koot M, van Rooijen MS, Schinkel J, Speksnijder AG, et al. Evaluation of a hepatitis C virus (HCV) antigen assay for routine HCV screening among men who have sex with men infected with HIV. J Virol Methods. 2015;213:147‒50. Vermehren J, Susser S, Berger A, Perner D, Peiffer KH, Allwinn R, et al. Clinical utility of the ARCHITECT HCV Ag assay for early treatment monitoring in patients with chronic hepatitis C genotype 1 infection. J Clin Virol. 2012;55(1):17‒22. Laperche S, Le Marrec N, Girault A, Bouchardeau F, Servant-Delmas A, Maniez-Montreuil M, et al. Simultaneous detection of hepatitis C virus (HCV) core antigen and anti-HCV antibodies improves the early detection of HCV infection. J Clin Microbiol. 2005;43(8):3877‒83. Nastouli E, Thomson EC, Karayiannis P, Main J, McClure M, Muir D. Diagnosing acute hepatitis C in HIVinfected patients: nucleic acid testing compared with antibody and antigen-antibody detecting methods. J Clin Virol. 2009;44(1):78‒80. Vermeersch P, Van Ranst M, Lagrou K. Evaluation of the use of a combined HCV antigen/antibody assay in routine laboratory practice. Acta Clin Belg. 2010;65(4):245‒7.

37. 38. 39.

40.

41.

42. 43.

44. 45.

46.

47. 48.

49.

50.

51.

52.

Page | 388

53. 54. 55.

Saito R, Yokota H, Takahashi E, Mashige F, Yoneyama A, Nakahara K, et al. Performance of an automated system for quantitation of hepatitis C virus core antigen. J Virol Methods. 2003;112(1-2):93‒7. Lu YC, Jiang ZY, Kuang YL, Chen WS, Tan YZ, Li DR. Clinical value of the detection of hepatitis C virus core antigen. Xi Bao Yu Fen Zi Mian Yi Xue Za Zhi. 2007;23(7):635‒7. Ouyang Y, Tan DM, Li TG, Zhou H, Tan C. Qualitative detection of hepatitis C virus core antigen in the serum in patients with chronic hepatitis C. Zhong Nan Da Xue Xue Bao Yi Xue Ban. 2006;31(6):894‒6, 905. Zhang HQ, Li SB, Wang GH, Chen K, Song XG, Feng XY. Detection of hepatitis C virus core antigen for early diagnosis of hepatitis C virus infection in plasma donor in China. World J Gastroenterol. 2007;13(19):2738‒42. Zhu H, Zhang T, Fan C, Xiao X, Zhang F. Clinical significance of testing hepatitis C virus core antigen. Journal of Dalian Medical University. 2010;32(2):211‒2, 8. El-Emshaty WM, Raafat D, Elghannam DM, Saudy N, Eltoraby EE, Metwalli AE. Diagnostic performance of an immunoassay for simultaneous detection of HCV core antigen and antibodies among haemodialysis patients. Braz J Microbiol. 2011;42(1):303‒9. Yang JF, Lin YY, Hsieh MH, Tsai CH, Liu SF, Yu ML, et al. Performance characteristics of a combined hepatitis C virus core antigen and anti-hepatitis C virus antibody test in different patient groups. Kaohsiung J Med Sci. 2011;27(7):258‒63. Alzahrani AJ. Simultaneous detection of hepatitis C virus core antigen and antibodies in Saudi drug users using a novel assay. J Med Virol. 2008;80(4):603‒6. Agha S, Tanaka Y, Saudy N, Kurbanov F, Abo-Zeid M, El-Malky M, et al. Reliability of hepatitis C virus core antigen assay for detection of viremia in HCV genotypes 1, 2, 3, and 4 infected blood donors: a collaborative study between Japan, Egypt, and Uzbekistan. J Med Virol. 2004;73(2):216‒22. el-Sayed Zaki M, el-Adrosy H. Recent approach for diagnosis of early HCV infection. Egypt J Immunol. 2004;11(1):123‒9. Letowska M, Brojer E, Mikulska M, Gronowska A, Rosiek A. Hepatitis C core antigen in Polish blood donors. Transfusion. 2004;44(7):1067‒71. Nubling CM, Unger G, Chudy M, Raia S, Lower J. Sensitivity of HCV core antigen and HCV RNA detection in the early infection phase. Transfusion. 2002;42(8):1037‒45. Feng B, Yang RF, Xie Q, Shang J, Kong FY, Zhang HY, et al. Hepatitis C virus core antigen, an earlier and stronger predictor on sustained virological response in patients with genotype 1 HCV infection. BMC Gastroenterol. 2014;14:47. Loggi E, Cursaro C, Scuteri A, Grandini E, Panno AM, Galli S, et al. Patterns of HCV-RNA and HCV core antigen in the early monitoring of standard treatment for chronic hepatitis C. J Clin Virol. 2013;56(3):207‒11. Moscato GA, Giannelli G, Grandi B, Pieri D, Marsi O, Guarducci I, et al. Quantitative determination of hepatitis C core antigen in therapy monitoring for chronic hepatitis C. Intervirology. 2011;54(2):61‒5. Fujino T, Nakamuta M, Aoyagi Y, Fukuizumi K, Takemoto R, Yoshimoto T, et al. Early decline of the HCV core antigen can predict SVR in patients with HCV treated by Pegylated interferon plus ribavirin combination therapy. J Dig Dis. 2009;10(1):21‒5. Takahashi M, Saito H, Higashimoto M, Atsukawa K, Ishii H. Benefit of hepatitis C virus core antigen assay in prediction of therapeutic response to interferon and ribavirin combination therapy. J Clin Microbiol. 2005;43(1):186‒91. Fytili P, Tiemann C, Wang C, Schulz S, Schaffer S, Manns MP, et al. Frequency of very low HCV viremia detected by a highly sensitive HCV-RNA assay. J Clin Virol. 2007;39(4):308‒11.

56.

57. 58.

59.

60. 61.

62. 63. 64. 65.

66.

67. 68.

69.

70.

Page | 389

Tables Table 1a. Characteristics of included studies for PICO 5a grouped alphabetically by index test type Author, year Country and income category Study design Study populati on Age group Number of subjects Proportion with HIV infection Proportion with HBV infection Proportion female Sample type Sample condition

Abbott ARCHITECT HCV Ag Buket, 2014 Chevaliez, 2014 Descamps, 2012 Kazakhstan (B) France (A) France (A) Cohort Cross-sectional Cross-sectional Cohort Cohort Cohort Cross-sectional Cohort Cross-sectional Cross-sectional Cohort Cross-sectional Cohort Cohort Cohort Cohort Broad Broad Broad Broad Broad Broad Broad Broad Broad Broad Broad Broad Healthy Broad Broad Broad Adults Adults Adults Adults Adults Mixed Adults Adults Unknown Unknown Adults Unknown Adults Unknown Adults Adults 115 514 22 114 189 272 76 292 304 105 596 154 212 32 405 92 Unknown Unknown Unknown 0% 44.9% Unknown 0% 100% Unknown Unknown Unknown Unknown Unknown Unknown Unknown 1.1% Unknown Unknown Unknown 0% 0% Unknown 0% 3.8% Unknown Unknown Unknown Unknown Unknown Unknown Unknown 2.2% 56.5% 36.6% 40.1% 43% 28.6% Unknown 75% 25.9% Unknown Unknown 43% Unknown 57.5% 45.5% 52.6% 41.3% Serum Serum Serum Serum Serum Serum Serum Serum Whole Serum Serum Serum Serum Serum Serum Serum Unknown Unknown Frozen Frozen Frozen Frozen Frozen Frozen Unknown Frozen Unknown Unknown Unknown Unknown Unknown Unknown

Durante-Mangoni, 2013 Italy (A) Duy Thong, 2015 Ergünay, 2011 Florea, 2014 Garbuglia, 2014 Gu, 2014 Hadziyannis, 2013 Heidrich, 2014 Kadkhoda, 2014 Kesli, 2011 Köroglu, 2012 Kuo, 2012 Li Cavoli, 2012 Thailand (B) Turkey (A) Romania (B) Italy (A) China (B) Greece (A) Germany (A) Canada (A) Turkey (A) Turkey (A) Taiwan (A) Italy (A)

Page | 390

Mederacke. 2009 Mederacke, 2012 Medici, 2011 Miedouge, 2010 Murayama, 2012 Ottiger, 2013 Park, 2010 Reyes-Méndez, 2014 Rouet, 2015 Russi, 2014 Tedder, 2013 van Helden, 2014 Vanhommerig. 2015 Vermehren, 2012

Germany (A) Germany (A) Italy, Spain (A) France (A) Japan (A) Switzerland (A) South Korea Mexico (B) Gabon (B) Italy (A) UK (A) Germany (A)

Cohort Cross-sectional Cross-sectional Cohort Cross-sectional Cross-sectional Cohort Cross-sectional Cross-sectional Cohort Cohort Cross-sectional

Broad Broad Broad Broad Broad Broad Broad Broad Broad Broad Broad Broad Broad Broad

Unknown Unknown Unknown Unknown Unknown Adults Adults Unknown Adults Adults Unknown Unknown Unknown Adults

118 237 1480 2850 80 97 282 211 54 102 54 3558 93 160

0% 49.50% Unknown Unknown Unknown 6% Unknown Unknown 100.00% 0% 0% 4.40% 100.00% 0%

0% 50.50% Unknown Unknown Unknown 0% Unknown Unknown Unknown 0% 0% 6.60% Unknown 0%

Unknown Unknown 52.6% Unknown Unknown 38.1% 49.3% Unknown 70.1% 78.4% Unknown Unknown 0% 54%

Serum Serum Serum Serum Plasma Plasma Serum Serum Plasma Serum Plasma Serum Serum Serum

Unknown Unknown Frozen Frozen Frozen Frozen Unknown Unknown Frozen Frozen Frozen Unknown Unknown Frozen

Netherlands (A) Cohort Germany (A) Cohort

Page | 391

Author, year

Country and income category

Study design

Study population

Age group

Number of subjects

Proportion with HIV infection

Proportion with HBV infection

Proportion female

Sample type

Sample condition

Bio-RAD MonolisaTM HCV Ag-Ab ULTRA Laperche, 2005 Nastouli, 2008 Schnuriger, 2006 Tuke, 2008 Vermeersch, 2010 France (A) UK (A) France (A) UK (A) Belgium (A) Cohort Cohort Cohort Cross-sectional Cross-sectional Broad Broad Broad Broad Broad Unknown Adults Adults Unknown Unknown 35 25 20 Unknown 337 Unknown 100% 100% Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown 0% Unknown Unknown Unknown Plasma Serum Serum Plasma Serum Frozen Frozen Frozen Frozen Unknown

EIKEN Lumispot HCV Ag Saito, 2003 Murayama, 2012 Japan (A) Japan (A) Cross-sectional Cross-sectional Broad Broad Unknown Unknown 155 80 Unknown Unknown Unknown Unknown Unknown Unknown Serum Plasma Frozen Frozen

Fujirebio Lumipulse Ortho HCV Ag Murayama, 2012 Japan (A) Cross-sectional Broad Unknown 80 Unknown Unknown Unknown Plasma Frozen

Hunan Jynda Bioengineering Group HCV Core Ag ELISA Lu, 2007 Ouyang, 2006 Zhang, 2007 Zhu, 2010 China (B) China (B) China (B) China (B) Cohort Cross-sectional Cohort Cross-sectional Broad Broad Healthy Broad Unknown Unknown Unknown Mixed 191 149 11 173 Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Serum Serum Serum Serum Unknown Unknown Frozen Unknown

DiaSorin S.A. Murex Ag/Ab EIA Alzahrani, 2008 El-Emshaty, 2011 Tuke, 2008 Yang, 2011 Saudi Arabia (A) Egypt (B) UK (A) Taiwan (A) Cohort Cohort Cross-sectional Cohort Broad Broad Broad Broad Adults Adults Unknown Adults 118 39 unknown 201 0.7% Unknown Unknown 25% 6.1% Unknown Unknown 0% 100% 69.2% Unknown 39% Serum Serum Plasma Serum Unknown Frozen Frozen Frozen

Page | 392

Ortho ELISA-Ag Agha, 2004 Egypt, Japan, Uzbekistan (AB) Egypt (B) Poland (A) USA (A) Japan (A) Japan (A) Cohort Broad Unknown Cross-sectional Cohort Cohort Cross-sectional Cohort Broad Healthy Broad Broad Broad Unknown Unknown Unknown Unknown Unknown 50 124 52 225 300 246 Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown 50.3% Serum Serum Plasma Serum Serum Frozen Unknown Frozen Unknown Unknown Serum Unknown

El-Sayed, 2004 Letowska, 2004 Nübling, 2002 Ohta, 2004 Okazaki, 2008

HCV: hepatitis C virus, HIV: human immunodeficiency virus, HBV: hepatitis B virus, Ag: antigen, Ab: antibody, ELISA: enzyme-linked immunosorbent assay, EIA: enzyme immunoassay, UK: United Kingdom, USA: United States of America, A: high-income countries, B: middle-income countries, C: low-income countries by World Bank List of Economies (July 2015)

Table 1b. Characteristics of included studies for PICO 9 grouped alphabetically by index test type Author, year Country and income category Study design Study Age group Number of populatio subjects n Proportion with HIV infection Proportion with HBV infection Proportion female Sample type Sample condition

Abbott ARCHITECT HCV Ag Feng, 2014 Loggi, 2013 Moscato, 2010 China (B) Italy (A) Italy (A) RCT Cohort Cohort Broad Broad Broad Adults Adults Unknown 32 35 23 0% 0% Unknown 0% 0% Unknown 50% 34.4% Unknown Serum Serum Serum Unknown Frozen Frozen

Fujirebio Lumipulse Ortho HCV Ag Fujino, 2009 Takahashi, 2005 Japan (A) Japan (A) Cohort Cohort Broad Broad Adults Adults 90 44 Unknown Unknown Unknown 0% 24% 31.8% Serum Serum Unknown Unknown

HCV: hepatitis C virus, HIV: human immunodeficiency virus, HBV: hepatitis B virus, Ag: antigen, Ab: antibody, RCT: randomized controlled trial, A: high-income countries, B: middle-income countries, C: low-income countries by World Bank List of Economies (July 2015)

Page | 393

Table 2. Diagnostic accuracy by HCVcAg index test type for diagnosis of active HCV infection compared to nucleic acid testing as the reference standard. Results from bivariate, univariate, range of studies, and single studies are all reported. Index Test Abbott ARCHITECT1 Abbott ARCHITECT2 HCV Ab status All All Known Ab positive Known Ab positive Known Ab negative All All All All All All All # Studies (# samples) 20 (11,820) 30 (12,788) 16 (5,246) 26 (6,214) 4 (3,458) 5 (1,177) 6 (1,423) 5 (525) 2 (235) 1 (80) 4 (524) 4 (770) Sensitivity 95% CI 93.4%1 (88.7, 96.2) 94.1%2 (92.4, 95.7) Abbott ARCHITECT1 Abbott ARCHITECT2 Abbott ARCHITECT1 Ortho Ortho ELISA-Ag1 ELISA-Ag2 92.5%1 (86.9, 95.8) 93.3%2 (91.2, 95.3) 74.4%1 (6.2, 99.2) 93.2%1 (81.6, 97.7) 90.8%2 (83.5, 98.2) Bio-RAD MonolisaTM HCV Ag-Ab ULTRA* EIKEN Lumispot HCV Ag Fujirebio Lumipulse Ortho HCV Ag3 Hunan Jynda Bioengineering Group HCV Core Ag ELISA1 Murex Ag/Ab EIA 28.6–95%* 97.5–98.1%* 95%** (90.2, 99.8) 59.5%1 (46, 71.7) 50–100%* 82.9%1 (58.6, 94.3) 83.8–100%* 3.5 (1.1, 12.6) NA 0.28 (0.2, 0.3) NA NA NA+ 94.9%** (89.9, 99.8) ND ND NA NA NA NA NA NA NA NA NA 98.8%1 (97.2, 99.5) 99.2%1 (87.9, 100) ND 62 (2, 198.5) 116.5 (6.7, 977) NA 0.25 (0.003, 0.94) 0.06 (0.02, 0.07) NA Sens: 8.4 (SE 16.6) Spec: 0.2 (SE 0.6) Sens: 1.4 (SE 1.0) Spec: 3.8 (SE 5.1); [–0.4] 122.0 97.8%1 (94.7, 99.1) ND 42 (16.4, 106.4) NA* 0.05 (0.03, 0.08) NA* Sens: 1.4 (SE 0.5) Spec: 1.7 (SE 1.0); [0.02] Sens: 19.4 Specificity 95% CI 98.7%1 (96.9, 99.4) ND 71.8 (28.6, 160.3) NA Positive LR 95% CI 0.07 (0.04, 0.12) NA Negative LR 95% CI Sens: 1.5 (SE 0.6) Spec: 2.3 (SE 1.0); [0.03] Sens: 14.1 τ2 [Covariance]

HCV: hepatitis C virus, cAg: core antigen, Ab: antibody, CI: confidence interval, LR: likelihood ratio, ELISA: enzyme linked immunosorbent assay, EIA: enzyme immunoassay, τ2: Tau squared, SE: standard error. 1. Determined by bivariate meta-analysis – “metandi” command in STATA, 2. Determined by univariate meta -analysis – “metan” command in STATA, *: Meta-analysis not possible, range of results seen across studies reported, **: results from one study only, ND: no data, NA: not applicable – if sensitivity and specificity results were not available from meta-analysis, likelihood ratios were not calculated; +: output of τ2 not interpretable given small number of studies

Page | 394

Table 3a. Available genotype information for studies included in PICO 5a grouped alphabetically by index test type Author, year Number of subjects % G1 % G1a % G1b %G2 %G3 %G4 %G5 %G6 % Other or unknown

Abbott ARCHITECT HCV Ag Chevaliez, 2014 Descamps, 2012 Durante-Mangoni, 2013 Duy Thong, 2015 Ergünay, 2011 Garbuglia, 2014 Hadziyannis, 2013 Kesli, 2011 Li Cavoli, 2012 Mederacke. 2009 Miedouge, 2010 Ottiger, 2013 Russi, 2014 Tedder, 2013 Vermehren, 2012 514 22 114 189 272 292 105 212 92 118 2850 97 102 54 160 19% 95% 45.8% 2% 30.9% 50% 40.7% 29% 22.2% 51% 8.2% 19.5% 17.3% 10.3% 59.3% 68.2% 49% 35.4% 0.8% 17.1% 36% 100% 5% 10% 15.3% 23.7% 48.1% 20.4% 19% 17.3% 15.5% 1.9% 16.7% 11.2% 8.2% 3.1% 24.6% 17.3% 30.9% 2.2% 14.5% 60.2% 9.4% 1% 6% 31% 0% 20% 44.9% 0.4% 27.6% 37% 0.4% 15.4% 21% 19.6% 35.8% 15% 5% 12.3% 19.2% 1% 1.9% 1.9% 31.8%

Page | 395

Bio-RAD MonolisaTM HCV Ag-Ab ULTRA Laperche, 2005 Nastouli, 2008 Schnuriger, 2006 Tuke, 2008* EIKEN Lumispot HCV Ag Saito, 2003 Ortho ELISA-Ag Agha, 2004 Nübling, 2002 246 52 37% 11.5% 42.3% 19.2% 9.8% 11.5% 5.8% 15.4% 47.3% 155 65.2% 35.80% 35 25 20 Unknown 20% 62% 5.7% 11.4% 68% 15% 34.3% 4% 5% 23% 25.7% 14.3% 4% 30% 15% 5.7% 16% 45% 2.9% 8%

HCV: hepatitis C virus, Ag: antigen, Ab: Antibody, ELISA: enzyme linked immunosorbent assay, G1: genotype 1, G1a: genotype 1a, G1b: genotype 1b, G2: genotype 2, G3: genotype 3, G4: genotype 4, G5: genotype 5, G6: genotype 6 * Data are the same for DiaSorin S.A. Murex Ag/Ab EIA

Page | 396

Table 3b. Genotype information for studies included in PICO 9 grouped alphabetically Author, year Number of subjects % G1 % G1a % G1b %G2 %G3 %G4 %G5 %G6 % Other or unknown

Abbott ARCHITECT HCV Ag Feng, 2014 Loggi, 2013 Moscato, 2010 32 35 23 100% 20% 4% 100% 80% 39.1% 26.1% 21.7% 8.7%

Fujirebio Lumipulse Ortho HCV Ag Fujino, 2009 Takahashi, 2005 90 44 66.7% 100% 33.3%

HCV: hepatitis C virus, Ag: antigen, Ab: antibody, G1 = genotype 1, G1a = genotype 1a, G1b = genotype 1b, G2 = genotype 2, G3 = genotype 3, G4 = genotype 4, G5 = genotype 5, G6 = genotype 6

Table 4. Sensitivity and specificity of Abbott ARCHITECT HCV Ag assay compared to nucleic acid testing (NAT) assessed at baseline, at week 4 of interferon based therapy (early viral response), and at week 24 after completion of treatment (sustained viral response) Baseline Author, Year Feng, 2014 Number of subjects 32 Se (95% CI) 100% NA Sp (95% CI) Early viral response (EVR) Se (95% CI) 100% Sp (95% CI) 88.9% (68.4%, 100%) Loggi, 2013 35 100% NA 73.5% (58.7%, 88.4%) Moscato, 2010 23 NA NA 100% 70% (41.6%, 98.4%) NA NA 100% 100% 94.1% (82.9%, 100%) Sustained viral response (SVR) Se (95% CI) 100% Sp (95% CI) 100%

HCV: hepatitis C virus, Ag: antigen, Se: sensitivity, Sp: specificity, CI: confidence interval, NA: not applicable as cannot be calculated from study data

Page | 397

Table 5. Sensitivity and specificity of HCV core antigen assay in prediction of sustained viral response (SVR) after initiation of interferon-based treatment Author, Year No. of subjects (no.to achieve SVR) Feng, 2014 Fujino, 2009 Takahashi, 2005 32 (21) 90 (57) 44 (10) Abbott ARCHITECT Fujirebio Lumipulse Fujirebio Lumipulse Index test Timing of test after treatment start 6 days 7 days 7 days Change in HCVcAg Sensitivity Specificity

Log 10 Absolute Absolute

95.2% 79.4% 57.1%

70% 88.5% 93.3%

HCV: hepatitis C virus, No.: number

Page | 398

Figures Fig. 1a. PRISMA diagram of studies excluded from screen one, and those full papers retrieved for more detailed evaluation Potentially relevant citations identified from electronic databases: 8146

Excluded screen one: 7833 Reason: Not relevant based on assessment of title and abstract

Full papers retrieved for more detailed evaluation: 313

Full studies relevant to PICO 5a 283

Full studies relevant to PICO 5b 11

Full studies relevant to PICO 9 44

Page | 399

Fig. 1b. PRISMA diagram of studies included in the review for PICO 5a Full papers retrieved for more detailed evaluation: 283 Excluded screen two: 229 Reasons:             Abstract or poster: 53 Duplicate data/study: 6 Editorial/Comment: 9 Inappropriate ref standard: 6 Less than 10 samples: 16 Non-commercial or off-market assay: 65 No core antigen, does not apply to study question: 25 Non-blood specimen: 20 Non-human specimens or commercial sera panels: 5 Review article: 21 Unable to translate: 3 Unable to retrieve full article: 1

Excluded for non-extractable data with no response from authors: 4

Papers (studies) included in the systematic review: 50

Page | 400

Fig. 1c. PRISMA diagram of studies included in the review for PICO 5b

Full papers retrieved for more detailed evaluation: 11

Excluded screen two: 10 Reasons:     Abstract or poster: 2 Inappropriate ref standard: 1 Does not apply to study question: 4 Review article: 3

Papers (studies) included in the systematic review: 1

Page | 401

Fig. 1d. PRISMA diagram of studies included in the review for PICO 9 Full papers retrieved for more detailed evaluation: 44 Excluded screen two: 35 Reasons:       Abstract or poster: 13 Duplicate data/study: 1 Less than 10 samples: 2 Non-commercial or off-market assay: 14 Does not apply to study question: 4 Review article: 1

Excluded for non-extractable data with no response from authors or authors unable to provide needed information: 4

Papers (studies) included in the systematic review: 5

Page | 402

QUADAS figures Fig. 2a. Risk of bias and applicability summary as judged by review authors about each QUADAS-2 domain presented as percentages across the 50 included studies for PICO 5a Risk of bias Flow and Timing Reference Standard Index Test Patient Selection

Applicability Concerns 0% 25% 50% 75% Unclear risk of bias 100%

Reference Standard Low risk of bias Index Test Patient Selection 0% Low risk of bias

High risk of bias

25%

50%

75% Unclear risk of bias

100%

High risk of bias

Page | 403

Fig. 2b. Risk of bias and applicability summary as judged by review authors about each QUADAS-2 domain presented as percentages across the 5 included studies for PICO 9 Risk of bias

Flow and Timing Reference Standard Index Test Applicability Patient Selection

Fig. 3a. Risk of bias and applicability summary as judged by review authors about each 0% 25% 50% 100% QUADAS-2 domain presented by individual study included in PICO 5a. 75% Reference Standard Low risk of bias Index Test Patient Selection 0% Low risk of bias 25% 50% 75% Unclear risk of bias 100% High risk of bias Unclear risk of bias

High risk of bias

Page | 404

Fig. 3a (cont). Risk of bias and applicability summary as judged by review authors about each QUADAS-2 domain presented by individual study for PICO 5a

Page | 405

Page | 406

Fig. 3a (cont). Risk of bias and applicability summary as judged by review authors about each QUADAS-2 domain presented by individual study for PICO 5a

Page | 407

Fig. 3 (cont). Risk of bias and applicability summary as judged by review authors about each QUADAS-2 domain presented by individual study for PICO 5a

Page | 408

Fig. 3b. Risk of bias and applicability summary as judged by review authors about each QUADAS-2 domain presented by individual study for PICO .

Page | 409

Fig. 4a. Forest plot of Abbott ARCHITECT HCV Ag assay sensitivity and specificity for the diagnosis of active HCV infection compared to NAT reference test for all samples regardless of HCV Ab status

HCV: hepatitis C virus, Ag: antigen, NAT: nucleic acid testing, Ab: antibody, TP: true positive, FP: false positive, FN: false negative, TN: true negative, CI: confidence interval

Page | 410

Fig. 4b. Univariate analysis of Abbott ARCHITECT HCV Ag Assay sensitivity for the diagnosis of active HCV infection compared to NAT reference test for all studies with sensitivity data regardless of HCV Ab status

HCV: hepatitis C virus, Ag: antigen, NAT: nucleic acid testing, Ab: Antibody, TP: true positive, FP: false positive, FN: false negative, TN: true negative, CI: confidence interval

Page | 411

Fig. 4c. Forest plot of Abbott ARCHITECT HCV Ag assay sensitivity and specificity for the diagnosis of active HCV infection compared to NAT reference test for known HCV antibody-positive samples

HCV: hepatitis C virus, Ag: antigen, NAT: nucleic acid testing, Ab: antibody, TP : true positive, FP: false positive, FN: false negative, TN: true negative, CI: confidence interval

Page | 412

Fig. 4d. Forest plot of Abbott ARCHITECT HCV Ag assay sensitivity and specificity for the diagnosis of active HCV infection compared to NAT reference test for known HCV antibody-negative samples

HCV: hepatitis C virus, Ag: antigen, NAT: nucleic acid testing, Ab: antibody, TP: true positive, FP: false positive, FN: false negative, TN: true negative, CI: confidence Interval

Page | 413

Fig. 5. Bivariate analysis of Abbot ARCHITECT HCV antigen assay sensitivity and specificity for diagnosis of active HCV infection compared to gold standard nucleic acid testing in (a) all samples regardless of HCV antibody (Ab) status, (b) HCV Ab-positive samples (c) HCV Abnegative samples. These plots show pooled summary estimates (red squares), the dashed red line represents the 95% confidence region and the dashed green line represents the 95% prediction region. The individual circles represent each study and the size of the circle is proportional to the total sample size. (a) (b)

Page | 414

Fig. 6. Forest plot Ortho ELISA-Ag sensitivity and specificity for diagnosis of active HCV infection compared to NAT reference test for all samples regardless of HCV Ab status

ELISA: enzyme linked immunosorbent assay, Ag: antigen, HCV: hepatitis C virus, NAT: nucleic acid testing, Ab: antibody, TP: true positive, FP: false positive, FN: false negative, TN: true negative, CI: confidence interval

Page | 415

Fig. 7. Bivariate analysis of Ortho ELISA-Ag sensitivity and specificity for diagnosis of active HCV infection compared to gold standard nucleic acid testing in all samples regardless of HCV antibody status. This plot shows pooled summary estimates (red squares), the dashed red line represents the 95% confidence region and the dashed green line represents the 95% prediction region. The individual circles represent each study and the size of the circle is proportional to the total sample size.

Page | 416

Fig. 8. Forest plot of Bio-RAD MonolisaTM HCV Ag-Ab ULTRA sensitivity for diagnosis of active HCV infection compared to NAT reference test for all samples regardless of HCV Ab status HCV: hepatitis C virus, Ag: antigen, Ab: antibody, NAT: nucleic acid testing, TP: true positive, FP: false positive,

FN: false negative, TN: true negative

Page | 417

Fig. 9. Forest plots of Hunan Jynda HCV Core Ag ELISA sensitivity and specificity for diagnosis of active HCV infection compared to NAT reference test for all samples regardless of HCV Ab status

HCV: hepatitis C virus , ELISA: enzyme linked immunosorbent assay, Ag: antigen, NAT: nucleic acid testing, Ab: Antibody, TP: true positive, FP: false positive, FN: false negative, TN: true negative, CI: confidence interval

Page | 418

Fig. 10. Bivariate analysis of Hunan Jynda Bioengineering Group HCV Core Ag ELISA sensitivity and specificity for diagnosis of active HCV infection compared to gold standard nucleic acid testing for all samples regardless of HCV antibody status. This plot shows pooled summary estimates (red squares), the dashed red line represents the 95% confidence region and the dashed green line represents the 95% prediction region. The individual circles represent each study and the size of the circle is proportional to the total sample size.

Page | 419

Fig. 11. Forest plot of DiaSorin S.A. Murex Ag/Ab EIA sensitivity for diagnosis of active HCV infection compared to NAT reference test for all samples regardless of HCV Ab status

Ag: antigen, Ab: antibody, EIA: enzyme immunoassay, HCV: hepatitis C infection, NAT: nucleic acid testing, TP: true positive, FP: false positive, FN: false negative, TN: true negative

Page | 420

Fig. 12. Non-parametric regression smoother of pooled quantitative data assessing correlation between Abbott ARCHITECT HCV core Ag measured in log fmol/L and HCV RNA measured in log IU/mL. The red line indicates the positivity threshold of the core antigen index test corresponding to 3 fmol/L.

HCV = hepatitis C virus, Ag = antigen, RNA = ribonucleic acid

Page | 421

Appendices Appendix A. Search report for systematic review on HCV Antigen use for diagnostics as well as treatment monitoring Date of search = March 2015. The following tables shows the sources that have been searched and the hits retrieved from those searches Source Date range searched Hits retrieved (before duplicate removal)

Electronic databases Medline (PubMed) Cochrane Embase Web of Science Scopus Final number of records in EndNote database after deleting duplicates All available All available All available All available All available 2820 127 5501 2635 3549 8146

Search strategy Embase 1. ‘hepatitis c antigen’/exp OR ‘hepatitis c antigen’ OR ‘hepatitis c’/exp OR ‘hepatitis c virus’/exp OR ‘hepatitis c virus’ 2. ‘hepatitis c antigen’ OR ‘hepatitis C’ OR hepatitis c virus’ or ‘hcv’ 3. #1 OR #2 4. ‘antigen’/exp OR ‘antigen’ OR ‘virus antigen’/exp OR ‘virus antigen’ 5. ‘antigen’ OR ‘virus antigen’ 6. #4 OR #5 7. ‘nucleic acid amplification’/exp OR ‘nucleic acid amplification’ OR ‘virus rna’/exp OR ‘virus rna’ OR ‘rna’/exp OR ‘rna’ 8. ‘nucleic acid amplification’ OR ‘virus rna’ OR ‘rna’ OR ‘nucleic acid test’ 9. #7 OR #8 10. #3 AND #6 AND #9 'hepatitis c antigen'/exp OR 'hepatitis c'/exp OR 'hepatitis c virus'/exp OR 'hepatitis c antigen' OR 'hepatitis c' OR 'hepatitis c virus' OR 'hcv' AND ('antigen'/exp OR 'virus antigen'/exp OR 'antigen' OR 'virus antigen') AND ('nucleic acid amplification'/exp OR 'virus rna'/exp OR 'rna'/exp OR 'nucleic acid amplification' OR 'virus rna' OR 'rna' OR 'nucleic acid test' Page | 422

Search strategy Web of Knowledge (SCI-expanded, SSCI, Conference Proceedings science, BIOSIS previews) 1. 2. 3. 4. Hepatitis C OR HCV (topic) Antigen* OR core antigen* (topic) RNA OR NAT or nucleic acid test* (topic) #1 AND #2 AND #3

Search strategy PubMed ((hepatitis C[MeSH Terms] OR "Hepacivirus"[Mesh] OR "HCV" OR "hepatitis C”) AND (antigen* OR antigens, viral/blood[MeSH Terms] OR hepatitis c antigens[MeSH Terms] ) AND ("Nucleic Acid Amplification Techniques"[Mesh] OR nucleic acid test*[tw] OR nucleic acid amplification[tw] OR RNA OR RNA, viral/blood[MeSH Terms])) Search strategy SCOPUS ( TITLE-ABS-KEY ( "Hepatitis C"OR hcv )AND TITLE-ABS-KEY ( antigen*OR "core antigen*" )AND TITLE-ABS-KEY ( rna OR"nucleic acid test" ) )AND( LIMIT-TO ( DOCTYPE ,"ar" )OR LIMIT-TO ( DOCTYPE ,"ip" ) ) Search strategy Cochrane 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. MeSH descriptor: Hepatitis C MeSH Descriptor: Hepacivirus MeSH Descriptor: Hepatitis C Antigens Hepatitis C HCV #1 or #2 or #3 or #4 or #5 MeSH Descriptor: Nucleic Acid Amplification Techniques RNA Nucleic acid test #7 or #8 or #9 Antigen “Core antigen”

13. #11 or #12 14. #6 AND #10 AND #13

Page | 423

Appendix B. Data extraction form ID First Author Corresponding author and email Was author contacted? 1 – Yes 2 – No If yes, dates(s)

Title Year (of publication) Year (study start date) Language 1 – English 2 – Other If other, specify: 1 – n, % Genotype 1 __ Genotype 1a __ Genotype 1b __ 2 – n, % Genotype 2 __ 3 – n, % Genotype 3 __ 4 – n, % Genotype 4 __ 5 – n, % Genotype 5 __ 6 – n, % Genotype 6 __

HCV Genotypes specified

% HIV positive % HBsAg + (chronic HBV infection) % Adults/children Age (mean SD, median IQR, range) Gender, % Female Country where study was conducted Country World Bank Classification (at time of study start date) Study design 1 – Middle/Low 2 – High 3 – Both middle/low and high 1 – Randomized trial 2 – Cross-sectional 3 – Cohort 4 – Case Control 5 – Other, specify 9 – Unk/NR If other, specify:

Page | 424

Participant selection

1 – Consecutive 2 – Random 3 – Convenience 4 – Other 9 –Unk/NR 1 – Broad 2 – Healthy persons only 3 – Unk/NR Comments: ____________________________________________ 1 – Prospective 2 – Retrospective 9 – Unk/NR

Study population

Direction of study data collection

Comments about study design

Were samples excluded based on prior testing of the sample?

1 – yes (specify below) 2 – no 9 – Unk/NR Comments: __________________________________________ _____ 9 – Unk/NR _____ 9 – Unk/NR 1 – One specimen per patient 2 – Multiple specimens per patient 3 – Unknown number of specimens per patient 9 – NR/Unclear Describe as in paper, if unclear: 1 – Serum ___ 2 – Plasma ___ 3 – Whole blood____ 1 – PCR 2 – bDNA 3 – TMA 9 – Unk/NR 1 – Yes 2 – No 9 – Unk/NR 1 – Abbott ARCHITECT

Number after screening by exclusion and inclusion criteria Sample size (total number included in 2/2 table) Unit of analysis

Types of specimen and number

HCV NAT method used

HCV NAT method quantitative?

HCV Ag test manufacturer

Page | 425

2 – Bio-RAD Mono Lisa 3 – Other: please specify 9 – Unk/NR HCV Ag method quantitative? 1 – Yes 2 – No 9 – Unk/NR 1 – Yes 2 – No 3 – Unk/NR 1 – yes (Specify below) 2 – No Specify timing of Ag Collection 1 – Baseline, prior to treatment 2 – EVR 3 – SVR 12 weeks 4 – SVR 24 weeks What treatment regimen was used? 1 – Interferon based therapy 2 – Interferon free direct acting antivirals 3- Unk/NR Did all patients NAT within the study? 1 – yes 2 – no 9 – Unk/NR Comments: __________________________________________ 1 – Yes 2 – No 9 – Unk/NR Comments: ___________________________________________

Were reference NAT test and HCV Ag test performed on specimen within 30 days Was Ag test obtained/repeated while subject was on treatment for HCV infection or after treatment completed?

Was index test performed per recommendation of the manufacturer?

Page | 426

Appendix C. QUADAS-2 protocol Domain 1. Patient selection Risk of bias: could the selection of patients have introduced bias?  Signalling question 1: Was a consecutive or random sample of patients or specimens enrolled? Score “yes” if the study enrolled a consecutive or random sample of eligible patients; “no” if the study selected patients by convenience, and “unclear” if the study did not report the manner of patient selection or unable to tell. Signalling question 2: Was a case–control design avoided? Rate “no” if case–control study, “yes” if prospective or cross-sectional study. Signalling question 3: Did the study avoid inappropriate exclusions? Score “no” if the study excluded samples based on prior testing of the sample and “unclear” if unable to tell.

 

Risk of Bias is scored as “low risk” if selection was done in a random or consecutive manner and the study was prospective and did not exclude samples based on prior testing; “high risk” if selection was by convenience, from case-control study or excluded samples; and ”unclear risk” if the manner of participant selection is unclear and no clinical information is provided. Applicability: Are there concerns that the included patients and setting do not match the review question? We are interested in how HCV AG test performs across HCV genotypes and among HIVinfected (immunocompromised) persons. If a study includes only very selected persons, only healthy or blood donors, it would not be relevant to the study question. Setting of testing is not relevant to the review question. We will score “low risk” if broad study population, “high risk” if population is blood donors or healthy persons only, and “unclear risk” if the population is not well characterized. Domain 2. Index test Risk of bias: could the conduct or interpretation of the index test have introduced bias?  Signalling question 1: Were the index test results interpreted without knowledge of the results of the reference standard? Rate “yes” if results of reference standard were blinded. Rate “no” if reference standard results were unblinded.  Signalling question 2: If a threshold was used, was it pre-specified? Answer “yes” for all studies as limit of detection for all commercially available HCV Ag tests are pre-specified. Score “low risk” for all tests interpreted with blinded results of reference standard. Score “high risk” for antigen tests interpreted with results from reference standard available. Score “unclear risk” if availability of reference test is not specified. Applicability: Are there concerns that the index test, its conduct, or its interpretation differ from the review question? Variations in test technology, execution, or interpretation may affect estimates of the diagnostic accuracy of a test. Page | 427



Score “low concern” if the test was done as per recommendation of the manufacturer. Score “high concern” if additional processing steps were added. Score “unclear” if not discussed in the study.

Domain 3. Reference standard Risk of bias: could the reference standard, its conduct, or its interpretation have introduced bias?  Signalling question 1: Is the reference standard likely to correctly classify the target condition? There are multiple methods of NAT, each with slightly varying sensitivity, however overall the tests are highly sensitive and the verification should be minimal. We will score “yes” for all studies Signalling question 2: Were the reference standard results interpreted without knowledge of the results of the index test? The reference standard in this case also does not allow for interpretation. Therefore it is unlikely to introduce bias even if reference standard was resulted with knowledge of the index test result. For risk of bias, score “low risk” for all studies. Applicability: Are there concerns that the target condition as defined by the reference standard does not match the question? Judge applicability to be of “low risk” for all studies as circulating virus is by definition associated with active infection and the specificity of the reference standard is high. While the reference standard is not able to differentiate between acute or chronic infection, the core antigen is also not expected to do so. The differentiation will be done based on the constellation of NAT results with serology results. This will be assessed in a stratified analysis. Domain 4. Flow and timing Risk of bias: Could the patient flow have introduced bias?  Signalling question 1: Was there an appropriate interval between the index test and reference standard? We will limit time between reference and index testing to <1 month. Score “yes” if time between tests is <1 month, score “no” if time between tests is more than 1 month. Signalling question 2: Did all patients in the study receive the same reference standard? Answer “yes” if all patients had NAT, answer “no” if reference standard NAT was not used for all patients, answer “unclear” if it is not specified. Signalling question 3: Were all patients included in the analysis? Determined the answer to this question by comparing the number of patients enrolled with the number of patients included in the two-by-two tables.



Page | 428

For risk of bias, score “low risk” if the number of participants enrolled was clearly stated and corresponded to the number presented in the analysis or if exclusions were adequately described. Score “high risk” if there were participants missing or excluded from the analysis and there was no explanation given; and “unclear risk” if not enough information was given to assess whether participants were excluded from the analysis; usually this means that the number of participants originally enrolled in the study was not explicitly stated. Appendix D. List of excluded studies and reasons for exclusion organized by PICO. PICO 5a: 1. Alados-Arboledas JC, Calbo-Torrecillas L, Lopez-Prieto MD, de Francisco-Ramirez JL, de MiguelSastre C. Clinical assessment of Monolisa HCV ag-ab ULTRA (Bio-Rad) in a general hospital. Enferm Infecc Microbiol Clin. 2007;25(3):172‒6. Inappropriate reference test Allain JP, Coghlan PJ, Kenrick KG, Whitson K, Keller A, Cooper GJ, et al. Prediction of hepatitis C virus infectivity in seropositive Australian blood donors by supplemental immunoassays and detection of viral RNA. Blood. 1991;78(9):2462‒8. No HCV core antigen performed Alzahrani AJ. Analysis of hepatitis C virus core antigenemia in Saudi drug users. Saudi Med J. 2005;26(10):1645‒6. Editorial or comment Alzahrani AJ, Obeid OE, Al-Ali A, Imamwardi B. Detection of hepatitis C virus and human immunodeficiency virus in expatriates in Saudi Arabia by antigen-antibody combination assays. J Infect Dev Ctries.2009;3(3):235‒8. Less than 10 independent samples Aoyagi K, Iida K, Ohue C, Matsunaga Y, Tanaka E, Kiyosawa K, et al. Performance of a conventional enzyme immunoassay for hepatitis C virus core antigen in the early phases of hepatitis C infection. Clin Lab. 2001;47(3‒4):119‒27. Non-commercial or off-market assay Aoyagi K, Ohue C, Iida K, Kimura T, Tanaka E, Kiyosawa K, et al. Development of a simple and highly sensitive enzyme immunoassay for hepatitis C virus core antigen. J Clin Microbiol. 1999;37(6):1802‒8. Non-commercial or off-market assay Araujo AC, Astrakhantseva IV, Fields HA, Kamili S. Distinguishing acute from chronic hepatitis C virus (HCV) infection based on antibody reactivities to specific HCV structural and nonstructural proteins. J Clin Microbiol. 2011;49(1):54‒7. No HCV core antigen performed Arrojo IP, Pareja MO, Orta MDR, Luque FN, Lamas MCH, Gordo FS, et al. Detection of a healthy carrier of HCV with no evidence of antibodies for over four years. Transfusion. 2003;43(7):953‒ 7. Less than 10 independent samples Attallah AM, Ismail H, Tabll AA, Shiba GE, El-Dosoky I. A novel antigen detection immunoassay for field diagnosis of hepatitis C virus infection. J Immunoassay Immunochem. 2003;24(4):395‒ 407. Non-commercial or off-market assay Attallah AM, Omran MM, Nasif WA, Ghaly MF, El-Shanshoury AERR, Abdalla MS, et al. Diagnostic performances of hepatitis C virus-NS4 antigen in patients with different liver pathologies. Arch Med Res. 2012;43(7):555‒62. Non-commercial or off-market assay Attallah AM, Shiha GE, Malak CAA, Hagras HE, Abdel-Razik WS, Ismail H. Utility of a novel HCVNS4 antigen detection immunoassay for monitoring treatment of HCV-infected individuals with pegylated interferon alpha-2a. Hepatol Res. 2004;28(2):68‒72. Non-commercial or off-market assay

3. 4.

10.

11.

Page | 429

12.

Attia MA, Zekri AR, Goudsmit J, Boom R, Khaled HM, Mansour MT, et al. Diverse patterns of recognition of hepatitis C virus core and nonstructural antigens by antibodies present in Egyptian cancer patients and blood donors. J Clin Microbiol. 1996;34(11):2665‒9. No HCV core antigen performed Baggett DW, Moroney S, Saewert M, Jaczko B, Zelechowski J, Bahi C, et al. Dynamics of circulating HCV core antigen and HCV RNA in the early phase of HCV infection. Transfusion. 2000;40(10):26S. Abstract or poster Ballardini G, Manzin A, Giostra F, Francesconi R, Groff P, Grassi A, et al. Quantitative liver parameters of HCV infection: relation to HCV genotypes, viremia and response to interferon treatment. J Hepatol. 1997;26(4):779‒86. Non-blood specimen Baranov AV, Maleev VV. Association between HCV RNA level and anti-HCV antibodies during chronic hepatitis C. Zh Mikrobiol Epidemiol Immunobiol. 2009(5):19‒22. No HCV Core antigen performed Bdour S. Hepatitis C virus infection in Jordanian haemodialysis units: serological diagnosis and genotyping. J Med Microbiol. 2002;51(8):700‒4. No HCV core antigen performed Beer N, Shinar E, Novack L, Safi J, Soliman H, Yaari A, et al. Accuracy of hepatitis C virus core antigen testing in pools among seroconverters. Transfusion. 2006;46(10):1822‒8. Noncommercial or off-market assay Berger A. Recent developments in hepatitis C infection (epidemiology, diagnosis and therapy). Laboratoriums Medizin. 2001;25(7‒8):218‒22. Review article Berger A, Doerr HW, Preiser W, Weber B. Lack of correlation between different hepatitis C virus screening and confirmatory assays. J Virol Methods. 1996;59(1‒2):141‒6. No HCV core antigen performed Bochkova G, Fomina S, Puzyrev V, Obriadina A, Burkov A, Ulanova T. The evaluation of the ELISA kit nullDS-EIA-anti-HCVSPECTR-GMnull as supplemental assay for confirmation of anti-HCV screening positive results. Clin Microbiol Infect. 2011;17:S664. Abstract or poster Bochkova G, Fomina S, Puzyrev V, Obriadina A, Burkov A, Ulanova T. The evaluation of the new ELISA kit “EIA-anti- HCV-SPECTRUM-M” intended for separate detection of anti-IgM to different HCV antigens. J Viral Hepat. 2012;19:5‒6. Abstract or poster Bouvier-Alias M, Patel K, Dahari H, Beaucourt S, Larderie P, Blatt L, et al. Clinical utility of total HCV core antigen quantification: a new indirect marker of HCV replication. Hepatology. 2002;36(1):211‒8. Non-commercial or off market assay Bouzgarrou N, Fodha I, Ben Othman S, Achour A, Grattard F, Trabelsi A, et al. Evaluation of a total core antigen assay for the diagnosis of hepatitis C virus infection in hemodialysis patients. J Med Virol. 2005;77(4):502‒8. Non-commercial or off market assay Brandao CPU, Marques BLC, Marques VA, Villela-Nogueira CA, Do O KMR, de Paula MT, et al. Simultaneous detection of hepatitis c virus antigen and antibodies in dried blood spots. J Clin Virol. 2013;57(2):98‒102. Inappropriate reference test Brody RI, Eng S, Melamed J, Mizrachi H, Schneider RJ, Tobias H, et al. Immunohistochemical detection of hepatitis C antigen by monoclonal antibody TORDJI-22 compared with PCR viral detection. Am J Clin Pathol. 1998;110(1):32‒7. Non-blood specimens Brojer E, Gronowska A, Medyńska J, Grabarczyk P, Mikulska M, Lȩtowska M, et al. The hepatitis C virus genotype and subtype frequency in hepatitis C virus RNA-positive, hepatitis C virus antibody-negative blood donors identified in the nucleic acid test screening program in Poland. Transfusion. 2004;44(12):1706‒10. No HCV core antigen performed

13.

14.

15.

16. 17.

18. 19.

20.

21.

22.

23.

24.

25.

26.

Page | 430

27.

Brojer E, Liszewski G, Niznik A, Rosiek A, Letowska M, Peterson JE, et al. Detection of HCV core antigen in HCV RNA positive, anti-HCV negative blood donations from Polish blood donors. Transfusion. 2001;41(2):304. Editorial or comment Brojer E, Gronowska A, Rosiek A, Mikulska M, Letowska M. HCV core antigen detection and quantification of viremia in the “window period” of donors identified by routine HCV RNA screening in Poland. Vox Sang. 2005;89:92. Poster or abstract Burek V. Hepatitis C viral infection – news in diagnostics. Infektoloski Glasnik. 2002;22(1):27‒9. Review article Busch MP, Wright DJ, Hirschkorn DF, Baggett D, Maret S, Lee SR, et al. Sensitivity of 1(st) and 2(nd) generation HCV antigen assays versus nucleic acid testing (NAT) for detection of ramp-up phase of HCV infection. Transfusion. 2001;41(9):3S. Abstract or poster Cagnon L, Wagaman P, Bartenschlager R, Pietschmann T, Gao TJ, Kneteman NM, et al. Application of the trak-C (TM) HCV core assay for monitoring antiviral activity in HCV replication systems. J Virol Methods. 2004;118(1):23‒31. Non-human subjects, non-commercial or off-market assay Cano H, Candela MJ, Lozano ML, Vicente V. Application of a new enzyme-linked immunosorbent assay for detection of total hepatitis C virus core antigen in blood donors. Transfus Med.. 2003;13(5):259‒66. Non-commercial or off-market assay Cao H, Zhang K, Shu X, Li G. The effect of hepatitis B virus infection on detection of hepatitis C virus core antigen. Hepatol Int 2012;6(1):199. Abstract or poster Cao H, Zhang K, Shu X, Xu QH, Li G. Detection of core antigen of hepatitis virus C in patients infected with hepatitis virus C and B. Zhonghua Gan Zang Bing Za Zhi. 2011;19(10):726‒8. Noncommercial or off-market assay Cao H, Zhang K, Shu X, Xu QH, Li G. Detection of hepatitis C core antigen in intravenous drug addictions. Zhonghua Shi Yan He Lin Chuang Bing Du Xue Za Zhi. 2011;25(4):304‒6. Noncommercial or off-market assay Carabaich A, Ruvoletto M, Bernardinello E, Tono N, Cavalletto L, Chemello L, et al. Profiles of HCV core protein and viremia in chronic hepatitis C: possible protective role of core antigen in liver damage. J Med Virol. 2005;76(1):55‒60. Non-commercial or of- market assay Carney R, Maranao D, Sudra R, Chaytor S, Labbett W, Johnson M, et al. A hepatitis C virus core antigen assay is a cost-effective, sensitive and specific test in the detection of acute hepatitis C in HIV infected subjects. HIV Med. 2014;15:8. Abstract or poster Cavazza S, Lagging M. Indeterminate third-generation hepatitis C recombinant immunoblot assay and HCV RNA analysis: isolated reactivity against NS5 associated with HCV viraemia in clinical patients but not blood donors. Scand J Infect Dis. 2005;37(6‒7):488‒92. No HCV core antigen performed Cetinić Balent N, Mikulić R, Đaković Rode O. Enzyme immunoassay for separate detection of anti-HCV antibodies to individual HCV antigens as a confirmatory assay in diagnostics of HCV infection. Infektoloski Glasnik. 2013;33(3):109‒15. No HCV core antigen performed Chakravarti A, Chauhan MS, Dogra G, Banerjee S. Hepatitis C virus core antigen assay: can we think beyond convention in resource limited settings? Braz J Infect Dis. 2013;17(3):369‒74. Non-commercial or off market assay Challine D, Dameron G, Laperche L, Larderie P, Rigot P, Claquin J, et al. Does HCV core antigen or nucleic acid testing in graft donors improve organ transplantation viral safety? Transfusion. 2001;41(9):3S‒4S. Abstract or poster

28.

29. 30.

31.

32.

33. 34.

35.

36.

37.

38.

39.

40.

41.

Page | 431

42. 43. 44.

Chaudhary R, Burres E. Detection of antibody to envelope (E2) antigen of hepatitis C virus. Can J Infect Dis. 1997;8(4):229‒31. No HCV core antigen performed Choi YJ, Hu Y, Goris J, Kim E. Histopathology, viral antigen and RNA in hepatitis C virusassociated hepatitis. Lab Invest. 1997;76(1):803. Cividini A, Cerino A, Muzzi A, Furione M, Rebucci C, Segagni L, et al. Kinetics and significance of serum hepatitis C virus core antigen in patients with acute hepatitis C. J Clin Microbiol. 2003;41(5):2144‒6. Less than 10 independent samples Courouce AM, Barin F, Botte C, Lunel F, Maisonneuve P, Maniez M, et al. A comparative evaluation of the sensitivity of 7 anti-hepatitis-C virus screening tests. Vox Sang. 1995;69(3):213‒6. No HCV core antigen performed Couroucé AM, Le Marrec N, Bouchardeau F, Razer A, Maniez M, Laperche S, et al. Efficacy of HCV core antigen detection during the preseroconversion period. Transfusion. 2000;40(10):1198‒ 202. Less than 10 independent samples Craxi A, Valenza M, Fabiano C, Magrin S, Fiorentino G, Diquattro O, et al. Third generation hepatitis C virus tests in asymptomatic anti-HCV-positive blood donors. J Hepatol. 1994;21(5):730‒4. No HCV core antigen performed Cresswell F, Shaw S, Hughes D, Youssef E, Homer G, Hassan-Ibrahim M, et al. Hepatitis C antigen testing: a reliable alternative for diagnosing acute hepatitis C infection. HIV Med. 2014;15:137. Abstract or poster Cucchietti A, Parker S, Oldfield L, Limas C, Garrod A, Mc Bryne L. Seroconversion sensitivity of elisa screening assays for HCV, HIV and HBV antigens and/or antibodies. Vox Sang. 2011;101:87. Abstract or poster Daniel HD, Vivekanandan P, Raghuraman S, Sridharan G, Chandy GM, Abraham P. Significance of the hepatitis C virus (HCV) core antigen as an alternative plasma marker of active HCV infection. Indian J Med Microbiol. 2007;25(1):37‒42. Non-commercial or off-market assay Daniel HDJ, Chandy GM, Abraham P. Quantitation of hepatitis C virus using an in-house real-time reverse transcriptase polymerase chain reaction in plasma samples. Diagn Microbiol Infect Dis. 2008;61(4):415‒20. No HCV core antigen performed Dawson G. The potential role of HCV core antigen testing in diagnosing HCV infection. Antivir Ther. 2012;17(7 PARTB):1431‒5. Review article Dawson GJ. HCV core antigen and combination (antigen/antibody) assays for the detection of early Seroconversion. J Med Virol. 2007;79:S54‒S8. Review article Dawson GJ. HCV core antigen detection in seropositive samples. J Med Virol. 2007;79:S52‒S3. Review article De Almeida Ponde RA. Enzyme-linked immunosorbent/chemiluminescence assays, recombinant immunoblot assays and nucleic acid tests in the diagnosis of HCV infection. Eur J Clin Microbiol. Infect Dis 2013;32(8):985‒8. Review article Dickson RC, Mizokami M, Orito E, Qian K, Lau JYN. Quantification of serum HCV core antigen by a fluorescent enzyme immunoassay in liver transplant recipients with recurrent hepatitis C – clinical and virologic implications. Transplantation. 1999;68(10):1512‒6. Non-commercial or offmarket assay Dow BC, Munro H, Buchanan I, Ferguson K, Davidson F, Lycett C, et al. Acute hepatitis C virus seroconversion in a Scottish blood donor: HCV antigen is not comparable with HCV nucleic acid amplification technology screening. Vox Sang. 2004;86(1):15‒20. Less than 10 independent samples

45.

46.

47.

48.

49.

50.

51.

52. 53. 54. 55.

56.

57.

Page | 432

58.

Dubrous P, Hugard L, Terrier F. Screening for HCV core antigen among HIV-positive patients infected by sexual transmission. Med Mal Infect. 2004;34(5):236‒8. Less than 10 independent samples Durante-Mangoni E, Vallefuoco L, Perna E, Iossa D, Andini R, Caianiello C, et al. Correlates and prognostic value of hepatitis C virus core antigen serum levels during chronic infection and treatment. Clin Microbiol Infect. 2010;16:S317. Abstract or poster Dwyer E, Riley P, Pakianathan M. Analysis of hepatitis C antigen testing in an urban sexual health clinic. HIV Med. 2014;15:111‒2. Abstract or poster Eiras A, Franco E, Montoro JA, Planelles D, Villaescusa R. HCV NAT (minipool RT-PCR) and HCV core antigen ELISA. Transfusion. 2003;43(1):118; author reply -9. Editorial or comment El Ekiaby M, Laperche S, Moftah F, Burnouf T, Lelie N. The impact of different HCV blood screening technologies on the reduction of transfusion transmitted HCV infection risk in Egypt. Vox Sang. 2009;96:23‒4. Abstract or poster Ellethy AT, Sliem HA, Hassan GMA. Updated molecular diagnosis of chronic hepatitis C. J Gastroenterol Hepatol Res. 2012;1(8):147‒52. Review article Ergunay K, Sener B, Alp A, Karakaya J, Hascelik G. Quantitative determination of HCV core antigen: A marker for monitoring hepatitis C viraemia. Clin Microbiol Infect. 2011;17:S663. Abstract or poster Fabrizi F, Lunghi G, Aucella F, Mangano S, Barbisoni F, Bisegna S, et al. Novel assay using total hepatitis C virus (HCV) core antigen quantification for diagnosis of HCV infection in dialysis patients. J Clin Microbiol. 2005;43(1):414‒20. Non-commercial or off-market assay Frank K, Karl A. Comparison of different confirmation test methods to anti-HCV used for repeatedly HCV-reactive samples. Infusionsther Transfusionsmed. 2001;28(Suppl. 1):47. No HCV core antigen performed Galli C. New strategies for the identification of active hepatitis C virus (HCV) infections. Biochim Clin. 2013;37:S137. Abstract or poster Galli C. New strategies for the screening of hepatitis C virus infection. Vox Sang. 2013;105:181. Abstract or poster Gaudy C, Thevenas C, Tichet J, Mariotte N, Goudeau A, Dubois F. Usefulness of the hepatitis C virus core antigen assay for screening of a population undergoing routine medical checkup. J Clin Microbiol. 2005;43(4):1722‒6. Non-commercial or off-market assay Gavarro A, Cebollero A, Martinez S, Veraguas A, Acosta A. Evaluation of a new method to determine core antigen of hepatitis C virus and its comparison with the determination of RNAHCV. Clin Chem Lab Med. 2011;49:S522‒S. Abstract or poster Grant PR, Sims CM, Tedder RS. Quantification of HCV RNA levels and detection of core antigen in donations before seroconversion. Transfusion. 2002;42(8):1032‒6. Non-commercial or offmarket assay Gu SJ, Liu J, Zhang HJ, Gu BL, Lai HJ, Zhou HL, et al. Core antigen tests for hepatitis C virus: a meta-analysis. Mol Biol Rep. 2012;39(8):8197‒208. Review article Hadziyannis E, Vassilopoulos D, Georgiou A, Spanou F, Koskinas J. Comparison of a new HCV core antigen automated immunoassay to quantitative HCV RNA measurement in patients with chronic hepatitis C. Hepatology. 2010;52(4):1230A. Abstract or poster Hayashi K, Hasuike S, Kusumoto K, Ido A, Uto H, Kenji N, et al. Usefulness of a new immunoradiometric assay to detect hepatitis C core antigen in a community-based population. J Viral Hepat. 2005;12(1):106‒10. Non-commercial or off-market assay

59.

60. 61. 62.

63. 64.

65.

66.

67. 68. 69.

70.

71.

72. 73.

74.

Page | 433

75.

Heidrich B, Pischke S, Kirschner J, Schneider J, Mederacke I, Raupach R, et al. Performance of HCV-core antigen testing in liver transplanted patients. J Hepatol. 2011;54:S223. Abstract or poster Higashimoto M, Takahashi M, Jokyu R, Syundou H, Saito H. Improvement of sensitivity in the second generation HCV core antigen assay by a novel concentration method using polyethylene glycol (PEG). Rinsho Byori. 2007;55(11):1008‒14. Less than 10 independent samples Hmaïed F, Ben Mamou M, Arrouji Z, Slim A, Ben Redjeb S. Use of combined detection of hepatitis C virus core antigen and antibodies to reduce the serological window-phase. Pathol Biol (Paris). 2007;55(2):121‒6. Inappropriate reference test Hosseini-Moghaddam SM, Iran-Pour E, Rotstein C, Husain S, Lilly L, Renner E, et al. Hepatitis C core Ag and its clinical applicability: potential advantages and disadvantages for diagnosis and follow-up? Rev Med Virol. 2012;22(3):156‒65. Review article Icardi G, Ansaldi F, Bruzzone BM, Durando P, Lee S, de Luigi C, et al. Novel approach to reduce the hepatitis C virus (HCV) window period: clinical evaluation of a new enzyme-linked immunosorbent assay for HCV core antigen. J Clin Microbiol. 2001;39(9):3110‒14. Less than 10 independent samples Icardi G, Bruzzone B, Gota F, Torre F, Giannini E, Massone L, et al. A new assay for hepatitis C virus (HCV) core antigen detection: an alternative to nucleic acid technologies in positive or indeterminate anti-HCV subjects? Ann Ig. 2003;15(6):863‒70. Non-commercial or off market assay Irshad M, Dhar I, Joshi YK. Significance of hepatitis C virus core protein in the diagnosis of hepatitis C virus infection in different liver diseases. J Investig Med. 2006;54(8):478‒83. Duplicate data Irshad M, Dhar I, Khushboo, Singh S, Kapoor S. Comparison of serological and nucleic Acid based assays used to diagnose hepatitis C virus (HCV) infection in acute and chronic liver diseases. Int J Health Sci (Qassim). 2007;1(1):3‒10. No extractable data, no response from author Ivanov YD, Kaysheva AL, Frantsuzov PA, Pleshakova TO, Krohin NV, Izotov AA, et al. Detection of hepatitis C virus core protein in serum by atomic force microscopy combined with mass spectrometry. Int J Nanomedicine. 2015;10:1597‒608. Non-commercial or off market assay Ivanyi-Nagy R, Makowska Z, Lastra ML, Darlix JL. Overview on the properties and functions of the core protein of hepatitis C cvirus (HCV). In: Bernhardt LV, editor. Nova Science Publishers; 2010:1‒47. Textbook chapter Kamili S, Drobeniuc J, Araujo AC, Hayden TM. Laboratory diagnostics for hepatitis C virus infection. Clin Infect Dis. 2012;55:S43‒S48. Review article Kashiwakuma T, Hasegawa A, Kajita T, Takata A, Mori H, Ohta Y, et al. Detection of hepatitis C virus specific core protein in serum of patients by a sensitive fluorescence enzyme immunoassay (FEIA). J Immunol Methods. 1996;190(1):79‒89. Non-commercial or off-market assay Kenfe FR, Urbaczek AC, Silva JC, Neó TA, Da Silva FH, Da Costa PI. Development of diagnostic methods and study of the immunoreactivity of a mixture of recombinant core and E2 proteins fused to GST with control serum positive for hepatitis C. Talanta. 2013;110:32‒8. Noncommercial or off-market assay Kesli R, Polat H, Terzi Y, Kurtoglu MG, Uyar Y. Comparison of a newly developed automated and quantitative hepatitis C virus core antigen test with the hepatitis C virus RNA assay for the clinical usefulness of confirming anti-hepatitis C virus results. Clin Microbiol Infect. 2012;18:676.

76.

77.

78.

79.

80.

81.

82.

83.

84.

85. 86.

87.

88.

Page | 434

89.

Kocazeybek B, Yuksel P, Saribas S, Caliskan R, Ergin S, Aslan M, et al. New approaches in in-vitro diagnosis of hepatitis C infections: the diagnostic performance of new hepatitis C virus core antigen detection test. Int J Infect Dis. 2010;14:e228. Abstract or poster Krajden M, Shivji R, Gunadasa K, Mak A, McNabb G, Friesenhahn M, et al. Evaluation of the core antigen assay as a second-line supplemental test for diagnosis of active hepatitis C virus infection. J Clin Microbiol. 2004;42(9):4054‒9. Non-commercial or off-market assay Kuo YH, Lu SN. Is HCV core antigen (HCV Ag) an adequate marker for community screening? Hepatol Int. 2012;6(1):201‒2. Abstract or poster Kups J, Wozniakowska-Gesicka T. Evaluation of specific humoral response to hepatitis C core antigen in children. Przegl Lek. 2003;60(12):802‒5. No HCV core antigen performed Kurtz JB, Boxall E, Qusir N, Shirley J, Coleman D, Chandler C. The diagnostic significance of an assay for “total” hepatitis C core antigen. J Virol Methods. 2001;96(2):127‒32. Non-commercial or off-market assay Lagging LM, Garcia CE, Westin J, Wejstål R, Norkrans G, Dhillon AP, et al. Comparison of serum hepatitis C virus RNA and core antigen concentrations and determination of whether levels are associated with liver histology or affected by specimen storage time. J Clin Microbiol. 2002;40(11):4224‒9. Non-commercial or off-market assay Lambert N. Value of HCV antigen-antibody combined HCV assay in hepatitis C diagnosis. Dev Biol (Basel). 2007;127:113‒21. Review article Laperche S, Elghouzzi MH, Morel P, Asso-Bonnet M, Le Marrec N, Girault A, et al. Is an assay for simultaneous detection of hepatitis C virus core antigen and antibody a valuable alternative to nucleic acid testing? Transfusion. 2005;45(12):1965‒72. Non-human subjects or commercial samples Combined detection of hepatitis C virus core antigen and antibody as an alternative to nucleic acid testing in blood screening. Vox Sang. 2005;89:21. Abstract or poster Laperche S, Le Marrec N, Simon N, Bouchardeau F, Defer C, Maniez-Montreuil M, et al. A new HCV core antigen assay based on disassociation of immune complexes: an alternative to molecular biology in the diagnosis of early HCV infection. Transfusion. 2003;43(7):958‒62. Noncommercial Laperche S, Rouger P, Smilovici W, Herve P, Lefrere JJ. Alternatives to nucleic acid testing in the blood transfusion service. Lancet. 2002;360(9344):1519. Editorial or comment

90.

91. 92. 93.

94.

95. 96.

97. 98.

99.

100. Larrat S, Bourdon C, Baccard M, Garnaud C, Hilleret MN, Quesada JL, et al. HCV screening without venipuncture: pointof-care versus laboratory-based antigenantibody combined assay. Hepatology. 2011;54:573A‒4A. Abstract or poster 101. Larrat S, Bourdon C, Baccard M, Garnaud C, Mathieu S, Quesada JL, et al. Performance of an antigen-antibody combined assay for hepatitis C virus testing without venipuncture. J Clin Virol. 2012;55(3):220‒5. Inappropriate reference test 102. Leary TP, Gutierrez RA, Muerhoff AS, Birkenmeyer LG, Desai SM, Dawson GJ. A chemiluminescent, magnetic particle-based immunoassay for the detection of hepatitis C virus core antigen in human serum or plasma. J Med Virol. 2006;78(11):1436‒40. Non-human subjects or commercial samples 103. Lee L, Perrett G, Howard P, Makanjuola D, Clark J. HCV p22 antigen test: Serological response and diagnostic advantages. Gut. 2012;61:A140. Abstract or poster

Page | 435

104. Lee S, Kim YS, Jo MJ, Jin M, Lee DK, Kim S. Chip-based detection of hepatitis C virus using RNA aptamers that specifically bind to HCV core antigen. Biochem Biophys Res Commun. 2007;358(1):47‒52. Non-commercial or off-market assay 105. Lee SR, Peterson J, Niven P, Bahl C, Page E, DeLeys R, et al. Efficacy of a hepatitis C virus core antigen enzyme-linked immunosorbent assay for the identification of “window-phase” blood donations. Vox Sang. 2001;80(1):19‒23. Non-commercial or off-market assay 106. Leon P, Lopez JA, Elola C, Lee SR, Calmann M, Echevarria JM. Use of overlapping synthetic peptides to characterize samples from blood donors with indeterminate results to hepatitis C virus core antigen. Vox Sang. 1998;75(1):32‒6. No HCV core antigen performed 107. Letowska M, Rosiek A, Gronowska A, Mikulska M, Grabarczyk P, Brojer E. Hepatitis C virus (HCV) core antigen detection in HCV RNA-positive/anti-HCV-negative Polish blood donors identified by nucleic acid testing. Transfusion. 2009;49(10):2241‒2. Editorial or comment 108. Li Cavoli G, Zagarrigo C, Schillaci O, Tralongo A, Rotolo U. Hepatitis C virus core antigen testing in the monitoring of patients on dialysis. Saudi J Kidney Dis Transpl. 2012;23(5):1056‒8. Editorial or comment 109. Li Cavoli G, Zagarrigo C, Tralongo A, Schillaci O, Rotolo U, Li Destri N, et al. Hepatitis C virus core antigen in virological monitoring of dialysis patients. NDT Plus. 2010;3:iii247. Duplicate data 110. Liao WS, Shen CH, Tung SY, Lee JH. Detection of NS3 antigen of hepatitis C virus in liver tissue and its relation to serum HCV-RNA. J Gastroenterol Hepatol. 2009;24:A182. Non-blood specimen 111. Long L, Shen T, Gao J, Duan Z, Liang H, Lu F. Effectiveness of HCV core antigen and RNA quantification in HCV-infected and HCV/HIV-1-coinfected patients. BMC Infect Dis. 2014;14:577. No extractable data, non-response from author 112. Long RX, Li H, Cui PF. Stability of three hepatitis C virus markers. Chinese Journal of Biologicals. 2009;22(5):468‒9, 473. No HCV core antigen performed 113. Lorenzo J, Castro A, Aguilera A, Prieto E, López-Calvo S, Regueiro B, et al. Total HCV core antigen assay: a new marker of HCV viremia and its application during treatment of chronic hepatitis C. J Virol Methods. 2004;120(2):173‒7. Non-commercial or off-market assay 114. Lozano ML, Candela MJ, Lozano ML, Cano H, Zuazu I, Vicente V. Detection of free hepatitis C virus core antigen by enzyme-linked immunosorbent assay is not suitable for screening of granulocyte colony-stimulating factor-mobilized hematopoietic progenitor donors. Transfusion. 2004;44(12):1755‒61. Non-commercial or off-market assay 115. Łucejko M, Grzeszczuk A, Jaroszewicz J, Flisiak R. [Serum HCV core antigen concentration in HCV monoinfection and HCV/HIV coinfection.] Pol Merkur Lekarski. 2013;35(206):72‒6. Unable to translate 116. Lunel F, Pivert A, Payan C, Comm RS. Comparison of HCV RNA and HCV core antigen kinetics in the follow up of a therapeutic protocol in HCV-HIV co-infected patients (RIBAVIC). J Hepatol. 2004;40:144‒5. Abstract or poster 117. Lunel F, Veillon P. Antigen and viral load. Transfus Clin Biol. 2003;10(2):74‒7. Non-commercial or off-market assay 118. Lunel F, Veillon P, Payan C. Evaluation of the ortho total HCV core antigen assay in comparison to methods of detection and quantification for HCV RNA. Hepatology. 2002;36(4):353A. Abstract or poster 119. Lunel-Fabiani F, Payan C. Virological tools for the diagnosis and follow up of hepatitis C: use and role of new tests. Gastroenterol Clin Biol. 2003;27(8‒9):718‒26. Review article

Page | 436

120. Ma CX, Xie GM, Zhang W, Liang M, Liu B, Xiang H. Label-free sandwich type of immunosensor for hepatitis C virus core antigen based on the use of gold nanoparticles on a nanostructured metal oxide surface. Mikrochim Acta. 2012;178(3‒4):331‒40. Non-commercial or off-market assay 121. Masalova OV, Atanadze SN, Samokhvalov EI, Petrakova NV, Kalinina TI, Smirnov VD, et al. Detection of hepatitis C virus core protein circulating within different virus particle populations. J Med Virol. 1998;55(1):1‒6. Non-commercial or off-market assay 122. Masalova OV, Vishnevskaia TV, Shkurko TV, Garanzha TA, Tupoleva TA, Filatov FP, et al. Comparative analysis of hepatitis C virus core protein in the plasma and serum samples from HCV-infected blood donors and patients with hepatitis C. Vopr Virusol. 2007;52(4):11‒7. Noncommercial or off-market assay 123. Masopust J, Kracíková J, Němeček V, Procházková R. [A combined antigen-antibody HCV detection assay.] Transfuze a Hematologie Dnes. 2005;11(4):141‒7. Unable to translate 124. Massaguer A, Forns X, Costa J, Feliu A, García-Retortillo M, Navasa M, et al. Performance of hepatitis C virus core antigen immunoassay in monitoring viral load after liver transplantation. Transplantation. 2005;79(10):1441‒4. Non-commercial or off-market assay 125. Mathur A. Comparative evaluation of anti-HCV-third-generation assay and anti-HCV-fourthgeneration assay in blood donor screening. Transfusion. 2014;54:216A‒7A. Abstract or poster 126. Mayerhofer T, Deimel M, Ambrus E, Bertsch AL, Aliskanovic V, Trubert-Exinger D, et al. Evaluation of the clinical performance of the Abbott ARCHITECT HCV core antigen assay. Clinical Chemistry and Laboratory Medicine. 2012;50(4):A86. Abstract or poster 127. Mederacke I, Meier M, Lüth JB, Schmidt-Gürtler H, Raupach R, Horn-Wichmann R, et al. Different kinetics of HBV and HCV during haemodialysis and absence of seronegative viral hepatitis in patients with end-stage renal disease. Nephrol Dial Transplant. 2011;26(8):2648‒56. Less than 10 independent samples 128. Mederacke I, Raupach R, Manns MP, Wedemeyer H, Tillmann HL. Quantitative HCV core antigen and HCV-RNA kinetics in HCV infected patients over a period of up to 8 years. Hepatology. 2009;50(4):1070A‒1A. Abstract or poster 129. Medhi S, Potukuchi SK, Polipalli SK, Swargiary SS, Deka P, Chaudhary A, et al. Diagnostic utility of hepatitis C virus core antigen in hemodialysis patients. Clin Biochem. 2008;41(7‒8):447‒52. Non-commercial or off market assay 130. Meng S, Li X, Yin H, Li DF. Establishment of detection test for hepatitis C virus antigen. Zhonghua Shi Yan He Lin Chuang Bing Du Xue Za Zhi. 2001;15(3):287‒90. Non-commercial or off-market assay 131. Miceli M, Agresti A, Palange M, Iudicone P, Paluzzi C, Perret M, et al. A new high sensitive HCV antigen assay for blood donors screening. Vox Sang. 2010;99:300. Abstract or poster 132. Mihaljevic I. Evaluation of the architect HCV antigen assay for use in organ and tissue donors. Vox Sang. 2009;96:24. Abstract or poster 133. Moini M, Ziyaeyan M, Aghaei S, Sagheb MM, Taghavi SA, Moeini M, et al. Hepatitis C virus (HCV) infection rate among seronegative hemodialysis patients screened by two methods; HCV core antigen and polymerase chain reaction. Hepat Mon. 2013;13(6). Non-commercial or off-market assay 134. Moini M, Ziyaeyan M, Aghaei S, Sagheb MM, Taghavi SA, Moeini M, et al. Hepatitis C virus (HCV) screening using HCV core antigen in hemodialysis patients. Hepatol Int. 2012;6(1):150. Abstract or poster

Page | 437

135. Moreno M, Pérez-Alvarez R, Rodrigo L, Pérez-López R, Suárez-Leiva P. Long-term evolution of serum and liver viral markers in patients treated for chronic hepatitis C and sustained response. J Viral Hepat. 2006;13(1):28‒33. Non-commercial or off market-assay 136. Moriya T, Sasaki F, Tanaka J, Mizui Ma, Nakanishi T, Takahashi K, et al. Comparison of HCV core antigen activity by ELISA and amount of HCV RNA by branched DNA assay. International Hepatology Communications. 1994;2(3):175‒7. Non-commercial or off-market assay 137. Morota K, Fujinami R, Kinukawa H, Machida T, Ohno K, Saegusa H, et al. A new sensitive and automated chemiluminescent microparticle immunoassay for quantitative determination of hepatitis C virus core antigen. J Virol Methods. 2009;157(1):8‒14. Non-human subjects or commercial samples 138. Muerhoff AS, Jiang L, Shah DO, Gutierrez RA, Patel J, Garolis C, et al. Detection of HCV core antigen in human serum and plasma with an automated chemiluminescent immunoassay. Transfusion. 2002;42(3):349‒56. Non-commercial or off-market assay 139. Mukomolov S, Barantsevich E, Mukomolova A, Barantsevich N, Schlyakhto E. Study of sensitivity and specificity of ELISA test system MONOLISA HCV Ag-Ab ULTRA. Clin Microbiol Infect. 2010;16:S298. Abstract or poster 140. Nayak NC, Sathar SA. Immunohistochemical detection of hepatitis C virus antigen in paraffin embedded liver biopsies from patients with chronic liver disease. Acta Histochem. 1999;101(4):409‒19. Non-blood specimens 141. Netski DM, Wang XH, Mehta SH, Nelson K, Celentano D, Thongsawat S, et al. Hepatitis C virus (HCV) core antigen assay to detect ongoing HCV infection in Thai injection drug users. J Clin Microbiol. 2004;42(4):1631‒6. Non-commercial or off-market assay 142. Nuovo GJ, Holly A, Wakely P, Frankel W. Correlation of histology, viral load, and in situ viral detection in hepatic biopsies from patients with liver transplants secondary to hepatitis C infection. Hum Pathol. 2002;33(3):277‒84. Non-blood specimens 143. Oliva JA, Ercilla G, Mallafre JM, Bruguera M, Carrio J, Pereira BJ. Markers of hepatitis C infection among hemodialysis patients with acute and chronic infection: implications for infection control strategies in hemodialysis units. Int J Artif Organs. 1995;18(2):73‒7. No HCV core antigen performed 144. Orito E, Mizokami M, Tanaka T, Lau JYN, Suzuki K, Yamauchi M, et al. Quantification of serum hepatitis C virus core protein level in patients chronically infected with different hepatitis C virus genotypes. Gut. 1996;39(6):876‒80. Non-commercial or off-market assay 145. Ottiger C, Gygli N, Huber AR. Detection limit of architect hepatitis C Core antigen assay in correlation with HCV-PCR viral load. Clinical Chemistry and Laboratory Medicine. 2012;50(5):A171. Abstract or poster 146. Pawelczyk A, Kubisa N, Jablonska J, Bukowska-Osko I, Caraballo Cortes K, Fic M, et al. Detection of hepatitis C virus (HCV) negative strand RNA and NS3 protein in peripheral blood mononuclear cells (PBMC): CD3+, CD14+ and CD19+. Virol J. 2013;10:346. No HCV core antigen performed 147. Pawlotsky JM. Use and interpretation of virological tests for hepatitis C. Hepatology. 2002;36(5):S65‒S73. Review article 148. Pawlotsky JM. Use and interpretation of hepatitis C virus diagnostic assays. Clin Liver Dis. 2003;7(1):127‒37. Review article 149. Pessôa MG, Alves VAF, Wakamatsu A, Gomes JG, Maertens G, van der Borght B, et al. Posttransplant recurrent hepatitis C: immunohistochemical detection of hepatitis C virus core

Page | 438

antigen and possible pathogenic implications. Liver Int. 2008;28(6):807‒13. Non-blood specimens 150. Peterson J, Green G, Iida K, Caldwell B, Kerrison P, Bernich S, et al. Detection of hepatitis C core antigen in the antibody negative “window” phase of hepatitis C infection. Vox Sang. 2000;78(2):80‒5. Non-commercial or off-market assay 151. Pham BN, Martinot-Peignoux M, Ripault MP, Boyer N, Levy V, Marcellin P. Quantitative measurement of hepatitis C virus core antigen is affected by the presence of cryoglobulins. Clin Exp Immunol. 2006;146(2):211‒17. Non-commercial or off-market assay 152. Pivert A, Payan C, Lunel F, Conseil scientifique R. Comparison of hepatitis C viral RNA and core antigen kinetics in the therapeutic follow up of hepatitis C virus and human immunodeficiency virus co-infected patients, treated by bitherapy interferon-ribavirin, within the framework of RIBAVIC protocol. Pathol Biol (Paris). 2004;52(9):522‒8. Duplicate data 153. Podesta MA, Cancarini G, Cucchiari D, Montanelli A, Badalamenti S, Graziani G. Diagnosis and follow-up of HCV infection in hemodialysis patients and renal transplant recipients: HCV core antigen and IgM anti-HCV. Nephrol Dial Transplant. 2014;29:iii514. Abstract or poster 154. Poljak M, Lepej SZ, Rode OA. Recent developments in serologic and molecular diagnosis of hepatitis B and C. Acta Med Croatica. 2013;67(4):281‒90. Review article 155. Pontisso P, Carabaich A, Bernardinello E, Boccato S, Chemello L, Gatta A, et al. Correlation of HCV RNA to total HCV core antigen in chronic hepatitis C patients. Hepatology. 2002;36(4):546A. Abstract or poster 156. Pradat P, Maynard M, Berthillon P, Bailly F, Bordes I, Tillmann HL, et al. Baseline HCV core antigen/HCV RNA ratio: association with treatment responses. Hepatology. 2004;40(4):326A. Abstract or poster 157. Raffaele L, Spreafico M, Foglieni B, Guarnori I, Alessandra B, Galli C, et al. Performance evaluation of fully automated quantitative HCV-core antigen assay. Transfusion. 2010;50:202A. Abstract or poster 158. Raker CA, Tabor E, Okayama A, Yu MYW, Kohara M, Mueller NE, et al. HCV core antigen as an alternative to NAT to detect HCV viremia. Transfusion. 2004;44(2):307‒8. Editorial or comment 159. Ravera G, Bottaro LC, Franceschini M, Morando A, De Polo M, Zare M, et al. Reliability and diagnostic use of a test for the search of the hepatitis C virus Ag (AgHCV). Hepatogastroenterology. 2006;53(71):753‒6. Non-commercial or off-market assay 160. Rebucci C, Cerino A, Cividini A, Timo L, Furione M, Mondelli MU. Monitoring response to antiviral therapy for patients with chronic hepatitis C virus infection by a core-antigen assay. J Clin Microbiol. 2003;41(8):3881‒4. Less than 10 independent samples 161. Reddy AK, Dakshinamurty KV, Lakshmi V. Utility of HCV core antigen ELISA in the screening for hepatitis C virus infection in patients on hemodialysis. Indian J Med Microbiol. 2006;24(1):55‒7. Less than 10 independent samples 162. Rehany U, Chaikovsky I, Kra-Oz Z, Satinger J, Bersudsky V, Cohen I. The expression of hepatitis-C virus antigen by immunohistochemical stain and polymerase chain reaction in corneas of seropositive corneal donors. Cell Tissue Bank. 2002;3(2):139‒44. Non-blood specimens 163. Romano L, Zanetti AR, Brunetto M, Ciccorossi P, Lee SR, Calmann M, et al. Correlation of total HCV core antigen ELISA to HCV RNA in liver transplant patients. Hepatology. 2002;36(4):654A. Abstract or poster

Page | 439

164. Ross RS, Viazov S, Salloum S, Hilgard P, Gerken G, Roggendorf M. Analytical performance characteristics and clinical utility of a novel assay for total hepatitis C virus core antigen quantification. J Clin Microbiol. 2010;48(4):1161‒8. No extractable data, no response from author 165. Sabry AA, Sobh MA, Irving WL, Grabowska A, Wagner BE, Fox S, et al. A comprehensive study of the association between hepatitis C virus and glomerulopathy. Nephrol Dial Transplant. 2002;17(2):239‒45. Non-blood specimens 166. Saeed M, Suzuki R, Kondo M, Aizaki H, Kato T, Mizuochi T, et al. Evaluation of hepatitis C virus core antigen assays in detecting recombinant viral antigens of various genotypes. J Clin Microbiol. 2009;47(12):4141‒3. Non-human subjects or commercial samples 167. Sansonno D, Iacobelli AR, Cornacchiulo V, Iodice G, Dammacco F. Detection of hepatitis C virus (HCV) proteins by immunofluorescence and HCV RNA genomic sequences by non isotopic in situ hybridization in bone marrow and peripheral blood mononuclear cells of chronically HCVinfected patients. Clin Exp Immunol. 1996;103(3):414‒21. Non-blood specimens 168. Sansonno D, Lauletta G, Dammacco F. Detection and quantitation of HCV core protein in single hepatocytes by means of laser capture microdissection and enzyme-linked immunosorbent assay. J Viral Hepat. 2004;11(1):27‒32. Non-blood specimens 169. Sansonno D, Lauletta G, Montrone M, Grandaliano G, Schena FP, Dammacco F. Hepatitis C virus RNA and core protein in kidney glomerular and tubular structures isolated with laser capture microdissection. Clin Exp Immunol. 2005;140(3):498‒506. Non-blood specimens 170. Sansonno D, Lauletta G, Nisi L, Gatti P, Pesola F, Pansini N, et al. Non-enveloped HCV core protein as constitutive antigen of cold-precipitable immune complexes in type II mixed cryoglobulinaemia. Clin Exp Immunol. 2003;133(2):275‒82. Non-commercial or off-market assay 171. Sansonno L, Lauletta G, Russi S, Dammacco F. In situ simultaneous detection of hepatitis C virus RNA and hepatitis C virus-related antigens in hepatocellular carcinoma. Cancer Treatment Reviews. 2010;36:S98. Abstract or poster 172. Sarov B, Novack L, Beer N, Safi J, Soliman H, Pliskin JS, et al. Feasibility and cost-benefit of implementing pooled screening for HCVAg in small blood bank settings. Transfus Med. 2007;17(6):479‒87. Does not apply to study question 173. Sarrazin C. Advances in diagnosis and management of virus c hepatitis. Biochim Clin. 2013;37:S60‒S1. Review article 174. Schiano TD, Gutierrez JA, Walewski JL, Fiel MI, Cheng B, Bodenheimer Jr H, et al. Accelerated hepatitis C virus kinetics but similar survival rates in recipients of liver grafts from living versus deceased donors. Hepatology. 2005;42(6):1420‒8. Does not apply to study question, noncommercial or off-market assay 175. Schiano TD, Gutierrez JA, Walewski JL, Fiel MI, Cheng B, Nelson C, et al. HCV RNA, core antigen, ALT values, and histology after liver transplantation: two year follow-up of CDLT versus LDLT. Hepatology. 2004;40(4):164A. Abstract or poster 176. Schmidt AJ, Falcato L, Zahno B, Burri A, Regenass S, Müllhaupt B, et al. Prevalence of hepatitis C in a Swiss sample of men who have sex with men: Whom to screen for HCV infection? BMC Public Health. 2014;14:3. Less than 10 independent samples 177. Schüttler CG, Thomas C, Discher T, Friese G, Lohmeyer J, Schuster R, et al. Variable ratio of hepatitis C virus RNA to viral core antigen in patient sera. J Clin Microbiol. 2004;42(5):1977‒81. Non-commercial or off-market assay

Page | 440

178. Seiskari T, Horsti J, Aittoniemi J. The value of HCV antigen test with reference to PCR in confirming infection in cases with indefinite antibody test result. J Clin Virol. 2011;51(1):90‒1. Editorial or comment 179. Seme K, Poljak M, Babic DZ, Mocilnik T, Vince A. The role of core antigen detection in management of hepatitis C: a critical review. J Clin Virol. 2005;32(2):92‒101. Review article 180. Sentjens RE, van Baal K, Verhoeven G, Lee S, Pieksma F, Niven P, et al. Comparison of first and second generation HCV core antigen test with qualitative and quantitative HCV-RNA PCR. Hepatology. 2000;32(4):548A. Abstract or poster 181. Serdarevic N. Detection of hepatitis C in serum using anti HCV and HCV Ag. Clinical Chemistry and Laboratory Medicine. 2014;52:S975. Abstract or poster 182. Shah DO, Chang C, Gutierrez R, Cheng KY, Jiang LX, Salbilla V, et al. Hepatitis C virus antibody test on a chemiluminescence automated analyzer as an aid in the diagnosis of HCV infection and for blood screening. Transfusion. 2010;50:203A‒4A. Abstract or poster 183. Shah DO, Chang CD, Jiang LX, Cheng KY, Muerhoff AS, Gutierrez RA, et al. Combination HCV core antigen and antibody assay on a fully automated chemiluminescence analyzer. Transfusion. 2003;43(8):1067‒74. Non-commercial or off-market assay 184. Shah O, Chang D, Gutierrez A, Cheng Y, Desai M, Dawson J. Hepatitis C virus antibody assay on a chemiluminescence automated analyzer as an AID in the diagnosis of HCV infection and for blood screening. Vox Sang. 2010;99:297‒8. Abstract or poster 185. Shaw SG, Cresswell F, Homer G, Hassan-Ibrahim M, Fisher M. Hepatitis C antigen testing: a reliable alternative for diagnosing acute HCV infection. Top Antivir Med. 2014;22(e‒1):312. Abstract or poster 186. Shen T, Chen X, Zhang W, Xi Y, Cao G, Zhi Y, et al. A higher correlation of HCV core antigen with CD4+ T cell counts compared with HCV RNA in HCV/HIV coinfected patients. Hepatol Int. 2012;6(1):150. Abstract or poster 187. Shepherd SJ, Aitken C, Walkowicz M, McOwan J, Cameron SO, Carman WF. HCV antigen testing in a busy diagnostic laboratory. Clin Microbiol Infect. 2012;18:676‒7. Abstract or poster 188. Shi HB, Xie L, Huang DZ, He LX. Detection of HCV antigen in the serum by immunodotting. Xi Bao Yu Fen Zi Mian Yi Xue Za Zhi. 2006;22(3):336‒8. Non-commercial or off market assay 189. Song D, Kang JE, Kim SY, Hwang SH, Kim HH, Lee EY, et al. [Evaluation of ARCHITECT HCV core antigen assay.] [Article in Korean] Korean J Lab Med. 2010;30(6):654‒9. Unable to translate 190. Syed SI, Sadiq S. Immunohistochemical detection of hepatitis C virus (HCV) in liver biopsies of hepatitis C patients. J Pak Med Assoc. 2011;61(12):1198‒201. Non-blood specimens 191. Tagny CT, Mbanya D, Murphy EL, Lefrère JJ, Laperche S. Screening for hepatitis C virus infection in a high prevalence country by an antigen/antibody combination assay versus a rapid test. J Virol Methods. 2014;199:119‒23. Inappropriate reference test 192. Takahashi K, Okamoto H, Kishimoto S, Munekata E, Tachibana K, Akahane Y, et al. Demonstration of a hepatitis C virus-specific antigen predicted from the putative core gene in the circulation of infected hosts. J Gen Virol. 1992;73(Pt 3):667‒72. No HCV core antigen performed 193. Tanaka E, Kiyosawa K, Matsumoto A, Kashiwakuma T, Hasegawa A, Mori H, et al. Serum levels of hepatitis C virus core protein in patients with chronic hepatitis C treated with interferon alfa. Hepatology. 1996;23(6):1330‒3. Non-commercial or off-market assay 194. Tanaka E, Ohue C, Aoyagi K, Yamaguchi K, Yagi S, Kiyosawa K, et al. Evaluation of a new enzyme immunoassay for hepatitis C virus (HCV) core antigen with clinical sensitivity approximating

Page | 441

that of genomic amplification of HCV RNA. Hepatology. 2000;32(2):388‒93. Non-commercial or off market assay 195. Tanoue Y, Maekawa H, Inoue T, Watanabe T, Shimoda H, Kuroda T, et al. The Cobas AmpliPrep/Cobas TaqMan real-time polymerase chain reaction assay produced false-negative results in two patients with hepatitis C virus genotype 2. Kanzo. 2013;54(7):507‒8. Less than 10 independent samples 196. Teo CG. Changing strategies for hepatitis C testing. Antivir Ther. 2012;17(7):1391‒5. Review article 197. Thong V, Akkarathamrongsin S, Avihingsanon A, Poovorawan Y, Tangkijvanich P. The correlation between hepatitis C core antigen and HCV RNA levels with respect to HIV status, HCV genotype and IFNL4 polymorphism. J Viral Hepat. 2014;21:21. Abstract or poster 198. Tillmann HL. Hepatitis C virus core antigen testing: role in diagnosis, disease monitoring and treatment. World J Gastroenterol. 2014;20(22):6701‒6. Review article 199. Tillmann HL, Wiegand J, Glomb I, Jelineck A, Picchio G, Wedemeyer H, et al. Diagnostic algorithm for chronic hepatitis C virus infection: role of the new HCV-Core antigen assay. Z Gastroenterol. 2005;43(1):11‒6. Non-commercial or off-market assay 200. Tobler LH, Stramer SL, Lee SR, Baggett D, Wright D, Hirschkorn D, et al. Performance of ORTHO (R) HCV core antigen and trak-C (TM) assays for detection of viraemia in pre-seroconversion plasma and whole blood donors. Vox Sang. 2005;89(4):201‒7. Non-commercial or off-market assay 201. Tokita H, Kaufmann GR, Matsubayashi M, Okuda I, Tanaka T, Harda H, et al. Hepatitis C virus core mutations reduce the sensitivity of a fluorescence enzyme immunoassay. J Clin Microbiol. 2000;38(9):3450‒2. Non-commercial or off-market assay 202. Tsutsumi M, Urashima S, Takada A, Date T, Tanaka Y. Detection of antigens related to hepatitis C virus RNA encoding the NS5 region in the livers of patients with chronic type C hepatitis. Hepatology. 1994;19(2):265‒72. Non-blood specimens 203. Valcavi P, Medici MC, Casula F, Arcangeletti MC, De Conto F, Pinardi F, et al. Evaluation of a total hepatitis C virus (HCV) core antigen assay for the detection of antigenaemia in anti-HCV positive individuals. J Med Virol. 2004;73(3):397‒403. Non-commercial or off-market assay 204. Vanhommerig JW, Van De Laar TJW, Van Rooijen MS, De Vries HJ, Speksnijder AGCL, Prins M, et al. Hepatitis c virus (HCV) core antigen assay for detection of active HCV infection among HIVinfected men who have sex with men attending an STI clinic. J Hepatol. 2014;60(1):S312. Duplicate data 205. Vargas V, Krawczynski K, Castells L, Martinez N, Esteban J, Allende H, et al. Recurrent hepatitis C virus infection after liver transplantation: Immunohistochemical assessment of the viral antigen. Liver Transpl Surg. 1998;4(4):320‒7. Non-blood specimens 206. Veillon P, Payan C, Picchio G, Maniez-Montreuil M, Guntz P, Lunel F. Comparative evaluation of the total hepatitis C virus core antigen, branched-DNA, and amplicor monitor assays in determining viremia for patients with chronic hepatitis C during interferon plus ribavirin combination therapy. J Clin Microbiol. 2003;41(7):3212‒20. Non-commercial or off-market assay 207. Vrielink H, Vanderpoel CL, Reesink HW, Lelie PN. Comparison of two anti-hepatitis C virus enzyme-linked immunosorbent assays: Wellcozyme-VK45 and Ortho-2.0. Infusionsther Transfusionsmed. 1995;22(3):164‒7. No HCV core antigen performed

Page | 442

208. Vucetic D, Trkuljic M, Balint B, Borovcanin N, Ljubenov M, Jovicic D. Evaluation of immunoassays for hepatitis C Virus antigen and antibody combined detection in Reactive serum samples among blood donors. Vox Sang. 2009;96:95. Abstract or poster 209. Waldenström J, Konar J, Ekermo B, Norder H, Lagging M. Neonatal transfusion-transmitted hepatitis C virus infection following a pre-seroconversion window-phase donation in Sweden. Scand J Infect Dis. 2013;45(10):796‒9. Less than 10 independent samples 210. Wang C, Zhang L, Shen X. Development of a nucleic acid lateral flow strip for detection of hepatitis C virus (HCV) core antigen. Nucleosides Nucleotides Nucleic Acids. 2013;32(2):59‒68. Non-commercial or off-market assay 211. Wang F, Wang S, Jin L. Detections of hepatitis C virus RNA and NS3 antigen and their relation to liver histopathology. Zhonghua Yi Xue Za Zhi. 1995;75(11):670‒2, 709‒10. Non-blood specimens 212. Wang FC, Shi ZY, Cai J, Su J. Evaluation on the use of detection of hepatitis C core antigen for screening blood donor. Zhonghua Shi Yan He Lin Chuang Bing Du Xue Za Zhi. 2007;21(4):389‒ 90. Less than 10 independent samples 213. Watanabe J, Matsumoto C, Fujimura K, Shimada T, Yoshizawa H, Okamoto H, et al. Predictive value of screening tests for persistent hepatitis C virus infection evidenced by viraemia. Japanese experience. Vox Sang. 1993;65(3):199‒203. No HCV core antigen performed 214. Widell A, Molnegren V, Pieksma F, Calmann M, Peterson J, Lee SR. Detection of hepatitis C core antigen in serum or plasma as a marker of hepatitis C viraemia in the serological window-phase. Transfus Med. 2002;12(2):107‒13. Non-commercial or off-market assay 215. Willems M, Lerut J, Roskams T, Donataccio M, Ciccarelli O, Habets W, et al. Detection of HCV antigen and HCV-RNA in liver biopsies of patients with chronic hepatitis-C undergoing livertransplantation ‒ a close correlation. Hepatology. 1993;18(4):A262‒A. Non-blood specimens 216. Wong JSJ, Karamalakis D. Evaluation of automated hepatitis C core antigen assay in occupational exposure. Pathology. 2013;45(5):529‒31. Editorial or comment 217. Wu RH. [Detection of hepatitis C virus antigen in hemodialysis patients]. Zhonghua Shi Yan He Lin Chuang Bing Du Xue Za Zhi. 2009;23(3):232‒4. Review article 218. Xie L, Guang YP, Liu F, Shi HB, Yan L, Lou JL. A magnetic microparticle-based immunoassay for hepatitis C virus NS3 antigen. Afr J Microbiol Res. 2011;5(1):28‒33. Non-commercial or offmarket assay 219. Xie L, Huang DZ, Chen HL, He LX, Wang J, Han DK. The clinical application and analysis of hepatitis C virus NS 3 antigen detection by ELISA in human serum. Chinese Journal of Microbiology and Immunology. 2009;29(1):88‒92. No extractable data, non-response from author 220. Xie L, Wu XD, Huang DZ, Chen HL, He LX, Wang J, et al. Clinical application and analysis of hepatitis C virus NS3 antigen detection by ELISA in human serum. Chin Med J (Engl). Duplciate data 2007;120(4):294‒9. 221. Yagci S, Padalko E. Comparison of monolisa HCV Ag/Ab ULTRA with two anti-HCV assays for the detection of HCV infection in hospital setting. Curr Microbiol. 2012;64(2):148‒51. Inappropriate reference test 222. Yan F, Hao F, Zhao L. Study of expression of hepatitis C virus antigens and viral replication in extrahepatic tissues. Zhonghua Gan Zang Bing Za Zhi. 2000;8(1):40‒2. Non-blood specimens 223. Yan FM, Chen AS, Hao F, Zhao XP, Gu CH, Zhao LB, et al. Hepatitis C virus may infect extrahepatic tissues in patients with hepatitis C. World J Gastroenterol. 2000;6(6):805‒11. Non-blood specimens

Page | 443

224. Yang RF, Wei L. Identifying factors that may influence utility of HCV core antigen assay. Hepatol Int. 2014;8(1):S200. Abstract or poster 225. Yang Z, Qi ZB, Yu Y. Evaluation of A kit for determination of total hepatitis C virus core antigen. Chinese Journal of Biologicals. 2008;21(1):51‒3. Non-commercial or off-market assay 226. Yao RN, Zhang JH, Huang XJ, Yang Q, Cao Q, Jiang XC. Technique of detection of hepatitis C core antigen used in safety blood transfusion. Zhongguo Shi Yan Xue Ye Xue Za Zhi. 2006;14(3):617‒ 8. Non-commercial or off-market assay 227. Yap SH, Willems M, Van den Oord J, Habets W, Middeldorp JM, Hellings JA, et al. Detection of hepatitis C virus antigen by immuno-histochemical staining: a histological marker of hepatitis C virus infection. J Hepatol. 1994;20(2):275‒81. Non-blood specimens 228. Yokosuka O, Kawai S, Suzuki Y, Fukai K, Imazeki F, Kanda T, et al. Evaluation of clinical usefulness of second-generation HCV core antigen assay: comparison with COBAS AMPLICOR HCV MONITOR assay version 2.0. Liver Int. 2005;25(6):1136‒41. Non-commercial or off-market assay 229. You E, Lee W, Lee M, Kim M, Kang S. Comparison of HCV RNA with the positive pattern of RIBA. J Mol Diagn. 2013;15(6):888. Abstract or poster 230. Yousaf MZ, Idrees M, Saleem Z, Rehman IU, Ali M. Expression of core antigen of HCV genotype 3a and its evaluation as screening agent for HCV infection in Pakistan. Virol J. 2011;8. Noncommercial or off-market assay 231. Zhang C, Zhou Y, Wang C. Detection of hepatitis C virus RNA and C33c antigen in the liver tissue from hepatitis C virus infection patients with chronic liver disease. Zhonghua nei ke za zhi. 1995;34(3):176‒9. Non-blood specimens

PICO 5b: 1. Bruhn R, Lelie N, Busch M, Kleinman S. Relative efficacy of nucleic acid amplification testing and serologic screening in preventing hepatitis C virus transmission risk in seven international regions. Transfusion. 2015;55(6):1195‒205. Does not apply to study question Galli C. New strategies for the identification of active hepatitis C virus (HCV) infections. Biochimica Clinica. 2013;37:S137. Abstract or poster Galli C. New strategies for the screening of hepatitis C virus infection. Vox Sang. 2013;105:181. Abstract or poster Hosseini-Moghaddam SM, Iran-Pour E, Rotstein C, Husain S, Lilly L, Renner E, et al. Hepatitis C core Ag and its clinical applicability: potential advantages and disadvantages for diagnosis and followup? Rev Med Virol. 2012;22(3):156‒65. Review article Kadkhoda K, Smart G. HCV antigen testing for the diagnosis of hepatitis C infection: a cost-efficient algorithm. Clin Lab. 2014;60(4):677‒80. Does not apply to study question Marwaha N, Sachdev S. Current testing strategies for hepatitis C virus infection in blood donors and the way forward. World J Gastroenterol. 2014;20(11):2948‒54. Review article Reyes-Mendez MA, Juarez-Figueroa L, Iracheta-Hernandez P, Medina-Islas Y, Ruiz-Gonzalez V. Comparison of two diagnostic algorithms for the identification of patients with HCV viremia using a new HCV antigen test. Ann Hepatol. 2014;13(3):337‒42. Does not apply to study question Tagny CT, Mbanya D, Murphy EL, Lefrère JJ, Laperche S. Screening for hepatitis C virus infection in a high prevalence country by an antigen/antibody combination assay versus a rapid test. J Virol Methods. 2014;199:119‒23. Inappropriate reference test

2. 3. 4.

5. 6. 7.

Page | 444

Teo CG. Changing strategies for hepatitis C testing. Antivir Ther 2012;17(7):1391‒5. Review article

10. Tillmann HL, Wiegand J, Glomb I, Jelineck A, Picchio G, Wedemeyer H, et al. Diagnostic algorithm for chronic hepatitis C virus infection: Role of the new HCV-Core antigen assay. Z Gastroenterol. 2005;43(1):11‒6. Does not apply to study question

PICO 9: 1. Alsio A, Lagenland N, Pedersen C, Farkkila MA, Buhl M, Morch K, et al. Importance of HCV core antigen below 0.2 PG/ML day 3 in short-term Peg-interferon therapy for chronic genotype2/3 infection. Hepatology. 2010;52:813A. Abstract or poster Attallah AM, Shiha GE, Malak CAA, Hagras HE, Abdel-Razik WS, Ismail H. Utility of a novel HCV-NS4 antigen detection immunoassay for monitoring treatment of HCV-infected individuals with pegylated interferon alpha-2a. Hepatol Res. 2004;28(2):68‒72. Non-commercial or off-market assay Buti M, Mendez C, Schaper M, Sauleda S, Valdes A, Rodriguez-Frias F, et al. Hepatitis C virus core antigen as a predictor of non-response in genotype 1 chronic hepatitis C patients treated with peginterferon α-2b plus ribavirin. J Hepatol. 2004;40(3):527‒32. Non-commercial or off-market assay Colombatto P, Moriconi F, Oliveri F, Cavallone D, Ciccorossi P, Coco B, et al. Early HCV core antigen kinetics predict sustained virologic response in chronic hepatitis C patients treated with standard of care. Hepatol Int. 2013;7:S433. Abstract or poster Feng B, Yang R, Kong F, Zhang H, Rao H, Jin Q, et al. Earlier kinetics of serum HCV core antigen predict sustained virologic response in genotype 1 CHC patients during antiviral treatment. Hepatol Int. 2013;7:S420‒1. Abstract or poster Gonzalez V, Padilla E, Diago M, Gimenez MD, Sola R, Matas L, et al. Clinical usefulness of total hepatitis C virus core antigen quantification to monitor the response to treatment with peginterferon alpha-2a plus ribavirin. J Viral Hepat. 2005;12(5):481‒7. Non-commercial or offmarket assay Gonzalez V, Planas R, Padilla E, Perez C, Gimenez D, Matas L, et al. Total HCV core antigen correlates with HCV RNA, and predicts sustained virological response (SVR) to peginterferon alfa 2-a (PEG-IFN) and ribavirin (RIB) therapy in chronic hepatitis C (CHC). J Hepatol. 2004;40:142. Abstract or poster Higashimoto M, Takahashi M, Jokyu R, Saito H. Clinical benefit of HCV core antigen assay in patients receiving interferon and ribavirin combination therapy. Rinsho Byori. 2006;54(2):111‒5. No extractable data, non-response from author Higashimoto M, Takahashi M, Jokyu R, Syundou H, Saito H. Improvement of sensitivity in the second generation HCV core antigen assay by a novel concentration method using polyethylene glycol (PEG). Rinsho Byori. 2007;55(11):1008‒14. Less than 10 independent samples

10. Kim MN, Park JY, Kim HS, Kim DY, Ahn SH, Han KH. Benefit of hepatitis C virus core antigen assay in prediction of therapeutic response. Hepatology. 2011;54:845A. Abstract or poster 11. Lindh M, Lagging M, Westin J, Wejstål R, Norkrans G. Monitoring treatment response by the hepatitis C virus core antigen assay. Eur J Clin Microbiol Infect Dis. 2005;24(3):230‒2. Noncommercial or off market assay 12. Lunel F, Veillon P. Antigen and viral load. Transfus Clin Biol. 2003;10(2):74‒7. Non-commercial or off market assay

Page | 445

13. Maynard M, Pradat P, Berthillon P, Picchio G, Voirin N, Martinot M, et al. Clinical relevance of total HCV Core antigen testing for hepatitis C monitoring and for predicting patients ” response to therapy. J Viral Hepat. 2003;10(4):318‒23. Non-commercial or off-market assay 14. Mederacke I, Ciesek S, Raupach R, Wursthorn K, Deterding K, Steinmann E, et al. Kinetics of HCV core antigen during antiviral treatment of acute and chronic hepatitis C as determined by a novel chemiluminescent microparticle immunoassay. J Hepatol. 2009;50:S129. Abstract or poster 15. Nakamuta M, Shimohashi N, Tada S, Kinukawa N, Enjoji M, Uchimura K, et al. Serum levels of HCV RNA and core protein before and after incubation at 37°C for 24 h. Hepatol Res. 2001;19(3):254‒ 62. Does not apply to study question 16. Padilla E, Planas R, Gonzalez V, Perez C, Gimenez D, Matas L, et al. Total HCV core antigen correlates with HCV RNA, and predicts response to PEGinterferon alfa 2-a (PEG-IFN) and ribavirin (Rib) therapy in chronic hepatitis C. Hepatology. 2003;38(4):626A. Abstract or poster 17. Pivert A, Payan C, Lunel F, Conseil scientifique R. [Comparison of hepatitis C viral RNA and core antigen kinetics in the therapeutic follow up of hepatitis C virus and human immunodeficiency virus co-infected patients, treated by bitherapy interferon-ribavirin, within the framework of RIBAVIC protocol.] Pathologie Biologie (Paris). 2004;52(9):522‒8. Duplicate data 18. Pivert A, Payan C, Morand P, Fafi-Kremer S, Deshayes J, Carrat F, et al. Comparison of serum hepatitis C virus (HCV) RNA and core antigen levels in patients coinfected with human immunodeficiency virus and HCV and treated with interferon plus ribavirin. J Clin Microbiol. 2006;44(2):417‒22. Non-commercial or off-market assay 19. Pradat P, Maynard M, Berthillon P, Bailly F, Bordes I, Tillmann HL, et al. Baseline HCV core antigen/HCV RNA ratio: association with treatment responses. Hepatology. 2004;40(4):326A. Abstract or poster 20. Pradat P, Maynard M, Buti M, Berthillon P, Picchio G, Tillmann HL, et al. The predictive value of core antigen testing for the management of hepatitis C patients receiving pegylated interferon/ribavirin treatment. J Med Virol. 2004;73(3):392‒6. Non-commercial or off-market assay 21. Rebucci C, Cerino A, Cividini A, Timo L, Furione M, Mondelli MU. Monitoring response to antiviral therapy for patients with chronic hepatitis C virus infection by a core-antigen assay. J Clin Microbiol. 2003;41(8):3881‒4. Less than 10 independent samples 22. Ricco G, Moriconi F, Colombatto P, Oliveri F, Cavallone D, Ciccorossi P, et al. Early prediction of sustained virologic response by HCV core ag kinetics in chronic hepatitis c patients treated with peginterferon plus ribavirin. Dig Liver Dis. 2014;46:e46. Abstract or poster 23. Ross RS, Viazov S, Salloum S, Hilgard P, Gerken G, Roggendorf M. Analytical performance characteristics and clinical utility of a novel assay for total hepatitis C virus core antigen quantification. J Clin Microbiol. 2010;48(4):1161‒8. No extractable data, non-response from author 24. Russi S, Sansonno D, Mariggio MA, Vinella A, Pavone F, Lauletta G, et al. Assessment of total hepatitis C virus (HCV) core protein in HCV-related mixed cryoglobulinemia. Arthritis Res Ther. 2014;16(2):R73. Does not apply to study question 25. Sasase N, Kim SR, Kim KI, Taniguchi M, Imoto S, Mita K, et al. Usefulness of a new immunoradiometric assay of HCV core antigen to predict virological response during PEGIFN/RBV combination therapy for chronic hepatitis with high viral load of serum HCV RNA genotype 1b. Intervirology. 2008;51:70‒5. Non-commercial or off-market assay 26. Schnuriger A, Dominguez S, Valantin MA, Tubiana R, Duvivier C, Ghosn J, et al. Early detection of hepatitis C virus infection using a new combined antigen-antibody detection assay: potential use in

Page | 446

HIV co-infected individuals. Pathologie Biologie (Paris). 2006;54(10):578‒86. Does not apply to study question 27. Shakado S, Tanaka T, Inomata S, Anan A, Matsumoto T, Hanano T, et al. [Serum HCV core antigen level is a useful prediction marker of serum HCV RNA negative at 24-week-treatment with peginterferon and ribavirin in patients with chronic hepatitis C.] [Article in Japanese] Kanzo. 2006;47(8):411‒2. Abstract or poster 28. Shiratori Y, Kato N, Yokosuka O, Hashimoto E, Hayashi N, Nakamura A, et al. Cut-off values of HCV core “antigen” and HCV RNA levels for prediction of viral eradication by interferon treatment. Gastroenterology. 1996;110(4):A1324. Abstract or poster 29. Soffredini R, Rumi MG, Parravicini ML, Ronchi G, Del Ninno E, Russo A, et al. Serum levels of hepatitis C virus core antigen as a marker of infection and response to therapy. Am J Gastroenterol. 2004;99(9):1738‒43. Non-commercial or off-market assay 30. Takayanagi M, Tanaka A, Komura N, Kuroda M, Miyazawa Y, Kurihara Y, et al. Availability of “highrange” HCV-RNA monitoring and HCV-core antigen quantification for the prediction of the efficacy of IFN and ribavirin combination treatment. Teikyo Medical Journal. 2005;28(3):191‒7. Does not apply to study question 31. Tanaka E, Kiyosawa K, Matsumoto A, Kashiwakuma T, Hasegawa A, Mori H, et al. Serum levels of hepatitis C virus core protein in patients with chronic hepatitis C treated with interferon alfa. Hepatology. 1996;23(6):1330‒3. Non-commercial or off-market assay 32. Tanaka E, Ohue C, Aoyagi K, Yamaguchi K, Yagi S, Kiyosawa K, et al. Evaluation of a new enzyme immunoassay for hepatitis C virus (HCV) core antigen with clinical sensitivity approximating that of genomic amplification of HCV RNA. Hepatology. 2000;32(2):388‒93. Non-commercial or offmarket assay 33. Tedder RS, Tuke P, Wallis N, Wright M, Nicholson L, Grant PR. Therapy-induced clearance of HCV core antigen from plasma predicts an end of treatment viral response. J Viral Hepat. 2013;20(1):65‒71. No extractable data, author unable to provide further data 34. Tillmann HL. Hepatitis C virus core antigen testing: role in diagnosis, disease monitoring and treatment. World J Gastroenterol. 2014;20(22):6701‒6. Review article 35. Veillon P, Payan C, Picchio G, Maniez-Montreuil M, Guntz P, Lunel F. Comparative evaluation of the total hepatitis C virus core antigen, branched-DNA, and amplicor monitor assays in determining viremia for patients with chronic hepatitis C during interferon plus ribavirin combination therapy. J Clin Microbiol. 2003;41(7):3212‒20. Non-commercial or off market assay 36. Vermehren J, Susser S, Berger A, Perner D, Peiffer KH, Zeuzem S, et al. Clinical utility of the architect HCV Ag assay for early treatment response monitoring in patients with chronic Hepatitis C genotype 1 infection. J Hepatol. 2012;56:S461. Does not apply to study question 37. Yang R, Rao H, Wei L. Hepatitis C virus core antigen was correlated with RNA load and had similar dynamics during treatment for chronic hepatitis C. Hepatol Int. 2011;5(1):220. Abstract or poster 38. Zanetti AR, Romano L, Brunetto M, Colombo M, Bellati G, Tackney C. Total HCV core antigen assay: a new marker of hepatitis C viremia for monitoring the progress of therapy. J Med Virol. 2003;70(1):27‒30. Non-commercial or off-market assay 39. Zhang YZ, Wu WJ, Jin H, Liu YY, Hu YW, Zhang YX. Clinical relevance of serum HCV core antigen level and antiviral therapy response. Int J Infect Dis. 2010;14:S68. Abstract or poster

Page | 447

Annex 5.9.1 PICO 7 - Dried blood spots Dried blood spots as a sample collection method for hepatitis B surface antigen serological testing A systematic review and meta-analysis

MSF Access Campaign Chauffour J, Gummadi N, Nagarantham A, Roberts T, Cohn J (Team lead) Médecins Sans Frontières, Geneva, Switzerland

August 2015

Page | 448

1. Abstract Introduction: Several expert organizations recommend screening of individuals living in high or even intermediate prevalence of hepatitis B. Despite this, few countries have or implement such screening recommendations. In high-prevalence low- and middle-income countries, there is a need for improved HBV screening, especially in decentralized settings. The use of dried blood spots (DBS) sent to centralized lab facilities may be useful in certain contexts. A systematic review and meta-analysis were performed to address the question: Among persons identified for hepatitis B testing, what is the diagnostic accuracy and impact of detecting HBsAg from DBS samples versus venous sample? Methods: Following an a priori protocol, PubMed, MEDLINE, WHO Global Index Medicus, Web of Science, MSF, Cochrane, EMBASE, CABS Abstracts and LILACS databases were searched by two reviewers in duplicate. Data were extracted with the primary outcome of HBsAg DBS test accuracy using the gold standard of a venous sample. For analysis of sensitivity and specificity, a bivariate analysis using a maximum likelihood estimate and 95% confidence intervals was used. Likelihood ratios were calculated directly from the pooled sensitivity and specificity. QUADAS-2 was used to assess bias and a GRADE evaluation was performed to evaluate the quality of included studies. Results: Two hundred forty studies were obtained for consideration, of which 10 met the criteria for inclusion in the review and 9 provided sufficient data for inclusion in the meta-analysis. The meta-analysis revealed a sensitivity of 92.9% (upper bound–lower bound 86.2–96.5) and specificity of 99.0% (upper bound–lower bound 96.2–99.7). From the pooled sensitivity and specificity, the positive likelihood ratio is 92.9 and the negative likelihood ratio is 0.072. As heterogeneity was identified on the forest plots, stratified analyses were done by storage temperature and test cut-off values. The analysis stratified by storage temperature revealed a sensitivity of 78.7% (upper bound–lower bound 70.3–85.2) and 96.1% (upper bound–lower bound 91.9–98.2) for cold chain versus ambient temperature or higher, and a specificity of 98.6% (upper bound–lower bound 68.0–100) and 99.7% (upper bound–lower bound 98.3–100) for cold chain versus ambient temperature or higher. The analysis stratified by test cut-off revealed a sensitivity of 88.0% (upper bound–lower bound 74.0–95.0) and 95.6% (upper bound–lower bound 91.2– 97.8) for standard and lowered cut-off, respectively, and a specificity of 98.6% (upper bound– lower bound 89.5–99.8) and 99.1% (upper bound–lower bound 96.6–99.8), respectively. The assessment for risk of bias revealed some risk due to patient selection and interpretation of index test. The GRADE analysis showed the included studies to provide evidence of moderate quality to assess sensitivity and specificity. Discussion: This review includes evidence of moderate quality that supports acceptable accuracy of DBS for testing HBSAg as compared to use of plasma samples and suggests that DBS may be used where there is limited access to venepuncture or inadequate technology to prepare and transport plasma samples. It is important to note that use of DBS may require changing cut-offs to

Page | 449

determine test positivity. As there are relatively little data on accuracy of DBS in real-life conditions (including high humidity), operational research will be needed in real-life conditions.

2. Introduction According to the World Health Organization’s latest data, an estimated 240 million people are chronically infected with the hepatitis B virus (HBV) and more than 780 000 people die every year due to complications of hepatitis B.1 In 20–30% of chronically infected individuals, hepatitis B progresses to liver cancer and/or cirrhosis.2 For some individuals, acute hepatitis shortly after infection can cause fatal liver failure. Hepatitis B disproportionately affects populations in lowand middle-income countries: adult chronic infection prevalence in sub-Saharan Africa and East Asia is 5–10% (similar high rates affect populations in the Amazon and south-east and southcentral Europe), and of 2–5% in the Middle East and India.3 Several organizations such as the US Centers for Disease Control and Prevention (CDC), the European Association for the Study of the Liver (EASL) and the American Association for the Study of Liver Diseases (AASLD) recommend screening of individuals living in high or even intermediate prevalence. Despite this, few countries have or implement such screening recommendations and a very large number of individuals living with HBV are not diagnosed. In high-prevalence low- and middle-income countries, there is a need for improved HBV screening, especially in decentralized settings. While commercial rapid tests exist, the use of dried blood spots (DBS) sent to centralized lab facilities may be useful in certain contexts. There are a number of HBsAg serological tests used, including rapid diagnostic tests (RDTs) and enzyme immunoassays (EIAs). Among RDTs, there are numerous options, such as the Determine HBsAg RDT. A recent meta-analysis of accuracy of hepatitis B RDTs demonstrated that the pooled sensitivity was 94.76% (95% credible interval [CrI] 90.08–98.23%) and specificity was 99.54% (95% CrI 99.03–99.95%).4 Use of DBS Hepatitis B surface antigen (HBsAg) is the screening test for HBV. Use of DBS for HBV screening may simplify sample collection and preparation (e.g. through collection of finger-prick blood samples) and improve the ability to store and transport samples for testing.5,6 Starting with the sample collection method, blood can easily be collected from pricking a finger or a heel, thus reducing the need for more highly-trained health-care workers. For DBS, less blood volume is required than in conventional venepuncture, and sample preparation is simple (it does not require electrical power, or a centrifuge), and inexpensive. Furthermore, once collected, the handling of the samples is rendered more easy: samples are less cumbersome, and can be transported in little space and at room temperature thus reducing or eliminating need for cold chain. Finally, individuals conducting the tests on the samples have a reduced risk of Page | 450

contamination once the blood has dried. This technique has already had success in diagnosing other infections such as HIV and is being developed for screening of other diseases such as hepatitis C.7 It has also been used since decades for the mass screening of congenital disorders and neonatal diseases, such as hypothyroidism and phenylketonuria (Guthrie test). In March 2015, WHO published the first guidelines for the prevention, care, and treatment of individuals with chronic HBV infection. These guidelines focused on assessment for treatment eligibility, initiation of first-line therapies, switching, and monitoring, and did not include screening recommendations. WHO is now undertaking guidelines for testing for chronic hepatitis B and C infection in low- and middle-income settings. A topic for consideration in these guidelines is the potential use of DBS for serological and molecular testing for HBV and HCV to facilitate access to and uptake of testing. In order to better evaluate the sensitivity and specificity of DBS for testing for HBsAg, the following PICOT question was developed for HBsAg. Among persons identified for hepatitis B testing, what is the diagnostic accuracy and impact of detecting HBsAg from DBS samples versus venous sample? Population: Samples for serology (HBsAg) for HBV Intervention: Using DBS samples Comparisons: Using plasma or serum from venous samples Outcomes: Diagnostic accuracy (Sensitivity, Specificity, Positive likelihood ratio, Negative likelihood ratio, TN, TP, FN, and FP)

3. Methods A protocol was prepared for the literature search, article selection, data extraction and assessment of methodological quality.

a. Search strategy and selection criteria Types of studies Case–control, cohort, and cross-sectional and randomized trials were included and articles were selected that compared DBS HBsAg testing against the gold standard of HBsAg testing using serum, and reported specificity and sensitivity or sufficient data to calculate sensitivity and specificity. Participants No date, geographical or population demographic exclusions were used. Patients of all age groups were included. Page | 451

Target conditions For use in screening for or diagnosing hepatitis B Reference standard Testing for HBsAg in serum using any commercially available test Outcome measures Sensitivity refers to the proportion of samples with true HBV infection diagnosed with a positive HBsAg test using DBS confirmed with a positive HBsAg in serum. Specificity refers to the proportion of samples with negative HBsAg using DBS and no evidence of HBV infection confirmed with a negative HBsAg in serum. Search methods We searched English language manuscripts from PubMed, MEDLINE, WHO Global Index Medicus, Web of Science, MSF, Cochrane, EMBASE, CABS Abstracts and LILACS databases using the search terms contained in Annex 1. The search was conducted between April and June 2015. Title, abstract and full-text review was done in duplicate using pre-defined eligibility criteria with a third reviewer serving as a tie-breaker for inclusion disagreements. The reference lists for articles selected for inclusion were also reviewed for additional manuscripts to review. Additional data and clarifications were sought by contacting study authors. Two articles were excluded because of inability to locate the full text of these articles.

b. Data extraction All the studies were subject to the same data extraction procedure and form based on the following parameters: author, publication and study dates, country and their World Bank economic category (high, middle/low income, or both high and middle/low income), percentage of HIV-positive participants, percentage of children and adults, age range, gender distribution, study design, participant selection, direction of study (prospective vs retrospective), type of specimen used for DBS, specimen used as gold standard (plasma or serum), test used, and whether or not the tests were administered following the recommendations from the manufacturer. Two reviewers independently extracted data. Studies without extractable sensitivity or specificity data were excluded from the meta-analysis.

Page | 452

c. Statistical data analysis Statistical analysis of the data was performed using OpenMeta [Analyst]. For analysis of sensitivity and specificity, a bivariate analysis using maximum likelihood estimate and 95% confidence intervals was used. Likelihood ratios were calculated directly from the pooled sensitivity and specificity. We used forest plots to visually assess heterogeneity. Stratified analysis was performed by studies that used the standard manufacturer-recommended cut-off value and those that used a lower cut-off, and by studies that stored DBS in the cold chain (refrigerated or frozen) and those that stored the DBS samples at ambient temperatures or higher.

d. Risk of bias and quality assessment The QUADAS-2 tool was used to assess risk of bias. A GRADE assessment performed by two reviewers in parallel to assess the quality of included studies.

4. Results a. Summary of included studies A search yielded 240 studies for consideration of which 108–17 met the criteria for inclusion in the review (Fig. 1 and Table 1) and 9 provided sufficient data for inclusion in the meta-analysis.8–11, 13, 14, 16, 17 The studies provided 370 positive samples among 1516 total samples. Four studies were from high-income countries,10, 14–16 four from middle-income countries8, 9, 12,17 and two from lowincome countries.11, 13 Five studies drew samples from broad populations, including pregnant women,8, 9, 13–15 two from inpatient populations,10, 12 two from attendees at liver clinics16, 17 and one from attendees of an HIV testing centre.11 Storage conditions including time and temperature varied among the studies, with 3 studies keeping samples in refrigerated or frozen storage,8, 9, 12 six studies with some or all samples at ambient temperature or higher10, 11, 13, 14, 16, 17 and one not specified.15 Finally, some studies lowered the test cut-off value for DBS samples while others used the manufacturer-recommended cut-off level.

b. Diagnostic performance In general, testing for HBsAg using DBS maintained good accuracy as compared with the reference test using plasma or serum. The meta-analysis revealed a sensitivity of 92.9% (upper bound–lower bound 86.2–96.5) and specificity of 99.0% (upper bound–lower bound 96.2–99.7) (Table 4 and

Page | 453

Figs 2 and 3). From the pooled sensitivity and specificity, the positive likelihood ratio is 92.9 and the negative likelihood ratio is 0.072.

c. Impact of storage conditions A range of storage conditions were evaluated in the included papers, including storage temperature ranging from –20 to 33°C and storage time ranging from overnight to 180 days. In general, storage at room temperature or higher (30–33°C) did not affect accuracy of testing as compared to storage in the cold chain. However, the two studies that evaluated storage at room temperature or higher with prolonged storage time did note a decrease in sensitivity when samples were stored for >15 days at room temperature16 or a decrease in sensitivity and specificity for samples stored at room temperature for more than 63 days.17 A stratified analysis of studies that stored samples in the cold chain (refrigerated at 2–8°C or frozen)8, 9 versus those that stored samples at ambient temperature or above (30–33°C)11, 13, 14 was performed for 5 studies (2 cold chain and 3 ambient temperature or higher). Other studies were excluded as they examined DBS samples in a range of temperature conditions, but did not provide sufficient information to determine sensitivity or specificity at a given temperature. The analysis revealed a sensitivity of 78.7% (upper bound–lower bound 70.3–85.2) and 96.1% (upper bound–lower bound 91.9–98.2) for cold chain versus ambient temperature or higher and a specificity of 98.6% (upper bound–lower bound 68.0–100) and 99.7% (upper bound–lower bound 98.3–100) for cold chain versus ambient temperature or higher (Figs 4 and 5).

d. Impact of cut-off Several papers noted that the ideal cut-off (e.g. as suggested by receiver operating characteristic [ROC] curves) for determining test positivity should be lower for DBS samples as compared to plasma or serum samples (Table 1).11–12, 15 Authors postulated this was due to the small blood volume used in DBS (commonly 50 µL for a circle of 12 mm diameter). Stratified analysis by cut-off value (standard versus raised) revealed a sensitivity of 88.0% (upper bound–lower bound 74.0– 95.0) and 95.6% (upper bound–lower bound 91.2–97.8) for standard8, 9, 13, 16 and lower cut-off,10, 11, 14, 17 respectively, and a specificity of 98.6% (upper bound–lower bound 89.5–99.8) and 99.1% (upper bound–lower bound 96.6–99.8), respectively (Figs 6 and 7).

e. Assessment of bias and quality assessment The assessment for risk of bias revealed that several studies did not use a random or consecutive sampling method or used a study population that is not consistent with the target screening Page | 454

population.8, 13, 14, 16, 17 As there are no current standards for test cut-offs for HBsAg using DBS, several studies used results of ROC curves to change the cut-off used for the DBS samples, adjusting the cut-off downward to achieve better sensitivity.10–12, 14, 15, 17 This adjustment may introduce bias and thus, these studies were judged to have a high risk of bias on the index test domain. The included studies had a low risk of bias on the reference test and flow and timing domains (Table 2). The GRADE analysis showed the included studies to provide evidence of moderate quality to assess sensitivity and specificity (Table 3).

5. Discussion Overall conclusions This systematic review and meta-analysis includes evidence of moderate quality that supports acceptable accuracy of DBS for testing HBsAg as compared to use of plasma samples, and suggests that DBS may be used for screening for HBV using HBsAg DBS where there is limited access to venepuncture or inadequate technology to prepare and transport plasma samples. Although studies are limited, DBS is likely stable and maintains good accuracy in conditions with higher temperatures and with higher humidity, although accuracy may be negatively affected when storing for prolonged durations (>14 days) at higher temperatures (room temperature and above). In a stratified analysis that examined studies that stored samples in a cold chain versus those that stored samples at ambient or higher temperatures did not reveal a detriment in accuracy for DBS stored at ambient or higher temperatures. As there are relatively little data on accuracy of DBS in real-life conditions (including high humidity), operational research will be needed in real-life conditions. It is important to note that use of DBS may require changing cut-offs to determine test positivity. As DBS uses a small sample of blood, in order to maintain sensitivity, a lower cut-off may be required as compared to when using plasma samples. In the stratified analysis performed examining standard or lower cut-off values, DBS that used a lower-cut off had a higher sensitivity than those that used the manufacturer-recommended cut-off without sacrificing significant specificity. Each individual test kit needs to be evaluated separately for its own ideal cut-off for DBS based on a ROC curve. This will usually differ from the serum test cut-off and usually be a lower cut-off to increase sensitivity. It will be important to validate use of DBS with commercially available HBsAg tests and determine cut-off values. Key limitations This review also has a number of limitations. Overall, the number of studies was small. In particular, there is a dearth of studies that systematically examine the effects of storage conditions on the accuracy of DBS and of those that did assess this, several did not specify the exact conditions (e.g. exactly what is room temperature). Thus, should DBS be adopted for screening for HBV, it will be important to pursue further operational research using field specimens in prepared and stored in real-life conditions. Ideally, to fit operational needs, these Page | 455

studies should use capillary whole blood to prepare the DBS, use a commonly available type of filter paper while comparing several commercial test kits and conditions of storage. Future work Expanded use of HBsAg testing will be critical to improving the diagnosis and treatment of HBV globally. While commercial HBsAg rapid diagnostic tests may be useful in certain contexts, the ability to use DBS for improved preparation, storage and transport of samples for decentralized testing may help to expand the reach of this important diagnostic test.

Fig. 1. Study flow diagram Potential relevant titles 378 PubMed: 53 Embase: 91 Web of Science: 154 De-duplication (total duplicates: 138) Excluded Screen 1: 194 Reason: Not relevant based on abstract Number of records in EndNote database: 240

Full papers retrieved: 46

Excluded Screen 2: 34 Reasons: Did not evaluate accuracy of DBS: 26 Did not include primary outcome: 4 Not a study: 3

Studies included for WHO analysis: 10

Page | 456

Table 1. Study-level characteristics of included studies Author Country and income category Brazil Upper–middle income Study design Study pop Sample size Storage conditions Reference test Sample for DBS Filter paper Specificity Sensitivity Cut-off value Effect of storage conditions

Boa-Sorte 2014

Crosssectional

Pregnant women

692

Temperature: refrigerated Time: less than 5 days

IMUNOSCREENHBSAG–SS (Mbiolog Diag.) and Murex HBsAg (MurexBioTechUnlmtd)

Venous

Schleicher and Schuell 903

100

Unchanged

Forbi 2010

Nigeria Lower–middle income

Crosssectional

Broad

300

Temperature: 4 °C Time: overnight

Shantest TM- HBsAg ELISA

Venous

Whatman no. 3

88.6

78.6

Unchanged

Gruner 2015

Germany High income

Crosssectional

Inpatients

299

Temperature: 20 °C, 4 °C or ambient temperature Time: up to 14 days

HBsAg assay ARCHITECHT system (Abbott Diagnostics)

Venous or capillary

PerkinElmer 99.8 226 and Whatman no. 3

91.7

0.15 IU/mL

Kania 2013

Burkina-Faso Low income

Crosssectional

Attendees of HIV testing centre

218

Temperature: ambient temperature Time: not specified

ETI-MAK-4 HBsAg EIA (DiaSorin S.p.A.)

Venous

Whatman 903

100

Optical density: MAPC (mean absorbance of positive control)/2 + 0.3 standard deviations= 0.825.

Lee 2011

Malaysia Upper middle income

Crosssectional

Patients at a tertiary hospital

150

Temperature: 20 °C Time: not specified

Abbott, kit not specified (Abbott Laboratories)

Venous

Whatman 903

97.8

96.5

Cut-off point of 1.72 Relative light units

Mendy 2005

The Gambia Low income

Cohort

Broad

166

Temperature: 30– 33 °C (humid conditions) Time: up to 4 weeks

Determine HBsAg (Abbott Laboratories)

Venous

Whatman, grade BFC 180

100

Unchanged

Page | 457

Mohamed 2013

France High income

Cohort

Broad

200

Temperature: Room temperature Time: 1, 3, 7 and 14 days

Abbott ARCHITECT HBsAg assay (Abbott Laboratories)

Venous

Whatman FTA DMPKC

100

0.30 +/-0.81 IU/mL

No significant change

Ross 2013

Germany High income

Crosssectional

Broad

299

Time and temperature not specified

Abbott ARCHITECT HBsAg (Abbott Laboratories)

Venous

Not specified

100

98.6

15 IU/mL (HBsAg)

Villa 1981

Italy High income

Crosssectional

Patients attending liver clinic

Temperature: -20 °C, Ausria II, RIA kit 4 °C and room (Abbott Laboratories) temperature. Time: 1,7, 15, 30, 60, and 180 days

Capillary

Not specified

100

Unchanged

Temperature: storage at room temperature resulted in no significant change compared to samples stored at 4 °C or -20 °C Time: storage longer than 15 days negatively affected sensitivity.

Villar 2011

Brazil Upper middle income

Cross sectional

Patients attending hepatitis clinic

133

Temperature: –20 °C, 4–8 °C, 22– 25 °C Time: 1, 7, 14, 21, 42, 63, 112, and 183 days

ETI-MAK-4 HBsAg (Diasorin)

Venous or capillary

Whatman 903

96.7

97.62

Absorbance value Accuracy of DBS 0.115 samples was stable over 63 days at all temperatures evaluated but after 63 days, accuracy diminished when stored at 22–25 °C

Page | 458

Table 2. QUADAS-2 risk of bias assessment Author Boa-Sorte Forbi Gruner Kania Lee Mendy Mohamed Ross Villa Villar Patient selection LR LR LR LR LR HR HR LR HR HR Index test LR LR HR HR HR LR HR HR LR HR Reference standard LR LR UR LR LR LR LR LR LR LR Flow and timing LR LR UR LR LR LR LR LR LR LR

HR: high risk of bias; UR: unknown risk of bias; LR: low risk of bias

Page | 459

Table 3. GRADE table Number of studies Type of study Directness Precision Consistency Risk of bias Overall quality

Sensitivity 92.9% (95% CI 86.2–96.5) 10 studies (370 HBsAg positive among 1516 samples) Specificity 99.0% (95% CI 97.6–100%) 10 studies (370 HBsAg positive among 1516 samples) Crosssectional or cohort No significant indirectness No significant imprecision No significant inconsistency Significant risk of bias (patient enrolment not consecutive or random in some studies; pre-specified cut-off not used in some studies) Moderate Cross-sectional or cohort No significant indirectness No significant imprecision Significant inconsistency (one paper reported lower sensitivity) Significant risk of bias (patient enrolment not consecutive or random in some studies; pre-specified cut-off not used in some studies) Moderate

Page | 460

Table 4. Meta-analysis: sensitivity and specificity Estimate Sensitivity Specificity 92.98 99.09 Upper bound–lower bound 6.2–96.5 6.2–99.7

Fig. 2. Forest plot sensitivity

Page | 461

Fig. 3. Forest plot specificity

Page | 462

Fig. 4. Forest plot, sensitivity at differing storage temperatures Maintained in cold chain (refrigerated or frozen)

Maintained at room temperature or higher

Page | 463

Fig. 5. Forest plot, specificity at differing storage temperatures. Maintained in cold chain (refrigerated or frozen)

Maintained at room temperature or higher

Page | 464

Fig. 6. Forest plot sensitivity, standard and lowered cut-off Standard cut-off

Lowered cut-off

Page | 465

Fig. 7. Forest plot specificity, standard and lowered cut-off Standard cut-off

Lowered cut-off

Page | 466

References 1. World Health Organization. Hepatitis B factsheet 204. (http://www.who.int/mediacentre/factsheets/fs204/en/, accessed on 2 July 2015). 2. Ganem D, Prince AM. Hepatitis B virus infection – natural history and clinical consequences. N Engl J Med. 2004;350(11):1118–29. 3. Ott JJ, Stevens GA, Groeger J, Wiersma ST. Global epidemiology of hepatitis B virus infection: new estimates of age-specific HBsAg seroprevalence and endemicity. Vaccine. 2012;30(12):2212–9. 4. Shivkumar S, Peeling R, Jafari Y, Joseph Y, Pai P. Rapid point-of-care first-line screening tests for hepatitis B infection: a meta-analysis of diagnostic accuracy (1980–2010). Am J Gastroenterol. 2012;107:1306–13. 5. Solmone M, Girardi E, Costa F, Pucillo L, Ippolito G, Capobianchi MR. Simple and reliable method for detection and genotyping of hepatitis C virus RNA in dried blood spots stored at room temperature. J Clin Microbiol. 2002;40(9): 3512. 6. McDade TW, Williams S, Snodgrass JJ. What a drop can do: dried blood spots as a minimally invasive method for integrating biomarkers into population-based research. Demography. 2007;44:899–925. 7. Sherman G, Stevens G, Jones SA, Horsfield P, Stevens WS. Dried blood spots improve access to HIV diagnosis and care for infants in low-resource settings. J AIDS. 2005;38(5):615–17. 8. Boa-Sorte N, Purificacao A, Amorim T, Assuncao L, Reis A, Galvao-Castro B. Dried blood spot testing for the antenatal screening of HTLV, HIV, syphilis, toxoplasmosis and hepatitis B and C: prevalence, accuracy and operational aspects. Braz J Infect Dis. 2014;18(6):618–24 9. Forbi JC, Obagu JO, Gyar SD, Pam CR, Pennap GR, Agwale SM. Application of dried blood spot in the sero-diagnosis of hepatitis B infection (HBV) in an HBV hyperendemic nation. Ann Afr Med. 2010;9:44–5. 10. Grüner N, Stambouli O, Ross RS. Dried blood spots – preparing and processing for use in immunoassays and in molecular techniques. J Vis Exp. 2015;97:e52619. doi:10.3791/52619 (2015). 11. Kania D, Bekale AM, Nagot N, Mondain AM, Ottomani L, Meda N, et al. Combining rapid diagnostic tests and dried blood spot assays for point-of-care testing of human immunodeficiency virus, hepatitis B and hepatitis C infections in Burkina Faso, West Africa. Clin Microbiol Infect. 2013;19:E533–E541. 12. Lee CE, Sri Ponnampalavanar S, Syed Omar SF, Mahadeva S, Ong LY, Kamarulzaman A. Evaluation of the dried blood spot (DBS) collection method as a tool for detection of HIV Ag/Ab, HBsAg, anti-HBs and anti-HCV in a Malaysian tertiary referral hospital. Ann Acad Med Singapore. 2011;40:448–53. 13. Mendy M, Kirk GD, van der Sande M, Jeng-Barry A, Lesi OA, Hainaut P, et al. Hepatitis B surface antigenaemia and alpha-foetoprotein detection from dried blood spots:

Page | 467

14.

15.

16.

17.

applications to field-based studies and to clinical care in hepatitis B virus endemic areas. J Viral Hepat. 2005;12:642–7. Mohamed S, Raimondo A, Pe´naranda G, Camus C, Ouzan D, Ravet S, et al. Dried blood spot sampling for hepatitis B virus serology and molecular testing. PLoS One. 2013;8(4): e61077. doi:10.1371/journal.pone.0061077. Ross RS, Stambouli O, Gruner N, Marcus U, Cai W, Zhang W, et al. Detection of infections with hepatitis B virus, hepatitis C virus, and human immunodeficiency virus by analyses of dried blood spots – performance characteristics of the ARCHITECT system and two commercial assays for nucleic acid amplification. Virol J. 2013;10:72. Villa E, Cartolari M, Bellentani S, Rivasi P, Casolo G, Manenti F. Hepatitis B virus markers on dried blood spots. A new tool for epidemiological research. J Clin Pathol. 1981 Jul; 34(7): 809–812. Villar LM, de Oliveira JC, Cruz HM, Yoshida CF, Lampe E, Lewis-Ximenez LL. Assessment of dried blood spot samples as a simple method for detection of hepatitis B virus markers. J Med Virol. 2011;83:1522–9.

Page | 468

Appendix 1: Search strategies Pubmed (("Hepatitis B" [Mesh] OR "Hepatitis B virus" [Mesh] OR “hepatitis B” OR “hepatitis B virus” OR “HBV”) OR (“Hepatitis B Antigens” [Mesh] OR “Hepatitis B Surface Antigens” [MESH] OR hepatitis b antigen* OR hepatitis b virus antigen* OR hepatitis b surface antigen* OR “HBsAg” OR “hbv sag” OR "hb sag" OR "HBsAg") AND (“Dried Blood Spot Testing”[Mesh] OR “dried blood spot testing” OR dried blood spot* OR dried plasma spot* OR “DBS” OR “dried blood” OR “dried plasma”)) Embase 1. ‘hepatitis b’/exp OR ‘hepatitis b antigen’/exp OR ‘hepatitis b surface antigen’/exp OR ‘hepatitis b virus’/exp 2. ‘hepatitis b’ OR ‘hepatitis b virus’ OR ‘hepatitis b surface antigen’ OR ‘hepatitis b antigen’ OR ‘hbv’ OR ‘HBsAg’ OR ‘hepatitis b sag’ OR ‘hbv sag’ OR ‘hbsag’ OR ‘hb sag’ 3. ‘dried blood spot testing’/exp OR ‘dried blood spot testing’ OR ‘dried blood spot’ OR ‘dried blood’ OR ‘dried plasma spot testing’ OR ‘dried plasma spot’ OR ‘dried plasma’ OR ‘dbs’ 4. #1 OR #2 5. #3 AND #4 Web of Knowledge (SCI-expanded, SSCI, Conference Proceedings science, BIOSIS previews) 1. TOPIC: (hepatitis b) OR TOPIC: (hepatitis b surface antigen) ORTOPIC: (hepatitis b antigen) OR TOPIC: (hepatitis b virus) OR TOPIC: (hbv) OR TOPIC: (HBsAg) OR TOPIC: (hbv sag) OR TOPIC: (hbsag) OR TOPIC: (hb sag) Indexes = SCI-EXPANDED, SSCI, A&HCI Timespan=All years 2. ((((((TOPIC: (dried blood spot testing) ORTOPIC: (dried blood spot)) ORTOPIC: (dried blood)) ORTOPIC: (dried plasma spot testing)) ORTOPIC: (dried plasma spot)) ORTOPIC: (dried plasma)) ORTOPIC: (dbs)) Indexes = SCI-EXPANDED, SSCI, A&HCI Timespan=All years 3. #2 AND #1 Indexes= SCI-EXPANDED, SSCI, A&HCI Timespan=All years Cochrane 1. 2. 3. 4. 5. MeSH descriptor: [Hepatitis B] explode all trees MeSH descriptor: [Hepatitis B Antigens] explode all trees MeSH descriptor: [Hepatitis B Surface Antigens] explode all trees MeSH descriptor: [Hepatitis B virus] explode all trees hbv or hepatitis b or hepatitis b virus or hepatitis b antigen or hepatitis b surface antigen or HBsAg or hbv sag or hbsag or hb sag 6. MeSH descriptor: [Dried Blood Spot Testing] explode all trees

Page | 469

7. dried blood spot testing or dried blood spot or dried blood or dried plasma spot testing or dried plasma spot or dried plasma or dbs 8. #1 or #2 or #3 or #4 or #5 9. #6 or #7 10. #8 and #9

Medline (OVID) 1. exp Hepatitis B/ or exp Hepatitis B virus/ or hepatitis B.mp. or hepatitis b virus.mp. or hbv.mp. 2. exp Hepatitis B Antigens/ or exp Hepatitis B Surface Antigens/ or hepatitis b antigen*.mp. or hepatitis b virus antigen*.mp. or hepatitis b surface antigen*.mp. or hbsag.mp. or hbv sag.mp. or HBsAg.mp. or hb sag.mp. 3. exp Dried Blood Spot Testing/ or dried blood spot testing.mp. or dried blood spot*.mp. or dried plasma spot*.mp. or dbs.mp. or dried blood.mp. or dried plasma.mp. 4. 1 or 2 5. 3 and 4

Page | 470

Annex 5.9.2 PICO 7 - Dried blood spots Dried blood spots as sample collection method for HCV antibody: a systematic review and meta-analysis

Lange B, Tuaillon E, Easterbrook P, van de Perre P, Ishizaki A, Denkinger C, Roberts T, Cohn J (Team lead) Médecins Sans Frontières, Geneva, Switzerland

22 September 2015

Page | 471

Executive summary

Introduction: Dried blood spots are a convenient diagnostic for viral diseases due to transport and logistical advantages over venous blood sampling. Their diagnostic accuracy for the detection of HCV antibody is not known. Methods: We conducted a systematic review and meta-analysis on the diagnostic accuracy of HCV antibody from DBS samples compared to venous samples in those persons identified for HCV testing. MEDLINE, EMBASE, Global Health and Cochrane library were searched with a sensitive search strategy and data was extracted based on a predefined extraction scheme. We described ranges of diagnostic accuracy outcomes as well as agreement of DBS against venous blood samples. For pooled analysis of sensitivity and specificity, a bivariate analysis using maximum likelihood estimate and 95% confidence interval was used and heterogeneity of results was assessed. PRISMA guideline was followed and the QUADAS tool was used to assess for risk of bias. Results: Eighteen studies of 485 abstracts were included in the qualitative review; 14 of those contributed to the quantitative analysis. Overall quality of studies was moderate. A pooled bivariate analysis of sensitivity and specificity revealed an overall sensitivity of 98% (CI95% 94–99) and an overall specificity of 99% (CI95% 97–100). Positive likelihood ratio was 171 and negative likelihood ratio 0.02. Heterogeneity was moderate with a tau2 of 0.1. In a separate univariate analysis, storage conditions did not explain heterogeneity of sensitivity or specificity, with all samples contributing to the quantitative analysis stored less than 24 hours at room temperature. However, two studies reported on false positive samples when DBS samples were stored at room temperature for longer than 3 days. Many different tests and cut-offs were used, so no stratified analysis on type of test or cut-off was performed. Discussion: This systematic review of diagnostic accuracy of HCV antibody in DBS compared to venous blood shows good diagnostic accuracy with a body of evidence of moderate quality. Manufacturers should validate their tests for the use of DBS and include instructions for this in their manuals. Future studies on diagnostic accuracy should focus on storage conditions common to field situations to be able to assess pragmatic use of this sample method.

Page | 472

2.Introduction i. Epidemiology More than 150 million people are infected with hepatitis C (HCV) worldwide but only a proportion of these are aware of their infection.1 Diagnosing HCV is done via serology by detecting antibodies to the core and non-structural antigens. Many guidelines from highincome settings recommend screening major high-risk groups including injecting drug users and persons living in high-prevalence settings.1 ii. Approach to HCV antibody testing HCV testing is traditionally done by serology using enzyme-linked immunosorbent assay (ELISA) and many validated commercially available tests exist for this, two of which are WHO prequalified. Confirmation of active infection is gained by doing a polymerase chain reaction (PCR) of HCV RNA, as 15–45% spontaneously clear the virus. A previous WHO guideline based on a systematic review performed found low-quality evidence that confirmation of chronic infections should directly follow and not be delayed.1 iii. Use of dried blood spot sampling Dried blood spots (DBS) are another way of obtaining blood samples, not requiring patients to undergo venous blood sampling if sourced from capillary blood. Storage and transportation are easier and risk of biohazard during transportation is reduced.2 That is why DBS has been used increasingly in recent years to diagnose viral diseases, including HIV and viral hepatitis.3 Disadvantages of using DBS include the fact that the commercial assays existing are not validated or regulatory approved for this method. The actual work in the laboratory is also more laborious (in terms of manual sample processing) than using serum samples.4 Some studies show the use of DBS increases uptake of hepatitis testing among several vulnerable risk groups,5-7 while others were not able to confirm this.8 The advantages of transport and storage make DBS a good choice for diagnosis of HCV in low-resource settings.3 Several programmes and studies have used DBS for HCV antibody screening without validation,9–11 and several recent studies have attempted to validate the use of DBS in diagnostic accuracy studies.12,13 Recent systematic reviews have been published on HCV RNA detection with DBS,14 on the uptake of interventions for HCV screening15 and the use of pointof-care tests in viral hepatitis testing.16 However, to our knowledge no attempt has been made to summarize the evidence on diagnostic accuracy for HCV antibody testing on DBS. iv. Systematic review as preparation for a new WHO guideline In March 2015, WHO published the first guidelines for the prevention, care and treatment of individuals with chronic HBV infection. These guidelines focused on assessment for treatment eligibility, initiation of first-line therapies, switching and monitoring. They did not include Page | 473

screening recommendations. WHO is now undertaking guidelines for testing for chronic hepatitis B and C infection in low- and middle-income settings. A topic for consideration in these guidelines is the potential use of DBS for serological and molecular testing for HBV and HCV to facilitate access to and uptake of testing. We conducted a systematic review and meta-analysis on the diagnostic accuracy of HCV antibody from DBS samples compared to venous samples in those persons identified for HCV testing. We looked at diagnostic accuracy outcomes as well as agreement of DBS against venous blood samples. In order to better evaluate the sensitivity and specificity of DBS for testing for HCV antibody, the following PICO question was developed for HCV Ab: Among persons identified for hepatitis C testing, what is the diagnostic accuracy and impact of detecting HCV Ab from DBS samples versus venous sample? Population: Samples for serology (HCV Ab) for HCV Intervention: Using DBS samples Comparisons: Using plasma or serum from venous samples Outcomes: Diagnostic accuracy (sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, TN, TP, FN and FP) and agreement (kappa, intra-class coefficients)

3. Method PRISMA guidelines were followed and QUADAS-2 was used to estimate quality of studies. i. ii. Search strategy and selection criteria

Types of studies Observational (including diagnostic accuracy studies) and interventional studies were included. We chose studies including comparisons of the index test HCV antibody in DBS against the reference test HCV antibody using serum and reported correlations, regression coefficients, specificity, sensitivity or predictive values. Only English language reports were included. Participants No date, geographical or population demographic exclusions were made. Patients of all age groups were included. Target conditions For use in screening, for diagnosing HCV.

Page | 474

Index test Testing for HCV antibody in DBS.

Reference standard Testing for HCV antibody in serum using any commercially available or in-house tests. Outcome measures Sensitivity refers to the proportion of samples with true HCV infection diagnosed with HCV antibody test using DBS confirmed with a positive HCV antibody in serum. Specificity refers to the proportion of samples with negative HCV antibody using DBS and no evidence of HCV antibody confirmed with a HCV antibody in serum. Any measures of agreement (kappa, intra-class coefficients) will also be included. iii. Search methods We searched English language manuscripts from PubMed, MEDLINE, Web of Science, EMBASE, Global Health and LILACS databases using a sensitive search strategy. The search was conducted during August–September 2015. Title, abstract and full-text review was done using predefined eligibility criteria. The reference lists for articles selected for inclusion were also reviewed for additional manuscripts to review. Additional data and clarifications were sought by contacting study authors. iv. Data extraction All the studies were subject to the same data extraction procedure by one reviewer (BL) and form based on the following parameters: author, publication and study dates, country, type of specimen used for DBS, specimen used as gold standard (plasma or serum), test used, storage conditions and effect of storage conditions and assay type. v. Statistical data analysis Statistical analysis of the data was performed using STATA 13. For analysis of sensitivity and specificity, a bivariate analysis using maximum likelihood estimate and 95% confidence interval was used. Likelihood ratios were calculated directly from the pooled sensitivity and specificity. We used forest plots to visually assess heterogeneity. If not all diagnostic values could be extracted from the study, univariate and bivariate analysis was compared. Stratified analysis was performed by type of assay and by storage conditions.

Page | 475

vi. Risk of bias and quality assessment The QUADAS-2 tool was used to assess risk of bias. A GRADE assessment was performed by two reviewers in parallel to assess the quality of included studies.

4. Results i. Search results and summary of included studies Our search yielded 485 abstracts for screening after deduplication (manually and by reference software). One hundred fifteen full texts were screened for potential inclusion and 18 studies were chosen to be included for the qualitative review.4,12,13,17–31 Fourteen of those contributed to the quantitative analysis. Of those not contributing, one did provide a ROC curve but no denominators for sensitivity and specificity,28 two did not provide any data for calculation of sensitivity or specificity30,32 and one was testing avidity in comparison to venous blood samples and not overall diagnostic accuracy.22 Of the 14 studies providing enough data to calculate sensitivity, one did not have any negative references so that no specificity could be calculated.29 (Fig. 1 and Table 1) Studies mainly stemmed from Europe, North America and Australia,33,34 two studies could be included from South America (Brazil)33,34 and three from South-East and Central Asia (India,23 Mongolia35 and Malaysia28). Studies were published from 1997 to 2014 and most used 50µL to 100µL of whole blood on filter paper to test for HCV antibody. Only one study included children;31 however, age ranges or gender for adult patients were rarely reported. ii. Diagnostic performance Diagnostic accuracy Of those studies included in the quantitative analysis, reported sensitivity of HCV-antibody in DBS ranged from 70% to 100% and specificity ranged from 95.1% to 100% (see Table 1). A pooled bivariate analysis of sensitivity and specificity revealed an overall sensitivity of 98% (CI95% 94–99) and an overall specificity of 99% (CI95% 97–100). From the pooled sensitivity and specificity, the positive likelihood ratio was 171 and the negative likelihood ratio was 0.02 (see Tables 3, 4 and 5). E(logitSE) was 3.6 and E(logitSp) was 5.1 with a covariance between estimates of E(logitSe) and E(logitSp) (tau2) of 0.11, showing moderate heterogeneity in the bivariate analysis of studies. Agreement Three of 18 included studies provided agreement measures with kappa ranging from 0.87 to 0.94 between DBS and venous blood samples.33, 34, 36

Page | 476

iii. Effect of test and cut-off used Fifteen different assays were used in HCV antibody detection. Cut-offs varied widely, and as no standardized cut-offs existed many studies devised their own cut-off via receiver operator characteristics. Nine of the included studies did report some threshold or cut-off used for DBS.12,13,20,21,25–28,34 No attempt was made to stratify by type of test or cut-off used as too many strata would have rendered results difficult to interpret. iv. Effect of storage conditions and type of test Four studies evaluated different storage conditions. In one study, three of three previously negative samples exceeded threshold values after 3 days at room temperature.27 Similarly, Tuaillon et al. showed that after 6 days of room temperature storage, threshold values were exceeded and previously negative samples would be seen as positive.37 In another study, stability was shown until 60 days at room temperature, but variation in quantitative values was less after storage at –20 °C.34 This was confirmed in another study that also tested different storage conditions and found lowest variation of results after storage at –20 °C.33 No study had left study samples at room temperature for longer than 24 hours. Therefore, in another pooled analysis, we stratified studies according to whether samples had been left at room temperature for longer than 4 hours or not. This did not change the high heterogeneity found in our meta-analysis (see Appendix). v. Assessment of study quality and risk of bias Concerning risk of bias, several studies did not report adequately on major issues; as such, rating risk of bias was difficult. Four of the included studies used case-control designs and only two reported consecutive sampling. However, the rest did provide some report on sampling. Only two studies blinded laboratory personnel to either reference or index test while performing the other one, however all studies performed index and reference tests consistently and reported on the protocol used. Overall, we graded the quality of studies to be moderate (Tables 2 and 6). b. Discussion

This systematic review and meta-analysis shows that there is evidence of moderate quality on the use of DBS for HCV antibody testing. The pooled analysis of the data available suggests high diagnostic accuracy of DBS samples for detection of HCV antibodies with good precision. The descriptive review also shows that studies looking at agreement found good agreement between DBS HCV antibody testing and testing on venous blood samples.

Page | 477

i. Impact of storage and other factors No study stored DBS samples at room temperature for longer than 24 hours. Additionally, those studies looking at variation of results after putting samples in different storage conditions found that samples could become false positive with longer exposure at ambient temperatures.13,27 In stratification on storage conditions of our pooled results, pooled sensitivity and specificity were only slightly different for samples stored longer at room temperature. However, as no study reported on storing samples for longer than a day, results are not generalizable to conditions often found in low-resource locations. ii. Key limitations This review has a number of limitations. We did not look at studies in languages other than English and no unpublished data from laboratories was included. While some studies had only a low risk of bias, overall the quality of studies was moderate at best. Another important limit of this review is its inability to suggest certain commercial tests over others to use for DBS testing of HCV antibody or to suggest a cut-off that should be used for DBS testing. As tests used were varied, no stratified analysis was done for the type of test. Additionally, to suggest a cut-off, individual-level patient data would have to have been available. iii. Future work We would therefore suggest that subsequent diagnostic accuracy studies concentrate on showing applicability of DBS under field transport and storage conditions and report on different cut-offs used with their tests. We would also urge manufacturers to validate their tests for the use of DBS, apply for regulatory approval for this sample type and include instructions for this in their manuals.

5. Conclusion In conclusion, while diagnostic accuracy of DBS for HCV antibody testing is good in those studies included in this review, uncertainty about the storage conditions needed and the cutoffs to use seriously limit its wider application in low-resource settings.

Page | 478

a. Figures and tables Fig. 1. PRISMA flowchart Records identified through database searching (n = 733)

Identification

Lilacs: 0, EMBASE: 209, MEDLINE: 109, Global Health: 75, Web of Science: 257

Screening

Records after duplicates removed (n =484) (65 duplicates manually removed)

Abstracts excluded (n =369) 41 no original paper

Abstracts screened (n = 485)

131 not on diagnosis of hepatitis B or C 97 studies on hepatitis B not C

Eligibility

22 studies on HCV but not on HCV antibody

Full-text articles assessed for eligibility (n =115)

Included

Full-text articles excluded, with reasons

Studies included in qualitative synthesis 18

23duplicates 4 no antibody 54 no diagnostic accuracy

Studies included in quantitative synthesis (meta-analysis) (n = 14)

Page | 479

Table 1. Study characteristics

Page | 480

Author

Title

Journal

Year

Country

Study pop, sample size 386 persons, 40 anti-HCV positive, 346 blood donors HCV nonreactive

Storage conditions DBS samples air dried at room temperature for 4 hours, stored at – 20°C

DBS collection method 75µL whole blood onto Whatman filter paper / alternatively 3–5 drops of capillary blood by finger-prick

Plasma antibody test MonolisaTM HCV

DBS antibody test Suggested cut-off

Specificity

Sensitivity

Correlation/agr Effect of storage eement conditions PPV and NPV calculated Kappa=0.99 (with ROC cutoff), Stability up to 60 days of storage at room temperature, but less variation at – 20°C

Brandao

Simultaneous Journal of detection of Clinical hepatitis C virus Virology antigen and antibodies in dried blood spots

2013

Brazil

ROC cut-off: 0.287 AgAb ULTRA, Bio- nm for Monolisa assay Rad (Marnes-laCoquette, France), and Murex ROC cut-off for HCV AgAb, Abbott HCV AgAb, Abbott Murex assay 0.238 (Kyalami, Republic of (Kyalami, Republic nm South Africa) of South Africa) MonolisaTM HCV AgAb ULTRA, BioRad (Marnes-laCoquette, France), and Murex Monolisa EIA confirmation test: Murex anti HCV (version 4.0), EIA Monolisa EIA confirmation test: Murex anti HCV (version 4.0), EIA NR

99.7 (98.4– 99.9)

97.5 (86.8– 99.9)

95.9 (93.3– 97.8)

97.5 (86.8– 99.9) NR NR

Croom

Commercial enzyme immunoassay adapted for the detection of antibodies to hepatitis C virus in dried blood spots

Journal of Clinical Virology

2006

Australia

103 samples from high- risk groups, negative samples from 94 individuals tested at Haematology Lab

Air dried at room temperature, storage at –20°C, plasma at –20°C, time of storage 1 week to 11 months

80 µL of each whole blood sample spotted onto Schleicher and Schuell cards (Grade 903)

100% (96–100) 100% (94– 108/108 100) 75/75

Chevaliez

Dried blood spots Conference 2014 (DBS), a abstracts promising tool for large-scale hepatitis C screening, diagnosis and treatment monitoring Comparison of Journal of hepatitis C virus Clinical RNA and antibody Virology detection in dried blood spots and plasma specimen 2014

France

529 patients, NR 183 HCV seronegative, 346 seropositive

EIA HCV assay

0.2

98.9 (96.1– 99.7)

99.1 (97.4– 99.7)

R=0.56

Dokubo

148 participants DBS airdried for 2 Fingerstick on Standard HCV TMA in a prospective hour, then sent to Whatman 903 cards (Novartis®) study of HCV another institute, 0.5 ml blood ELISA v3.0(Ortho®). then stored at – 70°C

Standard HCV TMA (Novartis®) ELISA v3.0(Ortho®).

100% (71/71)

70% (54/77)

Kappa 0.69

Page | 481

Larrat

Performance of an antigen– antibody combined assay for hepatitis C virus testing without venipuncture

Journal of Clinical Virology

2012

France

113 HCVpositive cases consecutively recruited

DSB dried 24 hours at room temperature

Finger-prick blood on Whatman card

Monolisa® HCV-AgAb-ULTRA, Bio-Rad

Oraquick HCV

0.1

100 (95.8–100) 97.4(92.5– 88/88 99.1) 110/113 100 (95.8–100) 88/88

ROC AUC OMT At 3 days room cEIABiorad: 0.99 temperature 3/3 HCV negative samples NR ROC AUC FSB cEIABiorad : 0.918

CEIA Biorad 0.2 cEIA

98.2 (93.8– 99.5) 111/113 100% 97.3%

Lee

Evaluation of the Ann Acad dried blood spot Med (DBS) collection Singapore method as a tool for detection of HIV Ag/Ab, HBsAg, anti-HBs and anti-HCV in a Malaysian tertiary referral hospital Simultaneous determination of HIV antibodies, hepatitis C antibodies and hepatitis B antigens in dried blood spots – a feasibility study using a multianalyte immunoassay

2011

Malaysia

600 samples overall, not quite clear how many used antiHCV

Left to dry 3 mL blood sample Abbott overnight at room by venous puncture temperature, then stored –20°C

Abbott

ROC cut-off 0.10 RLU

ROC curve AUC: NR 0.99 R=0.631

Lukacs

Clinical 2005 Chemistry and Laboratory Medicine

Germany

7 samples from known HCV patients

Luminex

100% 7/7

McCarron

Hepatitis C J Viral antibody Hepat detection in dried blood spots

1999

0.99 1.99

87.5% 100%

100% 97.2%

Page | 482

Marques

Dried blood spot Journal of samples: Medical optimization of Virology commercial EIAs for hepatitis C antibody detection and stability under different storage conditions

2012

Brazil

21 and 24 HCV reactive patients, 234 individual and 132 HCV negative

Serum stored at – 75 µL whole blood 20°C on Whatman paper

Two methods: HCV Ab Radim, Pomezzia, Italy and ETI-ABHCVK-4 DiaSorin, Vercelli, Italy

Two methods: HCV Ab Radim, Pomezzia, Italy and ETI-AB-HCVK-4 DiaSorin, Vercelli, Italy

Radimcut-off: 99.5% (98 – manufacturers cut- 99.9) off ROC curve for DiaSorin EIA

97.5% (86.84– 99.94)

2–8 °C, 20–25 °C, and –20°C were evaluated, –20 °C resulted in lowest variation Methods of cutoff determination: the receiver operating characteristic curve (AUROC)

98.9% (96.80– 99.55) 88.9% (75.95– 96.29)

Nandagopal

Evaluation of dried blood spot as an alternative specimen for the diagnosis of antiHCV in resourcelimited setting Detection of hepatitis C antibody with athome collection kits using an innovative laboratory algorithm

Indian 2014 Journal of Medical Microbiolog y

India

Murex

60 samples

50 µL of whole blood NR 903 Whatman card

100 (29/29)

100 (31/31)

Pearson NR correlation coefficient 0.98

O Brien

Infectious 2001 Diseases in Clinical Practice

1286 subjects enrolled in multicentre study

Air dry for 30 min, Self-collected with sent in FedEx at-home kit envelope

Hepatitis C check, Home Access Corp. self use DBS home kit

NR 100% Several 686/686 inconclusive and indeterminate results not included in diagnostic accuracy calculations

99.5% 402/404

Parker

A method for the Journal for detection and Virological confirmation of Methods antibodies to hepatitis C virus in dried blood spots

1997

80 anti HCV positive samples, 52 negative 569 DBS sample fields from South African neonates

Air dry at room Dried blood field temperature samples before storage at 4°C

In-house IgG ELISA, immunoblot RIBA 3.0

In-house IgG ELISA, T/N 5.0 immunoblot RIBA 3.0 T/N10.0

541/569 95.1% 78/80 98% 69/80 86.2%

Page | 483

Ross

Detection of Virology infections with hepatitis B virus, hepatitis C virus, and human immunodeficienc y virus by analyses of dried blood spots-performance characteristics of the ARCHITECT system and two commercial assays for nucleic acid amplification A hepatitis C Journal of avidity test for Clinical determining Virology recent and past infections in both plasma and dried blood spots

2013

Germany

339 samples

Dried overnight at 100 µL of whole room blood applied to temperature Whatman 903 filter paper

ARCHITECT system (Abbott Diagnostics, Delkenheim, Germany)

100% (97.7– 100) 160/160

97.8% (96–100) 175/179

Sheperd

2013

19 recently infected 300 chronic carrier 82 resolved infection

DBS stored at 4°C 50 µL on 903 until use Whatman Protein saver cards

ORTHO HCV 3.0 ELISA Test System with Enhanced SAVekit (Ortho Clinical Diagnostics) was used to detect antiHCV in DBS

Avidity cut-off AI<30

98.3% Arc

100%

Comparison to NR known carriers, not to venous blood samples

Journal of Tejada-Strop Disparate detection Virological outcomes for Methods anti-HCV IgG and HCV RNA in dried blood spots

2015

33 adult patients with chronic Hep C

-20°C until 5 years 75 µL of whole blood Two immunoassays, later on 12 mm DBS the VITROS anti-HCV IgG chemiluminescence assay (CIA) and the HCV 3.0 enzyme immunoassay (EIA), both from Ortho Clinical Diagnostics (Rochester, NY)

Two 3.26 CIA immunoassays, the VITROS anti-HCV 1.5 EIA IgG chemiluminescence assay (CIA) and the HCV 3.0 enzyme immunoassay (EIA), both from Ortho Clinical Diagnostics (Rochester, NY)

Not calculated CIA 48/52 92%

EIA 90% 47/52 For stored samples CIA: 100% (33/33) EIA: 32/33 97%

Page | 484

Tuaillon,

Dried blood spot Hepatology 2010 for hepatitis C virus serology and molecular testing

France

100 anti HCV positive serum samples and 100 anti HCV negative samples

18 hours dried at 50 µL of whole blood room on Whatman 12 mm temperature, paper discs stored at –20°C for 1–8 weeks

Ortho HCV 3.0 ELISA, immunoblot assay INNO-LIA HCV Score as confirmatory test

Ortho HCV 3.0 Threshold value ELISA, immunoblot 0.380 assay INNO-LIA HCV Score as confirmatory test

98% (97–100) 99% (97–99) NR

Stability of anti HCV and HCV RNA investigated by varying room temperature exposure 2–12 days until freezing, after 6 days at room temperature ODs > than cutoff values

Waterboer,

Dried blood spot Journal of samples for sero- Clinical epidemiology of Virology infections with human papilloma viruses, Helicobacter pylori, hepatitis C virus and JC virus

2011

Mongolia 1022 sexually active women from cross sectional study (response rate 69%)

Room temperature up to 8 hours, then – 20°C up to 1 month (serum + DBS)

Whole blood applied to 5 spots on DBS filter paper cards (Whatman 903)

The

HCV (strain H77, HCV (strain H77, subtype 1a) core and subtype 1a) core NS3 proteins and NS3 proteins

Sera 1492 (Core) 371 (NS3)

Not calculable Not from the data calculable from the data

98% agreement NR (kappa 0.94) for core 96.1% agreement (kappa 0.90) for NS3

DBS 967 (Core) 310 (NS3) c

Page | 485

Table 2. Risk of bias Patient selection Was a case–control design avoided? Consecutive or random sample of patients? Inappropriate exclusions? Bias Index test Blinded to reference standard Could the conduct or interpretation of the index test have introduced bias? Bias Reference standard Blinded to index? Could the reference standard have introduced bias? Bias Flow Is there an appropriate interval between the index test and reference standard? Do all patients receive the same reference standard? Are all patients recruited into the study included in the analysis? UR NR UR Bias

Brandao

No case–control design, consecutive sample, no exclusions No case–control, sampling from different cohorts NR No case–control, sampling from a prospective cohort

UR NR

Croom

UR NR

UR All patients included, same reference standard

Chevaliez Dokubo

UR NR LR NR

UR NR UR NR

UR NR UR Sampling reported, same reference standard

UR LR

Larrat

Consecutive recruitment LR

Blinded

Sampling reported, same reference standard

Lee

Consecutive recruitment LR

UR NR

UR Sampling reported, same reference standard UR Sampling reported, same reference standard UR NR UR Sampling reported, same reference standard UR NR LR Sampling partly reported, same reference standard

Lukacs

UR NR

McCarron Marques

Case–control No case–control design

HR NR LR NR

UR NR UR NR

UR LR

Nandagopal O Brien

NR No case–control design

UR NR LR Blinded

UR LR

Parker

Case–control design

HR NR

UR NR

UR Sampling partly reported, same reference standard UR NR UR NR UR NR

Ross Sheperd Tejada-Strop

No case control design No case–control design Case–control

LR LR

NR NR

UR NR UR NR UR NR

UR UR UR

HR NR

Tuaillon,

Case–control

HR NR

UR NR

UR Sampling reported, same reference standard UR NR

Waterboer,

No case–control

UR NR

Page | 486

Table 1. Sensitivities and specificities of included studies Study Estimate Sensitivity CI95% lower CI95% upper bound bound Estimate Specificity CI95% lower bound CI95% upper bound

Brandao/2013/MonolisaTMAgAb Brandao/2013/Murex Croom/2006/MonolisaTMAgAb Chevaliez/2014/EIA HCV Dokubo/2013/HCV TMA Novartis Larrat1/2012/Oraquick Larrat2/2012/cEIABiorad Lee/2011/Abott Lukacs/2005/Luminex McCarron/1999/ Marques1/2012/RadimPomezzia

0.98 0.98 1.00 0.99 0.70 0.97 0.98

0.87 0.87 0.95 0.97 0.59 0.92 0.94

1.00 1.00 1.00 1.00 0.80 0.99 1.00

1.00 0.96 1.00 0.98 1.00 1.00 1.00 –

0.98 0.93 0.97 0.95 0.85 0.96 0.96 – – – 0.98 0.94 0.88 0.99 0.93 0.93 0.98

1.00 0.98 1.00 1.00 1.00 1.00 1.00 – – – 1.00 0.98 1.00 1.00 0.97 0.97 1.00

1.00

0.59

1.00

– –

0.98

0.87 0.76 0.89 0.98 0.91 0.77 0.94

1.00 0.96 1.00 1.00 1.00 0.93 0.99

0.99 0.96 1.00 1.00 0.95 0.95 1.00

Marques2/2012/ETI-AB-HCVK-4-Diasorin 0.89 Nandagopal/2014/Murex OBrien/2001/Hepatitis C check Parker/1997/in-house Parker2/1997/in-house Ross/2013/Architect 1.00 1.00 0.98 0.86 0.98

Page | 487

Table 4. Forest plot of sensitivities and specificities

StudyId

SENSITIVITY (95% CI)

StudyId

SPECIFICITY (95% CI)

Waterboer/./

. [. - .] 0.99 [0.95 - 1.00] 0.90 [0.79 - 0.97] . [. - .] 0.98 [0.94 - 0.99] 0.86 [0.77 - 0.93] 1.00 [0.98 - 1.00] 1.00 [0.89 - 1.00] 0.89 [0.76 - 0.96] . [. - .] 1.00 [0.59 - 1.00] . [. - .] 0.98 [0.94 - 1.00] . [. - .] 0.70 [0.59 - 0.80] 0.99 [0.97 - 1.00] 1.00 [0.95 - 1.00] 0.98 [0.87 - 1.00]

Waterboer/./

. [. - .] 0.98 [0.93 - 1.00] 1.00 [0.93 - 1.00] . [. - .] 1.00 [0.98 - 1.00] 0.95 [0.93 - 0.97] 1.00 [0.99 - 1.00] 1.00 [0.88 - 1.00] 0.96 [0.94 - 0.98] . [. - .] . [. - .] . [. - .] 1.00 [0.96 - 1.00] . [. - .] 1.00 [0.95 - 1.00] 0.98 [0.95 - 1.00] 1.00 [0.97 - 1.00] 0.96 [0.93 - 0.98]

Tuaillon/2010/Ortho HCV

Tejada2/2015/HCV EIA

Sheperd/2013/OrthoHCV

Ross/2013/Architect

Parker2/1997/in house

OBrien/2001/Hepatitis C check

Nandagopal/2014/Murex

Marques2/2012/ETI-AB-HCVK-4-Diasorin

McCarron/1999/

Lukacs/2005/Luminex

Lee/2011/Abott

Larrat2/2012/cEIA Biorad

Judd/2003/

Dokubo/2013/HCV TMA Novartis

Chevaliez/2014/EIA HCV

Croom/2006/MonolisaTMAgAb

Brandao/2013/Murex

COMBINED

0.98[0.94 - 0.99] Q =277.18, df = 17.00, p = 0.00 I2 = 93.87 [91.98 - 95.75] 0.6 1.0

COMBINED

0.99[0.97 - 1.00] Q =207.39, df = 17.00, p = 0.00 I2 = 91.80 [89.05 - 94.55] 0.9 1.0

SENSITIVITY

SPECIFICITY

Table 5. Forest plot of likelihood ratios for included studies

StudyId

DLR POSITIVE (95% CI)

StudyId

DLR NEGATIVE (95% CI)

Waterboer/./ Tuaillon/2010/Ortho HCV Tejada2/2015/HCV EIA Sheperd/2013/OrthoHCV Ross/2013/Architect Parker2/1997/in house OBrien/2001/Hepatitis C check Nandagopal/2014/Murex Marques2/2012/ETI-AB-HCVK-4-Diasorin McCarron/1999/ Lukacs/2005/Luminex Lee/2011/Abott Larrat2/2012/cEIA Biorad Judd/2003/ Dokubo/2013/HCV TMA Novartis Chevaliez/2014/EIA HCV Croom/2006/MonolisaTMAgAb Brandao/2013/Murex

. [0.01 - 1000.00] 49.50 [12.55 - 195.21] 93.21 [5.90 - 1000.00] . [0.01 - 1000.00] 313.95 [19.72 - 1000.00] 17.53 [12.09 - 25.42] 1365.52 [85.50 - 1000.00] 59.06 [3.78 - 922.93] 23.24 [13.76 - 39.24] . [0.01 - 1000.00] . [0.01 - 1000.00] . [0.01 - 1000.00] 174.10 [10.97 - 1000.00] . [0.01 - 1000.00] 99.22 [6.24 - 1000.00] 60.47 [19.68 - 185.76] 216.57 [13.63 - 1000.00] 24.10 [14.39 - 40.35]

. [0.01 - 1.00] 0.01 [0.01 - 0.07] 0.10 [0.05 - 0.23] . [0.01 - 1.00] 0.03 [0.01 - 0.06] 0.14 [0.08 - 0.25] 0.01 [0.01 - 0.02] 0.02 [0.01 - 0.25] 0.12 [0.05 - 0.26] . [0.01 - 1.00] . [0.01 - 1.00] . [0.01 - 1.00] 0.02 [0.01 - 0.08] . [0.01 - 1.00] 0.30 [0.22 - 0.43] 0.01 [0.01 - 0.03] 0.01 [0.01 - 0.10] 0.03 [0.01 - 0.18]

COMBINED

171.48[36.48 - 806.04] Q =176.56, df = 17.00, p = 0.00 I2 = 86.99 [86.99 - 93.76] 0.0 1365.5

COMBINED

0.02[0.01 - 0.06] Q =254.66, df = 17.00, p = 0.00 I2 = 93.32 [91.22 - 95.43] 0 1

DLR POSITIVE

DLR NEGATIVE

Page | 488

Table 6. GRADE table Number of Type of study studies Directness Precision Consistency Risk of bias Overall quality

Sensitivity 98% (95% CI 94.0%–99.%) 14 studies (1549 HCV positive among 4304 samples) Diagnostic accuracy No significant indirectness No significant imprecision Significant inconsistency Moderate risk Moderate of bias (patient enrollment only partly consecutive or random; several case– control studies)

Specificity 99% (95% CI 97–100%) 13 studies (2756 HCV positive among 4304 samples) Diagnostic accuracy No significant indirectness No significant imprecision Significant inconsistency Moderate risk Moderate of bias (patient enrollment only partly consecutive or random; several casecontrol studies)

Page | 489

Appendix i. Sources of heterogeneity In a univariate analysis of sensitivity and specificity stratified by storage conditions sensitivity and specificity are similar among those reporting to have stored samples for less than 4 h compared to those reporting up to 24 h storage. Table A1. Sensitivities of included studies stratified on storage conditions

Study <4 hours, freeze Brandao (2013) Dokubo (2013) Croom (2006) Subtotal (I^2 = 98.18%, p = 0.00) NR Chevaliez (2014) Marques2 (2012) Lukacs (2005) Nandagopal (2014) Subtotal (I^2 = 98.18%, p = 0.00) 4-24 hours Larrat2 (2012) OBrien (2001) Parker2 (1997) Ross (2013) Tejada2 (2015) Tuaillon (2010) Subtotal (I^2 = 73.67%, p = 0.00) Heterogeneity between groups: p = 0.011 Overall (I^2 = 83.60%, p = 0.00);

ES (95% CI)

% Weight

0.98 (0.87, 1.00) 7.98 0.70 (0.59, 0.79) 2.73 (Excluded) . 0.92 (0.88, 0.97) 10.71

0.99 (0.97, 1.00) 17.08 0.89 (0.77, 0.95) 3.26 (Excluded) . (Excluded) . 0.99 (0.98, 1.00) 20.35

0.98 (0.94, 1.00) 13.65 1.00 (0.98, 1.00) 17.51 0.86 (0.77, 0.92) 4.44 0.98 (0.94, 0.99) 14.36 0.90 (0.79, 0.96) 4.06 0.99 (0.95, 1.00) 14.92 0.98 (0.96, 1.00) 68.94

0.97 (0.95, 0.99) 100.00

-.5

1.5

Page | 490

Table A2. Specificities of included studies stratified on storage conditions

Study <4 hours, freeze Brandao (2013) Croom (2006) Dokubo (2013)

ES (95% CI)

% Weight

0.96 (0.93, 0.98) 19.75 (Excluded) . (Excluded) .

NR Chevaliez (2014) Marques2 (2012) Nandagopal (2014) Subtotal (I^2 = 87.31%, p = 0.01) 4-24 hours Parker2 (1997) Tuaillon (2010) Larrat2 (2012) OBrien (2001) Ross (2013) Tejada2 (2015) Subtotal (I^2 = 87.31%, p = 0.01) Heterogeneity between groups: p = 0.000 Overall (I^2 = 49.23%, p = 0.10);

0.98 (0.95, 0.99) 0.96 (0.94, 0.98) (Excluded) 0.97 (0.96, 0.99)

22.22 20.88 . 43.10

0.95 (0.93, 0.97) 0.98 (0.93, 0.99) (Excluded) (Excluded) (Excluded) (Excluded) 0.96 (0.94, 0.97)

22.93 14.23 . . . . 37.16

0.97 (0.95, 0.98) 100.00

-.5

1.5

Page | 491

References 1. Guidelines for the screening, care and treatment of persons with hepatitis C infection. Geneva: World Health Organization; 2014. Available at: http://apps.who.int/iris/bitstream/10665/ 111747/1/9789241548755_eng.pdf (accessed on 06 June 2016) 2. 3. Hirtz C, Lehmann S. Blood sampling using "dried blood spot": a clinical biology revolution underway? Ann Biol Clin (Paris). 2015;73(1):25‒37(in French). Snijdewind IJ, van Kampen JJ, Fraaij PL, van der Ende ME, Osterhaus AD, Gruters RA. Current and future applications of dried blood spots in viral disease management. Antiviral Res. 2012;93(3):309‒21. Ross RS, Stambouli O, Gruner N, Marcus U, Cai W, Zhang W, et al. Detection of infections with hepatitis B virus, hepatitis C virus, and human immunodeficiency virus by analyses of dried blood spots – performance characteristics of the ARCHITECT system and two commercial assays for nucleic acid amplification. Virol J. 2013;10:72. Hickman M, McDonald T, Judd A, Nichols T, Hope V, Skidmore S, et al. Increasing the uptake of hepatitis C virus testing among injecting drug users in specialist drug treatment and prison settings by using dried blood spots for diagnostic testing: a cluster randomized controlled trial. J Viral Hepat. 2008;15(4):250‒4. McAllister G, Innes H, McLeod A, Dillon JF, Hayes PC, Fox R, et al. Uptake of hepatitis C specialist services and treatment following diagnosis by dried blood spot in Scotland. J Clin Virol. 2014;61(3):359‒64. Hutchinson S. Translating research into public health policy: the Scottish national hepatitis C action plan. In: 3rd International Symposium on Hepatitis Care in Substance Users, Munich, Germany, 5–6 September 2013. (var.pagings). 15 (4): 217. Craine N, Whitaker R, Perrett S, Zou L, Hickman M, Lyons M. A stepped wedge cluster randomized control trial of dried blood spot testing to improve the uptake of hepatitis C antibody testing within UK prisons. Eur J Public Health. 2015;25(2):351‒7. Bravo MJ, Vallejo F, Barrio G, Brugal MT, Molist G, Pulido J, et al. HCV seroconversion among neverinjecting heroin users at baseline: no predictors identified other than starting injection. Int J Drug Policy. 2012;23(5):415‒9.

10. Allen EJ, Palmateer NE, Hutchinson SJ, Cameron S, Goldberg DJ, Taylor A. Association between harm reduction intervention uptake and recent hepatitis C infection among people who inject drugs attending sites that provide sterile injecting equipment in Scotland. Int J Drug Policy. 2012;23(5):346‒52. 11. McLeod A, Weir A, Aitken C, Gunson R, Templeton K, Molyneaux P, et al. Rise in testing and diagnosis associated with Scotland's Action Plan on Hepatitis C and introduction of dried blood spot testing. J Epidemiol Community Health. 2014;68(12):1182‒8. 12. Brandao CP, Marques BL, Marques VA, Villela-Nogueira CA, Do OK, de Paula MT, et al. Simultaneous detection of hepatitis C virus antigen and antibodies in dried blood spots. J Clin Virol. 2013;57(2):98‒102. 13. Tuaillon E, Mondain AM, Meroueh F, Ottomani L, Picot MC, Nagot N, et al. Dried blood spot for hepatitis C virus serology and molecular testing. Hepatology. 2010;51(3):752‒8. 14. Greenman J, Roberts T, Cohn J, Messac L. Dried blood spot in the genotyping, quantification and storage of HCV RNA: a systematic literature review. J Viral Hepat. 2015;22(4):353‒61.

Page | 492

15. Jones L, Bates G, McCoy E, Beynon C, McVeigh J, Bellis MA. Effectiveness of interventions to increase hepatitis C testing uptake among high-risk groups: a systematic review. Eur J Public Health. 2014;24(5): 781‒8. 16. Khuroo MS, Khuroo NS, Khuroo MS. Diagnostic accuracy of point-of-care tests for hepatitis C virus infection: a systematic review and meta-analysis. PLoS One. 2015;10(3): e0121450. 17. O’Brien J, Kruzel K, Wandell M, Vinogradov I, Sheagren J, Frank A. Detection of hepatitis C antibody with at-home collection kits using an innovative laboratory algorithm. Infect Dis Clin Pract (Baltim Md). 2001;10(9):474‒80. 18. Judd A, Parry J, Hickman M, McDonald T, Jordan L, Lewis K, et al. Evaluation of a modified commercial assay in detecting antibody to hepatitis C virus in oral fluids and dried blood spots. J Med Virol. 2003;71(1):49‒55. 19. Croom HA, Richards KM, Best SJ, Francis BH, Johnson EI, Dax EM, et al. Commercial enzyme immunoassay adapted for the detection of antibodies to hepatitis C virus in dried blood spots. J Clin Virol. 2006;36(1):68‒71. 20. Marques BL, Brandao CU, Silva EF, Marques VA, Villela-Nogueira CA, Do OK, et al. Dried blood spot samples: optimization of commercial EIAs for hepatitis C antibody detection and stability under different storage conditions. J Med Virol. 2012;84(10):1600‒7. 21. Waterboer T, Dondog B, Michael KM, Michel A, Schmitt M, Vaccarella S, et al. Dried blood spot samples for seroepidemiology of infections with human papillomaviruses, Helicobacter pylori, hepatitis C virus, and JC virus. Cancer Epidemiol Biomarkers Prev. 2012;21(2):287‒93. 22. Shepherd SJ, Kean J, Hutchinson SJ, Cameron SO, Goldberg DJ, Carman WF, et al. A hepatitis C avidity test for determining recent and past infections in both plasma and dried blood spots. J Clin Virol. 2013;57(1):29‒35. 23. Nandagopal P, Iqbal HS, Saravanan S, Solomon SS, Mehta S, Selvakumar M, et al. Evaluation of dried blood spot as an alternative specimen for the diagnosis of anti-HCV in resource-limited settings. Indian J Med Microbiol. 2014;32(2):208‒10. 24. Dokubo EK, Evans J, Winkelman V, Cyrus S, Tobler LH, Asher A, et al. Comparison of hepatitis C virus RNA and antibody detection in dried blood spots and plasma specimens. J Clin Virol. 2014;59(4):223‒7. 25. Tejada-Strop A, Drobeniuc J, Mixson-Hayden T, Forbi JC, Le NT, Li L, et al. Disparate detection outcomes for anti-HCV IgG and HCV RNA in dried blood spots. J Virol Methods. 2015;212:66‒70. 26. Chevaliez S, Soulier A, Poiteau L, Pawlotsky JM. Dried blood spots (DBS), a promising tool for largescale hepatitis c screening, diagnosis and treatment monitoring. In: 49th Annual Meeting of the European Association for the Study of the Liver, International Liver Congress 2014 London, United Kingdom, 9–13 April 2014. [Abstract P765]. J Hepatol. 2014;60 (1 Suppl. 1): S325–S326. 27. Larrat S, Bourdon C, Baccard M, Garnaud C, Mathieu S, Quesada JL, et al. Performance of an antigen-antibody combined assay for hepatitis C virus testing without venipuncture. J Clin Virol. 2012;55(3):220‒5. 28. Lee CE, Sri Ponnampalavanar S, Syed Omar SF, Mahadeva S, Ong LY, Kamarulzaman A. Evaluation of the dried blood spot (DBS) collection method as a tool for detection of HIV Ag/Ab, HBsAg, anti-HBs and anti-HCV in a Malaysian tertiary referral hospital. Ann Acad Med Singapore. 2011;40(10):448‒ 53.

Page | 493

29. Lukacs Z, Dietrich A, Ganschow R, Kohlschutter A, Kruithof R. Simultaneous determination of HIV antibodies, hepatitis C antibodies, and hepatitis B antigens in dried blood spots – a feasibility study using a multi-analyte immunoassay. Clin Chem Lab Med. 2005;43(2):141‒5. 30. McCarron B, Fox R, Wilson K, Cameron S, McMenamin J, McGregor G, et al. Hepatitis C antibody detection in dried blood spots. J Viral Hepat. 1999;6(6):453‒6. 31. Parker SP, Cubitt WD, Ades AE. A method for the detection and confirmation of antibodies to hepatitis C virus in dried blood spots. J Virol Methods. 1997;68(2):199‒205. 32. Judd A, Parry J, Hickman M, McDonald T, Jordan L, Lewis K, et al. Evaluation of a modified commercial assay in detecting antibody to hepatitis C virus in oral fluids and dried blood spots. J Med Virol. 2003;71(1):49‒55. 33. Marques BLC, Brandao CU, Silva EF, Marques VA, Villela-Nogueira CA, Do O KMR, et al. Dried blood spot samples: optimization of commercial EIAs for hepatitis C antibody detection and stability under different storage conditions. J Med Virol. 2012;84(10):1600‒7. 34. Brandao CP, Marques BL, Marques VA, Villela-Nogueira CA, Do OKM, de Paula MT, et al. Simultaneous detection of hepatitis C virus antigen and antibodies in dried blood spots. J Clin Virol. 2013;57(2):98‒102. 35. Waterboer T, Dondog B, Michael KM, Michel A, Schmitt M, Vaccarella S, et al. Dried blood spot samples for seroepidemiology of infections with human papillomaviruses, helicobacter pylori, hepatitis C virus, and JC virus. Cancer Epidemiol Biomarkers Prev. 2012;21(2):287‒93. 36. Waterboer T, Dondog B, Michael KM, Michel A, Schmitt M, Vaccarella S, et al. Dried blood spot samples for seroepidemiology of infections with human papillomaviruses, helicobacter pylori, Hepatitis C Virus, and JC Virus. Cancer Epidemiol Biomarkers Prev. 2012;21(2):287‒93. 37. Tuaillon E, Mondain AM, Meroueh F, Ottomani L, Van De Perre P, Ducos J. Dried blood spot for hepatitis C virus molecular diagnosis: an alternative tool for hard-to-reach population. In: 45th Annual Meeting of the European Association for the Study of the Liver (EASL), International Liver Congress 2010 Vienna, Austria, 14–18 April 2010. J Hepatol. 2010;52:S408‒S409.

Page | 494

Annex 5.9.3 PICO 7 - Dried blood spots Dried blood spots as sample collection method for HBV DNA A systematic review and meta-analysis

Lange B, Tuaillon E, Eastbrook P, van de Perre P, Ishizaki A, Denkinger C, Roberts T, Cohen J (Team lead) Médecins Sans Frontières, Geneva, Switzerland

September 2015

Page | 495

1. Executive summary Introduction: Dried blood spots are a convenient diagnostic for viral diseases due to transport and logistical advantages over venous blood sampling. The diagnostic accuracy for the detection of hepatitis B (HBV DNA) on DBS samples is not known. Methods: We conducted a systematic review and meta-analysis on the diagnostic accuracy of HBV DNA from DBS samples compared to venous samples in those persons identified for HBV DNA testing. MEDLINE, EMBASE, Global Health, Web of Science and Cochrane library were searched with a sensitive search strategy and data were extracted following a predefined extraction scheme. We described ranges of diagnostic accuracy outcomes as well as correlation and regression coefficients of DBS against venous blood samples reported by included studies. For pooled analysis of sensitivity and specificity, a bivariate analysis using maximum likelihood estimate and 95% confidence intervals were used and heterogeneity of results was assessed. PRISMA guideline was followed and the QUADAS tool was used to assess for risk of bias. Results: Ten studies of 485 abstracts were included in the qualitative review, 8 of those contributed to the quantitative analysis. Overall quality of studies was low. A pooled bivariate analysis revealed an overall sensitivity of 96% (CI 95% 91–98) and an overall specificity of 100% (CI 95% 55–100). Positive likelihood ratio was 307 and the negative likelihood ratio was 0.04. Heterogeneity was moderate with an I2 of 18% (95% CI 0–100) and a tau2 of 0.018. One study reporting on thresholds reported a sensitivity of 98% for a cut-off of 2000 IU/mL, another reported a limit of detection of a commercial assay for DBS of 914 IU/ml. No study reported on storage conditions >24 hours at room temperature for DBS, but several studies varying storage conditions for individual samples did not find high variation of results. Discussion: This systematic review and meta-analysis show that data are scarce and of suboptimal quality on the use of DBS for HBV DNA testing. The pooled meta-analysis of data available suggests that sensitivity compared to serum is good while specificity estimate is adequate with high imprecision. Individual study reports seem to suggest that sensitivity of HBV DNA detection above 2000 IU/mL is good. Bigger and better-performed diagnostic accuracy studies reporting diagnostic accuracy at different thresholds are needed.

2. Introduction i. Epidemiology A large number of people are infected with hepatitis B (HBV) and nearly a third of those will develop liver cancer and cirrhosis. Prevalence of HBV is highest in sub-Saharan Africa, East Asia, in the Middle East and India.1 Few patients are aware of their infection until complications are present.

Page | 496

ii. Approach to HBV DNA testing While for a diagnosis of HBV, serology is often sufficient for follow up and for decisions on treatment, quantitative testing of HBV DNA is important.2 Most HBV DNA testing currently occurs on platforms in reference laboratories. Several guidelines recommend using a threshold of 2000 IU/mL for decisions on treatment in patients with chronic hepatitis B.2 iii. Use of DBS Dried blood spots (DBS) can facilitate sample transport and simplify logistics, as has been shown for HIV.3 They do not require venous blood to be taken, storage and transportation are less difficult and biohazard is reduced.4 Increasingly DBS has therefore been used in the management of viral diseases, including HIV and hepatitis.5 However, currently DBS protocols for hepatitis are not standardized for many applications and commercial tests and DBS is not yet regulatorily approved as a sample type.6 Several studies have shown potential of DBS to increase uptake of hepatitis testing among several vulnerable risk groups,7-9 while others have not been able to confirm this trend.10 Even if uptake is not affected, however, procedural advantages make DBS a good choice for diagnosis and follow up of patients with HBV in low-resource regions.5 Which is why DBS for serology has often been used in epidemiological studies, sometimes without validating it against serum.11,12 For HBV DNA testing, first efforts to test on DBS stem from more than 20 years ago,13 but only recently other studies followed.14,15 While several systematic reviews have been published on HBV diagnostics and on the use of DBS, no systematic review on HBV DNA and DBS exists to our knowledge. Recent reviews on DBS in HCV RNA16 and in tropical diseases17 did not include the diagnostic accuracy of HBV DNA in DBS. Recently, WHO published the first guidelines for the prevention, care, and treatment of individuals with chronic HBV infection.18 These guidelines focused on assessment for treatment eligibility, initiation of first-line therapies, switching, and monitoring and did not include screening recommendations. WHO is now preparing guidelines for testing for chronic hepatitis B and C infection in low- and middle-income settings. A topic for consideration in these guidelines is the potential use of DBS for serological and molecular testing for HBV and HCV to facilitate access to and uptake of testing. As preparation for this guideline on hepatitis B and C diagnosis, we present this systematic review on the diagnostic accuracy of HBV DNA DBS in comparison to HBV DNA in serum. In order to better evaluate the sensitivity and specificity of DBS for testing for HBV DNA the following PICO question was developed for HBV DNA. Among persons identified for HBV DNA testing, what is the diagnostic accuracy and impact of detecting HBV DNA from DBS samples versus venous sample? Population: Samples for HBV DNA detection Intervention: Using DBS samples Page | 497

Comparisons: Using plasma or serum from venous samples Outcomes: Diagnostic accuracy (Sensitivity, Specificity, Positive likelihood ratio, Negative likelihood ratio, TN, TP, FN, and FP), correlation and regression coefficients

3. Methods PRISMA guidelines were followed and QUADAS-2 was used to estimate quality of studies. i. Search strategy and selection criteria Types of studies Observational (including diagnostic accuracy studies) and interventional studies were included. We chose studies including comparisons of the index test HBV DNA in DBS against the reference test HBV DNA using serum and reported correlations, regression coefficients, specificity, sensitivity or predictive values. Only English language reports were included. Participants No date, geographical or population demographic exclusions were used. Patients of all age groups were included. Target conditions Hepatitis B diagnosis, patient follow up and treatment monitoring Index test Testing for HBV DNA in DBS for diagnosing HBV and for follow up of HBV patients Reference standard Testing for HBV DNA in serum using any commercially available or in house test Outcome measures Sensitivity refers to the proportion of samples with true HBV infection diagnosed with HBV DNA test using DBS confirmed with a positive HBV DNA in serum. Specificity refers to the proportion of samples with negative HBV DNA using DBS and no evidence of HBV infection confirmed with a HBV DNA in serum. Correlation refers to any bivariate quantitative correlation parameter between HBV DNA quantity amplified in DBS compared to serum. Regression coefficients refer to any linear regression coefficients describing the association between HBV DNA quantity in DBS compared to serum.

Page | 498

ii. Search methods We searched English language manuscripts from MEDLINE, Web of Science, EMBASE, Global Health and LILACS databases. The search was conducted in August 2015. Title, abstract and full-text review was done using predefined eligibility criteria. The reference lists for articles selected for inclusion were also reviewed for additional manuscripts to review. Additional data and clarifications were sought by contacting study authors. iii. Data extraction All the studies were subject to the same data extraction procedure and form based on the following parameters: author, publication and study dates, country, percentage of children and adults, age range, gender distribution, type of specimen used for DBS, specimen used as gold standard (plasma or serum), test used, storage conditions and effect of storage conditions. Data extraction was performed by one reviewer (BL). iv. Statistical data analysis Statistical analysis of the data was performed using STATA 13. For analysis of sensitivity and specificity, a bivariate analysis using maximum likelihood estimate and 95% confidence intervals was used if >4 studies contributed to the analysis. Likelihood ratios were calculated directly from the pooled sensitivity and specificity. We used forest plots to visually assess heterogeneity. If not all diagnostic values could be extracted from the study, univariate and bivariate analysis was compared. Stratified analysis was performed by studies that used the In house or commercially available tests and by storage conditions. v. Risk of bias and quality assessment The QUADAS-2 tool was used to assess risk of bias. A GRADE assessment was performed by two reviewers in parallel to assess the quality of included studies.

4. Results i. Search results and summary of included studies Our search yielded 485 abstracts for screening after de-duplication (manually and by reference software). Forty-eight full-text papers were screened for potential inclusion and 10 studies were chosen to be included for the qualitative review.13-15,19-25 Of the 10 studies, one of these studies was a conference abstract22 that was later published as a paper15 so this was not included in the quantitative analysis. Another study did not represent the data in analysable form24 (Fig. 1 and Table 1). So eight studies remained for analysis, out of which only 4 contributed to both sensitivity and specificity estimates. Of those studies included, three stemmed from Europe (France22, Germany15, Spain23) while three were from Africa (Congo26, Egypt21 and Zambia25) two from Asia (India,13 China24) Page | 499

and one from Mexico.14 No data from children were included, one study only included data on women13 and one study only included HIV-positive patients.25 Overall studies underreported on demographic characteristics. All studies used whole blood apart from one study that used plasma26 for preparation of DBS samples. ii. Diagnostic performance Diagnostic accuracy In the 4 studies that contributed both to sensitivity and specificity, sensitivity of HBV DNA in DBS ranged from 93% to 100% and specificity ranged from 86% to 100% (see Table 1). A pooled bivariate analysis revealed an overall sensitivity of 96% (CI95% 91–98) and an overall specificity of 100% (CI95% 55–100). From the pooled sensitivity and specificity, the positive likelihood ratio was 307 and the negative likelihood ratio was 0.04. Heterogeneity was low with a covariance between estimates of E(logitSe) and E(logitSp) (tau2) of 0.018 (E(logitSE) 3.3 and E(logitSP) 5.8) ( Tables 3, 4 and 5). In a univariate analysis of the 7 studies contributing to sensitivity, pooled sensitivity was 97% (CI95% 94–99). I2 in this analysis was 54% with a tau2 of 0.0006 (see Appendix). Association, correlation and agreement Five of the 10 included studies reported regression coefficients showing a high association between quantitative results of HBV DNA in DBS and in serum14,15,20,21,23 (regression coefficients between 0.61 and 0.96). In terms of correlation one study reported good Pearson correlation (0.93)14 and one study reported a good Spearman correlation coefficient of 0.84.27 One study reported an agreement of kappa >0.7 between binary results of HBV DNA on DBS and on serum. iii. Limit of detection and thresholds of HBV DNA of assays No guidelines for using HBV DNA assays on DBS exist and so several studies performed testing to understand the limit of detection of this method. Limit of detections of used assays (commercial and in-house) for a serum sample ranged from 10 to 100 IU/mL (see Table 1). On DBS the limit of detection in one study was 914 IU/mL for an often-used commercial assay (COBAS Taqman), with a plasma limit of detection of 20 IU/mL.15 Furthermore three studies reported that quantitation of HBV DNA in DBS below 3000–4000 IU/mL was difficult.6,23,28 A recent study from Zambia used diluted samples and reported that 13.8% (CI95% 7.7–23.7) of those detected in plasma were missed in DBS with a cut-off of 200 IU/mL, 1.8% (95% CI: 0.5– 6.6) with a cut-off of 2000 IU/mL and 0.2% (95% CI: 0.03–1.7) with a cut-off of 20 000 IU/mL (see Table 1). iv. Effect of storage conditions, type of test and DBS/DPS Two studies evaluating different storage conditions ranging from 4°C to 37°C for up to 7 days did not find a decline in diagnostic accuracy.14,23 For the diagnostic accuracy studies, all studies Page | 500

stored DBS samples at –20 °C, so we did not attempt a stratified meta-analysis on storage conditions. We stratified studies according to whether an in-house polymerase chain reaction (PCR) assay or a commercial assay was used in a univariate analyses for sensitivity and specificity. Pooled sensitivities were similar in both groups (95% and 98%) (see Table A3 and A4 in Appendix). Only one study used plasma instead of whole blood for preparation of dried samples, so no stratified analysis was undertaken to investigate heterogeneity. This study showed similar sensitivity and specificity to the other studies and the pooled estimate.21 v. Assessment of study quality and risk of bias Concerning risk of bias, most studies did not report adequately on major issues, so that rating risk of bias was limited. None of the studies blinded laboratory personnel to either the reference or the index test when performing the other one, or if so, this was not reported. However, as all studies used and reported a clear and consistent protocol for both reference and index test with an output that does not allow for interpretation we did not see a major cause of bias in this. We downgraded 4 of the included studies because they used a case– control design, while the other 5 did not adequately report on their sampling or the flow of participants. This leads to a general high risk of bias in these diagnostic accuracy studies and an overall low quality of evidence (Table 2 and Table 6).

5. Discussion This systematic review and meta-analysis shows that data are scarce and of suboptimal quality on the use of DBS for HBV DNA testing. The pooled meta-analysis of data available suggests that sensitivity compared to serum is good while the specificity estimate is adequate with high imprecision, and that different storage conditions do not effect DBS unduly. The descriptive review also shows a good correlation and high association between quantitative values for HBV DNA on DBS and in serum samples. i. Impact of storage and other factors Some studies did test samples at varying storage conditions and found no effect on the results of these tests. However, only one study stored dried plasma samples (also the only study using dried plasma samples and not DBS)28 at room temperature, while in all other studies those samples used in diagnostic accuracy calculations were stored at room temperature no longer than 24 hours and stored frozen afterwards. This means that all pooled results we present can only be considered valid under these conditions, which might limit applicability in field conditions tremendously.

Page | 501

ii. Key limitations This review has a number of limitations. Overall the number of studies was small. Only the English literature was looked at and no unpublished studies were examined. The few studies included were of limited quality and had small sample sizes. This review is not able to answer whether the sensitivity and specificity of certain HBV DNA thresholds – for example, 2000 IU/mL, the threshold below which treatment for HBV is not recommended – are high enough, as this could not be calculated due to not having individual sample data. This would be important, however, as guidelines suggest treatment not be started in certain cases below this threshold because it is considered inactive, immunecontrolled chronic hepatitis B. Therefore, any field test might not necessarily have to be able to detect HBV DNA below 2000 IU/mL.2 In our meta-analysis, sensitivity was high in general, with narrow confidence intervals. However, no conclusions regarding specificity were possible because the pooled meta-analysis of available data yielded very wide confidence intervals. Furthermore, no pooled receiver operator characteristics (ROC) to establish good overall cut-offs for the data could be performed. iii. Future works Bigger and better-performed diagnostic accuracy studies that avoid case–control designs, test HBV DBS under real-life storage and transport conditions and report diagnostic accuracy for different time, temperature and clinically relevant LOD thresholds are therefore called for. We would also advocate to perform a meta-analysis of individual-level patient (sample) data of published and unpublished studies on HBV DNA in DBS, as that approach might yield more data and would allow for quantitative analyses.

Page | 502

Prisma flowchart

Records identified through database searching (n = 734) Identification

Lilacs: 0, EMBASE: 209, MEDLINE: 109, Global Health: 75, Web of Science: 257

Records after duplicates removed (n = 484) (65 duplicates manually) removed) Screening

Abstracts excluded (n = 437 ) 41 no original paper

Abstracts screened (n = 485)

131 not on diagnosis of hepatitis B or C 12 on HBV and HCV but no diagnostic accuracy

Eligibility

58 studies on HBV using DBS, but not on HBV DNA Full-text articles assessed for eligibility (n = 48) 15 HBV studies, but no DBS Full-text articles excluded 22 studies on HCV but not on HCV antibody (n = 38) Hepatitis B: on HCV not on 158 abstracts HBV 7 further duplicates 2 no dried blood spots 15 No diagnostic accuracy Studies included in qualitative synthesis (n = 10) 4 No HBV DNA 4 Other 4 No original paper 2 HCV only

Included

Studies included in quantitative synthesis (metaanalysis) (n = 8) (for sensitivity, 4 for bivariate analysis and specificity) Fig. 2. Prisma Flowchart

Page | 503

Table 2. Characteristics of included studies Author Title Journal Year Country Study pop and sample size Storage conditions DBS collection Plasma method method PCR DBS method PCR Limit of detection Specificity Sensitivity Correlation/Association Effect of storage conditions Comments

Alidjinou

Detection of hepatitis B Diagnostic virus infection markers Microbiology and in dried plasma spots Infectious Disease among patients in Congo-Brazzaville

2014

Congo-Brazzaville 32 HBV patients

DBS at room 30 µL of temperature, plasma onto Frozen plasma filter paper samples at – 80 °C

COBAS Taqman/COBAS AmpliPrep

COBAS Detection limit for Taqman/COBA plasma was 12 S AmpliPrep IU/mL. In 3 patients, viral load in plasma was 152, 250, and 1727 IU/mL, respectively, whereas HBV DNA could not be quantified in DPS, but was detected.

96% (25/26)

Spearman correlation coefficient r=0.84

Alhusseini

Hepatitis B virus DNA can be amplified directly from dried blood spot on filter paper Direct detection of hepatitis B virus from dried blood spots by polymerase chain reaction amplification

American Journal 2012 of Biochemistry and Biotechnology

Egypt

50 HBs Ag pos Stored at – patients, 10 80 °C negative controls

Watman 903, 50 µL from venous blood sample

In house

No cut-off suggested

100%

Good correlation (r=0.88) between DBS and plasma viral load

Gupta

Journal of Clinical Microbiology

1992

India

Submitted for –20 °C for routine filter paper serological testing of HBs pos mothers 60 mothers with chronic HBV infection, 5 laboratory personell 100 Hbs Ag neg, 100 Hbs Ag pos with HBV DNA –20°C

In house (end- In house (endpoint) point)

Whatman paper

Limit of detection 86% (13/15) 10E4 virus particles in each 5-,u blood spot,

96% (43/45)

No diagnostic accuracy calculation but can be calculated from data

Halfon

Dried blood spot for Conference Hepatitis B virus abstract serology and molecular testing Usefulness of dried blood samples for quantification and molecular characterization of HBV-DNA Hepatology

2012

France

HBV Cobas 3 blood whole blood Taqman on paper card

HBV Cobas Taqman

1400 IU/mL 2000 IU/mL

100

98 (95–100) 91 (85–97)

100 2004 Spain 82 patients with chronic HBV infection (23 HBeAg pos, 39 HBeAg neg, 20 HBeAg inactive, 15 HBe neg under Room temperature for up to two hours, then – 20 °C 20 µL of In house capillary blood on 5mm paper disks (Scheicher) In house LOD 100 cop/mL among eight samples with serum HBV DNA between 103 and 104 copies/mL, seven Regression coefficient HBV DNA concentration in DBS versus serum samples r(2) = 0.96 (P<.001). Stability of DBS samples assessed by leaving samples for several days in differed

Conference abstract so no more data, same study as Mohamed below

Jardi

Page | 504

lamivudine therapy

tested positive using DBS samples, and among four samples with detectable serum HBV-DNA levels _103 copies/mL, none were positive using DBS samples.

conditions, no effect on HBV DNA levels

Lira R

Use of dried blood Virology Journal samples for monitoring hepatitis B virus infection

2009

Mexico

47 HBV Monoinfected patient

Plasma samples at – 70 °C, filter paper at – 20 °C

50 µL per card QIAamp® (Schleicher + Ultrasens® Virus kit (QIAGEN Schull) GMBH, Germany),

100%

The Pearson correlation 0.93 (p = 0.01)

No adverse effect by sample storage s from ten patients were stored at 4 °C, 25 °C, and 37 °C for 7 days

PLOS One Mohamed S Dried blood spot sampling for hepatitis B virus serology and molecular testing

2013

France

50 HBV-positive Dried for 18 h 15 µL on 12 patients, 10 HBV- in room mm discs negative patients temperature (Whatman)

Cobas AmpliPrep/Cobas Taqman HBV test,

Cobas AmpliPrep/Co bas Taqman HBV test,

Limit of detection of HBV DNA 20 IU/mL plasma, limit of detection DBS 914 IU/mL

50/50 100%

Correlation good between DBS and plasma (r2=0.86, P<0.001),

Ross

Detection of infections Virology with hepatitis B virus, hepatitis C virus, and human immunodeficiency virus by analyses of dried blood spots – performance characteristics of the ARCHITECT system and two commercial assays for nucleic acid Hepatitis B viral load in dried blood spots: a validation study in Zambia Clinical Journal of Virology

2013

Germany

299 samples

100 µL applied Artus HBV LC PCR to filter paper (Qiagen, Hilden, (Whatman, Germany) Schleicher+Sch üll)

Artus HBV LC PCR (Qiagen, Hilden, Germany)

Limit of detection 100 (96–100) 93 (92.9–93.1) 100 IU/mL in plasma, 7 samples with HBV concentrations of 409–3643 IU/mL missed

Vinikoor

2015

Zambia

68 HBs pos patients,

Dried for 12 50 µL applied COBAS hours at room to filter paper temperature

The probability of a undetectable DBS result at a plasma viral load of 200 IU/mL was 13.8% (95% CI:

91.2 62/68

Page | 505

7.7–23.7) but this dropped to 1.8% (95% CI: 0.5–6.6) when a 2000 IU/mL threshold was used and 0.2% (95% CI: 0.03–1.7) at 20,000 IU/ml. Zhang Detection of HBV-DNA in dried bloodstains on filter paper by nested polymerase chain reaction Laboratory Medicine 2010 China Hospital DBS samples patients;60 blood at –20 °C, samples whole blood samples at – 80 °C 10–20 µL on Whatman In house (endpoint) In house (end- All samples of 5 point) copies of HBV DNA per ml detected Kappa >0.7 for agreement between nested PCR and ELISA, 61% positive with whole blood sample, but only 51% with filter paper No diagnostic accuracy or agreement calculations for whole blood against filter paper

Page | 506

Table 3. Risk of bias of included studies Author Alidjinou Alhusseini Gupta Halfon Jardi R Lira R Mohamed S Ross Patient selection Was a case–control design avoided Consecutive or random sample of patients Inappropriate exclusions UR HR HR UR HR HR HR HR Bias Index test Blinded to reference standard Could the conduct or interpretation of the index test have introduced bias? Reference standard Blinded to index? Could the reference standard have introduced bias? Flow and timing Patient flow?

NR, but no case–control design Case–control design, sampling NR Case–control design, sampling NR NR Selection only of cases Selection of only cases Case–control design No case–control design, sampling NR

Not blinded, interpretation unbiased Not blinded, interpretation unbiased Not blinded, interpretation unbiased Not blinded, NR Not blinded, interpretation unbiased Not blinded, interpretation unbiased Not blinded, interpretation unbiased Not blinded, interpretation unbiased

LR LR LR UR LR LR LR LR

NR NR Partly reported NR Partly reported NR NR Flow reported

UR UR LR UR LR UR UR LR

Vinikoor

No case–control design, only cases

Not blinded, interpretation unbiased

Flow reported

Zhang

No case–control design, sampling NR

Not blinded, interpretation unbiased

Partly reported

HR: high risk of bias; LR: low risk of bias; UR: unclear risk of bias; NR: not reported

Page | 507

Table 4. Forest plot of sensitivities and specificities of included studies.

Page | 508

Table 5. Sensitivities and specificities of included studies Study Sensitivity Lower 95% confidence interval Alhusseini/2014/in house Alidjinou/2012/COBAS 1.00 0.96 0.93 0.80 Upper 95% confidence interval 1.00 1.00 1.00 Specificity Lower 95% confidence interval 0.69 – Upper 95% confidence interval 1.00 –

Gupta/1992/in house Jardi/2004/QuiaAMP

0.96 0.94

0.85 0.85

0.99 0.98

0.87

0.60 –

0.98 –

Lira/2009/QuiaAMP

1.00

0.92

1.00

–

Mohamed/2013/COBAS Ross/2013/Architect Vinikoor/2015/COBAS Zhang/2010/in house

0.98 0.93 0.91

0.89 0.86 0.82

1.00 0.97 0.96

1.00 1.00

0.93 0.93

1.00 1.00

–

Combined

0.96

0.92

0.98

1.00

0.39

1.00

Heterogeneity (Chi-square): Q = 9.68, df = 7.00, P = 0.21 Inconsistency (I-square): I = 27.69, 95% CI = [0.00–85.52] 2

Heterogeneity (Chi-square): Q = 16.89, df =7.00, P = 0.02 Inconsistency (I-square): I2 = 58.56, 95% CI = [26.14 – 90.98]

Page | 509

Table 6. Forest plot of likelihood ratios

Page | 510

Table 7. GRADE table Number of studies Type of study Directness Precision Consistency Risk of bias Overall quality

Sensitivity 96% (95% CI 92.0%–98.0%) 8 studies (463 HBV positive among 588 samples) Specificity 100% (95% CI 54–100%) 4 studies (125 HBV DNA neg among 588 samples) Diagnostic accuracy No significant indirectness Significant imprecision with small sample size Significant inconsistency High risk of bias (patient enrolment not consecutive or random in all studies; several case–control studies) Low Diagnostic accuracy No significant indirectness No significant imprecision Significant inconsistency High risk of bias (patient enrolment not consecutive or random in all studies; several case–control studies) Low

Page | 511

Appendix iv. Bivariate vs univariate analysis Comparing the analysis of bivariate sensitivity and specificity estimates and confidence intervals were similar. Heterogeneity was slightly lower in the univariate analysis. Univariate sensitivity

Table A3. Sensitivity of included studies in univariate analysis

Page | 512

Univariate specificity

Study

ES (95% CI)

Weight

Alhusseini (2014)

1.00 (0.72, 1.00)

2.33

Gupta (1992)

0.87 (0.62, 0.96)

1.19

Mohamed (2013)

1.00 (0.93, 1.00)

48.24

Ross (2013)

1.00 (0.93, 1.00)

48.24

Overall (I^2 = 0.00%, p = 0.52)

1.00 (0.98, 1.02)

100.00

-.5

1.5

Table A4. Specificity of included studies in univariate analysis v. Sources of heterogeneity – univariate analysis Looking at heterogeneity in the different assays used, sensitivity and specificity was not different in those with commercial or in-house assays. Different assays

Page | 513

% Study ES (95% CI) Weight

in house Alhusseini (2014) Gupta (1992) Zhang (2010) Subtotal (I^2 = 74.97%, p = 0.02) 1.00 (0.93, 1.00) 0.96 (0.85, 0.99) 0.84 (0.69, 0.92) 0.95 (0.88, 1.02) 18.41 8.41 2.84 29.66

commercial Alidjinou (2012) Jardi (2004) Lira (2009) Mohamed (2013) Ross (2013) Subtotal (I^2 = 49.24%, p = 0.10) 0.96 (0.81, 0.99) 1.00 (0.95, 1.00) 1.00 (0.92, 1.00) 0.98 (0.90, 1.00) 0.93 (0.86, 0.97) 0.98 (0.96, 1.01) 6.26 21.74 17.72 13.97 10.66 70.34

Heterogeneity between groups: p = 0.397 Overall (I^2 = 56.19%, p = 0.03); 0.98 (0.96, 1.00) 100.00

-.5

1.5

Table A5. Sensitivity of included studies stratified on type of assay

% Study ES (95% CI) Weight

in house Alhusseini (2014) Gupta (1992) Subtotal (I^2 = 56.14%, p = 0.13) 1.00 (0.72, 1.00) 0.87 (0.62, 0.96) 0.95 (0.85, 1.05) 2.33 1.19 3.52

commercial Mohamed (2013) Ross (2013) Subtotal (I^2 = 56.14%, p = 0.13) 1.00 (0.93, 1.00) 1.00 (0.93, 1.00) 1.00 (0.98, 1.02) 48.24 48.24 96.48

Heterogeneity between groups: p = 0.385 Overall (I^2 = 0.00%, p = 0.52); 1.00 (0.98, 1.02) 100.00

-.5

1.5

Table A6. Specificity of included assay stratified on type of assay

Page | 514

References 1. Schweitzer A, Horn J, Mikolajczyk RT, Krause G, Ott JJ. Estimations of worldwide prevalence of chronic hepatitis B virus infection: a systematic review of data published between 1965 and 2013. Lancet. 2015;386 (10003):1546–55. 2. NICE. Diagnosis and management of chronic hepatitis B in children, young people and adults. London: National Institute for Health and Care Excellence; 2013. 3. Lira R, Maldonado-Rodriguez A, Rojas-Montes O, Ruiz-Tachiquin M, Torres-Ibarra R, CanoDominguez C et al. Use of dried blood samples for monitoring hepatitis B virus infection. Virol J. 2009; 6:153. 4. Hirtz C, Lehmann S. [Blood sampling using "dried blood spot": a clinical biology revolution underway?] Ann Biol Clin. 2015;73(1):25–37. [Article in French]. 5. Snijdewind IJM, van Kampen JJA, Fraaij PLA, van der Ende ME, Osterhaus ADME, Gruters RA. Current and future applications of dried blood spots in viral disease management. Antiviral Res. 2012; 93(3): 309–21. 6. Ross RS, Stambouli O, Gruner N, Marcus U, Cai W, Zhang W et al. Detection of infections with hepatitis B virus, hepatitis C virus, and human immunodeficiency virus by analyses of dried blood spots--performance characteristics of the ARCHITECT system and two commercial assays for nucleic acid amplification. Virol J. 2013;10:72. 7. Hickman M, McDonald T, Judd A, Nichols T, Hope V, SKidmore S et al. Increasing the uptake of hepatitis C virus testing among injecting drug users in specialist drug treatment and prison settings by using dried blood spots for diagnostic testing: a cluster randomized controlled trial. J Viral Hepat. 2008;15(4):250–4. 8. McAllister G, Innes H, McLeod A, Dillon JF, Hayes PC, Fox R et al. Uptake of hepatitis C specialist services and treatment following diagnosis by dried blood spot in Scotland. J Clin Virol. 2014;61(3):359–64. 9. Hutchinson S. Translating research into public health policy: The scottish national hepatitis C action plan. Suchtmedizin in Forschung und Praxis 2013; Conference: 3rd International Symposium on Hepatitis Care in Substance Users, Munich, Germany. Conference Start: 20130905 Conference End: 20130906. Conference Publication: (var.pagings). 15 (4): 217. 10. Craine N, Whitaker R, Perrett S, Zou L, Hickman M, Lyons M. A stepped wedge cluster randomized control trial of dried blood spot testing to improve the uptake of hepatitis C antibody testing within UK prisons. Eur J Public Health. 2015;25(2):351–7. 11. Komas NP, Vickos U, Hubschen JM, Bere A, Manirakiza A, Muller CP et al. Cross-sectional study of hepatitis B virus infection in rural communities, Central African Republic. BMC Infect Dis. 2013;13:286. 12. Mahfoud Z, Kassak K, Kreidieh K, Shamra S, Ramia S. Prevalence of antibodies to human immunodeficiency virus (HIV), hepatitis B and hepatitis C and risk factors in prisoners in Lebanon. J Infect Dev Ctries. 2010;4(3):144–9. Page | 515

13. Gupta BP, Jayasuryan N, Jameel S. Direct detection of hepatitis B virus from dried blood spots by polymerase chain reaction amplification. J Clin Microbiol. 1992;30(8):1913–16. 14. Lira R, Maldonado-Rodriguez A, Rojas-Montes O, Ruiz-Tachiquin M, Torres-Ibarra R, CanoDominguez C et al. Use of dried blood samples for monitoring hepatitis B virus infection. Virol J. 2009;6:153. 15. Mohamed S, Raimondo A, Penaranda G, Camus C, Ouzan D, Ravet S et al. Dried blood spot sampling for hepatitis B virus serology and molecular testing. PLoS One. 2013;8(4):e61077. 16. Greenman J, Roberts T, Cohn J, Messac L. Dried blood spot in the genotyping, quantification and storage of HCV RNA: A systematic literature review. J Viral Hepat. 2015;22 (4):353–61. 17. Smit PW, Elliott I, Peeling RW, Mabey D, Newton PN. An overview of the clinical use of filter paper in the diagnosis of tropical diseases. Am J Trop Med Hyg. 2014;90 (2):195–210. 18. Guidelines for the prevention, care, and treatment of persons with chronic hepatitis B infection. Geneva: WHO; 2015. 19. Detection of infections with hepatitis B virus, hepatitis C virus, and human immunodeficiency virus by analyses of dried blood spots - performance characteristics of the ARCHITECT system and two commercial assays for nucleic acid amplification. Virol J 2013; 10(72). 20. Alidjinou EK, Moukassa D, Sane F, Twagirimana Nyenyeli S, Akoko EC, Mountou MV et al. Detection of hepatitis B virus infection markers in dried plasma spots among patients in CongoBrazzaville. Diagn Microbiol Infect Dis. 2014;78(3):229–31. 21. Alhusseini NF, Abadeer MZ, El-Taher SM. Hepatitis B virus DNA can be amplified directly from dried blood spot on filter paper. Am J Biochem Biotechnol. 2012;8 (2):143–9. 22. Halfon P, Raimondo A, Ouzan D, et al. Dried blood spot for hepatitis B virus serology and molecular testing. J Hepatol. 2012; Conference: 47th Annual Meeting of the European Association for the Study of the Liver, International Liver Congress 2012 Barcelona Spain. Conference Start: 20120418 Conference End: 20120422. Conference Publication: (var.pagings). 56: S62. 23. Jardi R, Rodriguez-Frias F, Buti M, Schaper M, Valdes A, Martinez M et al. Usefulness of dried blood samples for quantification and molecular characterization of HBV-DNA. Hepatology. 2004;40(1):133–9. 24. Zhang J, Zhang L, Song M, Wang W. Detection of HBV-DNA in dried bloodstains on filter paper by nested polymerase chain reaction. Lab Med. 2010;41 (9):535–9. 25. Vinikoor MJ, Zurcher S, Musukuma K, Kachuwaire O, Rauch A, Chi BH et al. Hepatitis B viral load in dried blood spots: a validation study in Zambia. J Clin Virol. 2015;72:20–4. 26. Alidjinou EK, Moukassa D, Sane F, et al. Detection of hepatitis B virus infection markers in dried plasma spots among patients in Congo-Brazzaville. Diagnostic Microbiology and Infectious Disease 2014; 78(3): 229-31.

Page | 516

27. Alidjinou EK, Moukassa D, Sane F, et al. Detection of hepatitis B virus infection markers in dried plasma spots among patients in Congo-Brazzaville. Diagn Microbiol Infect Dis 2014; 78(3): 229-31. 28. Alidjinou EK, Moukassa D, Sane F, et al. Detection of hepatitis B virus infection markers in dried plasma spots among patients in Congo-Brazzaville. Diagnostic Microbiology and Infectious Disease 2014; 78(3): 229-31.

Page | 517

Annex 5.9.4 PICO 7 - Dried blood spots Dried blood spots as sample collection method for hepatitis C virological testing: a systematic review and meta-analysis

MSF Access Campaign Cohn J (Team lead), Greenman J, Messac C, Roberts T Médecins Sans Frontières, Geneva. Switzerland

August 2015

Page | 518

Abstract b. Introduction The entry of new all-oral direct acting antiviral therapy for hepatitis C provides an opportunity to scale up HCV care in low- and middle-income countries. In HIV, use of dried blood spots (DBS) has facilitated the diagnosis and management of HIV in resource-poor settings. DBS may be used in a similar way to facilitate diagnosis and management of HCV. Here, we present a systematic review of the literature of DBS for HCV RNA detection to address the WHO PICOT 7 question. This is an update of the Greenman et al. 2014 paper addressing this question.

c. Methods Following an a priori protocol, PubMed, MEDLINE and Web of Science databases were searched by two reviewers duplicating each other’s efforts. Data was extracted with the primary outcome of HCV viral load DBS test accuracy using the gold standard of a venous sample. For analysis of sensitivity and specificity, a bivariate analysis using maximum likelihood estimate and 95% confidence intervals was used. Likelihood ratios were calculated directly from the pooled sensitivity and specificity. QUADAS-2 was used to assess bias and a GRADE evaluation was performed to evaluate quality of included studies.

d. Results The previous review found six papers eligible for inclusion and the update included three papers, one of which is in press, making for a total of nine papers eligible for inclusion. Eight studied DBS and one dried serum. The pooled sensitivity and specificity were 96.0% (upper-lower bounds 93.4–97.6) and 97.7% (upper-lower bounds 94.7–99.0), respectively. Heterogeneity was identified and a stratified analysis on capillary versus venous blood samples was performed which revealed a pooled sensitivity of 92.7% (lower-upper bounds 87.5–95.8%) and 97.4% (lower-upper bounds 95.9–98.3%) for capillary and venous samples, respectively, and pooled specificity of 97.4% (91.6– 99.2%) and 98.3% (93.9–99.5%) for capillary and venous samples, respectively. Although there were insufficient data on storage conditions to perform a quantitative analysis, several papers stored DBS at ambient temperature and for prolonged periods of time. While these storage conditions did not affect accuracy, two studies found deterioration of HCV RNA in DBS samples stored at room temperature, while two others failed to detect such deterioration.

Page | 519

e. Discussion These results support the potential use of DBS for HCV RNA detection. Further information is needed on the use of DBS for quantitative HCV RNA viral load when stored in DBS outside the cold chain.

1. Introduction Globally, there are approximately 130–150 million people living with hepatitis C virus (HCV) and the majority of these live in low- and middle-income countries (LMICs).1,2 In part because of the complexity and cost of the current algorithm for diagnosis and treatment using pegylated interferon, the majority of LMICs do not support HCV programming. As a result, without accessible and effective treatment, nearly half a million people die of HCV annually.2,3 However, the more effective and tolerable oral direct-acting antivirals (DAAs) offer the opportunity to significantly simplify both the treatment and diagnostic algorithm, enabling the implementation, decentralization and scale-up of HCV care in LMICs. In order to further simplify diagnosis and monitoring of HCV, dried blood spots (DBS) may be considered. DBS have been used to aid in the diagnosis of a wide variety of pathogens, including assessment of antibody to viral or bacterial infections such as HIV, hepatitis B, HPV and measles virus4,5 and qualitative and quantitative viral load detection in HIV.6 In particular, use of DBS in detecting and measuring HIV RNA has aided in decentralization of HIV services in low-resource areas and expanded the ability to perform early infant diagnosis for children at risk of vertical transmission.7 DBS for detection and monitoring of HCV viral load has a number of advantages. Especially salient for resource-poor settings is the possibility of using DBS to store and transport samples to a central laboratory without having to use refrigeration or dry ice, which is necessary for serum/plasma samples.8–10 Furthermore, plasma/serum should be processed within 6 h of venous blood draw, necessitating transportation to the laboratory within that time – an impossibility in most resource-limited settings, especially those outside urban areas. DBS can also be prepared using capillary blood, which obviates the need, seen in venipuncture, for centrifugation to separate blood cells from serum/plasma.12 In injecting drug users, venipuncture can be complicated by difficulty in finding an accessible vein and thus, capillary blood from fingerprick samples may be easier to obtain.11 DBS holds advantages over oral fluid sampling, which has also been used to detect HCV, as it has been shown that HCV RNA in saliva is independent of plasma viral load,13 and patients with low serum HCV RNA viral loads are less likely to have detectable HCV in saliva.14 Other advantages include lower cost, minimal storage facility and transportation requirements, decreased donor discomfort and decreased risk to health-care Page | 520

workers.10 If DBS could be used for HCV RNA detection and monitoring, it would facilitate the simplification and decentralization of HCV diagnosis and monitoring.8,11 Today a significant amount of literature supports the use of DBS as an alternative to serum/plasma obtained via venepuncture for the detection of HCV antibody,16–18 yet the use DBS in the detection and monitoring of HCV RNA and genotyping HCV, especially following storage at room temperature, has not been systematically reviewed.15 To address the question of whether DBS can be used in the diagnosis of HCV in resource-limited settings, we undertook a systematic review of use of DBS for HCV RNA detection and genotyping, examining a range of storage conditions. This review was published in the Journal of Viral Hepatitis in 2014.19 The previous review included nine papers – eight studied DBS and one studied dried serum. Two studies measured concordance between genotype and subtype determined by DBS and whole plasma and both found 100% concordance. Four studies measured end-point detection limits of HCV RNA positive samples by DBS and found sensitivity of 100% down to 250 IU/mL, 331 IU/mL, 2500 IU/mL and 24 160 IU/mL. Two studies found deterioration of HCV RNA in DBS samples stored at room temperature (10-fold reduction in HCV RNA at 4 weeks using dried serum spots and 3-fold reduction in RNA at 6 days with DBS), while two others failed to detect such deterioration. In 2014, WHO published the first guidelines for the prevention, care and treatment of individuals with HCV infection. These guidelines focused on assessment for treatment eligibility, initiation of first-line therapies and monitoring. WHO is now undertaking guidelines for testing for chronic hepatitis B and C infection in low- and middle-income settings. A topic for consideration in these guidelines is the potential use of DBS for serological and molecular testing for HBV and HCV to facilitate access to and uptake of testing. In order to inform these WHO guidelines, the following PICOT was developed: Among persons identified for HCV testing, what is the diagnostic accuracy and impact of detecting HCV NAT from DBS samples versus venous samples? Population: Samples for HCV NAT Intervention: Using DBS samples Comparisons: Using plasma or serum from venous samples Outcomes: Diagnostic accuracy (sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, TN, TP, FN and FP).

2. Methods The protocol used for this update is the same as that used in the Greenman et al. paper. This protocol is copied here from the Journal of Viral Hepatitis manuscript.

Page | 521

a. Search strategy Using a sensitive search strategy as part of a predefined protocol (Appendix 1), we searched MEDLINE, CAB abstracts and Web of Science (ISI Citation Index) published up to August 2013 for studies meeting our pre-specified inclusion criteria. No date or geographical exclusions were applied; only English language publications were included. Following an initial screening of abstracts by two separate reviewers, full-text copies of potential eligible articles were reviewed independently by two reviewers. A title search of references was performed on articles meeting the inclusion criteria to determine potential articles for inclusion not identified during the initial database search. After all articles meeting the inclusion criteria were identified, data was abstracted by two reviewers according to prespecified categories.

b. Inclusion/exclusion criteria We sought studies that reported on the use of dried DBS as a tool for monitoring and genotyping HCV RNA that included at least one of the following: the sensitivity and/or specificity of DBS in HCV RNA viral load quantification; the accuracy of DBS for genotyping HCV RNA; and the rate of degradation of HCV RNA during transport and storage at room temperature. Nine of the 73 articles met these criteria.

c. Data analysis For studies measuring HCV RNA presence and quantification, studies were analysed for viral load end-point detection, sensitivity, specificity, PPV and NPV of DBS compared to HCV RNA in whole plasma. For the question of HCV RNA deterioration at room temperature in DBS, reviewers extracted reports of the rate of deterioration and the definition of “room temperature” . The results of this analysis are displayed in Table 1. Studies determining HCV genotype were analysed for measures of genotype concordance between whole plasma and DBS, proportion of HCV RNApositive DBS samples successfully genotyped and genotypes observed. The updated review searched Medline and PubMed with the same search terms as from the original SR, but with the limit of 2013–present.

Page | 522

d. Statistical data analysis – update Statistical analysis of the data was performed using OpenMeta[Analyst]. For analysis of sensitivity and specificity, a bivariate analysis using maximum likelihood estimate and 95% confidence intervals was used. Likelihood ratios were calculated directly from the pooled sensitivity and specificity. We used forest plots to visually assess heterogeneity.

e. Risk of bias and quality assessment The QUADAS-2 tool was used to assess risk of bias (Table 1). A GRADE assessment was performed to assess quality of included studies (Table 2).

3. Results In the initial review search, 184 titles were identified. Seventeen were selected for full-text review. Overall, nine met eligibility criteria for the published review (Fig. 1) of which six reported on the outcomes for this PICOT, sensitivity and specificity of HCV NAT DBS versus venous samples.20–25 In the review update, 14 articles were identified of which 13 underwent title and abstract review (there was one article duplication in the search). Four articles were selected for full-text review. The remainder were not selected as they did not address HCV (n=4), did not address HCV viral load (n=4) or did not include outcome data (n=1). Of the four articles selected for full-text review, two were excluded as they did not deal with HCV viral load (VL). Two were retained.26,27 One further article that has been accepted for publication was identified by an expert in the field and was included in the update.28 Two of these articles26, 28 provided sufficient data to include in the meta-analysis (Fig. 2).

a. Study characteristics The selected studies were published between 1998 and 2015; all but two were published between 2007 and 2015. Two of the studies were located in the UK, two in Italy, two in France, two in the United States and one each in Lebanon, Brazil, Guinea-Bissau and Japan. Six studies reported outcomes in terms of end-point sensitivity of HCV RNA detection using DBS when compared to plasma.20–25 Four studies measured the stability of HCV RNA in DBS at room temperature (Table 1).20–21, 23–24

Page | 523

b. Sample characteristics The patient characteristics varied across studies. Five studies drew from patients attending liver clinics21, 25, 27, 28 or in HCV treatment studies.23 One study drew from a population of injecting drug users (IDUs), one from HIV-infected patients22 and one from a general population.24 One study’s23 participants were on PEG-RBV treatment. Finally, the preparation of DBS varied among the samples. Eight of the nine studies used DBS and one used dried plasma spot.20 Three studies used capillary blood23, 25, 26 and five used venous blood for preparation of DBS.21, 22, 24, 27, 28 The studies used a variety of RNA extraction and elution methods, and nucleic acid amplification assays. 21–28

c. Diagnostic accuracy values Overall, the pooled sensitivity was 96.0% (upper–lower bounds 93.4–97.6) and specificity was 97.7% (upper–lower bounds 94.7–99.0) (Figs 3 and 4). The negative and positive likelihood ratios were calculated to be 0.041 and 41.74, respectively. There was minor heterogeneity identified on the forest plots for sensitivity, with one study in particular contributing to the heterogeneity.26 This study used capillary blood; however, other studies with higher sensitivity (e.g. 25), also used capillary blood. Thus, a stratified analysis was performed on capillary versus venous samples. i.

d. Impact of sample type A stratified analysis looking at sample type (capillary versus venous) revealed a pooled sensitivity of 92.7% (lower–upper bounds 87.5–95.8%) and 97.4% (lower–upper bounds 95.9–98.3%) for capillary and venous samples, respectively (Fig. 5). The overall pooled specificity was 97.4% (91.6– 99.2%) and 98.3% (93.9–99.5%) for capillary and venous samples, respectively (Fig. 6).

e. Impact of storage conditions There was insufficient information to perform a meta-analysis on different storage conditions. However, from information provided in the included manuscripts, storage at room temperature or over time up to 1 year does not appear to affect sensitivity, though several studies did identify a reduction in viral yield.20, 25

Page | 524

f. Bias and quality assessment The QUADAS-2 revealed some risk of bias due to patient selection that was not consecutive or randomized in some studies or due to a case–control design. Otherwise, the included studies showed low risk of bias (Table 2). The GRADE table showed a moderate quality of data (Table 3).

4. Discussion a. Overall conclusions This systematic review and meta-analysis includes evidence of moderate quality that supports acceptable accuracy of DBS for testing HCV NAT as compared to use of plasma samples, and suggests that DBS may be used for diagnosis of HCV using HCV NAT DBS where there is limited access to venepuncture or inadequate technology to prepare and transport plasma samples. Although studies are limited, DBS is likely stable and maintains good accuracy in conditions with higher temperatures, although viral degradation may occur when storing for prolonged durations, which may affect quantitative assessment of viral load. As there are relatively little data on accuracy of DBS in real-life conditions (including high humidity), operational research will be useful to determine accuracy in such conditions. In a stratified analysis that examined studies that used capillary blood versus those that used venous blood for DBS, capillary blood sample use resulted in somewhat lower pooled sensitivity than venous blood sample, although bounds were wide and overlapping. One study that used capillary blood examined the discordant results and found that the signal to cut-off ratio for discordant results was lower than for concordant results, suggesting that a lower cut-off may help to improve accuracy for capillary blood samples.

b. Key limitations This review also has a number of limitations. Overall, the number of studies was small. In particular, there is a dearth of studies that systematically examines the effects of storage conditions on the accuracy of DBS; and of those that did assess this, several did not specify the exact conditions, e.g. exactly what is room temperature. This is particularly important as the majority of these studies came from high-income countries and room temperature in these areas will be very different than temperature in sub-Saharan Africa. Thus, should DBS be adopted for screening for HCV NAT, it will be important to pursue further operational research using field specimens prepared and stored in real-life conditions. Ideally, to fit operational needs, these Page | 525

studies should use capillary whole blood to prepare the DBS, using a commonly available type of filter paper while comparing several commercial test kits and conditions of storage.

Future work Expanded use of HCV NAT testing will be critical to improving the diagnosis and treatment of HCV globally. The use DBS for improved preparation, storage and transport of samples for decentralized testing may help to expand the reach of this important diagnostic test.

Tables and figures

Page | 526

c. Fig. 1. Flow diagram

Page | 527

d. Fig. 2. Additional search flow diagram

From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and MetaAnalyses: The PRISMA Statement. PLoS Med 6(6): e1000097. doi:10.1371/journal.pmed1000097

Page | 528

e. Table 1: Study-level characteristics Study Design Study site and population n =12 Japan DBS collection method Plasma on Whatman filter paper Storage conditions Assay Sensitivity/specificity Effect of storage conditions Cut-off value

Abe (1998)

Case–control

Room temperature and ABI Prism Dye Terminator tested at 1, 2, 3, 4 Cycle Sequencing Ready weeks Reaction Kit (Perkin-Elmer)

Sensitivity: 100% Specificity: 100%

10-fold reduction in virus yield in 4 weeks (in 6 of 8 samples, no samples lost HCV RNA positivity)

Bennett (2012)

Cross-sectional n=80 HCV antibody positive patients attending liver clinic, United Kingdom

Venous blood on Room temperature, Whatman 903 cards 4 °C, –20 °C, –80 °C

ABI 9700 and ABI 7500

Sensitivity: 100% (57/57) Specificity: 95.7% (22/23)

No significant variation in Suggested: cycle threshold over 1 year 150 –250 IU/mL (2 DBS from 1 patient)

De Cringis (2010)

Case–control

n=25 13 HIV–HCV coinfected 4 HIV monoinfected 3 HCV 5 healthy blood donors Italy

50 uL venous blood on Whatman No. 3 card

Stored at –80 °C

HIV-1/HCV multiplex SYBR Green real-time RT-PCR assay

Sensitivity: 93.8% (15/16) Specificity: 100% (9/9)

Dukobo (2014)

Cross sectional n=132 Adult (18–30 years) active IDU in San Francisco

Capillary blood on Whatman 903 cards

dHCV TMA (Norvatis)

Sensitivity: 89.6% (95% CI 77.8–95.5) Specificity: 100% (95% CI 95.6–100)

Page | 529

TejadaStrop (2015)

Case–control

Venous samples on Whatman 903 paper Adult patients with chronic hepatitis C at the (GE Healthcare) Liver Clinic of the University of California, San Francisco n=33

Control: DBS freshly prepared Long-term storage: 5 years at –20C

Superscipt III Platinum 1 Step qRT-PCR kit

Control Sensitivity: 88% Specificity: 100% Long-term storage Sensitivity: 33% Specificity: 100%

Soulier (2015)

Case–control

Venous samples on –80 °C with desiccant Whatman 903 paper package 315 adults with HCV-Ab- (GE Healthcare) positive chronic hepatitis C n=511 26 adults with HCV-Abpositive resolved infection 170 adults HCV-Abnegative France

CobasAmpliprep/CobasTaq Man HCV version 2 (CAP/CTM v2.0, Roche Molecular Systems) and m2000 platform (Abbott Molecular)

CAP/CTM v2 Sensitivity: 97.1% (94.7%–98.5%) Specificity: 100% (97.8%-100%) M200 Sensitivity: 98.1% (95.9%–99.1%) Specificity: 100% (97.8%–100%)

Santos

Cross-sectional n = 100 HCV chronically infected patients in a PEG-RBV study at 4 (n = 100) and 24 (n = 68) weeks posttreatment initiation Brazil

Capillary blood on Room temperature SS903 collection cards (Schleicher and Schuell, Keene, NH, USA)

pCR-II-TOPO plasmid (Invitrogen)

Sensitivity: 98.0% (99/101) Specificity: 94.0% (63/67)

Solmone (2002)

Cross-sectional n = 55 Residual samples from patients undergoing routine haematological

50 uL EDTA venous Subset of 16 paired In-house RT-PCR blood on SS903 card samples stored at room temperature and tested every 2–4 weeks over

Sensitivity: 100% (124/124) Specificity: 100% (24/24)

No loss of positivity after 11 months (16 paired samples)

Page | 530

controls (39 HCV antibody (Ab)–positive and 16 HCV Ab-negative patients) Italy

11 months

Tuallion (2009)

Case–control

n = 200 100 anti-HCV-positive and 100 anti-HCVnegative France

3 drops (50 lL) capillary blood on Whatman 903 card

Stored at –20 °C for 1–8 CobasAmpliprep Total weeks until use Nucleic Acid Isolation 100 kit (Roche); CobasTaqMan HCV test and real-time PCR COBAS TaqMan 48 instrument plus COBAS Ampliprep analyzer (Roche); One Step RT-PCR kit from Qiagen (Qiagen)

3-fold decrease in RNA in 6 days (No. of samples not reported)

Page | 531

f. Fig. 3. Forest plot of sensitivity

g. Fig. 4. Forest plot of specificity

Page | 532

h. Fig. 5. Forest plot of sensitivity of capillary and venous samples Capillary samples

Venous samples

i. Fig. 6. Forest plot of specificity of capillary and venous samples Capillary samples

Venous samples

Page | 533

j. Table 2: QUADAS-2 Author Abe Bennet De Crignis Dokubo Santos Solome Soulier Tejada-Strop Tuaillon Patient selection HR LR HR LR HR HR HR HR HR Index test LR LR LR LR HR LR LR HR LR Reference standard LR LR LR LR LR LR LR LR LR Flow and timing LR LR LR LR LR LR LR LR LR

HR: high risk; LR: low risk

k. Table 3. GRADE table Number of studies Type of study Directness Precision Consistency Risk of bias Overall quality

Sensitivity of DBS for HCV VL: 96.0% (upper-lower bounds 93.4–97.6) 9 studies 1335 samples Cross-sectional, case–control or cohort No significant indirectness No significant imprecision No significant inconsistency Significant risk of bias (non-randomized or consecutive patient recruitment or casecontrol design) Moderate

Specificity of DBS for HCV VL: 97.7% (upper-lower bounds 94.7-99.0) 9 studies 1335 samples Cross-sectional, case–control or cohort No significant indirectness No significant imprecision No significant inconsistency Significant risk of bias (non-randomized or consecutive patient recruitment or casecontrol design) Moderate

Page | 534

References 1. Mohd Hanafiah K, Groeger J, Flaxman AD, Wiersma ST. Global epidemiology of hepatitis C virus infection: new estimates of age-specific antibody to HCV seroprevalence. Hepatology. 2013;57(4):1333‒42. Hepatitis C factsheet. Geneva: World Health Organization; 2014 (http://www.who.int/mediacentre/factsheets/fs164/en/, accessed 07 June 2016). Global policy report on the prevention and control of viral hepatitis. Geneva: World Health Organization; 2013 (http://apps.who.int/iris/bitstream/10665/85397/1/ 9789241564632_eng.pdf, accessed 07 June 2016). Parker SP, Cubitt WD. The use of the dried blood spot sample in epidemiological studies. J Clin Pathol. 1999;52(9):633–9. Waterboer T, Dondog B, Michael KM, Michel A, Schmitt M, Vaccarella S, et al. Dried blood spot samples for seroepidemiology of infections with human papillomaviruses, Helicobacter pylori, and hepatitis C virus. Cancer Epidemiol Biomark Prev. 2012;21(2):287–93. Johannessen A, Garrido C, Zahonero N, Sandvik L, Naman E, Kivuyo SL, et al. Dried blood spots perform well in viral load monitoring of patients who receive antiretroviral treatment in rural Tanzania. Clin Infect Dis. 2009;49(6):976–81. Sherman GG, Stevens G, Jones SA, Horsfield P, Stevens WS. Dried blood spots improve access to HIV diagnosis and care for infants in low-resource settings. J Acquir Immune Defic Syndr. 2005;38(5):615–7. Solmone M, Girardi E, Costa F, Pucillo L, Ippolito G, Capobianchi MR. Simple and reliable method for detection and genotyping of hepatitis C virus RNA in dried blood spots stored at room temperature. J Clin Microbiol. 2002;40(9):3512–4. De Crignis E, Re MC, Cimatti L, Zecchi L, Gibellini D.HIV–1 and HCV detection in dried blood spots by SYBR Green multiples real-time PCR. J Virol Methods. 2010;165(1):51–6.

2. 3.

4. 5.

7. 8.

10. McDade TW, Williams S, Snodgrass JJ. What a drop can do: dried blood spots as a minimally invasive method for integrating biomarkers into population-based research. Demography. 2007;44(4):899–925. 11. Hickman M, McDonald T, Judd A, Nichols T, Hope V, Skidmore S, et al. Increasing the uptake of hepatitis C virus testing among injecting drug users in specialist drug treatment and prison settings by using dried blood spots for diagnostic testing: a cluster randomized controlled trial. J Viral Hepat. 2008;15(4):250–4. 12. Santos C, Reis A, Dos Santos CV, Damas C, Silva MH, Viana MV, et al. The use of real-time PCR to detect hepatitis C virus RNA in dried blood spots from Brazilian patients infected chronically. J Virol M.ethods 2012;179(1):17–20. 13. Lins L, Almeida H, Vitvisk L, Carmo T, Parana R, Reis MG. Detection of hepatitis C virus RNA in saliva is not related to oral health status or viral load. J Med Virol. 2005;77(2):216–20. 14. Suzuki T, Omata K, Satoh T, Miyasaka T, Arai C, Maeda M, et al. Quantitative detection of hepatitis C virus (HCV) RNA in saliva and gingival crevicular fluid of HCV-infected patients. J Clin Microbiol. 2005;43(9):4413–7. 15. Snijdewind I, van Kampen J, Fraaij P, van der Ende ME, Osterhaus AD, Gruters RA. Current and future applications of dried blood spots in viral disease management. Antiviral Res. 2012;93(3):309–21. 16. Parker SP, Cubitt WD, Ades AE. A method for the detection and confirmation of antibodies to hepatitis C in dried blood spots. J Virol Methods. 1997;68(2):199–205.

Page | 535

17. Judd A, Parry J, Hickman M, McDonald T, Jordan L, Lewis K, et al. Evaluation of a modified commercial assay in detecting antibody to hepatitis C virus in oral fluids and dried blood spots. J Med Virol. 2003;71(1):49–55. 18. Parker SP, Khan HI, Cubitt WD. Detection of antibodies to hepatitis C virus in dried blood spot samples from mothers and their offspring in Lahore, Pakistan. J Clin Microbiol. 1999;37(6):2061–3. 19. Greenman J, Roberts T, Cohn J, Messac L. Dried blood spot in the genotyping, quantification and storage of HCV RNA: a systematic literature review. J Viral Hepat. 2015;22(4):353–61. 20. Abe K, Konomi N. Hepatitis C virus RNA in dried serum spotted onto filter paper is stable at room temperature. J Clin Microbiol. 1998;36(10):3070–2. 21. Bennett S, Gunson R, McAllister G, Hutchinson SJ, Goldberg DJ, Cameron SO, et al. Detection of hepatitis C virus RNA in dried blood spots. J Clin Virol. 2012;54(2):106–9. 22. De Crignis E, Re MC, Cimatti L, Zecchi L, Gibellini D. HIV–1 and HCV detection in dried blood spots by SYBR Green multiples real-time PCR. J Virol Methods. 2010;165(1):51–6. 23. Santos C, Reis A, Dos Santos CV, Damas C, Silva MH, Viana MV, et al. The use of real-time PCR to detect hepatitis C virus RNA in dried blood spots from Brazilian patients infected chronically. J Virol Methods. 2012;179(1):17–20. 24. Solmone M, Girardi E, Costa F, Pucillo L, Ippolito G, Capobianchi MR.. Simple and reliable method for detection and genotyping of hepatitis C virus RNA in dried blood spots stored at room temperature. J Clin Microbiol. 2002;40(9):3512–4. 25. Tuaillon E, Montain A, Meroueh F, Ottomani L, Picot MC, Nagot N, et al. Dried blood spot for hepatitis C virus serology and molecular testing. Hepatology. 2010;51(3):752–8. 26. Dokubo EK, Evans J, Winkelman V, Cyrus S, Tobler LH, Asher A, et al. Comparison of hepatitis C Virus RNA and antibody detection in dried blood spots and plasma specimens. J Clin Virol. 2014;59(4):223–7. 27. Tejada-Strop A, Drobeniuc J, Mixson-Hayden T, Forbi JC, Le NT, Li L, et al. Disparate detection outcomes for anti-HCV IgG and HCV RNA in dried blood spots. J Virol Methods. 2015;212:66–70. 28. Soulier A, Poiteau L, Rosa I, Hezode C, Roudot-Thoraval F, Pawlotsky JM, et al. Dried blood spots: a tool to ensure broad access to hepatitis C screening, diagnosis and treatment monitoring. J Infect Dis. 2016;213(7):1087–95.

Page | 536

Appendix 1: Terms used in database searches MEDLINE-International Database for Medical Literature: Dried blood spot [Ab] AND Hepatitis C [Ab] Dried plasma spot [Ab] AND Hepatitis C [Ab] Dried blood spot [Ab] AND HCV RNA [Ab] Dried plasma spot [Ab] AND HCV RNA [Ab] Dried blood spot [Ab] AND HCV [Ab] AND genotype [Ab] Dried plasma spot [Ab] AND HCV [Ab] AND genotype [Ab] Dried blood spot [Ab] AND HCV RNA [Ab] AND storage [Ab] Dried plasma spot [Ab] AND HCV RNA [Ab] AND storage [Ab] Web of Science (ISI Citation Index) Dried blood spot [topic] AND Hepatitis C [topic] Dried plasma spot [topic] AND Hepatitis C [topic] Dried blood spot [topic] AND HCV RNA [topic] Dried plasma spot [topic] AND HCV RNA [topic] Dried blood spot [topic] AND HCV [topic] AND genotype [topic] Dried plasma spot [topic] AND HCV [topic] AND genotype [topic] Dried blood spot [topic] AND HCV RNA [topic] AND storage [Ab] Dried plasma spot [topic] AND HCV RNA [topic] AND storage [Ab] CAB abstracts Dried blood spot [all fields] AND Hepatitis C [all fields] Dried plasma spot [all fields] AND Hepatitis C [all fields] Dried blood spot [all fields] AND HCV RNA [all fields] Dried plasma spot [all fields] AND HCV RNA [all fields]

Page | 537

Annex 5.10 PICO 8 - Diagnostic accuracy of HBsAg/HBeAg test versus NAT to confirm successful treatment response: a meta-analysis and review of the literature

London School of Hygiene and Tropical Medicine team *Olivia Varsaneux, *Ali Amini, Weiming Tang, Wen Chen, Debi Boeras, Jane Falconer, Helen Kelly, Joseph Tucker, Rosanna Peeling (Team lead) London School of Hygiene and Tropical Medicine team *Co-leaders of this Review

September 2015 Page | 538

1. Executive summary Background: Advances in hepatitis B virus detection technology create new opportunities for enhancing screening, referral and treatment. The purpose of this review was to determine the diagnostic accuracy of HBsAg or HBeAg test versus nucleic acid testing (NAT) to confirm successful treatment response among patients receiving treatment for HBV. Method: A literature search was conducted focused on hepatitis B, diagnostic tests and diagnostic accuracy. Studies were included if they evaluated an assay to determine the sensitivity and specificity of a HBsAg or HBeAg test compared to a quantitative HBV RNA reference among humans. Two reviewers performed a quality assessment of the studies and extracted data for estimating test accuracy. Results: It was found that, despite HBV NAT being considered the gold standard in confirming response to treatment, both HBsAg and HBeAg were useful in monitoring patients receiving treatment as in many resource-limiting settings NAT is not readily available. Studies showed that the kinetics of HBsAg and HBV DNA followed similar profiles during treatment with pegylated interferon (PEG-IFN) and follow up in patients who developed sustained virological response (SVR). Further studies determined that this correlation was present for all four genotypes. It was also reported that HBsAg quantification can allow for detection of active cases of chronic HBV from true inactive carriers, therefore reducing the need to rigorously monitor HBV DNA levels. Studies showed that HBeAg was capable of differentiating late responders from non-responders to HBV DNA after 24 weeks of treatment. Conclusions: There is limited evidence for the sole use of HBsAg or HBeAg compared to HBV DNA for monitoring treatment response. More studies are needed to determine which tests for HBV antigen detection may be useful as a marker of treatment response for which therapeutic agent.

2.Background An estimated 240 million individuals worldwide1 are chronically infected with hepatitis B virus (HBV) and there are an estimated 4 million acute HBV infections each year. Of those with chronic hepatitis B infection, 20–30% will develop cirrhosis2 or hepatocellular carcinoma,3 leading to approximately 650 000 deaths each year.4 However, most individuals with chronic HBV infection are not aware of their serostatus, contributing to delayed diagnosis and complications from advanced disease.5 HBV testing is critically important in order to refer infected individuals to HBV treatment and care, to refer uninfected individuals to vaccination and to mobilize prevention and control efforts. The introduction of NAT is an integral step in the control of the disease as it allows for rapid diagnosis and early treatment of HBV. The virus can be transmitted by blood from asymptomatic donors with acute HBV infection before the development of HBsAg or an anti-HBc response. Therefore, NATs can used to detect HBV DNA in a donor’s blood before antigen or

Page | 539

antibody response are detected.6 Though NAT testing has been proven to be more sensitive in detecting viral infections, serological testing is better suited for the detection of active infections.7 Treatment with tenofovir or entecavir is effective for HBV. Their efficacy can be measured by a sustained reduction in viral load, but the quantitative HBsAg response may remain high. The data for measuring HBsAg quantitatively largely works for interferon-based agents. Locarini reviewed the literature on quantitative HBsAg in hepatology and highlighted some of the challenges of quantitative HBsAg testing in using other therapeutic agents. In March 2015, the World Health Organization (WHO) published the first guidelines for the prevention, care and treatment of individuals with chronic HBV infection.5 These guidelines focused on assessment for treatment eligibility, initiation of first-line therapies, switching and monitoring. These initial guidelines did not include screening recommendations. Given the large burden of HBV in low- and middle-income settings where there are limited or no existing HBV testing guidelines, there is a substantial need for HBV testing guidelines. Advances in HBV detection technology create new opportunities for enhancing screening, referral and treatment. Previous systematic reviews on hepatitis B infection have focused on immunological responses,7 surveillance of cirrhosis8 and treatment.9 Existing systematic reviews10‒ 13 on hepatitis B testing focused on point-of-care (POC) tests and included tests with unclear reference standards. No systematic reviews have examined the diagnostic accuracy of using HBsAg/HBeAg compared to HBV DNA detection to monitor treatment response.

PICO 8

Among patients receiving treatment for HBV, what is the diagnostic accuracy of HBsAg/HBeAg test versus NAT to confirm successful treatment response? Patients receiving treatment for HBV HBsAg/HBeAg testing NAT for HBV DNA detection Diagnostic accuracy True negatives (TNs) – who are screen negative and have cleared the HBV infection. False negatives (FN) – who are screen negative but have HBV infection. These will be misclassified and treatment will be stopped resulting in disease progression leading to liverrelated morbidity (fibrosis, cirrhosis, end-stage liver disease, hepatocellular carcinoma), progression of liver disease and mortality. True positives (TP) – who are screen positive and truly have HBV infection. This will increase the number of treated cases and cure rate. False positives (FP) – who are screen positive, but do not have HBV infection. (These will continue treatment inappropriately, and and will have unnecessary referral). Costs – cost of testing strategy, including lab reagents and running costs, cost of further evaluation of a false positive. Cost–effectiveness

P I C O

Page | 540

Acceptability to health-care workers and patients Other outcomes (missed cases of liver disease because of false-negative results, unnecessary referral, investigations and/or treatment in false positives).

3. Objectives The purpose of this review was to identify evidence on the sensitivity and specificity of HBsAg/HBeAg compared to HBV DNA detection for HBV treatment monitoring and to summarize the key test characteristics associated with detection of HBsAg/HBeAg.

Selection criteria

i. Types of studies We included observational studies and randomized controlled trials (RCTs) that provide original data from patient specimens, including cross-sectional and case–control studies, and studied HBsAg/HBeAg testing compared to a reference standard of HBV DNA detection.

ii. Participants Little information on participants was provided in the selection of papers included in the systematic review; therefore, we set a wide inclusion criterion. We included patients of all age groups from all settings and countries as well as all types of specimens.

iii. Index tests Studies that utilized commercially available HBsAg/HBeAg and HBV DNA assays were eligible for inclusion. The following four are the index tests included:     Architect HBsAg assays, Abbott COBAS AMPLICOR TM HBV Test v2.0 assay. Roche Diagnostics Systems IMx HBeAg assay, Abbott Iprobe, Abbott

Page | 541

iv. Reference standard The reference standards acceptable for a definitive diagnosis included tests for detection of HBV by the following HBV DNA detection techniques—polymerase chain reaction (PCR), branchedchain DNA (bDNA), or transcription-mediated amplification (TMA) and DNA hybridization assays.

Outcome measures Sensitivity refers to the proportion of samples with true HBV infection diagnosed with positive HBsAg/HBeAg test confirmed with a positive HBV DNA detection method. Specificity refers to the proportion of samples with negative HBsAg/HBeAg test confirmed with a negative HBV DNA detection method.

Search methods A database search of LILACS, MEDLINE, EMBASE, PubMed, Scopus, Web of Science, Cochrane and WHO Global Index Medicus was performed through April 2015. No language restriction was applied. The references of published articles found in the above databases were searched for additional pertinent materials. Study selection proceeded in three stages. First, titles/abstracts were screened by a single reviewer according to standard inclusion and exclusion criteria. Second, full manuscripts were obtained and assessed against inclusion criteria. Papers were accepted or rejected and reasons for rejection were specified. Third, two independent reviewers assessed each manuscript and differences were resolved by a third independent reviewer.

Data extraction Information on the following variables were extracted by a reviewer if the study met the exclusion and inclusion criteria—first author, total sample size, country (and city) of sampling, sample type (oral fluid, finger-prick, venous blood, etc.), point-of-care (Y/N), eligibility criteria, reference standard, manufacturer, raw cell numbers (true positives, false negatives, false positives, true negatives), sources of funding and reported conflicts of interest. We define point of care as being able to give a result within 60 min and having the results guide clinical management at the same encounter.

Assessment of methodological quality

Page | 542

Study quality was evaluated using the QUADAS-2 tool,16 the STARD checklist17 and the GRADE method.18 QUADAS includes domains to evaluate bias in the following categories—risk of bias (patient selection, index test, reference standard, flow and timing); applicability concerns (patient selection, index test, reference standard). The GRADE method evaluates the strength of evidence by assessing the risk and probability of bias, imprecision and inconsistency as well as dose– respondent gradient and residual confounding.18

5. Results PRISMA flowchart Fig. 1. PRISMA flow diagram outlining study selection examining diagnostic accuracy of HBV antibody tests compared to HBV DNA in confirming successful treatment response

Page | 543

Characteristics of included studies Only two of the studies analysed met the PICO criteria and data were extracted from both these studies. These studies took place in Sweden and the United States of America. The patient population for the Larsson 2013 study was derived from a clinical setting; there was no information on the patient population for the Perrillo 1993 study. The assays evaluated for this Page | 544

systematic review were Architect HBsAg assays, COBAS AMPLICOR TM HBV Test v2.0 assay, IMx HBeAg assay and Iprobe. Of these two studies, only one reported sensitivity/specificity of HBsAg and one reported sensitivity/specificity of HBeAg. The lack of information on these diagnostic accuracy measures is a large limitation in the quality of the studies. Other issues with the quality of these studies were insufficient information on the populations studied, randomization and sample collection. Table 1. Description of study design, study population and setting of all studies (n=2) First author Sample Country type & size Treatment Study population Eligibility criteria Index Reference Sensitivity diagnostic test test Architect HBsAg assays, Abbott COBAS 34% AMPLICORT M HBV Test v2.0 assay. Roche Iprobe, Abbott 95% Specificity

Larrson, 2013

Liver tissue Sweden and blood sample N= 160

INF

Infectious Patients Disease clinic with chronic N =160 HBV

89%

Perrillo, 1993

Plasma N= 34

United States of America

INF

29 patients on treatment and 5 neg. controls

IMx HBeAg assay, Abbott

44%

Larsson et al. (2013) monitored HBsAg levels (Architect assays, Abbott) and HBV DNA quantitation (COBAS AMPLICORTM HBV monitor, Roche) in 160 patients treated for chronic HBV infection at the Infectious Disease Clinic at Sahlgrenska University Hospital between 1993 and 1995. Sensitivity of HBsAg compared to HBV DNA was 34% and the specificity was 89%. A correlation between HBsAg and HBV DNA in serum samples (R2 = 0.39; P< 0.0001) was also noted, in that a 90% reduction of HBV DNA corresponded to a 48% decline in HBsAg. The authors also measured HBeAg levels and found that HBeAg-positive patients had a 300 times higher HBV DNA/HBsAg ratio compared to those who were HBeAg-negative. These results indicate that HBsAg quantification could be complementary to HBV DNA quantification for treatment monitoring and confirming successful treatment response. Perrillo et al. (1993) evaluated whether the HBeAg assay (Abbott IMX) was capable of providing comparable information to HBV DNA assays (Iprobe, Abbott) during and after IFN therapy in 29 consecutive, IFN-treated patients and five untreated controls. The authors found that decremental and incremental changes in HBeAg concentration during and after therapy mirrored those observed with HBV DNA with a significant correlation (R = 0.768 P>0.0001). Only 56% of HBV DNA-negative patients tested positive for HBeAg but 95% of HBV DNA-positive samples were also positive for HBeAg. Though this information allows us to understand that HBeAg concentrations can provide similar clinically relevant information compared to HBV DNA Page | 545

assays, it is difficult to state the accuracy of these tests against each other as they are traditionally used to measure different indicators.

Narrative summary of each systematic review’s findings Monitoring response to treatment is an essential mechanism in the control of HBV and requires both the sustained disappearance of HBV DNA and the clearance of HBsAg/ HBeAg from the blood. The systematic review showed that monitoring of HBeAg concentration can provide clinically relevant information, though only two of the 6464 studies identified for screening (Larrson et al. 2013 and Perrillo et al. 1993) were included in the systematic review as they were the sole articles that met both the inclusion and exclusion criteria for PICO 8 (showed sensitivity and specificity of assays).

i. Diagnostic accuracy of nucleic acid testing HBV DNA is essential when determining the presence of the virus as it is quantitatively expressed and allows for prompt detection of HBV. With the advent of reverse transcriptase-polymerase chain reaction (RT-PCR), it quickly became regarded as the gold standard or confirming response to therapy due to its accuracy and cost–effectiveness. However, in many resource-limited settings, such assays are not widely available, therefore it is important to determine if HBsAg or HBeAg can be used for monitoring response to treatment.19

ii. Diagnostic accuracy of HBsAg test compared to HBV DNA Although they did not include specific accuracy values, three supplemental studies provided useful information on the quantitation of HBsAg for treatment monitoring in chronic HBV patients. Martinot-Peignoux et al. (2015) reported that the kinetics of HBsAg and HBV DNA followed similar profiles during treatment with PEG-IFN and follow up in patients who developed SVR (solid line) (see Fig. 2).18 Fig. 2. Serum HBV DNA and HBsAg kinetics during treatment with PEG-IFN and follow up in patients who developed SVR (solid line) 20

Page | 546

This was confirmed by Ganji et al. (2011) who that showed HBsAg had strong correlation with HBV DNA (r =0.69; P<0.01) for both genotypes investigated.20 Larsson et al. 2014 further proved that that there was a correlation between HBsAg and HBV DNA for all four genotypes (Fig. 3).19 This highlights the potential for HBsAg to be a useful serological marker to predict response to treatment.

Fig. 3 (A–D). Correlation between HBsAg and HBV DNA in genotypes A–D Fig. 3 (E). Box plot of HBsAg levels in HBeAg-positive and -negative patients by genotype (no significant differences)19

Page | 547

Another important use for HBsAg assays is in monitoring treatment response for HBeAg-negative chronic patients with low HBV DNA levels. Sonneveld et al. (2011) reported that when monitoring PEG-IFN treatment in patients with chronic hepatitis B, HBsAg reduction is most pronounced in patients who achieve a response to therapy at 6 months post treatment.19 This suggests that HBsAg quantification can allow for detection of active cases of chronic HBV from true inactive carriers, thereby reducing the need to rigorously monitor HBV DNA levels.

Page | 548

Fig. 4. Hepatitis B surface antigen decline during PEG-IFN treatment of hepatitis B e antigen (HBeAg)-positive and HBeAg-negative patients26

iii. 4.5.3 Diagnostic accuracy of HBeAg test compared to HBV DNA Monitoring HBeAg has been shown to be important due to its association with the disappearance of replicative viral intermediates and its persistence in the blood once HBV DNA has cleared. 19 Using PEG-IFN alfa-2a, Fried et al. (2008) showed that HBeAg levels proved to be a stronger indicator of non-response compared to HBV DNA after 24 weeks of treatment. Lower levels of HBV DNA were seen to closely predict seroconversion (Table 2). Table 2. Serum HBV DNA at weeks 12 and 24 of treatment: relationship to HBeAg seroconversion

Page | 549

It was also shown that those who reached HBeAg seroconversion had a consistent decline in their levels of HBeAg and remained at the lowest levels while under the follow-up period. This was in contrast to those who failed to achieve seroconversion after treatment was discontinued, as a rebound was observed allowing for better determining of seroconversion and a higher negative predictive value (Fig. 5). This highlights the importance of HBeAg in differentiating late responders from non-responders and is an important aspect of treatment.22 Fig. 5. HBV DNA levels: responders versus non-responders at 24 weeks post treatment – HBeAg seroconversion 23

However, monitoring treatment response using HBeAg can be complicated as the response may vary with the therapy used. Non-interferon agents rarely cause HBeAg loss or might cause only a transient HBeAg loss while on therapy. Interferon agents are toxic and might convert ~35% of

Page | 550

HBeAg positives to negatives but only in a subset of people with high alanine aminotransferase (ALT).23 Monitoring HBV treatment response remains a challenge. Guidelines for chronic HBV management and treatment state that the ideal end-point of treatment should be dictated by a lack of detectable HBsAg. The use of HBsAg/HBeAg as a marker to detect sustained virological response is essential, because on-treatment decrease in HBV DNA shows similar patterns for both sustained responders and relapsers (Fig. 3).20 Due to the infrequency of obtaining this point with the current anti-HBV agents, the primary goal of antiviral therapy is defined as viral remission, PCR non-detectability (<300 copies/mL [57 IU/mL]).19,21 For the time being, NAT can be used as the reference standard to confirm this response to therapy. More studies are needed to determine which tests for HBV antigen detection may be useful as effective markers of treatment response for therapeutic agents.

References A. Reference list of studies that met criteria for inclusion in the analysis 1. 2. Larsson R, Eilard A, Malmström S, Hannoun C, Dhillon AP, Norkrans G, et al. HBsAg quantification for identification of liver disease in chronic hepatitis B virus carriers. Liver Int. 2014;34(7):238–45. Perrillo R, Mimms L, Schechtman K, Robbins D, Campbell C. Monitoring of antiviral therapy with quantitative evaluation of HBeAg: a comparison with HBV DNA testing. Hepatology. 1993;18(6):1306– 12.

B. Reference list from background 1. 2. 3. 4. Ott JJ, Stevens GA, Groeger J, Wiersma ST. Global epidemiology of hepatitis B virus infection: new estimates of age-specific HBsAg seroprevalence and endemicity. Vaccine. 2012;30(12):2212–19. Ganem D, Prince AM. Hepatitis B virus infection – natural history and clinical consequences. N Engl J Med. 2004;350(11):1118–29. Fattovich G, Stroffolini T, Zagni I, Donato F. Hepatocellular carcinoma in cirrhosis: incidence and risk factors. Gastroenterology. 2004;127(5 Suppl 1):S35–S50. Lozano R, Naghavi M, Foreman K, Lim S, Shibuya K, Aboyans V, et al. Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380(9859):2095–128. Guidelines for the Prevention, Care, and Treatment of Persons with Chronic Hepatitis B Infection. Geneva: World Health Organization; 2015 (http://apps.who.int/iris/bitstream/10665/154590/1/9789241549059_eng.pdf?ua=1&ua=1, accessed 06 June 2016). Krajden M, McNabb G, Petric M. The laboratory diagnosis of hepatitis B virus. Can J Infect Dis Med Microbiol. 2005;16(2):65–72. Stramer SL, Wend U, Candotti D, Foster GA, Hollinger FB, Dodd RY, et al. Nucleic acid testing to detect HBV infection in blood donors. N Engl J Med. 2011;364(3):236–47.

6. 7.

Page | 551

10. Qin XK, Li P, Han M, Liu JP. [Xiaochaihu Tang for treatment of chronic hepatitis B: a systematic review of randomized trials]. [Article in Chinese]. Zhong Xi Yi Jie He Xue Bao. 2010;8(4):312–20. 11. Hwang SH, Oh HB, Choi SE,Kim HH,Chang CL,Lee EY, et al. [Meta-analysis for the pooled sensitivity and specificity of hepatitis B surface antigen rapid tests]. [Article in Korean]. Korean J Lab Med. 2008;28(2):160–8. 12. Shivkumar S, Peeling R, Jafari Y, Joseph L, Pai NP. Rapid point-of-care first-line screening tests for hepatitis B infection: a meta-analysis of diagnostic accuracy (1980–2010). Am J Gastroenterol. 2012;107(9):1306–13. 13. Khuroo MS, Khuroo NS, Khuroo MS. Accuracy of rapid point-of-care diagnostic tests for hepatitis B surface antigen – systematic review and meta-analysis. J Clin Exp Hepatol. 2014;4(3):226–40. 14. CRD’s guidance for undertaking reviews in health care. York: Centre for Reviews and Dissemination ; 2008. 15. Pai M, Mcculloch M, Gorman, JD, PaiN, Enanoria W, Kennedy G, et al. Systematic reviews and metaanalyses: an illustrated, step-by-step guide. Natl Med J India. 2004;17(2):86–95. 16. Whiting PF, Rutjes AW, Westwood ME, Mallet S, Deeks JJ, Reitsma JB, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155(8):529–36. 17. Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziu PP, Irwig LM, et al. The STARD statement for reporting studies of diagnostic accuracy: explanation and elaboration. 2003;138(1):W1–W12. 18. Larsson SB, Eilard A, Malmström S, Hannoun C, Dhillon AP, Norkrans G, et al. HBsAg quantification for identification of liver disease in chronic hepatitis B virus carriers. Liver Int. 2014;34(7):e238–e45. 19. Perrillo R, Mimms L, Schechtman K, Robbins D, Campbell C. Monitoring of antiviral therapy with quantitative evaluation of HBeAg: a comparison with HBV DNA testing. Hepatology. 1993;18(6):1306– 12. 20. Ganji A, Esmaeilzadeh A, Ghafarzadegan K, Helalat H, Rafatpanah H, Mokhtarifar A. Correlation between HBsAg quantitative assay results and HBV DNA levels in chronic HBV. Hepat Mon. 2011;11(5):342–5. 21. Martinot-Peignoux M, Asselah T, Marcellin P. HBsAg quantification to optimize treatment monitoring in chronic hepatitis B patients. Liver Int. 2015;35(1):82–90. 22. Fried MW, Piratvisuth T, Lau GK, Marcellin P, Chow WC, Cooksley G, et al. HBeAg and hepatitis B virus DNA as outcome predictors during therapy with peginterferon alfa-2a for HBeAg-positive chronic hepatitis B. Heptaology. 2008;47(2):428–34. 23. Viganò M, Lampertico P. ClinicaliImplications of HBsAg quantification in patients with chronic hepatitis B. Saudi J Gastroenterol. 2012;18(2):81–6.

Page | 552

Annex 5.11 PICO 10 - Interventions to optimize the chronic viral hepatitis care continuum: a systematic review and metaanalysis of interventions to improve hepatitis B and C screening, linkage to care, treatment uptake, treatment adherence, and viral suppression

Team Leaders: Kali Zhou, MD (kalizhou26@gmail.com); Thomas Fitzpatrick (tomsfitzpatrick@gmail.com); Nick Walsh, MD PhD (walshn@wpro.who.int)

Team Members: JI Young Kim, Julia Scott, Ying-ru Lo, Xiaoping Tang, Cai Weiping, Joseph Tucker

Corresponding author: Joseph Tucker (jdtucker@med.unc.edu)

September 2015 Page | 553

Abstract Background: Recent advances in hepatitis B virus (HBV) and hepatitis C virus (HCV) therapeutics have ushered in a new era of effectively addressing chronic viral hepatitis. To optimize the real-world effectiveness of these medicines requires engaging and retaining individuals from screening through care and ultimately viral suppression. We carried out a systematic review of operational interventions to enhance chronic viral hepatitis screening, linkage to care, treatment uptake, treatment adherence, and ultimately viral suppression. Methods: The review was registered in PROSPERO (42014015094) and carried out according to PRISMA guidelines. We searched PubMed, EMBASE, the WHO library, Clinicaltrials.gov, International Clinical Trials Registry Platform, Psychinfo, and Cinahl. We included randomized controlled trials (RCT) or controlled non-randomized studies (NRS) targeting one or more steps along the chronic viral hepatitis (HBV, HCV) continuum of care. We used the Cochrane risk of bias tool and GRADE methodology to assess the quality of included studies. Pool data from studies of similar interventions were included in the meta-analyses for each specific step of the care continuum. Results: We identified 7581 citations and included 54 studies. All studies except one were from high-income countries. Studies reported outcomes for chronic viral hepatitis screening, linkage to care, treatment uptake, treatment adherence, or viral suppression. No studies evaluated interventions to improve HBV treatment uptake or treatment adherence. Six randomized controlled trials (RCTs) showed that lay health worker HBV test promotion interventions increased HBV testing rates (RR = 2.68 [1.82–3.93], moderate quality evidence). Two NRS and one RCT found clinician reminders to prompt HCV screening during clinical visits increased HCV testing rates (RR = 3.70 [1.81–7.57], very low-quality evidence). Three RCTs demonstrated that interventions facilitating referral and scheduling to specialist sites increased patient attendance at HCV specialist visits (RR = 1.57 [1.03–2.41], moderatequality evidence). Coordinated care between mental health and treatment specialists along with psychological therapy and counselling for patients with mental health and/or substance use comorbidities increased HCV treatment initiation (OR = 3.03 [1.24–7.37]), improved treatment completion (RR = 1.22 [1.05–1.41]), and increased sustained virological response (SVR) (RR = 1.21 [1.07–1.38]) compared to usual care (very low-quality evidence). Nurse-led therapeutic educational interventions improved treatment completion (RR = 1.14 [1.05–1.23], low-quality evidence) and increased SVR (OR = 1.93 [1.44–2.59], low-quality evidence).

Page | 554

Conclusion: A range of relatively simple, inexpensive operational interventions can significantly improve engagement and retention along the chronic viral hepatitis care continuum. In addition, integrated approaches to hepatitis screening, care and treatment for specific vulnerable populations are effective. In the era of highly effective antiviral therapies, further implementation science research specifically carried out in low and middle-income settings, is needed to optimize engagement and retention for people living with HBV and HCV in the chronic viral hepatitis continuum of care.

Systematic review and meta-analysis manuscript 1. Background Globally, 250 million people are chronically infected with hepatitis B virus (HBV),1 and 80–140 million are chronically infected with hepatitis C virus (HCV)2,3 resulting in 1.45 million annual deaths – the seventh leading cause of mortality worldwide. Chronic HBV and HCV responsible for over 90% of these deaths.4 Recent advances therapeutic in HBV and HCV therapeutics are now providing the impetus for substantial changes in the clinical management of chronic viral hepatitis. Optimizing the high efficacy of these new medicines will require engagement and retention across the care continuum, ranging from initial screening to viral suppression (HBV) or cure (HCV) (Fig. 1). Similar to the HIV continuum of care, each step of the chronic viral hepatitis continuum of care is contingent on the previous steps. This importance of the entire HBV care continuum is underlined in Australia where although 57% of the estimated population living with HBV are diagnosed, only 8% receive viral load testing and only 5% are on treatment.5 Operational interventions may enhance engagement and retention at each step along the continuum of care: screening, linkage to care, treatment uptake, treatment adherence, and viral suppression. Population-level data on the viral hepatitis treatment care continuum for viral hepatitis is limited, though even in high-income countries only a small fraction of the estimated population living with HBV or HCV are ultimately treated and achieve viral suppression.5,6 Indeed, large proportions of people living with viral hepatitis B and C are unaware of their infection, especially those from vulnerable groups and those living in low- and middle-income countries.7,8–10 To investigate the potential of operational strategies to facilitate engagement and retention in the care continuum, we conducted a systematic review to identify interventions in adults living with chronic HBV or HCV infection. We further quantified the effect size of these interventions and highlighted gaps in knowledge on progression through the care continuum.

Page | 555

2. Methods We searched PubMed, EMBASE, WHO library, International Clinical Trials Registry Platform, Psychinfo, and Cinahl for full-text or abstract entries published before 31 December 2014. Accepted scientific conference abstracts and clinical trials registered on Clinicaltrials.gov were also screened. References of articles selected for inclusion were searched for additional citations. Only peer-reviewed English language randomized controlled trials (RCT) or controlled nonrandomized studies (NRS) were included. Where study details were unclear, we contacted authors directly. Details on our search strategy can be found in Supplementary Materials. Briefly, we included studies on operational interventions at any point in the chronic viral hepatitis continuum for people living with diagnosed or undiagnosed chronic viral hepatitis. Only studies where the primary or secondary outcomes were engagement, retention, or progression along the care continuum were included. Exclusion criteria included study designs lacking a comparator or control, dissertations, studies enrolling only pediatric populations, and publications failing to report the outcome data necessary for extraction. Titles, abstracts, and full-texts were sequentially screened for inclusion by two authors independently, with a third author consulted where there was disagreement. Reasons for excluding abstracts and full-texts were recorded. The PRISMA flowchart for included studies is shown in Fig. 2. Data extraction was also performed independently by two reviewers. Differences in the data extracted by the two reviewers was first attempted to be reconciled through discussion. A third reviewer was consulted if disagreement remained. The following variables were extracted: authors, journal of publication, publication year, study design, studied population, inclusion criteria, exclusion criteria, participant characteristics, sample size, study context, intervention description, control description, duration of intervention, results, and conclusions. Data was extracted and analyzed according to an intention-to-treat approach, even if individual authors reported results or conclusions based on per-protocol analyses. Following data extraction, risk of bias was assessed for each included RCT and NRS using the Cochrane Collaboration’s risk of bias tool. Risk of bias was ranked along six domains: selection bias, performance bias, detection bias, attrition bias, reporting bias, and other bias. Detailed risk of bias tables for each stage of the chronic viral hepatitis continuum of care are included in Supplementary Materials. All included publications were assessed for comparability on the basis of intervention type, control condition, and outcome. Studies determined to be similar for intervention, control, and outcome were included in meta-analyses to determine pooled effect size. Pooled risk or odds ratios with confidence intervals and forest plots were generated using a random-effects model in Review Manager 5.3. The degree of heterogeneity between studies in a comparison was assessed by calculating I2. When studies only reported odds ratios with confidence intervals, data was pooled using the generic inverse variance method. If a portion of the studies included in a comparison reported Page | 556

outcomes that had been adjusted using matching or statistical modeling (e.g. regression modelling), a sub-analysis was generated using only adjusted results. Funnel plots were used to screen for reporting bias. The strength of evidence was assessed according to the methodology described by the GRADE working group, and a GRADE table was generated for each meta-analysis and sub-analysis. Where specific interventions were directed at specific populations (e.g. mental health or substance use patients living with hepatitis C) we did not downgrade for indirectness as any recommendation from this analysis would be relevant to that specific population. For imprecision, the pooled sample size for each meta-analysis was compared against the optimal information size (OIS), which was calculated using an alpha of 0.05 and power of 80%. Comparisons determined to be most relevant to current viral hepatitis treatment guidelines were included in the results section. All other comparisons can be found in Supplementary Materials. 3. Results A total of 11 806 titles were identified through database searches, and 19 additional titles were identified through searching article references and contacting authors. After duplicates were removed, 7581 titles were screened according to standard inclusion and exclusion criteria. 469 abstracts were selected for further screening, 353 of which were excluded. Based on the results of abstract screening, 116 articles were selected for full-text review. Ultimately, 54 studies were selected for inclusion in this systematic review,11–64 including 45 full-text publications,11,13,14,16–31,33– 39,41,45–47,49–55,57–64 5 abstracts,12,15,32,40,48 and 4 clinical trials42–44,56 (Table 1). Of the 54 total included studies, 31 were included a meta-analysis that calculated a pooled effect size (Fig. 2). Of the total 54 included studies, 37 reported an outcome along the HCV continuum of care, 15 reported an outcome along the HBV continuum of care, and 2 studies reported outcomes involving both HBV and HCV. Interventions to improve retention along the HBV continuum of care were limited to screening and linkage to care, while interventions to improve retention along the HCV continuum of care addressed all five steps. The most commonly reported outcomes among included studies were HCV treatment adherence (including treatment completion) and HCV viral suppression, with 21 and 20 studies, respectively, followed by HBV and HCV screening, with 15 and 11 studies, respectively. HBV and HCV linkage to care and treatment uptake were comparatively less well studied (Table 1). None of the interventions to improve HBV or HCV screening specifically targeted symptomatic or asymptomatic individuals. All included studies were conducted in high-income countries, except a single study conducted in Turkey.31 Of the included studies, 44.4% (24/54) were RCTs, with seven of those being cluster RCTs, and the remaining 55.5% (30/54) of included studies were NRS. Sample sizes ranged from 21 to 36,987 (Table 1). Thirteen meta-analyses were performed where studies were determined to be similar in terms of intervention, control and outcome so that reported data could be pooled. The number of Page | 557

studies included in a meta-analysis ranged from two to six. The limited number of studies in each comparison prevented us from performing stratified analyses for risk of bias, intervention intensity, and other relevant factors. Funnel plots did not detect reporting bias for any of the 13 meta-analyses. Nearly half (7/15) of the interventions to improve HBV screening were lay health worker (LHW) HBV test promotion interventions.16,29,38,53–55,64 Six of the seven LHW-led interventions were one-time activities that delivered educational content tailored to a particular community’s cultural and social context.16,29,53–55,64 All six studies targeted Asian communities in the United States or Canada. Results from these six studies were pooled in a meta-analysis. Self-reported HBV testing rates were higher among groups that received a single LHW educational intervention to improve HBV knowledge and promote testing compared to groups that received no or unrelated educational interventions (RR = 2.68, CI 95; 1.82–3.93). All studies found LHW-led interventions had a positive effect on HBV screening; however, these results were moderately heterogeneous (I2 = 56%). This heterogeneity in effect size may be due to differing study design (three were RCTs randomized by cluster, and three were RCTs randomized by individual) and intervention setting (three were home visits, and three were delivered at a community-based organization). Unlike interventions to improve HBV screening, which were primarily delivered in community settings, all 11 of the interventions to improve HCV screening either targeted healthcare providers or took place at an established health-care facility.18,19,21,23–25,32,36,41,50,51 At this stage in the HCV continuum of care, the two most common interventions were clinician reminders to prompt HCV screening during clinical visits (three studies)21,32,36 and pre-test counselling with onsite HCV testing at a health-care facility serving high-risk populations (three studies).19,41,50 Clinician reminders to prompt HCV screening during clinical visits consistently increased HCV testing rates compared to no clinician reminders (RR = 3.70, CI 95; 1.81–7.57). While all three studies found clinician reminders to have a positive effect on HCV screening rates, there was a large degree of heterogeneity between reported effect sizes (I2 = 99%). Providers were prompted by reminders to order HCV tests if patients belonged to a high-risk birth cohort (one study),32 reported risk behaviour (one study),21 or both (one study).36 Two studies used physical reminder stickers attached to patient charts,21,36 while one study incorporated reminders into an electronic medical records system.32 All three studies examined patients seeing primary care providers in New York City clinics. Three interventions had a facilitated referral component where staff at a site of established care actively assisted HCV+ patients with a history of substance use in scheduling specialist visits.19,39,50 Interventions that provided facilitated referral increased patient attendance to HCV specialist visits compared to no facilitated referral in all three studies (RR = 1.57, CI 95; 1.03–2.41). However, there was significant heterogeneity between reported effect sizes (I2 = 74%). Some of the interventions in this meta-analysis provided patient education and case management at varying degrees of intensity in addition to facilitated referral, which may partially explain why the three interventions reported widely varying effect sizes. Page | 558

In certain contexts HCV+ patients had been deemed ineligible for HCV treatment because of ongoing mental health and/or substance use comorbidities. Individually tailored mental health counselling and motivational therapy to treat mental health and/or substance use issues increased the number of patients who were referred eligible to treatment compared to usual care (OR = 3.43, CI 95; 1.81–6.49). There was little heterogeneity in this meta-analysis (I2 = 0%) despite the fact standards for treatment eligibility differed between the two included studies.22,30 Along the treatment uptake, treatment adherence, and viral suppression steps of the HCV continuum of care, six interventions provided “integrated” or “multidisciplinary” care.11,20,30,46,62,63 These interventions involved regular contact between mental health and specialist treatment providers throughout treatment, and also arranged regular psychological therapy and counselling for patients with mental health and/or substance use comorbidities. Some interventions also provided varying degrees of patient education and case management before and during treatment. Coordinated care between mental health and treatment specialists along with psychological therapy and counselling for patients with mental health and/or substance use comorbidities increased HCV treatment initiation (OR = 3.03, CI 95; 1.24–7.37), improved treatment completion (RR = 1.22, CI 95; 1.05–1.41), and increased SVR (RR = 1.21, CI 95; 1.07– 1.38) compared to usual care. Studies included in these meta-analyses differed in terms of additional services provided beyond coordinated care, percentages of patients with mental health and/or substance use comorbidities, and patient genotype. However, little heterogeneity in effect size was found for treatment adherence and viral suppression. An additional six studies investigated the impact of educational activities about HCV infection, treatment, side-effects of therapy, and the importance of treatment adherence for HCV+ patients beginning or maintained on interferon-based therapy.14,33,37,42,48,52 Nurse-led therapeutic educational sessions were found to improve treatment completion (RR = 1.14, CI 95; 1.05–1.23) and increase SVR (OR = 1.93, CI 95; 1.44–2.59). Meta-analyses for interventions to improve HBV screening Single culturally tailored LHW educational session to improve HBV knowledge and promote testing vs no or unrelated educational session for self-reported HBV screening.

Page | 559

Quality assessment No. of Study design studies Risk of bias Inconsistency Indirectness Imprecision Other considerations

No. of patients Single LHW educational session No or unrelated educational session Relative (95% CI)

Effect Absolute (95% CI)

Quality

HBV screening 6 Randomized Serious trials 1

Not serious

None

255/1344 (19.0%)

92/1413 (6.5%)

RR 2.68 (1.82–3.93)

109 more per 1000 (from 53 more to 191 more)

⨁ ⨁ ⨁ ◯ Moderate

6.6%

110 more per 1000 (from 54 more to 192 more)

II. Meta-analyses for interventions to improve HCV screening Clinician reminder to prompt HCV screening during clinical visits with or without supplementary provider education vs no clinician reminder for HCV screening.

Page | 560

Quality assessment No. of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations

No. of patients Clinical testing reminder No reminder Relative (95% CI)

Effect Absolute (95% CI)

Quality

HCV screening 3 Other 1 design Serious 2

Serious

Not serious

Serious

None

5185/33253 (15.6%)

976/19694 (5.0%)

RR 3.70 (1.81–7.57)

134 more per 1000 (from 40 more to 326 more) 161 more per 1000 (from 48 more to 393 more)

⨁ ◯◯◯ very low

6.0%

1. 2.

This meta-analysis includes 1 cluster RCT and 2 NRS. Drainoni (2012) is at high risk of performance bias and did not employ methods to adjust for confounding potentially introduced by its non-randomized study design. Krauskopf (2014) did not report comparability of randomized clusters and therefore was at high risk of bias. All included studies report a risk ratio >1.0. However, I² = 99%. The high degree of heterogeneity may be due to differences between HCV screening algorithms used in each intervention. Although the pooled sample size exceeds the calculated OIS, the confidence interval is wide. Additionally, Krauskopf (2014) was a cluster RCT that did not account for clustering in its analysis. Consequently, this metaanalysis commits a unit-of-analysis error and produces overprecise results. No ICC was reported, so statistical methods could not be used to reduce the effective sample size of the cluster RCT. All included studies report a risk ratio >2.0. However, the pooled results have not been upgraded for large effect because the non-randomized design of 2/3 studies introduces a significant possibility of confounding.

3. 4.

Page | 561

Quality assessment No. of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations

No. of patients Facilitated referral to specialist visit at site of established care No facilitated referral Relative (95% CI)

Effect Absolute (95% CI)

Quality

Attended HCV specialist visit 3 Randomized Not 1 trials serious Serious 2

Not serious

None

151/243 (62.1%)

72/194 (37.1%)

RR 1.57 (1.03– 2.41)

212 more per 1000 ⨁ ⨁ ⨁ ◯ (from 11 more to Moderate 523 more) 212 more per 1000 (from 11 more to 525 more)

37.2%

Rosenberg (2010) relied on self-reported HCV status and self-reported attendance to an HCV specialist visit, putting the study at high risk of detection bias. However, because this study had a relatively small sample size it was not judged to put the entire metaanalysis at high risk of bias. I² = 85%. This high degree of heterogeneity may be due to differences between the intensity of interventions in the included studies. The confidence interval is not wide. The OIS was calculated to be 124, and the pooled sample size exceeded the OIS.

2. 3.

Unadjusted results

Adjusted results

Page | 562

Quality assessment No. of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations

No. of patients Individually tailored mental health counselling and motivational therapy Usual care Relative (95% CI)

Effect Absolute (95% CI)

Quality

Physician referral to initiate treatment 2 Other design1 Serious 2

Not serious

None

66/120 (55.0%)

35/130 (26.9%)

RR 2.04 (1.48– 2.80)

280 more per 1000 (from 129 more to 485 more) 263 more per 1000 (from 121 more to 455 more)

⨁ ◯◯◯ Very low

25.3%

Adjusted physician referral to initiate treatment 2 Other design1 Serious5 Not serious Not serious3 Serious6 None –/120 –/165 OR 3.43 (1.81– 6.49) 0 fewer per 1000 (from 0 fewer to 0 fewer) ⨁ ◯◯◯ Very low

1. 2.

Evon (2011) is a RCT, while Knott (2006) is a NRS. Knott (2006) is at high risk of detection bias because the outcome was subjective and determined by the physician overseeing treatment who was not blinded. Unadjusted results from Knott (2006) were used in this meta-analysis that did not employ methods to adjust for confounding potentially introduced by its non-randomized study design. The decision to not downgrade for indirectness assumes guidelines are applied to other contexts where mental health or substance use comorbidities are also contraindications to recommending HCV+ patients for treatment. The confidence interval is not wide. The OIS was calculated to be 94, and the pooled sample size exceeded the OIS. Knott (2006) is at high risk of detection bias because the outcome was subjective and determined by the physician overseeing treatment who was not blinded. The confidence interval for the pooled adjusted outcomes is wide.

3. 4. 5. 6.

IV. Meta-analyses for interventions to improve HCV treatment initiation Coordinated care between mental health and treatment specialists with psychological therapy and counselling for patients with mental health and/or substance use comorbidities vs usual care for treatment initiation

Unadjusted results

Page | 563

Adjusted results

Quality assessment No. of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations

No. of patients Coordinated care with psychological therapy and counselling Usual care Relative (95% CI)

Effect Absolute (95% CI)

Quality

Treatment initiation 3 Other design1 Serious2 Serious3 Not serious Serious4 None 116/253 (45.8%) 94/263 (35.7%) RR 1.36 129 more per (0.94–1.97) 1000 (from 21 fewer to 347 more) 194 more per 1000 (from 32 fewer to 522 more) ⨁ ◯◯◯ Very low

53.9%

Adjusted treatment initiation 2 Other design5 Not serious Serious6 Not serious Serious7 None –/252 –/295 OR 3.03 (1.24– 7.37) 0 fewer per 1000 (from 0 fewer to 0 fewer) ⨁ ◯◯◯ very low

1. 2. 3.

Ho (2015) is a RCT, while Ahmed (2013) and Knott (2006) are NRS. Unadjusted results from Ahmed (2013) and Knott (2006) were used in this meta-analysis that did not employ methods to adjust for confounding potentially introduced by their non-randomized study design. I² = 77%. This high degree of heterogeneity may be due to differences between the populations under investigation. Ahmed (2013) included general HCV+ patients, while Ho (2015) and Knott (2006) only included patients with mental health and/or substance use comorbidities. Interventions also differed between studies. Ho (2015) provided additional case management, and Ahmed (2013) provided participants with therapeutic education and community support programmes in addition to coordinated care and psychological support. The confidence interval is not wide. However, the OIS was calculated to be 741, and the pooled sample size did not meet this threshold. Ho (2015) is a RCT, while Knott (2006) is a NRS. I² = 74%. This high degree of heterogeneity may be due to differences between the interventions under investigation. Ho (2015) provided case management in addition to coordinated care and psychological support. The confidence interval for pooled adjusted results is wide.

4. 5. 6. 7.

V. Meta-analyses for interventions to improve HCV-sustained virological response Coordinated care between mental health and treatment specialists with psychological therapy and counselling for patients with mental health and/or substance use comorbidities vs usual care for SVR

Page | 564

Quality assessment No. of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations

No. of patients Coordinated care with Usual care psychological therapy and counselling

Effect Relative (95% CI) Absolute (95% CI)

Quality

SVR 5 Other design Serious1 Not serious Not serious Not serious2 None 330/514 (64.2%) 168/332 (50.6%) RR 1.21 (1.07–1.38) 106 more per 1000 (from 35 more to 192 more) 86 more per 1000 (from 29 more to 156 more) ⨁ ◯◯◯ Very low

41.2%

Curcio (2010) is at high risk of performance bias because of differences between the treatment received by the two cohorts besides the intervention under examination. Knott (2006) did not employ methods to adjust for confounding potentially introduced by its non-randomized study design. The confidence interval is not wide. The OIS was calculated to be 434, and the pooled sample size exceeded the OIS.

Nurse-led therapeutic educational sessions with information on HCV infection, treatment, sideeffects, and/or adherence vs no therapeutic education for SVR

Page | 565

Unadjusted results

Adjusted results Quality assessment No. of studies SVR 6 Other design1 Serious2 Serious3 Not serious Not serious4 None 491/844 (58.2%) 452/913 (49.5%) RR 1.23 (1.03–1.47) 114 more per 1000 ⨁ ◯◯◯ (from 15 more to 233 Very low more) 115 more per 1000 (from 15 more to 235 more) Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations No. of patients Therapeutic education No specific education Relative (95% CI) Effect Absolute (95% CI) Quality

50.0%

Adjusted SVR 4 Other design5 Not serious6 Not serious Not serious Not serious None –/651 –/809 OR 1.93 (1.44– 2.59) 2 fewer per 1000 (from 1 fewer to 3 fewer) ⨁ ⨁ ◯◯ Low

1. 2.

Larrey (2011) is a RCT. Cacoub (2008), Lubega (2013), Renou (2009), Tait (2009), and Merck (2007) are NRS. Tait (2009) and Merck (2007) did not employ methods to adjust for confounding potentially introduced by nonrandomized study design. Tait (2009) and Cacoub (2008) are both at high risk of performance bias, and Merck (2007) is at high risk of attrition bias.

Page | 566

3. 4. 5. 6.

Tait (2009) reports a risk ratio that is <1.0 while all other included studies report a risk ratio >1.0. Additionally, I² = 68%. The confidence interval is not wide. The OIS was calculated to be 1029, and the pooled sample size exceeded the OIS. Larrey (2011) is a RCT. Cacoub (2008), Lubega (2013), and Renou (2009) are NRS. Cacoub (2008) is at high risk of performance bias because there was no standardization of intervention procedures. However, this risk of bias is not sufficient to downgrade the quality of evidence.

4. Discussion This systematic review and meta-analysis demonstrated that operational interventions can improve engagement, retention, or progression through the chronic viral hepatitis care continuum. Specifically, LHW HBV screening promotion activities increased HBV test uptake, while clinician reminders to prompt HCV screening during clinical visits increase HCV testing rates. Coordinated care between hepatitis and mental health specialists along with psychological therapy and counselling for patients with mental health and/or substance use comorbidities can increase HCV treatment initiation, improve treatment completion, and result in higher SVR rates. Our review covered all 5 major steps along the continuum of care for HBV and HCV. Previous systematic reviews examining progression along the care continuum have either limited their analysis to specific components of the care continuum,65 or limited analysis to either HBV or HCV.66 Additionally, these reviews did not focus on studies that had comparison groups. We used GRADE methodology to rigorously evaluate the quality of reported evidence and our data substantially extend the 2012 NICE guidelines – Hepatitis B and C: ways to promote and offer testing to people at increased risk of infection. Our results demonstrate that culturally appropriate lay health workers educational programs to promote HBV testing are effective. All six studies had consistently favorable results with relatively strong study designs and were graded as moderate quality evidence. Although all these interventions were conducted among Asian immigrant populations in the US, this intervention may be relevant in a range of other settings. Our findings are consistent with the growing body of evidence demonstrating that lay health workers can effectively perform a range of interventions that would otherwise be undertaken by trained medical personnel. The lay health workers in the six studies received training in order to help tailor the educational intervention and this training component was relatively simple and of low cost. Qualitative investigation supports these types of interventions as being feasible and acceptable to both those individuals screened and lay health workers themselves.67 While the importance of cultural context in developing interventions to improve HBV screening in high-risk populations and the usefulness of community settings in the delivery of healthcare is important, the low-cost nature of this intervention could facilitate its used in resource limited settings. Task shifting to LHW are well documented as strengthening service delivery capacity in a variety of clinical settings in low- and middle-income countries.68–71 Our results show that clinician reminders to prompt HBV and HCV screening during clinical visits increased HCV testing rates. While of obvious use in electronic medical records, one included study used a clinical “risk screening” sticker placed on a print charts.21 Clinician reminders are consistent with the broader shift towards standardizing clinical practice, including provider initiated screening and systems-based approaches to improving clinical outcomes. While this style of intervention does not operate through the lens of addressing patient barriers, implementation is Page | 567

relatively easy and similar systems have demonstrated effectiveness in multiple disease modalities, such as breast72 and colorectal cancer screening.73 We found that coordinating care between mental health and treatment specialists along with psychological therapy and counselling for patients with mental health and/or substance use comorbidities was effective in promoting HCV treatment initiation, treatment completion, and achieving SVR. Hepatitis C disproportionally affects individuals with comorbid mental health or substance use issues. Traditionally, services for hepatitis, mental health and addiction have been provided by separate clinicians or teams often located in different health facilities. This may contribute to HCV treatment dropout and/or treatment failure.74 While the interventions addressing multidisciplinary or integrated care in this review were diverse, a likely key contributor to improved outcomes was co-location and coordination of services. Integrating HCV screening and treatment with mental health and addiction services is feasible and acceptable to the targeted clients.75,76 Our funding also builds on the limited literature regarding integration of HIV and mental health services, which can also improve treatment outcomes.77 We identified substantial gaps in current knowledge examining progress along the chronic viral hepatitis continuum of care. Implementation science in viral hepatitis will become increasingly important as access to effective HBV and HCV medicines expands across the world. High quality evidence provides a strong basis for forming guidelines recommendations for program managers, clinicians and others working in the field. Most of our included studies were graded low or very low. The lack of studies with robust design in particular in HCV screening and HBV treatment are a significant gap. Our analysis found no studies investigating HBV treatment uptake, adherence, or viral suppression. While we did not provide an economic analysis of the value of incorporating operational interventions in the viral hepatitis care continuum, mathematical modelling in HCV suggests that imperfect follow-up during the HCV care continuum greatly reduces the real-world effectiveness of HCV therapy.78 It follows that interventions impacting on multiple steps along the care continuum are more resource efficient. Future research should focus quantifying the costs and effectiveness of elements or combinations of interventions to optimize treatment outcome. There are several limitations to our review. First, outcomes that were studied were intermediate outcomes related to diagnosis and treatment, not disease end-points such as morbidity and mortality associated with HBV and HCV. However, it is well known that treatment of HBV or HCV infection reduces liver-related deaths, hepatocellular carcinoma incidence and all-cause mortality.79,80 Second, almost all studies addressing treatment uptake, adherence, and viral suppression in HCV were carried out with interferon-based therapies. Current DAA-based regimens are simpler to administer, more effective, and better tolerated. In the era of DAAs, with near 100% efficacy, retention and progression along the care continuum is likely to become an important determinant of achieving SVR. Finally, all included studies were carried out in high-income settings, with the exception of one study from Turkey. More implementation science research is needed in low- and middle-income contexts where the majority of people living with chronic viral hepatitis live. Our systematic review demonstrates that a range of relatively simple, inexpensive operational interventions can significantly improve engagement and retention along the chronic viral hepatitis care continuum. In addition, we identified the importance of integrated approaches to hepatitis screening, care and treatment for specific vulnerable populations. Further implementation Page | 568

science research, robust in design and specifically carried out in low and middle-income settings, is needed to evaluate current gaps in our knowledge to improve engagement and retention for people living with HBV and HCV in the chronic viral hepatitis continuum of care.

Page | 569

1. Figures and tables

People living with undiagnosed asymptomatic chronic viral hepatitis

Screening

Patients who have received a positive HBVsAg or HCV Ab test result

Linkage to care

Patients who have received a positive viral load confirmatory test and liver disease staging

Treatment uptake

Patients who have initiated treatment for chronic viral hepatitis

Treatment adherence Patients who have completed HCV treatment or Patients who are maintained on HBV treatment

Viral suppression

Patients who have achieved HCV SVR or Patients who have achieved HBV virologic suppression

Fig. 1. Overview of the stages comprising the viral hepatitis treatment continuum, including testing, linkage to care, enrolment in care, treatment uptake, treatment adherence, and viral suppression.

Page | 570

Identification

Titles identified through database searching (n = 11,806)

Titles identified through other sources (n = 19)

Titles screened after duplicates removed (n = 7,581)

Screening

Abstracts screened (n = 469)

Abstracts excluded (n = 353) Wrong methodology (119) Review article, editorial, or no results (107) Wrong outcome (97) Wrong population (30) Full-text articles excluded (n = 62) No control (17) Missing necessary data/results (12) Duplicate abstract of publication (9) Non-English language (7) National-level campaign (4) Only identifies risk factors/associations (4) Outcome not along cascade (3) No intervention (2) Only describes state of hepatitis management (2) Dissertation/thesis (2)

Eligibility

Full-text articles assessed for eligibility (n = 116)

Included

Studies included in qualitative synthesis (n = 54) Full-text publication (45) Abstracts (5) Clinical trial (4)

Studies excluded from quantitative synthesis because of unique outcomes or intervention types (n = 23) Studies included in quantitative synthesis (meta-analysis) (n = 31)

Fig. 2. PRISMA flow diagram outlining study selection for this systematic review of interventions to optimize retention across the chronic viral hepatitis continuum of care.

Page | 571

Page | 572

Page | 573

Table 1. Summary table of characteristics of included studies

5. Supplementary materials 1. 1. Additional comparisons, forest plots and GRADE tables

New institutional testing protocol for at-risk populations with education and testing promotion activities for providers vs. previous standard of care for HBV screening

Quality assessment No. of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations

No. of patients New institutional testing protocol and provider education Previous standard of care

Effect Relative (95% CI) Absolute (95% CI)

Quality

HBV screening 3 Beforeafter studies Serious1 Serious2 Not serious Serious3 None4 18391/18526 (99.3%) 3126/19021 (16.4%) RR 3.77 (2.04– 6.97) 455 more ⨁ ◯◯◯ per 1000 Very low (from 171 more to 981 more) 456 more per 1000 (from 171 more to 982 more)

16.4%

1. 2.

3/3 included studies did not employ methods to adjust for confounding potentially introduced by non-randomized study design. All included studies report a risk ratio >1.0. However, I² = 91%. The high degree of heterogeneity may be due to the different at-risk populations under investigation between studies (pregnant women, patients starting chemotherapy, and patients with substance use comorbidities) or differences in the intensity of interventions between studies.

Page | 574

3. 4.

Although the pooled sample size exceeds the calculated OIS, the confidence interval is wide. All included studies report a risk ratio >2.0. However, the pooled results have not been upgraded for large effect because the non-randomized design of the included studies introduces a significant possibility of confounding.

HCV educational sessions, pretest counselling, and on-site testing at healthcare facilities serving high-risk populations vs. no educational and counselling sessions for HCV screening

Quality assessment No. of Study design studies Risk of bias Inconsistency Indirect Imprecision ness Other considerations

No. of patients Facility-based No education educational and and counselling counselling sessions Relative (95% CI)

Effect Absolute (95% CI)

Quality

HCV screening 3 Randomized trials Not serious1 Serious2 Not Serious3 serious None 250/383 (65.3%) 162/358 (45.3%) RR 2.15 (0.80– 5.79) 520 more per 1000 ⨁ ⨁ ◯◯ (from 91 fewer to Low 1000 more) 312 more per 1000 (from 54 fewer to 1000 more)

27.2%

1. 2.

Rosenberg (2010) is at high risk of detection bias because the control group outcome was self-reported. The introduced systematic bias would not necessarily exaggerate the reported effect size, so the quality of evidence was not downgraded. One reported risk ratio is at 1.0 while two are > 1.0, and I² = 96%. The high degree of heterogeneity may be due to different study contexts. Cullen (2006) and Rosenberg (2010) examined interventions at facilities where patients had established care. Merchant (2014) examined an intervention at a hospital emergency department. Additionally, Merchant (2014) offered onsite testing to both intervention and control groups, while the other included studies did not. Although the pooled sample size exceeds the calculated OIS of 193, the confidence interval is wide. Additionally, Cullen (2006) was a cluster RCT that did not account for clustering in its analysis. Consequently, this meta-analysis commits a unitof-analysis error and produces over-precise results. No ICC was reported, so statistical methods could not be used to reduce the effective sample size of the cluster RCT.

Coordinated care between mental health and treatment specialists with psychological therapy and counselling for patients with mental health and/or substance use comorbidities vs usual care for treatment completion

Page | 575

Quality assessment No. of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations

No. of patients Coordinated care with psychological therapy and counselling Usual care Relative (95% CI)

Effect Absolute (95% CI)

Quality

Treatment completion 4 Other 1 design Serious2 Not serious Not serious Serious3 None 144/214 (67.3%) 103/185 (55.7%) RR 1.22 122 more per (1.05– 1.41) 1000 (from 28 more to 228 more) 115 more per 1000 (from 26 more to 215 more) ⨁ ◯◯◯ Very low

52.5%

1. 2.

Neri (2010) and Ho (2015) are RCTs. Knott (2006) and Curcio (2010) are NRS. Curcio (2010) was at high risk of performance bias because intervention and control treatment were delivered at significantly different institutions. Knott (2006) did not employ methods to adjust for confounding potentially introduced by its non-randomized study design. The confidence interval is not wide. However, the OIS was calculated to be 553, and the pooled sample size did not meet this threshold.

Page | 576

Nurse-led therapeutic educational sessions with information on HCV infection, treatment, side effects, and/or adherence vs no therapeutic education for treatment adherence

Nurse-led therapeutic educational sessions with information on HCV infection, treatment, sideeffects, and/or adherence vs no therapeutic education for treatment completion

Quality assessment No. of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations

No. of patients Therapeutic education No specific education Relative (95% CI)

Effect Absolute (95% CI)

Quality

Treatment adherence 3 Cohort studies Serious1 Serious2 Not serious Serious3 None 215/490 (43.9%) 157/401 (39.2%) 31 more per 1000 0.87–1.34) (from 51 fewer to 133 more) RR 1.08 47 more per 1000 (from 77 fewer to 200 more) ⨁ ◯◯◯ Very low

59.0%

Treatment completion 4 Other design4 Serious5 Not serious Not serious Not serious6 None 321/414 (77.5%) 380/551 (69.0%) RR 1.14 (1.05 to 1.23) 97 more per 1000 ⨁ ◯◯◯ (from 34 more to 159 Very low more) 95 more per 1000 (from 34 more to 157 more)

68.2%

3/3 studies did not employ methods to adjust for confounding potentially introduced by their non-randomized study design. Cacoub (2008) was also at high risk of performance bias because the intervention was not standardized, and at high risk of detection bias because the outcome was assessed through self-report. Included studies report risk ratios on both sides of 1.0 and I² = 64%.

Page | 577

3. 4. 5.

The confidence interval is not wide. However, the OIS was calculated to be 3,481, and the pooled sample size did not meet this threshold. Larrey (2011) is a RCT. Merck (2007), Tait (2009) and Renou (2009) are NRS. Tait (2009) and Merck (2007) both did not employ methods to adjust for confounding potentially introduced by their non-randomized study design. Additionally, Tait (2009) was at high risk of performance bias due to study design. The confidence interval is not wide. The OIS was calculated to be 861, and the pooled sample size exceeded the OIS.

Directly observed interferon therapy vs self-administered interferon therapy for SVR

Unadjusted results

Adjusted results

Quality assessment No. of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations

No. of patients Directly observed therapy Self-administered therapy Relative (95% CI)

Effect Absolute (95% CI)

Quality

SVR

Page | 578

Quality assessment No. of studies Study design Risk of bias Inconsistency Indirectness Imprecision Other considerations

No. of patients Directly observed therapy 63/133 (47.4%) Self-administered therapy Relative (95% CI)

Effect Absolute (95% CI)

Quality

Other design

Serious

Not serious

Serious

None

35/86 (40.7%)

RR 1.14 (0.84–1.56)

57 more per 1000 ⨁ ◯◯◯ (from 65 fewer to Very low 228 more) 47 more per 1000 (from 53 fewer to 187 more)

33.3%

Adjusted SVR 3 Other design1 Serious4 Not serious Not serious Serious5 None –/133 –/86 OR 1.49 (0.72–3.08) 1 fewer per 1000 ⨁ ◯◯◯ (from 1 fewer to 3 Very low fewer)

1. 2.

Bruce (2012) and Bonkovsky (2008) are RCTs. Cioe (2013) is a NRS. Bruce (2012) is at high risk of attrition bias and reporting bias. Bruce (2012) also had significant baseline differences between the intervention and control groups but methods were not used to adjust for potential confounding. Unadjusted results from Cioe (2013) were used in this meta-analysis that did not employ methods to adjust for confounding potentially introduced by its non-randomized study design. The confidence interval is not wide. However, the OIS was calculated to be 1801, and the pooled sample size did not meet this threshold. Bruce (2012) is at high risk of attrition bias and reporting bias. Bruce (2012) also had significant baseline differences between the intervention and control groups but methods were not used to adjust for potential confounding. The confidence interval for the pooled adjusted results is wide.

3. 4. 5.

Page | 579

Risk of bias tables Risk of bias assessment for included studies with interventions for HBV screening

Risk of bias assessment for included studies with interventions for HCV screening

Page | 580

Risk of bias assessment for included studies with interventions for HBV linkage to care

Risk of bias assessment for included studies with interventions for HCV linkage to care

Risk of bias assessment for included studies with interventions for treatment uptake

Page | 581

Risk of bias assessment for included studies with interventions for treatment adherence and completion

Page | 582

Risk of bias assessment for included studies with interventions for viral suppression

2. 3. Domain

Cochrane risk of bias tool rubric Support for judgement Review authors’ judgement

Selection bias Random sequence generation Describe the method used to generate the allocation sequence in sufficient detail to allow an assessment of whether it should produce comparable groups. Describe the method used to conceal the allocation sequence in sufficient detail to determine whether intervention allocations could have been foreseen in advance of, or during, enrolment. Selection bias (biased allocation to interventions) due to inadequate generation of a randomized sequence.

Allocation concealment

Selection bias (biased allocation to interventions) due to inadequate concealment of allocations prior to assignment

Performance bias Blinding of participants and Describe all measures used, if any, to blind Performance bias due to knowledge of

Page | 583

Domain personnel: Assessments should be made for each main outcome (or class of outcomes)

Support for judgement study participants and personnel from knowledge of which intervention a participant received. Provide any information relating to whether the intended blinding was effective.

Review authors’ judgement the allocated interventions by participants and personnel during the study

Detection bias Blinding of outcome assessment: Assessments should be made for each main outcome (or class of outcomes) Describe all measures used, if any, to blind outcome assessors from knowledge of which intervention a participant received. Provide any information relating to whether the intended blinding was effective. Detection bias due to knowledge of the allocated interventions by outcome assessors

Attrition bias Incomplete outcome data: Assessments should be made for each main outcome (or class of outcomes) Describe the completeness of outcome data for each main outcome, including attrition and exclusions from the analysis. State whether attrition and exclusions were reported, the numbers in each intervention group (compared with total randomized participants), reasons for attrition/exclusions where reported, and any re-inclusions in analyses performed by the review authors. Attrition bias due to amount, nature or handling of incomplete outcome data

Reporting bias Selective reporting State how the possibility of selective outcome reporting was examined by the review authors and what was found. Reporting bias due to selective outcome reporting

Other bias Other sources of bias State any important concerns about bias not addressed in the other domains in the tool. If particular questions/entries were prespecified in the review’s protocol, responses should be provided for each question/entry. Bias due to problems not covered elsewhere in the table

Page | 584

PICO table PICO P Individuals living with chronic hepatitis B or C (diagnosed or undiagnosed) or providers caring for these patients. Operational interventions delivered in conjunction with screening, care, or treatment of hepatitis Standard of care or no intervention Retention and progression along the continuum of care

C O

Search strategy Potential search terms: Population: Hepatitis B OR HBV OR CHB Hepatitis C OR HCV OR CHC Chronic viral hepatitis Intervention: Intervention Counselling Education or educate Teach Training Program Engagement Alcohol and reduce, reduction, cessation Outcome: Screen OR screened OR screening Test OR tested OR testing Linking OR linkage Refer OR referral Uptake Retain OR retained OR retention Adherence OR adhere Compliance OR comply

PubMed search strategy: Page | 585

(Hepatitis B OR HBV OR CHB[tiab] OR Hepatitis C OR HCV OR CHC[tiab] OR chronic viral hepatitis[tiab] OR chronic viral hepatitis[mh]) AND (Intervention[tiab] OR counselling[tiab] OR education[tiab] OR educate[tiab] OR teach[tiab] OR training[tiab] OR program[tiab] OR Engagement[tiab] OR (alcohol[tiab] AND (reduce OR reduction OR cessation OR decrease)) AND (Uptake[tiab] OR Adherence[tiab] OR adhere[tiab] OR Compliance[tiab] OR comply[tiab] OR retain[tiab] OR retained[tiab] OR Retention[tiab] OR Screen[tiab] OR screened[tiab] OR screening[tiab] OR test[tiab] OR tested[tiab] OR testing[tiab] OR Linkage[tiab] OR linking[tiab] OR refer[tiab] OR Referral[tiab]) Additional information on study selection, inclusion criteria and exclusion criteria Authors of included abstracts were contacted to determine whether the same data had been later published as a full-text article in a peer-reviewed journal, in which case the abstract would be excluded and full-text article included. Study selection proceeded in three stages. First, two reviewers screened titles obtained from the initial search strategy according to standard inclusion and exclusion criteria. Second, abstracts for all titles identified for further review were assessed independently by two reviewers for inclusion. If there was disagreement, a third reviewer determined final inclusion. Finally, full texts for all abstracts identified for further review were assessed independently by two reviewers for inclusion. If there was disagreement, a third reviewer determined final inclusion.

References 1. 2. 3. Ott JJ, Stevens GA, Groeger J, Wiersma ST. Global epidemiology of hepatitis B virus infection: new estimates of age-specific HBsAg seroprevalence and endemicity. Vaccine. 2012;30(12):2212–19. Gower E, Estes C, Blach S, Razavi-Shearer K, Razavi H. Global epidemiology and genotype distribution of the hepatitis C virus infection. J Hepatol. 2014;61(1 Suppl):S45 –S57. Hanafiah KM, Groeger J, Flaxman AD, Wiersma ST. Global epidemiology of hepatitis C virus infection: new estimates of age-specific antibody to hepatitis C virus seroprevalence. Hepatology. 2012;57(4):1333–42. Global Burden of Disease Study C. Global, regional, and national incidence, prevalence, and years lived with disability for 301 acute and chronic diseases and injuries in 188 countries, 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet. 2015;386(9995):743 –800. Allard NL, MacLachlan JH, Cowie BC. The cascade of care for Australians living with chronic hepatitis B: measuring access to diagnosis, management and treatment. Aust N Z J Public Health. 2015;39(3):255 – 9. Yehia BR, Schranz AJ, Umscheid CA, Lo Re V, 3rd. The treatment cascade for chronic hepatitis C virus infection in the United States: a systematic review and meta-analysis. PLoS One. 2014;9(7):e101554. Brouard C, Le Strat Y, Larsen C, Jauffret-Roustide M, Lot F, Pillonel J. The undiagnosed chronicallyinfected HCV population in France. Implications for expanded testing recommendations in 2014. PLoS One. 2015;10(5):e0126920. Hagan LM, Wolpe PR, Schinazi RF. Treatment as prevention and cure towards global eradication of hepatitis C virus. Trends Microbiol. 2013;21(12):625–33. Viner K, Kuncio D, Newbern EC, Johnson CC. The continuum of hepatitis C testing and care. Hepatology. 2015;61(3):783–9.

6. 7.

8. 9.

Page | 586

10. 11.

Dore GJ, Ward J, Thursz M. Hepatitis C disease burden and strategies to manage the burden (Guest Editors Mark Thursz, Gregory Dore and John Ward). J Viral Hepat. 2014;21 Suppl 1:1 –4. Ahmed I, Habibi AN, Iqbal J, Niaz Z, Naqvi AA. Improving outcome in hepatitis C management: A need for dedicated multi-disciplinary service to improve compliance with treatment. J Gastroenterol Hepatol. 2013;2(8):737–9. Asthana AK, Choong J, Lubel JS. Education does not improve hepatitis B screening uptake in those receiving cytotoxic chemotherapy-time for alternative strategies. J Gastroenterol Hepatol. 2012;27:162. Bruce RD, Eiserman J, Acosta A, Gote C, Lim JK, Altice FL. Developing a modified directly observed therapy intervention for hepatitis C treatment in a methadone maintenance program: implications for program replication. Am J Drug Alcohol Abuse 2012;38(3):206–12. Cacoub P, Ouzan D, Melin P, Lang JP, Rotily M, Fontanges T, et al. Patient education improves adherence to peg-interferon and ribavirin in chronic genotype 2 or 3 hepatitis C virus infection: a prospective, real-life, observational study. World J Gastroenterol. 2008;14(40):6195–203. Chakrabarty G, Rice P, Forton DM. Randomized controlled trial of home-based self-administered dried blood spot testing versus written advice for community screening of hepatitis B contacts. Hepatology 2013;58(4):616A. Chen MS, Jr., Fang DM, Stewart SL, Ly MY, Lee S, Dang JH, et al. Increasing hepatitis B screening for hmong adults: results from a randomized controlled community-based study. Cancer Epidemiol Biomarkers Prev.. 2013;22(5):782–91. Chen WL, Chiu WT, Wu MS, Hsu MH, Tsai SH. Translational research of telecare for the treatment of hepatitis C. Biomed Res Int. 2014;2014:195097. Craine N, Whitaker R, Perrett S, Zou L, Hickman M, Lyons M. A stepped wedge cluster randomized control trial of dried blood spot testing to improve the uptake of hepatitis C antibody testing within UK prisons. Eur J Public Health. 2015;25 (2):351 –7. Cullen W, Stanley J, Langton D, Kelly Y, Staines A, Bury G. Hepatitis C infection among injecting drug users in general practice: a cluster randomised controlled trial of clinical guidelines' implementation. Br J Gen Pract. 2006;56(532):848–56. Curcio F, Di Martino F, Capraro C, Angelucci F, Bulla F, Caprio N, et al. Together – to take care: multidisciplinary management of hepatitis C virus treatment in randomly selected drug users with chronic hepatitis. J Addict Med. 2010;4(4):223–32. Drainoni ML, Litwin AH, Smith BD, Koppelman EA, McKee MD, Christiansen CL, et al. Effectiveness of a risk screener in identifying hepatitis C virus in a primary care setting. Am J Public Health. 2012;102(11):e115–e121. Evon DM, Simpson K, Kixmiller S, Galanko J, Dougherty K, Golin C, et al. A randomized controlled trial of an integrated care intervention to increase eligibility for chronic hepatitis C treatment. Am J Gastroenterol. 2011;106(10):1777–86. Hagedorn H, Dieperink E, Dingmann D, Durfee J, Ho SB, Isenhart C, et al. Integrating hepatitis prevention services into a substance use disorder clinic. J Subst Abuse Treat. 2007;32(4):391 –8. Helsper CW, van Essen GA, Bonten MJ, de Wit NJ. A support programme for primary care leads to substantial improvements in the effectiveness of a public hepatitis C campaign. Fam Pract. 2010;27(3):328–32. Hickman M, McDonald T, Judd A, Nichols T, Hope V, Skidmore S, et al. Increasing the uptake of hepatitis C virus testing among injecting drug users in specialist drug treatment and prison settings by using dried blood spots for diagnostic testing: a cluster randomized controlled trial. J Viral Hepat. 2008;15(4):250–54.

12.

13.

14.

15.

16.

17. 18.

19.

20.

21.

22.

23. 24.

25.

Page | 587

26.

Hirsch AA, Lawrence RH, Kern E, Falck-Ytter Y, Shumaker DT, Watts B. Implementation and evaluation of a multicomponent quality improvement intervention to improve efficiency of hepatitis C screening and diagnosis. Jt Comm J Qual Patient Saf. 2014;40(8):351–7. Hsu L, Bowlus CL, Stewart SL, Nguyen TT, Dang J, Chan B, et al. Electronic messages increase hepatitis B screening in at-risk Asian American patients: a randomized, controlled trial. Dig Dis Sci. 2013;58(3):807–14. Hussein M, Benner JS, Lee D, Sesti AM, Battleman DS, Brock-Wood C. Propensity score matching in the evaluation of drug therapy management programs: an illustrative analysis of a program for patients with hepatitis C virus. Qual Manag Health Care. 2010;19(1):25 –33. Juon HS, Lee S, Strong C, Rimal R, Kirk GD, Bowie J. Effect of a liver cancer education program on hepatitis B screening among Asian Americans in the Baltimore-Washington metropolitan area, 2009– 2010. Prev Chronic Dis. 2014;11:130258. Knott A, Dieperink E, Willenbring ML, Heit S, Durfee JM, Wingert M, et al. Integrated psychiatric/medical care in a chronic hepatitis C clinic: effect on antiviral treatment evaluation and outcomes. Am J Gastroenterol. 2006;101(10):2254–62. Koruk I, Koruk ST, Copur AC, Simsek Z. A intervention study to improve HBsAg testing and preventive practices for hepatitis B in an obstetrics hospital. Türk Silahl Kuvvetleri Koruyucu Hekim Bül. 2011;10(3):287–92. Krauskopf K, Kil N, Sofianou A, et al. Evaluation of an electronic health record prompt for hepatitis C antibody screening of baby boomers in primary care – a cluster randomized control trial. Journal of General Internal Medicine 2014;29:S88–S89. Larrey D, Salse A, Ribard D, Boutet O, Hyrailles-Blanc V, Niang B, et al. Education by a nurse increases response of patients with chronic hepatitis C to therapy with peginterferon-alpha2a and ribavirin. Clin Gastroenterol Hepatol. 2011;9(9):781–5. Le Lan C, Guillygomarc'h A, Danielou H, Le Dréau G, Lainé F, Védeilhié C, et al. A multi-disciplinary approach to treating hepatitis C with interferon and ribavirin in alcohol-dependent patients with ongoing abuse. J Hepatol. 2012;56(2):334–40. Lee R, Vu K, Bell CM, Hicks LK. Screening for hepatitis B surface antigen before chemotherapy: current practice and opportunities for improvement. Curr Oncol. 2010;17(6):32 –8. Litwin AH, Smith BD, Drainoni ML, McKee D, Gifford AL, Koppelman E, et al. Primary care-based interventions are associated with increases in hepatitis C virus testing for patients at risk. Dig Liver Dis. 2012;44(6):497–503. Lubega S, Agbim U, Surjadi M, Mahoney M, Khalili M. Formal hepatitis C education enhances HCV care coordination, expedites HCV treatment and improves antiviral response. Liver Int. 2013;33(7):999 – 1007. Ma GX, Gao W, Tan Y, Chae WG, Rhee J. A community-based participatory approach to a hepatitis B intervention for Korean Americans. Prog Community Health Partnersh. 2012;6(1):7 –16. Masson CL, Delucchi KL, McKnight C, Hettema J, Khalili M, Min A, et al. A randomized trial of a hepatitis care coordination model in methadone maintenance treatment. Am J Public Health. 2013;103(10):e81–e88. Matthews HC, McLeod MA, Oakes K, McCurdy G, Zuckerman M. Carey I, et al. Perinatal hepatitis B in a high prevalence inner city population: direct electronic referral improves care. Gut. 2012;61 (Suppl 2):A79–A80. Merchant RC, Baird JR, Liu T, Taylor LE, Montague B, Nirenberg T. Does a brief intervention increase HIV/HCV screening among drug-using emergency department patients? Acad Emerg Med. 2014;21(5):S305–S306.

27.

28.

29.

30.

31.

32.

33.

34.

35. 36.

37.

38. 39.

40.

41.

Page | 588

42.

ClinicalTrials.gov [website]. Compliance of HCV genotype 1 infected patients receiving PegIntron/Rebetol and a Patient Assistance Program (Study P04671). ClinicalTrials.gov identifier NCT00728494 (https://clinicaltrials.gov/ct2/show/NCT00728494, accessed 08 June 2016). ClinicalTrials.gov [website]. Adherence in patients receiving pegintron pen/Rebetol for hepatitis C in conjunction with a patient assistance program (Study P04281)(COMPLETED). 2009. ClinicalTrials.gov identifier NCT00723892 (https://clinicaltrials.gov/ct2/show/NCT00723892, accessed 08 June 2016). ClinicalTrials.gov [website]. Adherence in patients receiving pegintron/Rebetol for hepatitis C in conjunction with a psychotherapy support program (Study P04252) (COMPLETED). 2009. ClinicalTrials.gov identifier NCT00723892 (https://clinicaltrials.gov/ct2/show/NCT00723892, accessed 08 June 2016). Mostert MC, Richardus JH, de Man RA. Referral of chronic hepatitis B patients from primary to specialist care: making a simple guideline work. J Hepatol. 2004;41(6):1026 –30. Neri S, Bertino G, Petralia A, Giancarlo C, Rizzotto A, Calvagno GS, et al. A multidisciplinary therapeutic approach for reducing the risk of psychiatric side effects in patients with chronic hepatitis C treated with pegylated interferon alpha and ribavirin. J Clin Gastroenterol. 2010;44(9):e210 –e217. Reimer J, Schmidt CS, Schulte B, Gansefort D, Gölz J, Gerken G, et al. Psychoeducation improves hepatitis C virus treatment during opioid substitution therapy: a controlled, prospective multicenter trial. Clin Infect Dis. 2013;57 Suppl 2:S97–S104. Renou C, Lahmek P, Pariente A, et al. Impact of therapeutic education on the outcome of chronic hepatitis C treatment. Hepatology. 2009;50:729A. Rifai MA, Moles JK, Lehman LP, Van der Linden BJ. Hepatitis C screening and treatment outcomes in patients with substance use/dependence disorders. Psychosomatics. 2006;47(2):112 –21. Rosenberg SD, Goldberg RW, Dixon LB, Wolford GL, Slade EP, Himelhoch S, et al. Assessing the STIRR model of best practices for blood-borne infections of clients with severe mental illness. Psychiatr Serv. 2010;61(9):885–91. Sahajian F, Excler G, Bailly F, Caillat-Vallet E, Trépo C, Sepetjan M, et al. Hepatitis C screening practices among private practitioners: impact of an information campaign. Gastroenterol Clin Biol. 2004;28(8– 9):714–19. Tait JM, McIntyre PG, McLeod S, Nathwani D, Dillon JF. The impact of a managed care network on attendance, follow-up and treatment at a hepatitis C specialist centre. J Viral Hepat. 2010;17(10):698 – 704. Taylor VM, Bastani R, Burke N, Talbot J, Sos C, Liu Q, et al. Evaluation of a hepatitis B lay health worker intervention for Cambodian Americans. J Community Health. 2013;38(3):546 –53. Taylor VM, Gregory Hislop T, Bajdik C, Teh C, Lam W, Acorda E, et al. Hepatitis B ESL education for Asian immigrants. J Community Health. 2011;36(1):35–41. Taylor VM, Hislop TG, Tu SP, Teh C, Acorda E, Yip MP, et al. Evaluation of a hepatitis B lay health worker intervention for Chinese Americans and Canadians. J Community Health. 2009;34(3):165 –72. ClinicalTrials.gov [website]. Impact of physician directed education on patient compliance with hepatitis C therapy (OPTIMAL) (COMPLETED). 2014. ClinicalTrials.gov identifier NCT01405027 (https://clinicaltrials.gov/ct2/show/NCT01405027, accessed 08 June 2016). van der Veen YJ, van Empelen P, de Zwart O, Visser H, Mackenbach JP, Richardus JH. Cultural tailoring to promote hepatitis B screening in Turkish Dutch: a randomized control study. Health Promot Int. 2014;29(4):692–704.

43.

44.

45. 46.

47.

48. 49. 50.

51.

52.

53. 54. 55. 56.

57.

Page | 589

58. 59.

Arora S, Thornton K, Murata G, Deming P, Kalishman S, Dion D, et al. Outcomes of treatment for hepatitis C virus infection by primary care providers. N Engl J Med. 2011;364(23):2199 –2207. Bonkovsky HL, Tice AD, Yapp RG, Bodenheimer HC Jr, Monto A, Rossi SJ, et al. Efficacy and safety of peginterferon alfa-2a/ribavirin in methadone maintenance patients: randomized comparison of direct observed therapy and self-administration. Am J Gastroenterol. 2008;103(11):2757–65. Ramsey SE, Engler PA, Stein MD, Brown RA, Cioe P, Kahler CW, et al. Effect of CBT on depressive symptoms in methadone maintenance patients undergoing treatment for hepatitis C. J Addict Res Ther. 2011;2(2):2–10. Cioe PA, Stein MD, Promrat K, Friedmann PD. A comparison of modified directly observed therapy to standard care for chronic hepatitis C. J Community Health. 2013;38(4):679–84. Ho SB, Brau N, Cheung R, Liu L, Sanchez C, Sklar M, et al. Integrated care increases treatment and improves outcomes of patients with chronic hepatitis C virus infection and psychiatric illness or substance abuse. Clin Gastroenterol Hepatol. 2015;13(11):2005‒14.e1‒3. Carrion JA, Gonzalez-Colominas E, Garcia-Retortillo M, Cañete N, Cirera I, Coll S, et al. A multidisciplinary support programme increases the efficiency of pegylated interferon alfa-2a and ribavirin in hepatitis C. J Hepatol. 2013;59(5):926 –33. Bastani R, Glenn BA, Maxwell AE, Jo AM, Herrmann AK, Crespi CM, et al. Cluster-randomized trial to increase hepatitis B testing among Koreans in Los Angeles. Cancer Epidemiol Biomarkers Prev.. 2015;24(9):1341–9. Shah HA, Abu-Amara M. Education provides significant benefits to patients with hepatitis B virus or hepatitis C virus infection: a systematic review. Clin Gastroenterol Hepatol. 2013;11(8):922–33. Meyer JP, Moghimi Y, Marcus R, Lim JK, Litwin AH, Altice FL. Evidence-based interventions to enhance assessment, treatment, and adherence in the chronic Hepatitis C care continuum. The Int J Drug Policy. 2015;26(10):922–35. Glenton C, Colvin CJ, Carlsen B, et al. Barriers and facilitators to the implementation of lay health worker programmes to improve access to maternal and child health: qualitative evidence synthesis. Cochrane Database Syst Rev. 2013;(10):CD010414. Mwai GW, Mburu G, Torpey K, Frost P, Ford N, Seeley J. Role and outcomes of community health workers in HIV care in sub-Saharan Africa: a systematic review. J Int AIDS Soc.. 2013;16:18586. Patel V, Weiss HA, Chowdhary N, Naik S, Pednekar S, Chatterjee S, et al. Lay health worker led intervention for depressive and anxiety disorders in India: impact on clinical and disability outcomes over 12 months. Br J Psychiatry. 2011;199(6):459 –466. Joshi R, Alim M, Kengne AP, Jan S, Maulik PK, Peiris D, et al. Task shifting for non-communicable disease management in low and middle income countries – a systematic review. PLoS One. 2014;9(8):e103754. Kredo T, Adeniyi FB, Bateganya M, Pienaar ED. Task shifting from doctors to non-doctors for initiation and maintenance of antiretroviral therapy. Cochrane Database Syst Rev 2014;(7):CD007331. Mandelblatt JS, Gold K, O'Malley AS, Taylor K, Cagney K, Hopkins JS, et al. Breast and cervix cancer screening among multiethnic women: role of age, health, and source of care. Prev Med. 1999;28(4):418–25. Green BB, Wang CY, Anderson ML, Chubak J, Meenan RT, Vernon SW, et al. An automated intervention with stepped increases in support to increase uptake of colorectal cancer screening: a randomized trial. Ann Intern Med. 2013;158(5 Pt 1):301 –11. Willenbring ML. Integrating care for patients with infectious, psychiatric, and substance use disorders: concepts and approaches. AIDS 2005;19 (Suppl 3):S227 –S237.

60.

61. 62.

63.

64.

65. 66.

67.

68. 69.

70.

71. 72.

73.

74.

Page | 590

75.

Norman J, Walsh NM, Mugavin J, Stoové MA, Kelsall J, Austin K, et al. The acceptability and feasibility of peer worker support role in community-based HCV treatment for injecting drug users. Harm Reduct J. 2008;5:8. Sylvestre D, Loftis J, Hauser P, Genser S, Cesari H, Borek N, et al. Co-occurring hepatitis C, substance use, and psychiatric illness: treatment issues and developing integrated models of care. J Urban Health. 2004;81(4):719–34. Hoang T, Goetz MB, Yano EM, Rossman B, Anaya HD, Knapp H, et al. The impact of integrated HIV care on patient health outcomes. Med Care. 2009;47(5):560–7. Linas BP, Barter DM, Leff JA, Assoumou SA, Salomon JA, Weinstein MC, et al. The hepatitis C cascade of care: identifying priorities to improve clinical outcomes. PloS One. 2014;9(5):e97317. van der Meer AJ. Value anti-hepatitis C virus therapy by its clinical efficacy. Hepatology. 2015;62(2):334–6. Hosaka T, Suzuki F, Kobayashi M, Seko Y, Kawamura Y, Sezaki H, et al. Long-term entecavir treatment reduces hepatocellular carcinoma incidence in patients with hepatitis B virus infection. Hepatology. 2013;58(1):98–107.

76.

77. 78. 79. 80.

Page | 591

ANNEX 6. Predictive modelling analysis 6.1. 6.2. Testing strategies for hepatitis B and C infection - predictive modelling Decision analysis of strategies to identify hepatitis C – predictive modelling

Diagnostic accuracy of a one or two serological test strategy

Effect of prevalence and sensitivity of assay used on outcomes using a single serological testing strategy for HBV infection. Table 7.1 Scenario 1 Population 1000; Assay A: Sensitivity 98%, Specificity 99% Prevalence True positive 98 20 4 False negative 2 0 0 False positives 9 20 20 True negatives 891 970 986 Total tests 1000 1000 1000 PPV NPV Ratio of TP:FP 10.89 2.00 0.39

10.0% 2.0% 0.4%

91.59% 66.67% 28.24%

99.78% 99.96% 99.99%

Scenario 2 Population 1000; Assay A: Sensitivity 90%, Specificity 99% Prevalence 10.0% 2.0% 0.4% True Positive 90 18 4 False negative 10 2 0 False positives 9 10 10 True negatives 891 970 986 Total tests 1000 1000 1000 PPV 90.91% 64.75% 26.55% NPV 98.89% 99.79% 99.96% Ratio of TP:FP 10.00 1.84 0.36

Scenario 3 Population 1000; Assay A: Sensitivity 98%, Specificity 98% Prevalence 10.0% 2.0% 0.4% True Positive 98 20 4 False negative 2 0 0 False positives 18 20 20 True negatives 882 960 976 Total tests 1000 1000 1000 PPV 84.48% 50.00% 16.44% NPV 99.77% 99.96% 99.99% Ratio of TP:FP 5.44 1.00 0.20

Page | 592

Scenario 4 Population 1000; Assay A: Sensitivity 80%, Specificity 99% Prevalence 10.0% 2.0% 0.4% True Positive 80 16 3 False negative 20 4 1 False positives 9 10 10 True negatives 891 970 986 Total tests 1000 1000 1000 PPV 89.89% 62.02% 24.32% NPV 97.80% 99.59% 99.92% Ratio of TP:FP 8.89 1.63 0.32

Effect of prevalence and sensitivity of assay used on outcomes using a two serological testing strategy for HBV infection. Table 7.2.Scenario 5 Population 1000; Assay A: Sensitivity 90.0%, Specificity 99%; Assay B: Sensitivity 90.00%, Specificity 99% Prevalence True Positive 81 16 3 False negative 19 4 1 False positive 0 0 0 True negative 900 980 996 Total tests (A+B) 1099 1028 1014 PPV NPV Ratio of TP:FP 900.00 165.31 32.53 Observed sensitivity 81.00% 81.00% 81.00% Observed specificity 99.99% 99.99% 99.99%

10.0% 2.0% 0.4%

99.89% 99.40% 97.02%

97.93% 99.61% 99.92%

Scenario 6 Population 1000; Assay A: Sensitivity 98.0%, Specificity 99%; Assay B: Sensitivity 98.00%, Specificity 99% Prevalence True Positive 96 19 4 False negative 4 1 0 False positive 0 0 0 True negative 900 980 996 Total tests (A+B) 1107 1029 1014 PPV NPV Ratio of TP:FP 1067.1 1 196.00 38.57 Observed sensitivity 96.04% 96.04% 96.04% Observed specificity 99.99% 99.99% 99.99%

10.0% 2.0% 0.4%

99.91% 99.49% 97.47%

99.56% 99.92% 99.98%

Condition: Assay B was performed on all individuals reactive in assay A. Assay B performance is considered independent of assay A. Outcome of two test strategy: when assay A+ and B+ = positive; assay A- or A+ B- = negative

Page | 593

Effect of prevalence and sensitivity of assay used on outcomes using a single serological testing strategy for HCV infection. Table 8.2. Scenario 1 Population 1000; Assay A: Sensitivity 99%, Specificity 99% Prevalence 40.0% 10.0% 2.0% 0.4% True Positive 396 99 20 4 False negative 4 1 0 0 False positives 6 9 10 10 True negatives 594 891 970 986 Total tests 1000 1000 1000 1000 PPV 98.51% 91.67% 66.89% 28.45% NPV 99.33% 99.89% 99.98% 100.00% Ratio of TP:FP 66.00 11.00 2.02 0.40

Scenario 2 Population 1000; Assay A: Sensitivity 98%, Specificity 99% Prevalence 40.0% 10.0% 2.0% 0.4% True Positive 392 98 20 4 False negative 8 2 0 0 False positives 6 9 10 10 True negatives 594 891 970 986 Total tests 1000 1000 1000 1000 PPV 98.49% 91.59% 66.67% 28.24% NPV 98.67% 99.78% 99.96% 99.99% Ratio of TP:FP 65.33 10.89 2.00 0.39

Scenario 3 Population 1000; Assay A: Sensitivity 99.5%, Specificity 98.0% Prevalence 40.0% 10.0% 2.0% 0.4% True Positive 396 99 20 4 False negative 4 1 0 0 False positives 12 18 20 20 True negatives 588 882 960 976 Total tests 1000 1000 1000 1000 PPV 97.06% 84.62% 50.25% 16.58% NPV 99.32% 99.89% 99.98% 100.00% Ratio of TP:FP 33.00 5.50 1.01 0.20

Page | 594

Effect of prevalence and sensitivity of assay used on outcomes using a two serological testing strategy for HCV infection. Table 8.3. Scenario A Population 1000; Assay A: Sensitivity 98%, Specificity 99%; Assay B Sensitivity 98%, Specificity 98% Prevalence True Positive 384 96 19 4 False negative 16 4 1 0 False positive 0 0 0 0 True negative 600 900 980 996 Total tests (A+B) 1398 1107 1029 1014 PPV NPV Ratio of TP:FP 3201.3 3 533.56 98.00 19.29 Observed sensitivity 96.04% 96.04% 96.04% 96.04% Observed specificity 99.98% 99.98% 99.98% 99.98%

40.0% 10.0% 2.0% 0.4%

99.97% 99.81% 98.99% 95.07%

97.43% 99.56% 99.92% 99.98%

Page | 595

Annex 6.1 Testing strategies for hepatitis B and C infection Predictive Modelling

Professor John V. Parry Public Health England, London, United Kingdom

September 2015

Page | 596

Hepatitis C

1. Diagnostic background to HCV infection Unlike HIV, infection with HCV does not inevitably lead to long-term chronic infection, the body spontaneously clearing infection in a substantial minority of cases. Estimates of clearance vary, but appear to occur in 15–25% of HCV infections. The majority of spontaneous clearances appear to arise in the first few months of infection. Clearance of infection, whether spontaneous or as a result of antiviral therapy, does not appear to signify protection from reinfection. The exact time course of virological and immunological markers of infection in HCV, particularly during the first months of infection, has not been easy to define because of the difficulties of pinpointing the time of exposure in community cases and of obtaining suitable specimens at frequent intervals. Nonetheless, several studies on plasma/blood donors, recipients and exposed health-care workers have provided estimates that should be sufficient to inform testing strategies. As illustrated below (Fig. 1), following an initial “eclipse phase” of 1–2 weeks when no virological or serological markers of infection may be detected, there is typically a substantial viræmia that is present in the absence of an anti-HCV antibody response for a further 6–10 weeks. During this “serological window” it has been shown that free (i.e. not complexed with anti-HCV antibody) core antigen of HCV (HCV Ag) can be detected in a proportion of cases employing combined anti-HCV/HCV Ag immunoassays. Fig. 1. Approximate time course for HCV virological and serological markers in chronic HCV infection. Serological window Eclipse pase Seroconversion/acute phase Chronic phase

Anti-HCV HCV RNA

HCV Ag

6 months

 decades

Following the development of the anti-HCV response, HCV Ag becomes complexed with anti-HCV antibody such that its presence cannot reliably be detected without a dissociation step to release HCV Ag from immune complexes prior to an immunoassay for HCV Ag. A majority of HCV infections become established as long-term chronic infections, persisting for decades, with an increased risk of Page | 597

severe liver disease. Such individuals also present a source of infection to others. There are rare reports of persistent “occult” HCV infection, i.e. HCV RNA positive/anti-HCV negative. In a substantial minority of cases, usually considered to be 15–25%, HCV fails to establish as a long-term chronic infection and the individual spontaneously clears infection. It is thought that this mostly happens during the first 6 months of infection; active infection persisting beyond that time usually associated with long-term carriage. Individuals who clear the infection usually also exhibit slowly waning anti-HCV antibody levels, presumably due to the removal of the antigenic stimulus to the immune system (Fig. 2). In turn, this leads to low-level anti-HCV antibody levels that may not be consistently detected by all anti-HCV test devices. It is likely that such low-level reactivities are in the main associated with waning anti-HCV levels following virus clearance rather than with a relatively brief period during the early stages of anti-HCV seroconversion. Fig. 2. Hypothetical time course for HCV virological and serological markers in self-resolving HCV infection. Timings are ill-defined and may vary widely. Serological window Eclipse phase SCVN phase Virus clearance Waning anti-HCV Seroreversion

HCV RNA

Anti-HCV

HCV Ag

6 months

 decades

These two likely outcomes in untreated individuals complicate the matter of identifying those who would benefit from HCV-specific care, including antiviral therapy. Unlike HIV, a simple and inexpensive anti-HCV antibody test will not alone identify those with active infection. Furthermore, there is at present no simple and immediate prognostic test to differentiate between those who are in the first months of infection, a substantial minority of whom will eradicate their infection without therapeutic intervention, and those who are already in an established chronic state of HCV infection. There is evidence that, as with HIV, the avidity of anti-HCV antibody in the first months after seroconversion is low, and thus those with high avidity antibody would be more predictive of chronic infection. However, there is currently no commercial anti-HCV avidity method available.

Page | 598

2. Aim of testing The preferred testing strategy will depend on the aims of the testing programme as well as factors such as assay performance, technical feasibility, access and cost. 1. Identify all anti-HCV positive individuals This scenario employs the most widely available and least expensive diagnostics. It would not, however, detect recent infections in the period prior to anti-HCV seroconversion. If the ultimate aim is to offer antiviral treatment to those with an active HCV infection, anti-HCV positive individuals, or at least a blood specimen, would then have to be referred to a facility able to carry out the necessary procedures to determine whether HCV treatment was needed, and then to monitor its success. Strategy: Anti-HCV test (RDT or EIA) with or without a different anti-HCV test

2. Identify all active HCV infections This approach would ascertain whether individuals are carrying replicating HCV. The current cost and complexity of tests for HCV RNA would probably preclude their use as a screening tool in all but the highest prevalences. Consequently, the primary screening test would likely be an antiHCV test, which would fail to detect acute infections in the seroconversion window phase. This would be partially offset by the use of a combined anti-HCV/HCV Ag assay. Identification of active HCV infection would require the application of a test for HCV RNA, though an immunoassay that is able to detect immune-complexed HCV Ag may be sufficiently accurate. Early infections, a minority of which would have gone on to be spontaneously eradicated by the host would also be treated under this approach. Strategy: Anti-HCV test (RDT or EIA) or combined anti-HCV + HCV Ag test followed by an HCV RNA test or an HCV Ag test

3. Identify chronic HCV infections This approach would need to go a further step to aim 2, by incorporating the means to distinguish chronic from acute infection. Such a test would permit immediate treatment of those identified as having a chronic infection and deferral of those with a recent acute HCV infection. At present, though, there is no appropriate and validated laboratory marker to distinguish chronic from acute HCV infection. While a combination of clinical and non-virological laboratory tests might identify those in need of immediate treatment, others may have to be deferred for 3–6 months to allow a re-test to exclude the minority whose infection cleared spontaneously. Dependent on the costs of antiviral drugs, the risks of unnecessary delay to treatment in those that would benefit from it and of loss to follow up would probably exclude this approach at this time.

Page | 599

3.HCV infections in a community Many factors will influence the status of HCV infection in any given community, including prevalence, incidence, risk behaviours, harm reduction interventions, access to health care, and specific HCV diagnosis and treatment. If anti-HCV prevalence is known and incidence has been estimated it may be reasonable to make crude estimates of the numbers of cases at the several stages of HCV infection. Taking a reasonably straightforward example of a concentrated epidemic, such as has occurred widely in persons who inject drugs (PWID), it is commonplace to find an ongoing outbreak with anti-HCV prevalence in the region of 40–60% and an annual incidence of approximately 10%. Applying this to a population of PWID of 100 000 one can deduce the following approximations: Anti-HCV positive (45%) Acute HCV (RNA/Ag only) Eclipse cases (no marker) Seropos HCV (Ab + RNA/Ag) Chronic HCV (Ab + RNA + >6 m) =N×P = (N – (N × P)) × (W × I) = (N – (N x P)) × (E × I) = (N × P × (1 – C)) = 100k × 0.45 = (100k – 45k) × (0.2 × 0.1) = (100k – 45k) × (0.04 × 0.1) = (100k × 0.45 × (1 – 0.2)) = 36k – 9k – 2.25k = = = = = 45 000 1 100 220 36 000 24 750

= (N × P × (1 – C)) – (N × P × C) – (N × P × I × 0.5)

where: N = population; P = anti-HCV prevalence; W = seroconversion window (estimated 10 weeks); E = eclipse phase (estimated 2 weeks); I = annual HCV incidence; C = proportion of resolved HCV infection

In the above model population, there are at least 46 320 individuals who have been infected with HCV. This does not include a further group with resolved infection who will have seroreverted to an anti-HCV negative status. Employing the above estimates, one can then approximate the effectiveness of different strategies to detect HCV infection. Strategies and algorithms It was recognized soon after HIV diagnostics became available in the mid-1980s that the application of a combination of HIV tests would need to be employed to minimize the risk of inaccurate results. At that time, the diagnosis was almost entirely based on the detection of anti-HIV antibodies, and false-positive reactions were common. Within several years a very wide range of anti-HIV diagnostics became available, often employing a range of constructions (e.g. indirect, competitive, immunometric) and different sources of HIV antigens (e.g. viral, peptides, recombinants). While many countries adopted an approach of EIA screen followed by an immunoblot EIA (western blot) or indirect immunofluorescence, which were both expensive and imperfect, others recognized that a simpler and more economic approach employing a carefully selected combination of screening tests could achieve equivalent or better accuracy. This led to the development of WHO guidance that outlined several testing strategies. Each strategy was based on the prevalence of HIV in the community being tested and the expected accuracy of the individual tests employed. The strategies were generic, not specifying individual diagnostic products or types, but providing a decision flow chart and guidance about how to select a combination of diagnostic devices to establish and validate a local algorithm employing specific diagnostic tests.

Page | 600

4.Strategies under consideration to identify individuals infected with HCV Aim 1: Identify anti-HCV positive individuals The two strategies employed to generate the outcomes in the following tables are as follows:

Fig. 3: Single anti-HCV test Test Kit A

Fig. 4. Two independent anti-HCV tests in sequence Test Kit A

Reactive Reactive Non-reactive

Non-reactive

Test Kit B Anti-HCV detected Anti-HCV not detected

Reactive

Non-reactive

Anti-HCV detected

Anti-HCV not detected

Employing the simple Single-test strategy for the detection of anti-HCV antibodies, Table 1 demonstrates hypothetical outcomes based on a range of typical performance characteristics for anti-HCV tests employed by trained individuals within an appropriate quality system. The estimates have been prepared for a range of prevalences, reflecting a concentrated epidemic such as is found among PWIDs (45%) through to a low endemicity setting among a general population, say of northern Europe (0.4%). Table 1. Outcomes for single-test strategy based on typical estimates of test accuracy

Page | 601

The outcomes illustrate the strong influence of prevalence and assay specificity on positive predictive value (PPV). At high prevalences, the use of a highly specific single screening test yields a PPV in excess of 0.99, and the ratio of true positive (TP) : false positive (FP) results is high (164), but nonetheless there would be a small but substantial number of false-positive diagnoses. At the lowest prevalences (e.g. 0.4%), PPV might be expected to fall below 0.5, yielding more false-positive diagnoses than true ones, particularly if the test specificity falls below 0.995. Even in middling prevalences, the PPV will struggle to exceed 0.9 should the specificity performance of the test employed fall below 0.99. The negative predictive value (NPV) is generally high except in highprevalence populations (e.g. 45%) when a test with relatively poor sensitivity (<99%) is employed. As discussed above (Strategies and algorithms), the use of a second diagnostic test is able to improve the diagnostic accuracy. While repeating testing of a reactive specimen by the original test can provide some gains in specificity, in a quality-controlled environment, these gains are generally modest. This is because in such a context false-positive reactions are generally due to real and reproducible reactions between factors in the specimen and materials incorporated in the diagnostic device. In circumstances in which initial reactions are not reproducible it is usually a failure of the quality system such as training, mistakes, failure to follow defined procedures, and instrument calibration and maintenance. In such circumstances the solution is to invest in training and the quality system; repeating unreliable tests is no substitute. The choice of the second test product employed in the local algorithm is critical to satisfactory outcomes. It is not uncommon for different products from the same manufacturer to use assay components in common; also, different manufacturers or distributors can source the same raw materials, or even the entire device, from the same third party manufacturer/supplier. In such cases, there is an increased likelihood that false reactions in the primary screening test will arise also in the second assay, and this will erode the PPV of the applied algorithm (discussed in more detail below). Consequently, test devices employed should be sourced from a reputable accredited manufacturer, and most critically the translation of the Testing strategy, which is generic, into a Local algorithm that employs specific test devices, needs carefully to be piloted and validated before being put into routine clinical use. (It is understood) that the Guidelines Group’s current thinking is that a two-test strategy would follow that outlined in Fig. 4. Unlike HIV testing in which discordant results would generally be considered “indeterminate”, and undergo further testing, the finding of a negative reaction in the second test (Test Kit B), performed only on those individuals reactive in the initial test (Test Kit A), would be interpreted as an absence of anti-HCV. Such findings are most likely to be non-specific, but even if the reaction did reflect the presence of true, but low level, anti-HCV reactivity, as discussed above, this is more likely to reflect the waning antibodies associated with cleared HCV infection than with a brief phase early in HCV infection, during seroconversion. Hypothetical outcomes of the two-test strategy are given in Table 2, employing the same notional anti-HCV prevalences as in Table 1. To keep the tables relatively simple, only two of the four sets of illustrative sensitivity and specificity employed for Test Kit A in Table 1 have been employed. Similarly, only two sets of sensitivity and specificity measures have been employed for Test Kit B when calculating the expected outcomes. Page | 602

Table 2: Outcomes for two-test strategy based on typical estimates of test accuracy (see Fig. 4)

Across all prevalences, the introduction of a second test kit to be applied to all specimens reactive in the initial test kit provides substantial potential gains in the PPV of this strategy. Even at the lowest prevalence shown here (0.4%), employing two test kits with relatively low specificity (0.98), the PPV was vastly improved, converting a PPV of 0.164 in the one-test strategy to 0.906. In all but the highest population prevalences, the use of less sensitive test kits in the two-test strategy had only a modest impact on NPVs. In all but the highest prevalences, the calculations demonstrate the potential for substantial gains in diagnostic accuracy in return for a modest investment in a second, independent, test kit to be applied to initially reactive individuals. For example, taking a model population of 100 000 with an anti-HCV prevalence of 2%, the PPV of an algorithm employing two independent test kits with modest sensitivities and specificities of 0.98 is improved from 0.500 to 0.980 in return for the supplementary use of <4000 of Test Kit B. Selection of test kits for local algorithm As discussed above, it is of critical importance that the two test kits selected for a local algorithm are chosen with great care. It must be remembered that the sensitivity and specificity estimates claimed by manufacturers and even those of independent evaluations should be based on, or approximate to, the testing of a panel of samples representative of an unbiased and unselected population of positives and negatives. This can be difficult to achieve, and sometimes some bias may be introduced either unintentionally or intentionally, such as including only strongly positive samples. There can also be regional differences due to prevalent genotypes, transmission patterns, the maturity of an outbreak or age at acquisition. However, when applying a testing strategy such as Page | 603

that illustrated in Fig. 4, it is critical to recognize that the population of samples that is selected for testing in Test Kit B has been highly selected by the application of Test Kit A. Under such circumstances, the performance claims made by the manufacturer of Kit A, or described in an independent evaluation, are unlikely to apply. In particular, the occurrence of false-positive reactions in Test Kit A will generally increase the risk of a false-positive reaction also in Test Kit B. Consequently, in this algorithmic approach the specificity claims are likely to be misleading and better than will be achieved when applied to individuals who were falsely reactive on Test Kit A. This phenomenon has been observed when applying HIV testing algorithms, with shared false-positive reactions observed frequently between particular diagnostic products. Similarly, when employing two kits in an algorithm, each with a sensitivity of 0.995 on a “random” sample, it is probable that the 1 in 200 true positives not detected by Test Kit A will be the same sample that would be missed by Test Kit B, rather than one of the other 199, and therefore the sensitivity on the sample selected by Test Kit A is likely to be closer to 1.00 than 0.995. The outcomes illustrated in Table 2 are based on an assumption that the two test kits employed act independently and that false reactions arise at random, such a reaction in Test Kit A not increasing the likelihood of a false reaction in Test Kit B. In Table 3, the performance characteristics of Test Kit B have been adjusted to take into account the potential impact of selection by Test Kit A of the individuals to be tested by Test Kit B. The modified sensitivity and specificity of Test Kit B have been adjusted to 0.999 and 0.850, reflecting the selection bias, including a 15% increased risk of false-positive reactions in Test Kit B. Table 3: Outcomes for a two-test strategy taking into account the theoretical impact on the performance characteristics Test Kit B introduced by selecting the individuals to be tested by first applying Test Kit A

While PPV holds up well for the highest prevalences of anti-HCV, at lower prevalences the PPV will fall substantially, particularly if the specificity of Test Kit A is relatively poor. On the other hand, recognizing the consequence to the sensitivity of Test Kit B when testing the subpopulation selected by the use of Test Kit A, gives rise to improved NPVs, even at higher prevalences. In summary, particularly for resource-poor settings, testing strategies must be simple, inexpensive, practically feasible and produce clear categorical results. The use of a two-test strategy Page | 604

has the potential to provide accurate anti-HCV results, substantially reducing the numbers of individuals who would be inappropriately referred on to more specialist services, with the concomitant burdens on scarce health-are resources, of inappropriate interventions and on individuals well-being. The two tests employed in such a strategy, to form part of the local algorithm, must be selected with great care, validated locally, and their output monitored for adverse events. The use of combinations of assays without such care may provide a false sense of security, but will lead to greater costs both to the health-care system and patients, but also to a potential loss of credibility of any associated programme. Aim 2: Detection of active HCV infection While the use of a two-test strategy has the potential to improve the accuracy of anti-HCV detection, it is recognized that for a programme whose aims are primarily to treat and cure individuals with chronic HCV infection, further diagnostic tests would be needed before initiating an expensive course of antiviral therapy. Estimates vary, and may depend on differences in the pattern, timing and maturity of the epidemic in different regions, but as discussed earlier, 10–25% of HCV infections spontaneously resolve, as evidenced by the proportion of anti-HCV positive individuals that lack HCV RNA. If such individuals can be identified at the time of initial testing this would reduce the burden on more expensive parts of the health-care system as well as reduce unnecessary anxiety among those who no longer have active HCV infection. Furthermore, one could take the view that if the aim of any HCV screening programme is to identify individuals who need antiviral therapy then even the most sensitive and specific anti-HCV test strategy might, with that goal in mind, achieve a specificity of only 0.75–0.90, begging the question as to whether there is any value in undertaking a second anti-HCV test to improve the PPV of anti-HCV screening. In view of the relatively small number of second anti-HCV tests that would be needed at lower prevalences, there may be a case for including the second anti-HCV test kit, should a test for HCV RNA, or possibly HCV Ag, not be feasible. It should be borne in mind that even this approach would fail to identify those in the pre-seroconversion phases of HCV infection, a majority of whom would, without therapeutic intervention, go on to develop a long-term chronic HCV infection, and be a potential source of new infections in their contacts. It is likely, though, that the opportunity to detect such incident infections will be rare with the exception of populations with higher prevalences of infection practising high-risk behaviours. The premium cost of a test for HCV RNA, or even that for sequestered HCV Ag, plus the equipment requirements for such assays that prevail at present, would preclude testing for HCV RNA or HCV Ag as a primary test in low- and middle-income settings. Therefore, this aim would probably be best met by a two-test strategy, the first being to screen for the presence of anti-HCV, and the second to test the anti-HCV reactive specimens for the presence of HCV RNA or HCV Ag, and this should prove feasible once the promise of low-technology tests for HCV RNA or HCV Ag become a reality. At present, though, in many resource-poor settings, testing for HCV RNA or HCV Ag may not be available in all settings in which anti-HCV screening might be desirable. A potentially important benefit of a test Fig. 5. Two-test strategy to identify active HCV strategy that detects different markers of infection infection, anti-HCV and HCV RNA, and most importantly, employs technologies, immunoassay and nucleic acid amplification, that there will be minimal risk of confounding between the Reactive

Anti-HCV Test

distinct Non-reactive

is two

HCV RNA/Ag Test

Anti-HCV Page | 605 not detected

diagnostic tests, providing added assurance of the reliability of testing. As with the two-test anti-HCV strategy, tables can generate hypothetical outcomes based on expected Reactive

be Non-reactive

prepared to performance charac

teristics of the assays employed. The principles and No evidence of active Active HCV underlying calculations are the same. However, the infection HCV infection aim of this strategy is to detect active HCV infection, and therefore performance characteristics of the test kits involved should be adjusted to reflect this application. For example, although a test kit for anti-HCV antibodies might have an associated specificity claim of 0.995, when detecting individuals who have active HCV infection its specificity might be more modest, in the range 0.75–0.9. The influence of several of these variables is summarized in Table 4. To limit the number of variables, Table 4: Impact of model anti-HCV prevalence, HCV and for the purposes of the model clearance & assay performance on PPV & NPV of a outcomes given in Table 4, it is assumed 2-test strategy employing an anti-HCV test kit, that the anti-HCV screening test has the reactive persons then being tested by a test kit for same performance characteristics a marker of ongoing HCV infection (sensitivity 0.995; specificity 0.990) for each of the three hypothetical algorithms. In each algorithm “virus” test kits would subsequently be applied to the samples that were reactive in the anti-HCV screening test. Three model performance characteristics have been applied: the first reflecting the hypothetical performance of a “licensed” laboratory nucleic acid amplification test (NAT) kit for HCV RNA, with sensitivity 0.995 and specificity 0.990; the second a notional “field-based” NAT kit or a laboratorybased HCV Ag kit, with sensitivity 0.960 and specificity 0.970; and the third, a rapid diagnostic test (RDT) for HCV More comprehensive findings from which Table 4 is drawn can be antigen (HCV Ag). The effect of anti-HCV found in Appendix 1. prevalence in the population screened (0.45; 0.1; 0.02; 0.004) and that of spontaneous viral clearance (anti-HCV +ve/HCV RNA or Ag –ve), employing a low and a high estimate (10%; 25%), on predictive values in each of the algorithms were explored. As would be expected, as prevalence falls the PPV also falls. Conversely, NPVs increase as prevalences fall. PPVs were slightly worse when a larger proportion of HCV had spontaneously resolved but, on the other hand, NPVs were marginally better. The application of a two-test algorithm ensures that PPV was maintained well above 0.9, except for the lowest prevalences (≤0.004).

Page | 606

The model shows that the accuracy of the “virus” test (HCV RNA/ HCV Ag) would have a substantial influence both on PPV and NPV, e.g. the most accurate algorithm would be expected to achieve a PPV of 0.990 in a population with a 2% anti-HCV prevalence, of which 25% had resolved infection, yet the least accurate approach might achieve only 0.964. Translating this into diagnoses, employing the better algorithm, 10 false-positive diagnoses of active HCV infection might arise in a population of 1000 in which there are actually 15 real active HCV infections; the worse algorithm might yield 36 false diagnoses of active HCV, nearly 2.5 times the number of real active infections. At high anti-HCV prevalences, the use of an insensitive device (e.g. sensitivity 0.8) to identify active HCV infection, hypothetically possibly an HCV Ag RDT (to the best of the author’s knowledge no example is yet on the market), could fail to identify a considerable proportion of individuals with active HCV, the NPV falling to 0.878 in a population with a 45% anti-HCV prevalence, of which 10% had resolved infection, translating to approximately 12% of active infection going undetected. At lower prevalences, the NPV holds up well, but this masks the finding that the use of such a low sensitivity device would lead to approximately 20% of the active infections going undetected whatever the prevalence.

Hepatitis B 1.Diagnostic background to HBV infection Despite the small size of the hepatitis B virus (HBV, the infectious virion, “Dane Particle”, is 42 nm diameter) and of its genome (just over 3200 base pairs), it is a persistent and pervasive pathogen that causes a wide range of public health, disease, therapeutic and diagnostic problems. The genome is used very efficiently to generate several gene products, mostly utilizing some overlapping regions that control replication and provide the structural proteins for a new infectious virus, as well as help suppress host immune surveillance, and after many years of chronic infection, cause liver disease and hepatic cancer. The viral proteins, and the host response to them, together provide the basis for diagnostic tests that permit the identification of acute and chronic HBV infection, as well as resolved infection. An almost universal marker of active HBV infection is hepatitis B surface antigen (HBsAg) (Figs 1 & 2), although in “occult” or “cryptic” HBV infections, HBsAg levels can dip below the level of detection of HBsAg immunoassays. Fig. 1. Acute HBV infection with recovery Fig. 2. Chronic HBV infection

Page | 607

The appearance of HBsAg in blood is soon followed by hepatitis B e antigen (HBeAg), which is a marker of high levels of viral replication. In acute HBV infection that resolves, HBeAg seroconverts relatively early to anti-HBe, but in chronic HBV infection, seroconversion to anti-HBe may be delayed for many years, HBeAg may persist, or neither may be detectable. An anti-HBc response occurs relatively soon, often within a week or two after the emergence of HBsAg, and is typified by a profound IgM anti-HBc response, which may wane approximately 6 months later. Differentiating between primary acute Fig. 3. Risk of chronicity by age of HBV infection and chronic HBV infection is important because in general the risk of HBV infection becoming chronic is inversely proportional to age, with approximately 90% of children infected in the neonatal period going on to develop chronic HBV, falling to 10% or less by school age (Fig. 3). While IgM anti-HBc has long been employed as a means of differentiating between acute and chronic HBV infection, its reappearance during “flares” in chronic infection (Fig. 2) make it an unreliable indicator of recent primary HBV infection. A potential consequence of misinterpreting a reactive IgM anti-HBc result is that a patient with chronic infection, experiencing flares in liver disease, may not be offered timely treatment and, although they could be re-examined several months later to confirm the original diagnosis of acute HBV, a substantial proportion of such patients may be lost to follow up. Anti-HBc avidity has also been used as a means of distinguishing acute from chronic HBV infection, the early anti-HBc response being of low avidity; however, there is no licensed commercial avidity test currently available. In recent years, the detection and quantification of HBV DNA has been increasingly used either to supplement or replace HBeAg testing as a more direct and accurate measure of active HBV replication. Sensitive nucleic acid tests (NATs) for HBV DNA are able to detect infection early in the acute phase through to resolution, and are now commonly used as a supplement to HBsAg testing for blood and plasma donors, detecting HBV DNA in occasional HBsAg-negative donors. The widespread need for and provision of NAT facilities for the care of HIV patients should facilitate access to NAT for HBV.

2.Aim of testing As for HCV, the preferred testing strategy will depend on the aims of the testing programme as well as factors such as assay performance, technical feasibility, access and cost. As a rule, it should be feasible to translate the strategy into a simple and inexpensive local algorithm. 1. Identify all HBsAg-positive individuals This scenario employs the most widely available and least expensive diagnostics. It would not detect recent infections in the brief period prior to HBsAg seroconversion. However, it is expected that a large majority of such infections in adults will be self-limiting. If the ultimate aim is to offer antiviral treatment to those with an active HBV infection, anti-HBV-positive individuals, or at least a blood specimen, would then have to be referred to a facility able to carry out the necessary procedures to determine whether HBV treatment was needed, and then to monitor its success. Page | 608

Strategy: HBsAg test (RDT or EIA) with or without a different HBsAg test. 2. Identify chronic HCV infections This approach would ascertain whether individuals are carrying replicating HCV. The current cost and complexity of tests for HCV RNA would probably preclude their use as a screening tool in all but areas with the highest prevalence. Consequently, the primary screening test would likely be an anti-HCV test, which would fail to detect acute infections in the seroconversion window phase. This would be partially offset by the use of a combined anti-HCV/HCV Ag assay. Identification of active HCV infection would require the application of a test for HCV RNA, though an immunoassay that is able to detect immune-complexed HCV Ag may be sufficiently accurate. Early infections, a minority of which would have gone on to be spontaneously eradicated by the host, would also be treated under this approach. Anti-HCV test (RDT or EIA) or combined anti-HCV + HCV Ag test followed by an HCV RNA test or an HCV Ag test. 3. Identify chronic HCV infections This approach would need to go a further step to aim 2, by incorporating the means to distinguish chronic from acute infection. Such a test would permit immediate treatment of those identified as having a chronic infection and deferral of those with a recent acute HCV infection. At present, though, there is no appropriate and validated laboratory marker to distinguish chronic from acute HCV infection. While a combination of clinical and non-virological laboratory tests might identify those in need of immediate treatment, others may have to be deferred for 3–6 months to allow a re-test to exclude the minority whose infection cleared spontaneously. Dependent on the costs of antiviral drugs, the risks of unnecessary delay to treatment in those who would benefit from it and of loss to follow up would probably exclude this approach at this time. Strategy:

Appendix 1: Outcomes from Two-Test Strategy to Identify Active HCV Infection Population: 100,000 % Resolved Anti-HCV HCV prevalence 10% 0.45 0.10 0.02 0.004 0.45 0.10 0.02 0.004 0.45 0.10 0.02 0.004 25% 0.45 0.10 0.02 0.004 0.45 0.10 0.02 0.004 0.45 0.10 0.02 0.004 Anti-HCV Test Effective anti-HCV PPV 0.889 0.825 0.603 0.257 0.889 0.825 0.603 0.257 0.889 0.825 0.603 0.257 0.740 0.687 0.502 0.214 0.740 0.687 0.502 0.214 0.740 0.687 0.502 0.214 HCV RNA or Ag Test Sens Spec Strategy Outcomes False Neg 404 90 18 4 1,814 403 81 16 8,262 1,836 367 73 337 75 15 3 1,512 336 67 13 6,885 1,530 306 61 Total Tests 145,348 110,855 102,971 101,394 145,348 110,855 102,971 101,394 145,348 110,855 102,971 101,394 145,381 110,863 102,973 101,395 145,381 110,863 102,973 101,395 145,381 110,863 102,973 101,395 Ratio of TP:FP 794 469 151 34 255 151 49 11 213 126 40 9 283 218 100 27 91 70 32 9 76 59 27 7 Overall sens. 0.990 0.990 0.990 0.990 0.955 0.955 0.955 0.955 0.796 0.796 0.796 0.796 0.990 0.990 0.990 0.990 0.955 0.955 0.955 0.955 0.796 0.796 0.796 0.796 Overall spec. 0.999 1.000 1.000 1.000 0.997 0.999 1.000 1.000 0.997 0.999 1.000 1.000 0.998 1.000 1.000 1.000 0.995 0.999 1.000 1.000 0.995 0.999 1.000 1.000

Sens

Spec

True Pos

False Pos True neg

PPV

NPV

Population

0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995

0.990 0.990 0.990 0.990 0.990 0.990 0.990 0.990 0.990 0.990 0.990 0.990 0.990 0.990 0.990 0.990 0.990 0.990 0.990 0.990 0.990 0.990 0.990 0.990

0.995 0.995 0.995 0.995 0.960 0.960 0.960 0.960 0.800 0.800 0.800 0.800 0.995 0.995 0.995 0.995 0.960 0.960 0.960 0.960 0.800 0.800 0.800 0.800

0.990 0.990 0.990 0.990 0.970 0.970 0.970 0.970 0.970 0.970 0.970 0.970 0.990 0.990 0.990 0.990 0.970 0.970 0.970 0.970 0.970 0.970 0.970 0.970

40,096 8,910 1,782 356 38,686 8,597 1,719 344 32,238 7,164 1,433 287 33,413 7,425 1,485 297 32,238 7,164 1,433 287 26,865 5,970 1,194 239

51 19 12 10 152 57 35 31 152 57 35 31 118 34 15 11 354 102 44 33 354 102 44 33

59,450 90,981 98,188 99,630 59,349 90,943 98,165 99,609 59,349 90,943 98,165 99,609 66,132 92,466 98,485 99,689 65,896 92,398 98,456 99,667 65,896 92,398 98,456 99,667

0.999 0.998 0.993 0.972 0.996 0.993 0.980 0.917 0.995 0.992 0.976 0.902 0.996 0.995 0.990 0.964 0.989 0.986 0.970 0.897 0.987 0.983 0.964 0.879

0.993 0.999 1.000 1.000 0.970 0.996 0.999 1.000 0.878 0.980 0.996 0.999 0.995 0.999 1.000 1.000 0.978 0.996 0.999 1.000 0.905 0.984 0.997 0.999

100000 100000 100000 100000 100000 100000 100000 100000 100000 100000 100000 100000 100000 100000 100000 100000 100000 100000 100000 100000 100000 100000 100000 100000

Page | 609

Annex 6.2 Decision analysis of strategies to identify hepatitis C Predictive Modelling

Jake R. Morgan, MS, Benjamin P. Linas, MD, MPH (Team Lead) Boston University School of Medicine, Boston, USA

September 2015

Page | 610

1.Executive summary We use a decision model to evaluate the positive predictive value (PPV), negative predictive value (NPV), percentage of HCV cases identified, cost of screening, and cost per case of HCV identified of 22 HCV testing strategies in three prevalence settings (2.3%, 10% and 40). The analysis answers the following four questions about HCV testing strategies 1. What is the best test for exposure to HCV, an EIA or RDT? In general, either RDT or EIA provide similar testing outcome and cost. Choice of which test to use should depend on in-country costs and previous investments. In very high prevalence settings, however, loss to follow up between tests for exposure and active infection results in missed cases of HCV. In such settings, it may be best to not use either test, and to instead employ a one-test strategy that uses nucleic acid amplification testing (NAT) or HCV core antigen (HCV core Ag) to directly assess for active HCV infection without previous tests for exposure. 2. Is it beneficial to use two tests for exposure to minimize false-positive results? When it is possible to reflexively conduct repeat testing for HCV exposure using the same specimen, it can be cost saving to employ two tests for exposure in lower (2.3%) and high (10%) prevalence settings. If reflex testing is not possible, however, then there is little value added by the two tests for exposure strategy. In very high prevalence settings (40%), loss to follow up results in strategies that test directly for active HCV infection (no test for exposure) being preferred. 3. Among those with a positive test for exposure to HCV, what is the best test for assessing active infection, qualitative HCV RNA, quantitative HCV RNA, or HCV core Ag? At current cost, qualitative RNA emerges as the preferred test for active HCV infection. Quantitative RNA is slightly less sensitive than qualitative RNA, and therefore results in slightly more false-negative results at similar cost. The difference in testing outcomes is small, however, and countries with previous investment in quantitative RNA platforms, as well as those that can leverage existing HIV testing infrastructure to conduct quantitative HCV RNA testing should do so. Further, if countries anticipate needing quantitative RNA for monitoring HCV therapy, then quantitative RNA should likely be used for HCV testing as well. At its current approximate cost, HCV core Ag results in worse testing accuracy and higher cost than either quantitative or qualitative RNA. 4. How does HCV prevalence impact the choice of testing strategy? Prevalence is central to the choice of testing strategy. There is no one strategy that is appropriate to all prevalence settings. In lower-prevalence settings, the preferred testing strategies limit false-positive results. It can be efficient to use two tests for HCV exposure (assuming that they can be done reflexively and on one specimen). In contrast, in very highprevalence settings, preferred strategies are sensitive and minimize false-negative results. In such settings, the high probability of being infected, as well as the potential for missed cases due to loss to follow up, results in strategies that directly test for active infection (no test for exposure) being preferred.

Page | 611

2.Introduction Hepatitis C virus (HCV) is a global public health burden and major cause of morbidity and mortality, including liver failure and hepatocellular carcinoma.1, 2 With current global HCV seroprevalence estimated to be 2.8%, or more than 185 million infected individuals worldwide,3 and new high-efficacy, low-duration therapy available,4 there is a need to optimize testing strategies for the detection of HCV. Because 10–25% of persons infected with HCV clear the virus without treatment, diagnosis of active HCV infection is typically a two-step process. First, a person is screened for exposure to HCV, typically with an HCV antibody (Ab) test. Those who test positive for exposure then undergo assessment for HCV viral replication, usually with nucleic acid testing (NAT), to confirm the diagnosis of active HCV. This two-step process has multiple barriers to implementation such as loss to follow up between testing for exposure and confirmation of active viral replication, missed pre-seroconversion infection, and limited access to costly NAT techniques, all of which allow for potential missed diagnosis and treatment. There are a variety of modalities for screening for HCV exposure, including enzymelinked immune assays (EIA), and rapid diagnostic tests (RDT). There are also a variety of tests for assessing viral replication, including qualitative and quantitative HCV RNA, as well as a newer testing modality, HCV core antigen (HCV core Ag). It is possible to combine these tests in a large number of permutations to create strategies with either one or two steps in the process of assessing HCV exposure and active infection. The sequence of testing has major implications for the likelihood of missing cases of HCV and of having false-positive test results, as well as directly impacting costs. The large number of potential strategies for testing generates four important questions about how to best use available assays to identify chronic HCV infection: 1. What is the best test for exposure to HCV, an EIA or RDT? 2. Is it beneficial to use two tests for exposure to minimize false-positive results? 3. Among those with a positive test for exposure to HCV, what is the best test for assessing active infection, qualitative HCV RNA, quantitative HCV RNA, or HCV core Ag? 4. How does HCV prevalence impact the choice of testing strategy? In this report, we use decision analytic methods to demonstrate the positive and negative predictive values of a wide array of possible testing strategies in three prevalence settings. We also estimate the average cost of each testing strategy, and the cost per case of HCV detected. The data are useful in making decisions about investment in HCV screening and can provide guidance in choosing the appropriate testing strategy for a given testing venue. Clarification of terminology used in this report Test for exposure – These are assays that test for the presence of anti-HCV antibody in a specimen. A positive test result indicates that a person has been previously exposed to HCV virus, but cannot confirm that the person has active infection. Patients with positive test for exposure require additional testing for active HCV infection. Test for active infection – although most people exposed to HCV establish a chronic infection, approximately 25% spontaneously clear their infection during the acute phase. Such patients will have a positive test for exposure, even though they do not have active HCV. Tests for active

Page | 612

infection assay either nucleic acids (viral load) or presence of HCV core Ag to determine whether a person has active HCV replication and therefore chronic HCV. EIA – enzyme-linked immune assay is one modality for testing for the presence of HCV antibody (HCV Ab) in a specimen. EIA can be used as a test for exposure. RDT – rapid diagnostic tests are another modality for testing for the presence of HCV Ab in a specimen. RDT can be used as a test for exposure. Qualitative RNA – a nucleic acid amplification test (NAT) can be used as a test for active infection. The results of a qualitative RNA are binary (positive/negative) and do not provide detail about the degree of viraemia. Quantitative RNA – a nucleic acid amplification test (NAT) that can be used as a test for active infection. The results of a quantitative RNA provide detail about the number of copies of HCV RNA present in every ml of specimen. HCV core Ag – a test that assays the presence of HCV core antigen (HCV core Ag) in a specimen. HCV core Ag is a product of viral replication. The test can be used as a test of active infection. Two-step strategy – a testing strategy that first employs tests for exposure to identify patients who are likely HCV-infected and rule out the majority of patients who are not infected. Next, patients with a positive test for exposure undergo testing for active HCV infection. This strategy is used most frequently around the world, as it reduces the need for more costly tests for active infection. One-step strategy – a strategy that does not employ any test for exposure, but instead directly assesses for active HCV infection in all patients. Tests in series – when conducting two tests for HCV exposure, it is possible to perform them in series or in parallel. Conducting tests in series means that the first test is performed and if the result is negative, the test is complete and it is considered a negative test. If the first test is positive, only then is the second test performed. If the second test is positive, then the two-test procedure is interpreted as positive.

Considerations for testing strategies Several dynamics impact the outcomes of screening and should inform decision-making about HCV testing: 1. Prevalence in a population – prevalence is central in thinking about HCV testing. In lowprevalence settings, the potential for false-positive tests for exposure to HCV is high. In such settings, it may be beneficial to require two tests for exposure in an effort to minimize false positive test results. In high-prevalence settings, however, the larger risk is missing cases of active HCV infection. In those venues, it is likely optimal to choose strategies that maximize sensitivity and HCV case detection. In some cases, it may even be efficient to forgo all screening for exposure and move directly to confirmation of active viral replication. The role of prevalence in determining outcomes and cost is so great, that it is not possible to consider how to test for HCV without contextualizing the decision to a specific prevalence setting. 2. Cost of diagnostic assays – tests for exposure to HCV are substantially less costly than tests for confirmation of active infection. As a result, it is typically useful to use a two-step strategy for HCV testing. First, a low cost test for exposure rules out the majority of patients who have never been exposed to HCV. Second, those with a reactive test for exposure Page | 613

undergo more costly testing for active infection. In some low-prevalence settings, the relative cost of testing for exposure and active replication may result in it being efficient to use two tests for exposure before acquiring testing for active infection. Presence of existing laboratory infrastructure – previous investments in testing infrastructure can be large. Further, some countries have existing infrastructure for HIV testing that they could leverage for HCV testing. This analysis does not consider previous investment or start-up costs. Small differences in the test characteristics of given strategies should not lead countries to abandon existing platforms for testing or duplicate testing infrastructure. Loss to follow up rates and rates of linkage to care – the cascade of care is important for the cost–effectiveness of efforts to treat HCV, and poor follow up can reduce reach and effectiveness of testing.6 Reducing the number of steps required to diagnose HCV limits the number of times an individual must present and re-present to care and likely improves diagnosis rates. When it is possible to acquire the results of testing for HCV exposure quickly, such that the patient can undergo testing for exposure and replication at the same visit, follow up will improve and more cases will be identified. Further, in some very highprevalence settings, a one-step strategy may be appropriate. Availability of highly effective new therapies – whereas historically it was very difficult to treat HCV, new therapies are oral and provide cure rates approaching 100% with very low toxicity.7–9 The ability to treat and cure HCV in the vast majority of cases identified and started on therapy increases the urgency of designing testing strategies that maximize case identification. Implementation of HCV screening and treatment could reduce the global prevalence of chronic HCV infection. The cost of new therapies – while new treatments to cure HCV are exciting, they are costly and are not available in all settings.10 Some low- and middle-income countries have had success negotiating the cost of new therapy, with 12-week treatment course priced at <$1000 in Egypt and $150 in India, for example.11 Still, the cost of therapy makes accurate diagnosis essential. False-positive screening results will lead to patients who are not infected with HCV being treated with costly drugs.

The optimal screening strategy for a given setting will balance the costs associated with falsenegative and false-positive test results.

3.Methods Overview This decision analysis demonstrates the test characteristics of a variety of strategies for testing for HCV and estimates costs of screening. We consider 22 testing strategies evaluated in three prevalence settings: 2.3%, 10%, and 40%. The data demonstrate the trade-offs between falsepositive and false-negative results, provide projections of the direct cost of HCV testing in different settings, and are useful for decision-making about investment in screening platforms and modalities. We seek to answer the following four questions: 1. What is the best test for exposure to HCV, an EIA or RDT? 2. Is it beneficial to use two tests for exposure to minimize false-positive results? 3. Among those with a positive test for exposure to HCV, what is the best test for assessing active infection, qualitative HCV RNA, quantitative HCV RNA, or HCV core Ag? 4. How does HCV prevalence impact the choice of testing strategy? Page | 614

Strategies A number of testing strategies have been evaluated in the literature, and a recently completed systematic review provides an overview of different strategy configurations. 12 We include three broad types of strategies in our model: (1) those with a single test of HCV exposure followed by testing for active HCV infection, (2) those with two tests of exposure followed by testing for active HCV infection, and (3) those that use a single test to simultaneously assess for both exposure and active infection (Table 1 and Fig. 1). Strategies that employ a single test of exposure first test for the presence of HCV Ab using either EIA or RDT. If non-reactive, we conclude that there is no serological evidence for exposure to HCV, and no further testing occurs. If the assay is reactive, the patient advances to the next step of testing and undergoes an additional test for the presence of active HCV replication, which can be evaluated with a qualitative RNA assay, a quantitative RNA assay, or an HCV core Ag. Strategies that employ two tests for exposure typically use those tests in series. First, test the sample with an anti-HCV assay such as EIA or RDT. If the result is non-reactive, there is no serological evidence of exposure to HCV and no further testing occurs. If the first test for exposure is positive, then the laboratory reflexively conducts a second anti-HCV assay on the same specimen (in our model we assume a single sample is collected for the two tests for exposure strategies and so can be tested again without the patient needing to return). If the second anti-HCV assay is non-reactive, we conclude no evidence of exposure to HCV and no further testing occurs. If positive, the patient proceeds to testing for the detection of active HCV infection with either a core antigen, qualitative RNA, or quantitative RNA assay. Strategies that use a single test to assess for both exposure and active infection employ one of the three tests for active HCV replication (qualitative HCV RNA, quantitative HCV RNA, or HCV core Ag) without previous testing. Positive results indicate both exposure and active infection. Negative results do not provide information about exposure, but confirm that there is no active infection.

Page | 615

Table 1. HCV testing strategies Test for exposure Test for active infection

Fig. 1. HCV testing strategies Single test for exposure

Single test for exposure EIA EIA EIA EIA EIA RDT RDT RDT RDT RDT Antigen Antigen/Qualitative RNA Antigen/Quantitative RNA Qualitative RNA Quantitative RNA Antigen Antigen/Qualitative RNA Antigen/Quantitative RNA Qualitative RNA Quantitative RNA

Two tests for exposure

Two tests for exposure EIA/EIA EIA/EIA EIA/EIA EIA/RDT EIA/RDT EIA/RDT RDT/RDT Antigen Qualitative RNA Quantitative RNA Antigen Qualitative RNA Quantitative RNA

Model structureAntigen We use a deterministic decision RDT/RDT Qualitative RNA model (Excel, Microsoft Corporation, Seattle, WA) to assess the performance of 22 testing strategies in three prevalence settings. The model simulates the RDT/RDT Quantitative RNA of a cohort of patients characterized by a given prevalence of testing experience and outcomes active HCV infection. Among those with chronic HCV, the probability of a positive test result is a No test for exposure function of test sensitivity, whereas among those without active HCV, the probability of a positive test result is a function of 1-specificity. Outcomes from the model include positive Antigen predictive value (PPV) of the entire testing strategy, negative predictive value (NPV) of the strategy, the percentage ofRNA all HCV cases identified, the cost of testing, and the cost per case of Qualitative HCV identified. Quantitative RNA RDT=rapid diagnostic test, EIA=enzymelinked immunoassay

Page | 616

Model inputs Inputs for the model include estimates of screening and diagnostic test characteristics, the costs of screening tests, and the probability that patients return to care when asked to navigate a multiple-step screening strategy (Table 2). Diagnostic performance: whenever possible, test characteristics in this decision analysis reflect the pooled sensitivity and specificity reported by the recently completed systematic reviews and meta-analyses commissioned ahead of the global HCV screening guidance meeting.12–14 When data are not available from meta-analyses, we identify parameter estimates from the medical literature6, 15, 16 and the expert opinion of the research. Table 2. Selected model inputs Variable Test sensitivity RDT EIA Antigen Quantitative RNA Qualitative RNA Test Specificity RDT EIA Antigen Quantitative RNA Qualitative RNA Cascade of Care Accept test 90% – – 17

Value

Sensitivity range

Source

93.0 99.5 93.4 96.0 98.0

91.0–95.0 98.0–99.9 88.7–96.2 90.0–99.0 92.9–99.4

98.0 89.1 98.7 99.0 99.0

97.0–99.0 78.0–95.0 96.9–99.4 95.0–99.9 96.9–99.2

Return for results of test for exposure 74% Test costs RDT $2

$1–$3

Page | 617

EIA Quantitative RNA Qualitative RNA Antigen Fixed costs Laboratory screening Point-of-care screening

$2 $25 $25 $35

$1–$3 $13–$38 $13–$38 $18–$53

Assumption* 18

$15 $20

$8–$23 $10–$30

Assumption*

RDT: rapid diagnostic test, EIA: enzyme-linked immunoassay *Assumptions based on expert opinion of the research team

Costs of testing: the specific cost of each testing strategy in every country likely differs and is changing over time. To estimate costs for various screening and diagnostic platforms we use the recently published Médecins Sans Frontières (MSF) product guide,18 as well as expert opinion of the research team. Costs in this analysis are intended to provide a general idea of the relative cost per testing strategy and cost per identified case for different testing strategies. Actual costs will be specific to individual countries based on purchase volume, currently held testing equipment, and other factors. Consequently, the preferred testing strategy may differ between countries. Analytic approach We first established a base case scenario, which is meant to reflect the current standard of care: 1. We assume that testing is performed in a centralized reference laboratory with a delay of at least one day in obtaining the results of tests for HCV exposure such that patients must represent to the site of screening for testing for active HCV infection. Some individuals are lost to follow up when they fail to return for testing. 2. Those who are lost to follow-up are counted as negative results, affecting the NPV of testing. Additionally, those lost do not accrue costs related to testing for active HCV infection. 3. For strategies that use two tests for HCV exposure, we assume that the tests are performed in series, and that they are performed on the same specimen. As a result, patients who have a negative first test for exposure accrue no additional testing costs. We used the model to estimate PPV, NPV, percentage of cases identified, and costs in three prevalence settings. Next, we considered a scenario in which the results of testing for exposure to HCV are available quickly, such that the person being screened can undergo testing for active HCV infection at the same visit, eliminating the need to return for a second step in the testing cascade.

Page | 618

Finally, we conducted a series of one-way sensitivity analyses, in which we ranged test characteristics and costs to explore critical dynamics that influence decision-making, as well as assess for the potential impact of uncertainty in the input parameters on the results. Because the large number of possible testing strategies makes interpretation difficult, we first eliminate strategies that are clearly not a good use of limited testing resources. First, we eliminate all strategies with test performance below a threshold of 95% PPV and NPV. Next we identify strategies that have both lower NPV and PPV than an alternative, as well as a higher cost of testing. Such strategies would result in a greater number of false positives AND false negatives than another option, while increasing cost. No matter what the cost of HCV therapy, or of untreated HCV, such strategies are never preferred and can be considered “inefficient”. When we eliminate strategies with unacceptable performance, as well as those that were inefficient, the remaining strategies represent tradeoffs between PPV, NPV, and cost. Without formal cost–effectiveness results, it is not possible to make quantitative conclusions about which strategy is preferred. We provide tables, however, that clearly illustrate those trade-offs, likely allow for qualitative conclusions, and that can inform rational discussion and policy-making.

4.Results Which screening test should be preferred, RDT or EIA? Summary: in general, either RDT or EIA provide similar testing outcomes and cost. Choice of which test to use should depend on in country costs and previous investments. In very highprevalence settings, however, it may be best to not use either test, and to instead employ a onetest strategy that uses NAT or HCV core Ag to directly assess for active HCV infection without previous tests for exposure. EIA is a slightly more sensitive test than RDT, but is also less specific. In lowerprevalence settings (2.3%), strategies that utilize EIA result in slightly more false-positive results (lower PPV) and slightly fewer false-negative results (higher NPV) than strategies that employ RDT, at a similar cost for testing. Without formal cost–effectiveness assessments, it is not possible to quantitatively conclude whether screening strategies should focus on minimizing false-positive or false-negative results. Generally speaking, however, there are multiple strategies using either RDT or EIA with PPV and NPV >99% that are likely appropriate for testing. The choice of EIA or RDT should be based largely on cost, in-country feasibility, and previous investments in technologies that may be leveraged for HCV screening. In the higher-prevalence setting (10%), strategies using either RDT or EIA had similar performance and cost. The choice of which platform to use for two tests for exposure should be a function of specific, in-country costs, and previous investments in technology that could be leveraged for HCV screening, while a single EIA may not be an efficient option at this higher prevalence. At very high prevalence (40%), assumptions about loss to follow up between testing for HCV exposure and active infection have an increasingly large impact on NPV. With our assumed loss to follow up (26%), only strategies that employ a single test for active infection have NPV >95%. Further, such one-step strategies identify a substantially greater proportion of HCV infections. What is the value added by performing a second test for exposure? Page | 619

Summary: when it is possible to reflexively conduct repeat testing for HCV exposure using the same specimen, it can be cost saving to employ two tests for exposure in lower (2.3%) and highprevalence (10%) settings. If reflex testing is not possible, however, then there is little value added by the two tests for exposure strategy. In very high-prevalence settings (40%), loss to follow up results in strategies that test directly for active HCV infection (no test for exposure) being preferred. In low-prevalence settings (2.3%), employing a second test for exposure improves PPV slightly and reduces costs by minimizing the need for more costly testing for active HCV infection. Similarly, at higher prevalence (10%), the addition of a second test for exposure improves PPV, has minimal impact in NPV, and reduces cost. It is important to note, however, that the model assumes reflexive testing on one specimen such that patients who have a negative first test for exposure accrue no cost for additional testing. If in reality, a two-test strategy for exposure requires the collection of two specimens, or if there are overhead costs associated with the second test for exposure, then the two-test for exposure strategy likely increases cost with little gain in performance. Finally, in the 40% prevalence setting, with 26% of individuals not returning for results of the test for exposure and confirmatory testing, our model demonstrates that the only strategies with NPV >95% are strategies that test directly for active HCV infection. What diagnostic test is preferred, quantitative HCV RNA, qualitative HCV RNA, or HCV core Ag? Summary: at current cost, qualitative RNA emerges as the preferred test for active HCV infection. Quantitative RNA is slightly less sensitive than qualitative RNA, and therefore results in slightly more false-negative results at similar cost. The difference in testing outcomes is small, however, and countries with previous investment in quantitative RNA platforms, as well as those that can leverage existing HIV testing infrastructure to conduct quantitative HCV RNA testing should do so. Further, if countries anticipate needing quantitative RNA for monitoring HCV therapy, then quantitative RNA should likely be used for HCV testing as well. At its current approximate cost, HCV core Ag results in worse testing accuracy and higher cost than either quantitative or qualitative RNA. HCV core Ag has lower sensitivity and specificity than HCV RNA and is costly. As a result, under the base case assumptions about core Ag cost, all strategies that used core Ag as a test for active infection had lower PPV and NPV at a higher cost than alternative strategies that utilized either quantitative or qualitative HCV RNA. Thus, assuming current cost of core Ag ($35),18 it is not an efficient test for active HCV infection. If, however, HCV core Ag becomes less expensive, it could be a very economically attractive option. In the base case analysis, core Ag was more costly than HCV RNA ($35 vs $25). When core Ag became less costly than RNA ($15 vs $25) it became very appealing, despite having slightly worse test characteristics. For example, with a less costly core Ag test in a 2.3% prevalence setting, two RDTs followed by confirmatory diagnosis using HCV core Ag had the highest PPV (100%) and outstanding NPV (99.1%) with the lowest cost of screening ($17.42). Thus, if countries are able to negotiate substantially lower core Ag prices, or the technology evolves to a lower cost platform, core Ag could be an excellent choice for screening.

Page | 620

Comparing quantitative RNA to qualitative RNA reveals no large differences in screening performance or cost. As a result, choice between these tests should likely be based entirely by in-country feasibility and price. At higher HCV prevalence such as 10%, it is clearly preferred to use the most sensitive diagnostic test possible. Using a less sensitive, more specific diagnostic test leads to falsenegative results while preventing very few false-positive outcomes. Therefore, qualitative HCV RNA emerges as preferred. With the base case cost assumptions, strategies that used quantitative RNA or HCV core Ag provided worse test performance often with higher cost compared to using qualitative HCV RNA. In a high 40% prevalence setting, nearly every individual with positive tests for HCV is a “true positive”. As a result, diagnostic testing should be as sensitive as possible in order to avoid false-negative results. This dynamic is clear in the results of the decision analysis, in which a single qualitative RNA (which is the most sensitive RNA test) is preferred to a single quantitative RNA. How does HCV prevalence impact the choice of testing strategy? Summary: as demonstrated above, prevalence is central to the choice of testing strategy. There is no one strategy that is appropriate to all prevalence settings. In lower-prevalence settings, the preferred testing strategies limit false-positive results. It can be efficient to use two tests for HCV exposure (assuming that they can be done reflexively and on one specimen). In contrast, in very high-prevalence settings, preferred strategies are sensitive and minimize false-negative results. In such settings, the high probability of being infected, as well as the potential for missed cases due to loss to follow up, results in strategies that directly test for active infection (no test for exposure) being preferred. Scenario in which results of tests for exposure are available quickly such that the patient can undergo testing for exposure and active infection at the same visit Same-day results reporting reduces loss to follow up because patients with positive tests for exposure do not need to return to the screening site for testing for active infection. This dynamic is most apparent in very high-prevalence settings (40%). With base case assumptions about loss to follow up, all of the two-step strategies resulted in missed cases of HCV and low NPV. As a result, strategies that test directly for active HCV infection were preferred. When patients can undergo testing for exposure and active infection without the need to leave the testing site and return at another time, many more strategies become viable with PPV and NPV greater than 95% and lower cost per case identified (Tables 6–8). Whenever possible, but especially in very high-prevalence settings, it is always preferred to use same-day tests that allow for testing for exposure and active infection at the same visit.

How to screen summary Integrating all of the scenarios and data above reveals that while not every HCV testing approach is efficient, there are multiple possible strategies for testing for HCV exposure and active infection that countries can employ to test for HCV. The choice of EIA or RDT is not highly significant and should be based on in-country cost and feasibility considerations. In lowerprevalence settings, it is potentially cost saving to use two tests for exposure before testing for active infection, but only when the two tests can be conducted reflexively and on the same specimen. In the highest-prevalence settings, avoiding missed cases becomes the dominating Page | 621

consideration. If possible, it is best to test for exposure and active replication at one visit. If such same-day testing is not feasible, then it is best to employ a one-step strategy that tests directly for active HCV infection without testing for exposure. Table 3. Base case analysis, testing for HCV in a 2.3% prevalence setting Test(s) for exposure RDT/RDT RDT/RDT RDT/RDT RDT/EIA RDT/EIA RDT/EIA EIA/EIA EIA/EIA EIA/EIA RDT RDT RDT RDT RDT EIA EIA EIA EIA EIA Qualitative RNA Quantitative RNA Antigen Test for active infection Qualitative RNA Quantitative RNA Antigen Qualitative RNA Quantitative RNA Antigen Qualitative RNA Quantitative RNA Antigen Qualitative RNA Quantitative RNA Antigen Antigen/Qualitative RNA Antigen/Quantitative RNA Qualitative RNA Quantitative RNA Antigen Antigen/Qualitative RNA Antigen/Quantitative RNA PPV 100.0% 100.0% 100.0% 99.9% 99.9% 99.9% 99.5% 99.5% 99.3% 99.1% 99.1% 98.7% 98.0% 97.9% 95.4% 95.4% 93.9% 90.4% 90.4% 69.7% 69.3% 62.8% NPV 99.1% 99.1% 99.1% 99.2% 99.1% 99.1% 99.0% 99.0% 99.0% 98.7% 98.7% 98.7% 98.8% 98.8% 96.5% 96.5% 96.5% 96.6% 96.6% 99.9% 99.9% 99.8% Cost/ strategy $17.42 $17.42 $17.57 $17.48 $17.48 $17.65 $17.70 $17.70 $17.96 $17.76 $17.76 $18.06 $18.44 $18.44 $19.39 $19.39 $20.35 $22.32 $22.32 $40.00 $40.00 $50.00 Cases identified 56.2% 55.3% 53.8% 60.1% 59.2% 57.6% 64.3% 63.3% 61.6% 60.4% 59.5% 57.9% 61.8% 61.8% 64.6% 63.6% 61.9% 66.2% 66.1% 87.8% 86.4% 84.1% Cost/ identified $1213 $1232 $1278 $1138 $1156 $1200 $1078 $1094 $1141 $1151 $1169 $1222 $1167 $1168 $1175 $1193 $1287 $1320 $1322 $1784 $1812 $2328

HCV: hepatitis C virus, RDT: rapid diagnostic test, EIA: enzyme-linked immunoassay

Page | 622

Table 4. Base case analysis, testing for HCV in 10% prevalence setting Test(s) for exposure RDT/EIA RDT/EIA RDT/EIA RDT/RDT RDT/RDT RDT/RDT EIA/EIA EIA/EIA EIA/EIA RDT RDT RDT RDT RDT EIA EIA EIA EIA EIA Qualitative RNA Quantitative RNA Antigen Test for active infection Qualitative RNA Quantitative RNA Antigen Qualitative RNA Quantitative RNA Antigen Qualitative RNA Quantitative RNA Antigen Qualitative RNA Quantitative RNA Antigen Antigen/Qualitative RNA Antigen/Quantitative RNA Qualitative RNA Quantitative RNA Antigen Antigen/Qualitative RNA Antigen/Quantitative RNA PPV 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 99.9% 99.9% 99.8% 99.8% 99.8% 99.7% 99.6% 99.6% 99.0% 99.0% 98.6% 97.8% 97.8% 91.5% 91.4% 88.9% NPV 96.4% 96.3% 96.1% 96.0% 95.9% 95.7% 96.6% 96.5% 96.3% 96.0% 95.9% 95.7% 96.1% 96.1% 94.2% 94.1% 93.9% 94.4% 94.4% 99.7% 99.6% 99.3% Cost/ strategy $18.94 $18.94 $19.64 $18.78 $18.78 $19.42 $19.25 $19.25 $20.06 $19.05 $19.05 $19.87 $20.32 $20.32 $20.66 $20.66 $22.12 $24.03 $24.03 $40.00 $40.00 $50.00 Cases identified 60.1% 59.2% 57.6% 56.2% 55.3% 53.8% 64.3% 63.3% 61.6% 60.4% 59.5% 57.9% 61.8% 61.8% 64.6% 63.6% 61.9% 66.2% 66.1% 87.8% 86.4% 84.1% Cost/ identified $284 $288 $307 $301 $306 $325 $269 $274 $293 $284 $288 $309 $296 $296 $288 $292 $322 $327 $327 $410 $417 $535

HCV: hepatitis C virus, RDT: rapid diagnostic test, EIA: enzyme-linked immunoassay

Page | 623

Table 5. Base case analysis, testing for HCV in 40% prevalence setting Test(s) for exposure Qualitative RNA Quantitative RNA Antigen EIA/EIA EIA/EIA EIA EIA EIA/EIA RDT RDT RDT/EIA EIA RDT RDT/EIA EIA RDT RDT/EIA EIA RDT RDT/RDT RDT/RDT RDT/RDT Qualitative RNA Quantitative RNA Antigen/Qualitative RNA Antigen/Quantitative RNA Antigen Antigen/Qualitative RNA Antigen/Quantitative RNA Qualitative RNA Qualitative RNA Qualitative RNA Quantitative RNA Quantitative RNA Quantitative RNA Antigen Antigen Antigen Qualitative RNA Quantitative RNA Antigen Test for active infection PPV NPV Cost/ strategy $40.00 $40.00 $50.00 $25.26 $25.26 $30.68 $30.68 $28.25 $27.62 $27.62 $24.61 $25.57 $24.10 $24.61 $25.57 $24.10 $27.36 $29.00 $26.95 $24.10 $24.10 $26.66 Cases identified 87.8% 86.4% 84.1% 64.3% 63.3% 66.2% 66.1% 61.6% 61.8% 61.8% 60.1% 64.6% 60.4% 59.2% 63.6% 59.5% 57.6% 61.9% 57.9% 56.2% 55.3% 53.8% Cost/ identified $103 $104 $134 $88 $90 $104 $104 $103 $100 $101 $92 $89 $90 $94 $90 $91 $107 $105 $105 $97 $98 $111

98.5% 98.5% 98.0% 100.0% 100.0% 99.6% 99.6% 100.0% 99.9% 99.9% 100.0% 99.8% 100.0% 100.0% 99.8% 100.0% 100.0% 99.8% 100.0% 100.0% 100.0% 100.0%

98.3% 97.4% 95.7% 83.7% 83.2% 82.6% 82.5% 82.4% 82.3% 82.3% 81.8% 81.8% 81.6% 81.4% 81.3% 81.1% 80.6% 80.4% 80.3% 80.0% 79.5% 78.8%

HCV: hepatitis C virus, RDT: rapid diagnostic test, EIA: enzyme-linked immunoassay

Table 6. Scenario assuming same-day results for tests for exposure (decreased loss to follow up) in a 2.3% prevalence setting

Page | 624

Test(s) for exposure RDT/RDT RDT/RDT RDT/RDT RDT/EIA RDT/EIA RDT/EIA EIA/EIA EIA/EIA EIA/EIA RDT RDT RDT RDT

Test for active infection Qualitative RNA Quantitative RNA Antigen Qualitative RNA Quantitative RNA Antigen Qualitative RNA Quantitative RNA Antigen Qualitative RNA Quantitative RNA Antigen Antigen/Qualitative RNA Antigen/Quantitative RNA Qualitative RNA Quantitative RNA Antigen Antigen/Qualitative RNA Antigen/Quantitative RNA

PPV

NPV

Cost/ strategy $17.55 $17.55 $17.75 $17.63 $17.63 $17.87 $17.93 $17.93 $18.27 $18.02 $18.02 $18.43 $18.95

Cases identified 84.3% 83.0% 80.8% 90.2% 88.8% 86.4% 96.5% 95.0% 92.5% 90.7% 89.3% 86.9% 92.8%

Cost/ identified $905 $919 $955 $850 $863 $899 $808 $820 $859 $864 $878 $923 $887

100.0% 100.0% 100.0% 99.9% 99.9% 99.9% 99.5% 99.5% 99.3% 99.1% 99.1% 98.7% 98.0%

99.6% 99.6% 99.5% 99.8% 99.7% 99.7% 99.9% 99.9% 99.8% 99.8% 99.7% 99.7% 99.8%

RDT

97.9%

99.8%

$18.95

92.8%

$888

EIA EIA EIA EIA

95.4% 95.4% 93.9% 90.4%

99.9% 99.9% 99.8% 100.0%

$20.23 $20.23 $21.53 $24.19

97.0% 95.5% 92.9% 99.3%

$907 $921 $1 007 $1 059

EIA

90.4%

100.0%

$24.19

99.2%

$1 060

Qualitative RNA Quantitative

69.7%

99.9%

$40.00

97.5%

$1 784

69.3%

99.9%

$40.00

96.0%

$1 812

Page | 625

RNA Antigen 62.8% 99.8% $50.00 93.4% $2 328

HCV: hepatitis C virus, RDT: rapid diagnostic test, EIA: enzyme-linked immunoassay

Table 7. Scenario assuming same-day results for tests for exposure (decreased loss to follow up) in a 10% prevalence setting Test(s) for Exposure RDT/EIA RDT/EIA RDT/EIA RDT/RDT RDT/RDT RDT/RDT EIA/EIA EIA/EIA EIA/EIA RDT RDT RDT RDT RDT EIA EIA EIA EIA EIA Qualitative RNA Quantitative RNA Test for active infection Qualitative RNA Quantitative RNA Antigen Qualitative RNA Quantitative RNA Antigen Qualitative RNA Quantitative RNA Antigen Qualitative RNA Quantitative RNA Antigen Antigen/Qualitative RNA Antigen/Quantitative RNA Qualitative RNA Quantitative RNA Antigen Antigen/Qualitative RNA Antigen/Quantitative RNA PPV 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 99.9% 99.9% 99.8% 99.8% 99.8% 99.7% 99.6% 99.6% 99.0% 99.0% 98.6% 97.8% 97.8% 91.5% 91.4% NPV 98.9% 98.8% 98.5% 98.3% 98.1% 97.9% 99.6% 99.5% 99.2% 99.0% 98.8% 98.6% 99.2% 99.2% 99.7% 99.5% 99.2% 99.9% 99.9% 99.7% 99.6% Cost/ Strategy $19.55 $19.55 $20.50 $19.34 $19.34 $20.21 $19.96 $19.96 $21.06 $19.78 $19.78 $20.89 $21.48 $21.48 $21.94 $21.94 $23.92 $26.50 $26.50 $40.00 $40.00 Cases Identified 90.2% 88.8% 86.4% 84.3% 83.0% 80.8% 96.5% 95.0% 92.5% 90.7% 89.3% 86.9% 92.8% 92.8% 97.0% 95.5% 92.9% 99.3% 99.2% 97.5% 96.0% Cost/ Identified $217 $220 $237 $229 $233 $250 $207 $210 $228 $218 $221 $240 $231 $232 $226 $230 $257 $267 $267 $410 $417

Page | 626

Antigen

88.9%

99.3%

$50.00

93.4%

$535

HCV: hepatitis C virus, RDT: rapid diagnostic test, EIA: enzyme-linked immunoassay

Table 8: Scenario assuming same-day results for tests for exposure (decreased loss to follow up) in a 40% prevalence setting Test(s) exposure EIA/EIA EIA/EIA EIA/EIA RDT RDT EIA EIA EIA EIA EIA Qualitative RNA Quantitative RNA Antigen RDT RDT/EIA RDT RDT/EIA RDT RDT/EIA RDT/RDT RDT/RDT RDT/RDT Qualitative RNA Qualitative RNA Quantitative RNA Quantitative RNA Antigen Antigen Qualitative RNA Quantitative RNA Antigen for Test for active infection Qualitative RNA Quantitative RNA Antigen Antigen/Qualitative RNA Antigen/Quantitative RNA Qualitative RNA Quantitative RNA Antigen Antigen/Qualitative RNA Antigen/Quantitative RNA PPV 100.0% 100.0% 100.0% 99.9% 99.9% 99.8% 99.8% 99.8% 99.6% 99.6% 98.5% 98.5% 98.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% NPV 97.7% 96.8% 95.2% 95.4% 95.4% 98.0% 97.1% 95.5% 99.6% 99.5% 98.3% 97.4% 95.7% 94.1% 93.9% 93.3% 93.1% 91.9% 91.7% 90.5% 89.8% 88.6% Cost/ Strategy $27.88 $27.88 $31.92 $31.35 $31.35 $28.59 $28.59 $33.22 $35.49 $35.49 $40.00 $40.00 $50.00 $26.60 $27.03 $26.60 $27.03 $30.44 $30.74 $26.35 $26.35 $29.81 Cases Identified 96.5% 95.0% 92.5% 92.8% 92.8% 97.0% 95.5% 92.9% 99.3% 99.2% 97.5% 96.0% 93.4% 90.7% 90.2% 89.3% 88.8% 86.9% 86.4% 84.3% 83.0% 80.8% Cost/ Identified $72 $73 $86 $84 $84 $74 $75 $89 $89 $89 $103 $104 $134 $73 $75 $74 $76 $88 $89 $78 $79 $92

HCV: hepatitis C virus, RDT: rapid diagnostic test, EIA: enzyme-linked immunoassay

Page | 627

References 1. de Oliveria Andrade LJ, D'Oliveira A, Melo RC, De Souza EC, Costa Silva CA, Paraná R. Association between hepatitis C and hepatocellular carcinoma. J Global Infect Dis. 2009;1(1):33–7. doi: 10.4103/0974-777X.52979. PubMed PMID: PMC2840947. Cooke GS, Lemoine M, Thursz M, Gore C, Swan T, Kamarulzaman A, et al. Viral hepatitis and the Global Burden of Disease: a need to regroup. J Viral Hepat. 2013;20(9):600 –1. doi: 10.1111/jvh.12123. PubMed PMID: 23910643. Hanafiah KM, Groeger J, Flaxman AD, Wiersma ST. Global epidemiology of hepatitis C virus infection: new estimates of age-specific antibody to HCV seroprevalence. Hepatology. 2013;57(4):1333 –42. doi: 10.1002/hep.26141. Younossi Z, Henry L. Systematic review: patient-reported outcomes in chronic hepatitis C – the impact of liver disease and new treatment regimens. Aliment Pharmacol Ther. 2015;41(6):497–520. Epub 2015/01/24. doi: 10.1111/apt.13090. PubMed PMID: 25616122. Aoyagi K, Iida K, Ohue C, Matsunaga Y, Tanaka E, Kiyosawa K, et al. Performance of a conventional enzyme immunoassay for hepatitis C virus core antigen in the early phases of hepatitis C infection. Clin Lab. 2001;47(3 – 4):119-27. PubMed PMID: 11294574. Linas BP, Barter DM, Leff JA, Assoumou SA, Salomon JA, Weinstein MC, et al. The hepatitis C cascade of care: identifying priorities to improve clinical outcomes. PLoS One. 2014;9(5):e97317. doi: 10.1371/journal.pone.0097317. PubMed PMID: 24842841; PubMed Central PMCID: PMC4026319. Afdhal N, Reddy KR, Nelson DR, Lawitz E, Gordon SC, Schiff E, et al. Ledipasvir and sofosbuvir for previously treated HCV genotype 1 infection. N Engl J Med. 2014;370(16):1483–93. doi: doi:10.1056/NEJMoa1316366. PubMed PMID: 24725238. Poordad F, Hezode C, Trinh R, Kowdley KV, Zeuzem S, Agarwal K, et al. ABT-450/r–Ombitasvir and dasabuvir with ribavirin for hepatitis C with cirrhosis. N Engl J Med. 2014;370(21):1973–82. doi: 10.1056/NEJMoa1402869. PubMed PMID: 24725237. Zeuzem S, Dusheiko GM, Salupere R, Mangia A, Flisiak R, Hyland RH, et al. Sofosbuvir and ribavirin in HCV genotypes 2 and 3. N Engl J Med. 2014;370(21):1993–2001. doi: 10.1056/NEJMoa1316145. PubMed PMID: 24795201.

2. 3.

10. Reau NS, Jensen DM. Sticker shock and the price of new therapies for hepatitis C: is it worth it? Hepatology. 2014;59(4):1246–9. doi: 10.1002/hep.27039. PubMed PMID: 24493069. 11. Lemoine M, Thursz M. Viral hepatitis: scaling up HCV treatment in resource-limited countries. Nat Rev Gastroenterol Hepatol. 2015;12(4):193–4. doi: 10.1038/nrgastro.2015.31. PubMed PMID: 25708046. 12. Boeras D, Amini A, Falconer J, Kelly H, Peeling R, Tang W, et al. PICO 4: diagnostic strategies for hepatitis C antibody detection: a meta-analysis and review of the literature 2015; In press. 13. Tang W, Chen W, Amini A, Boeras D, Falconer J, Kelly H, et al. PICO 2: Diagnostic accuracy of tests to detect hepatitis C antibody: a meta-analysis and review of the literature 2015; In press. 14. Freiman JM, Tran TM, Schumacher SG, White LF, Cohn J, Linas BP, et al. HCV core antigen testing for presence of active HCV infection and monitoring for treatment response and cure: a systematic review. 2015; In press. 15. Linas BP, Wong AY, Schackman BR, Kim AY, Freedberg KA. Cost-effective screening for acute hepatitis C virus infection in HIV-infected men who have sex with men. Clin Infect Dis. 2012;55(2):279–90. doi: 10.1093/cid/cis382. PubMed PMID: 22491339; PubMed Central PMCID: PMCPMC3403839. 16. Schirm J, van Loon AM, Valentine-Thon E, Klapper PE, Reid J, Cleator GM. External quality assessment program for qualitative and quantitative detection of hepatitis C virus RNA in diagnostic virology. J Clin Microbiol. 2002;40(8):2973–80. PubMed PMID: 12149361; PubMed Central PMCID: PMCPMC120662. 17. Rein DB, Smith BD, Wittenborn JS, Lesesne SB, Wagner LD, Roblin DW, et al. The cost-effectiveness of birthcohort screening for hepatitis C antibody in US primary care settings. Ann Intern Med. 2012;156(4):263 –70. doi: 10.7326/0003-4819-156-4-201202210-00378.

Page | 628

18. Médecins Sans Frontières. Putting HIV and HCV to the test: a product guide for point-of-care CD4 and laboratorybased and point-of-care virological HIV and HCV tests. Geneva, Switzerland: MSF; 2015. 19. Mohd Hanafiah K, Groeger J, Flaxman AD, Wiersma ST. Global epidemiology of hepatitis C virus infection: new estimates of age-specific antibody to HCV seroprevalence. Hepatology. 2013;57(4):1333–42. doi: 10.1002/hep.26141. PubMed PMID: 23172780. 20. Mohamoud YA, Mumtaz GR, Riome S, Miller D, Abu-Raddad LJ. The epidemiology of hepatitis C virus in Egypt: a systematic review and data synthesis. BMC Infect Dis. 2013;13:288. doi: 10.1186/1471-2334-13-288. PubMed PMID: 23799878; PubMed Central PMCID: PMCPMC3702438. 21. Hope VD, Eramova I, Capurro D, Donoghoe MC. Prevalence and estimation of hepatitis B and C infections in the WHO European Region: a review of data focusing on the countries outside the European Union and the European Free Trade Association. Epidemiol Infect. 2014;142(2):270–86. doi: 10.1017/S0950268813000940. PubMed PMID: 23714072; PubMed Central PMCID: PMCPMC3891474. 22. Nelson PK, Mathers BM, Cowie B, Hagan H, Des Jarlais D, Horyniak D, et al. Global epidemiology of hepatitis B and hepatitis C in people who inject drugs: results of systematic reviews. Lancet. 2011;378(9791):571 –83. doi: 10.1016/S0140-6736(11)61097–0. PubMed PMID: 21802134; PubMed Central PMCID: PMCPMC3285467.

Page | 629

ANNEX 7. Summary of Declared Interests Participants Isabelle Andrieux-Meyer Declared interests Personal: None Institutional: Employer (MSF) has received a UNITAID grant HIV-HCV grant to the value of 14 Million USD (current). Personal: None Institutional: Participation in the meeting covered by ECDC The outcomes of the meeting will substantially impact ECDC’s work on the subject. Personal: Travel support from MSD and Roche Institutional: None Personal: None Institutional: President of World Hepatitis Alliance that receives grants and other support from BMS, Boehringer Ingelheim, Janssen, Roche, Roche diagnostics, Merck, GSK, Gilead, AbbVie, Abbott, Achillion, Bayer, Novartis Personal: None Institutional: None Personal: Research support from Gilead and Abbvie to support an investigator initiated independent research study. Provided expert opinion to Australian Parliamentary committee on hepatitis. Institutional: None Personal: None Institutional: None Personal: None Institutional: None Personal: None Institutional: None Personal: Expert advisor on other WHO panels (diagnostics and testing technologies) Institutional: Past research on DBS testing supported by Gilead (no longer current, study complete). Personal: None Institutional: None £30,000 (no longer current) Declared amount US$14 million Level of participation Full

Lara Tavoschi

Full

Manal El-Sayed

Not declared

Full

Charles Gore

not declared; but all the funding for the organization comes from these firms

Full

Richard Njouom Margaret Hellard (Chair)

Full 3.5 million plus drugs (research support was not for income) Full

Niklas Luhmann Teri Roberts Jacinto Amandua John Parry

Full Full Full Full

Khwairakpam Giten Singh

Full

Page | 630

Roger Chou (methodologist) Michael Ninburg Trevor Peter

Personal: None Institutional: None Personal: None Institutional: None Personal: None Institutional: None

Full Full Full

Page | 631

ANNEX 8. Guideline Development group, Guideline Steering Group, Systematicreview teams, External Review Group GUIDELINES DEVELOPMENT GROUP Jacinto Amandua Ministry of Health Uganda Isabelle Andrieux-Meyer Médecins Sans Frontières Geneva, Switzerland Manal Hamdy El-Sayed Egypt National Hepatitis Committee Egypt Charles Gore World Hepatitis Alliance London, United Kingdom Margaret Hellard (Chair) Centre for Population Health Burnet Institute Melbourne, Australia Niklas Luhmann Médecins du Monde Paris, France Michael Ninburg Hepatitis Education Project Seattle, USA Richard Njouom Centre Pasteur of Cameroon Cameroon, Africa John Parry Public Health England London, United Kingdom Trevor Peter Clinton Health Access Initiative New York, USA Teri Roberts Foundation for Innovative New Diagnostics Geneva, Switzerland

Giten Khwairakpam Singh TREAT Asia/amFAR Bangkok, Thailand Lara Tavoschi European Center for Disease Prevention and Control Stockholm, Sweden Roger Chou Oregon Health & Science University Portland, USA Methodologist Richard Tedder (Unable to attend) Public Health England London, United Kingdom

Page | 632

GUIDELINES STEERING COMMITTEE GROUP Rachel Baggaley Coordinator, HIV Key Populations and Innovative Prevention Nicolas Campion Clark Medical Officer, HQ/MSB Management of Substance Shaffiq Essajee Medical Officer, Treatment and Care, HIV/AIDS Cheryl Johnson Consultant, HIV Key Populations and Innovative Prevention Anita Sands Technical Officer, Essential Medicines & Health Products Willy Urassa Essential Medicines & Health Products Marco Vitoria Medical Officer, Treatment and Care, HIV/AIDS Junping YU Technical officer Blood & Transfusion Safety (BTS) Service Delivery and Safety (SDS)

Page | 633

SYSTEMATIC REVIEW TEAMS Rosanna Peeling London School of Hygiene and Tropical Medicine London, United Kingdom Team Lead Claudia Denkinger Foundation for Innovative New Diagnostics Geneva, Switzerland Team Lead Jennifer Cohn Médecins Sans Frontières Geneva, Switzerland Team Lead Timothy Hallet Imperial College London, United Kingdom Team Lead Benjamin P Linas Boston University School of Medicine Boston, USA Team Lead Kali Zhou University of California, Department of Medicine San Francisco, USA Team Lead Thomas Fitzpatrick University of Washington, School of Medicine Seattle, USA Team Lead Nick Walsh WHO Regional Office for the Western Pacific Team Lead Roger Chou Oregon Health & Science University Portland, USA Methodologist Jake Morgan Boston University School of Medicine Boston, USA Shevanthi Nayagam Imperial College London, United Kingdom Berit Lange Center for Chronic Immunodeficiency & Division for Infectious Diseases, Medical Department II, University Hospital Freiburg, Germany Ali Amini London School of Hygiene and Tropical Medicine London, United Kingdom Helen Kelly London School of Hygiene and Tropical Medicine London, United Kingdom Debra Boeras London School of Hygiene and Tropical Medicine London, United Kingdom Wen Chen London School of Hygiene and Tropical Medicine, London, United Kingdom Jane Falconer London School of Hygiene and Tropical Medicine London, United Kingdom Weiming Tang London School of Hygiene and Tropical Medicine London, United Kingdom Olivia Varsaneux London School of Hygiene and Tropical Medicine London, United Kingdom Teri Roberts Foundation for Innovative New Diagnostics Geneva, Switzerland Edouard Tuaillon Page | 634

Montpellier Teaching Hospital Montpellier, France Joseph Tucker UNC Project -China University of North Carolina, USA Philippe Van de Perre Université Montpellier – EFS & CHU Montpellier, France

University of Utah Salt Lake City, USA Timothy Hallet Imperial College London, United Kingdom Team Lead Shevanthi Nayagam Imperial College London, United Kingdom

Mellanye Lackey Ji Young Kim WHO Regional Office for the Western Pacific Julia Scott WHO Regional Office for the Western Pacific

Page | 635

MODELLING TEAM Benjamin P Linas Boston University School of Medicine Boston, USA Jake Morgan Boston University School of Medicine Boston, USA John Parry Public Health England London, United Kingdom

VALUES AND PREFERENCES SURVEY Elena Ivanova Foundation for Innovative New Diagnostics Geneva, Switzerland Alessandra Trianni Foundation for Innovative New Diagnostics Geneva, Switzerland Teri Roberts Foundation for Innovative New Diagnostics Geneva, Switzerland FEASIBILITY SURVEY Niklas Luhmann Médecins du Monde Paris, France Julie Bouscaillou Médecins du Monde Paris, France Azumi Ishizaki WHO headquarters

Page | 636

EXTERNAL PEER REVIEW GROUP Jilian Sacks Clinton Health Access Initiative USA Tanya Applegate Kirby Institute Australia Cami Graham Beth Israel Deaconess Medical Center USA Gilles Wandeler University of Bern Switzerland Mark Sonderup University of Cape Town South Africa Ponsiano Ocama Makerere University Uganda Stephen Locarnini Doherty Institute Australia Yvan Hutin WHO Headquarters Hande Harmanci WHO Headquarters Stefan Wiktor WHO Headquarters

Alaa Gad Hashish Al Shams University Egypt Jean-Bosco Ndinokubwayo WHO Regional Office for Africa Nick Walsh WHO Regional Office for the Western Pacific Yap Boum Epicentre, Medecin sans Frontiere France Susan Best National Serology Reference Laboratory Australia Joumana Hermez WHO, Egypt Jules Mugabo Semahore WHO, Rwanda

Page | 637

Page | 638

Page | 639

Page | 640

Informations clés

Organisation Organisation mondiale de la santé (OMS)

Type de document Publications

Date d'adoption 16 février 2017

Source Organisation mondiale de la santé