Structured approaches for the screening and diagnosis of childhood tuberculosis in a high prevalence region of South Africa

Bull World Health Organ 2010;88:312–320 | doi:10.2471/BLT.09.062893312 Introduction Despite the scale of the worldwide tuberculosis epidemic, the disease remains very difficult to diagnose in children, especially in regions with limited resources.1 Childhood tuberculosis is often paucibacillary and the diagnosis rests on interpretation of chest radiograph findings and non-specific symptoms and signs.1 Improving diagnostic accuracy and reliability is key to integrating childhood tuberculosis into national control programmes, and the World Health Organization (WHO) has thus prioritized diagnostic criteria for childhood tuberculosis.2 Objective, repro- ducible tuberculosis diagnosis will also be pivotal for defining end-points in trials of new tuberculosis vaccines.3 The need for accurate diagnosis is felt most acutely among younger children, who contribute substantially to the burden of tuberculosis in high prevalence regions.4–7 Routine clinical use of a structured diagnostic approach that is unsuited to a particular setting can result in systematic errors in estimating the burden of tuberculosis and in patient management. It follows that regional guidelines for screening and diagnosis of childhood tuberculosis should be tailored to their epidemiological context. The relative merits of existing structured diagnostic ap- proaches are debatable.5,8–17 Hesseling et al. reviewed 16 such approaches and noted that few of the scoring systems, algorithms and classifications for the screening and diagnosis of childhood tuberculosis have been validated against a gold standard. Most have been developed for hospital-based studies and their useful- ness in community settings is relatively unknown.5,18–21 Some have suggested that structured diagnostic approaches should be used only as screening tools to select children for further investi- gation,9,10 while others have proposed a simplified case definition of childhood tuberculosis, based on cardinal symptoms, as an alternative to complex diagnostic systems.1,22 Existing structured approaches to childhood tuberculosis provide a logical and reproducible basis for diagnosis based on clinical acumen, which Cundall termed “the art of the possible”.23 However, we hypothesized that commonly used, structured ap- proaches for screening and diagnosing childhood tuberculosis may show poor agreement and yield highly variable case fre- quency results. The objectives of this paper were to quantify the tuberculosis case frequencies obtained by means of nine different diagnostic systems, to assess agreement between systems, and to offer possible explanations for discordant findings. Methods This analysis is based on data collected during a bacille Calmette-Guérin (BCG) vaccine trial conducted by the South African Tuberculosis Vaccine Initiative (SATVI) from March Une traduction en français de ce résumé figure à la fin de l’article. Al final del artículo se facilita una traducción al español. .ةلاقلما هذهل لماكلا صنلا ةياهن في ةصلاخلا هذهل ةيبرعلا ةمجترلا Objective To measure agreement between nine structured approaches for diagnosing childhood tuberculosis; to quantify differences in the number of tuberculosis cases diagnosed with the different approaches, and to determine the distribution of cases in different categories of diagnostic certainty. Methods We investigated 1445 children aged < 2 years during a vaccine trial (2001–2006) in a rural South African community. Clinical, radiological and microbiological data were collected prospectively. Tuberculosis case status was determined using each of the nine diagnostic approaches. We calculated differences in case frequency and categorical agreement for binary (tuberculosis/not tuberculosis) outcomes using McNemar’s test (with 95% confidence intervals, CIs) and Cohen’s kappa coefficient (Κ ). Findings Tuberculosis case frequency ranged from 6.9% to 89.2% (median: 41.7). Significant differences in case frequency (P < 0.05) occurred in 34 of the 36 pair-wise comparisons between structured diagnostic approaches (range of absolute differences: 1.5–82.3%). Kappa ranged from 0.02 to 0.71 (median: 0.18). The two systems that yielded the highest case frequencies (89.2% and 70.0%) showed fair agreement (Κ : 0.33); the two that yielded the lowest case frequencies (6.9% and 10.0%) showed slight agreement (Κ : 0.18). Conclusion There is only slight agreement between structured approaches for the screening and diagnosis of childhood tuberculosis and high variability between them in terms of case yield. Diagnostic systems that yield similarly low case frequencies may be identifying different subpopulations of children. The study findings do not support the routine clinical use of structured approaches for the definitive diagnosis of childhood tuberculosis, although high-yielding systems may be useful screening tools. Structured approaches for the screening and diagnosis of childhood tuberculosis in a high prevalence region of South Africa Mark Hatherill,a Monique Hanslo,a Tony Hawkridge,b Francesca Little,c Lesley Workman,a Hassan Mahomed,a Michele Tameris,a Sizulu Moyo,a Hennie Geldenhuys,a Willem Hanekom,a Lawrence Geiterd & Gregory Husseya a School of Child and Adolescent Health, University of Cape Town, Anzio Road, Cape Town, 7925, South Africa. b Aeras Global TB Vaccine Foundation, Rockville, United States of America (USA). c Department of Statistical Sciences, University of Cape Town, Cape Town, South Africa. d Otsuka Pharmaceutical Development and Commercialization Inc., Rockville, USA. Correspondence to Mark Hatherill (e-mail: mark.hatherill@uct.ac.za). (Submitted: 9 January 2009 – Revised version received: 11 September 2009 – Accepted: 7 October 2009 – Published online: 29 December 2009 ) Bull World Health Organ 2010;88:312–320 | doi:10.2471/BLT.09.062893 313 Mark Hatherill et al. Screening and diagnosis of childhood tuberculosis in South Africa Research 2001 to August 2006 near Cape Town, South Africa (clinical trials identifier: NCT00242047).8 In the Boland-Over- berg region of South Africa, tuberculosis incidence among children aged < 2 years was estimated as > 3000 cases per 100 000 in 2006.6,8,24 In the trial, which compared the vaccine efficacy obtained with percu- taneous versus intradermal Tokyo-172 BCG, 11 680 neonates were followed up for a minimum of 2 years after vac- cination.8 Children in the community suspect- ed of having tuberculosis due to a history of contact with an adult case or to the presence of symptoms compatible with the disease were identified by a regional surveillance system. All such children un- derwent comprehensive radiological and bacteriological investigation, even if they had no symptoms. The presence and dura- tion of cough, wheezing, fever or weight loss; the response to antibiotics; and the proximity of contact with an adult hav- ing tuberculosis (mother, other person within the household, person outside of the household), were recorded. Human immunodeficiency virus (HIV) status was determined by a rapid antibody test and, if the result was positive, confirmatory polymerase chain reaction (PCR) was performed as well. Tuberculin skin tests included both Mantoux and Tine. Chest radiographs (anteroposterior and lateral) were reviewed by three paediatricians and classified in terms of the likelihood of tuberculosis (Table 1). Two consecu- tive, paired gastric lavages and induced sputum samples were obtained for smear microscopy and culture of Mycobacterium tuberculosis using mycobacteria growth indicator tubes (Becton Dickinson and Co., Sparks, MD, United States of America). A diagnostic algorithm was developed, based on approaches described by Cundall and WHO, for objective post hoc determination of tuberculosis status as the trial end-point.21,23 The decision to start tuberculosis treatment was made on discharge by the attending clinician on the basis of all available results, inde- pendent of the assigned trial end-point. A protocol-specified objective was to compare the structured approaches used to diagnose childhood tuberculosis in de- veloping countries with a high prevalence of tuberculosis and limited resources. Diagnostic approaches relevant to sub-Sa- haran Africa, dating from 1990 onwards, were selected by literature review and expert consultation. Recent modifications were preferred over versions predating the HIV era. Eight structured approaches were compared with the SATVI trial al- gorithm for tuberculosis case frequency.8 The country of origin, lineage and type of approach are summarized in Table 2. Structured diagnostic approaches were categorized as follows: i) binary, with the diagnosis being sim- ply positive or negative (yes = tuber- culosis; no = not tuberculosis);12,15 ii) hierarchical, with stratification into categories of diagnostic certainty, such as “definite”, “probable”, “pos- sible”, “unlikely” or “not tuberculo- sis”;8,14,16 or iii) numerical, with a score obtained by adding the weighted values assigned to each variable (score ≥ x = tubercu- losis).9–11,13 Data for the variables used in these diagnostic approaches were collected prospectively during the trial. Missing variables were assigned a zero value. Ref- erenced threshold values were used for the analysis unless cut-off thresholds were unspecified, and trial algorithm values were used as the default.8 To standardize reporting, the terms for the hierarchi- cal categories of diagnostic certainty were “unlikely/not”, “possible”, “prob- able”, and “definite” tuberculosis.11,13,14,16 Details of the various diagnostic ap- proaches are provided in Appendix A (Available at: http://vacfa.com/index. php?option=com_content&view=secti on&layout=blog&id=10&Itemid=10). The variables required by each sys- tem to compute a tuberculosis outcome for each child were programmed using STATA version 10 (StataCorp, Inc., College Station, TX, USA). Tuberculosis cases were defined by: i) “positive” classification for binary (tuberculosis/not tuberculosis) sys- tems; ii) “definite”, “probable” or “possible” classification for hierarchical sys- tems; or iii) score ≥ the specified cut-off for nu- merical scoring systems. The analysis of binary outcomes compared the nine diagnostic approaches in terms of the number and percentage of tuberculosis cases diagnosed among the children investigated. McNemar’s test was used to compare the paired proportions of tuberculosis cases diagnosed with each system. P-values were not manipulated to adjust for multiple comparisons. Co- hen’s kappa coefficient (Κ) was used to examine agreement between individual observations for each system. Weighted Κ Table 1. Results of chest radiograph assessment by three independent paediatric reviewers, grouped by certainty of tuberculosis diagnosis, South Africa, 2001–2006 Diagnostic certaintya Reviewer 1 Reviewer 2 Reviewer 3 Final classification No. % No. % No. % No. % Highly likely to have tuberculosis 16 1.1 29 2.0 171 11.8 Likely to have tuberculosis 20 1.4 38 2.6 323 22.4 Suspected of having tuberculosis 124 8.6 145 10.0 242 16.7 Positive 160 11.1 212 14.6 736 50.9 271 18.8 Inconclusive 45 3.1 35 2.4 82 5.7 Abnormal but not tuberculosis 102 7.1 139 9.6 312 21.6 Normal 1038 71.8 778 53.9 59 4.1 Negative 1185 82.0 952 65.9 453 31.4 1174 81.2 Not read 100 6.9 281 19.5 256 17.7 Total 1445 100 1445 100 1445 100 1445 100 a “Highly likely to have tuberculosis”, “likely to have tuberculosis” and “suspected of having tuberculosis” were classified as positive; “inconclusive”, “abnormal but not tuberculosis” and “normal” were classified as negative. Final chest radiograph classification was determined by agreement of at least two reviewers. Bull World Health Organ 2010;88:312–320 | doi:10.2471/BLT.09.062893314 Mark Hatherill et al.Screening and diagnosis of childhood tuberculosis in South Africa Research statistics were calculated for systems with hierarchical classifications. The degree of agreement was defined by the following values of Κ: 0–0.2 = slight; 0.2–0.4 = fair; 0.4–0.6 = moderate; 0.6–0.8 = substan- tial; and 0.8–1.0 = nearly perfect.25 In total, 1869 case episodes involving 1654 children were investigated, and one case episode was selected for each child. Since children older than 2 years were excluded, 1445 children were included in this analysis. Results The median age at investigation was 11.4 months (interquartile range: 6.0–17.4). Contact with an adult with tuberculosis was reported for 952 children (65.9%), and 628 children (43.5%) had cough lasting > 2 weeks. Weight was recorded as being 60–80% of expected weight–for– age in 316 (21.9%) children and as being < 60% of expected weight–for–age in 29 children (2.0%). Of the 1445 children studied, 54 (3.7%) tested positive for HIV with enzyme-linked immunosorbent assay, and 28 of these children (1.9%) were confirmed positive for HIV by polymerase chain reaction (PCR) assay. The chest radiograph was compatible with tuberculosis in 271 children (18.8%) and Mycobacterium tuberculosis was cultured from induced sputum or gastric lavage in 172 children (11.9%). Treatment for tuberculosis was started by the attending clinician in 611 children (42.3%). Comparison of binary outcomes Fig. 1 illustrates the number and percent- age of tuberculosis cases diagnosed with each system. The median tuberculosis case frequency was 41.7% (602 of the 1445 children investigated). Differences in tuberculosis case frequency are shown in Table 3. The dif- ferences were significant (P < 0.05) in 34 of 36 possible pair-wise comparisons between the various structured diagnos- tic approaches. Only the comparisons between the Stegen–Toledo and SATVI approaches and between the Stoltz– Donald and Fourie approaches yielded non-significant differences. The pair-wise Table 2. Nine structured approaches for diagnosing childhood tuberculosis Approach Year Origin Source data Classification Purpose Lineage WHO–Harries10 1996 International Clinical Numerical Diagnosis Based on Keith Edwards criteria18 (Papua New Guinea) Fourie9 1998 International Clinical Numerical Screening High tuberculosis prevalence areas9 Osborne14 1995 Zambia Clinical Radiological Bacteriological Hierarchical Screening Adapted from Cundall23 (Kenya) and WHO21 Migliori12 1992 Uganda Clinical Radiological Bacteriological Binary Diagnosis Derived from Ghidey and Habte19 (Ethiopia) Stegen–Toledo13 2003 Peru Clinical Radiological Bacteriological Numerical Diagnosis Adapted from Stegen–Jones20 (Chile) MASA15 1996 South Africa Clinical Radiological Binary Diagnosis Clinical practice guideline, MASA Stoltz–Donald16 1990 South Africa Clinical Radiological Bacteriological Hierarchical Screening Adapted from Cundall23 (Kenya) and WHO21 Kibel11 1999 South Africa Clinical Radiological Bacteriological Numerical Diagnosis Clinical practice guideline, UCT SATVI vaccine trial algorithm8 2006 South Africa Clinical Radiological Bacteriological Hierarchical Diagnosis Adapted from Cundall23 (Kenya) and WHO21 MASA, Medical Association of South Africa; SATVI, South African Tuberculosis Vaccine Initiative; UCT, University of Cape Town; WHO, World Health Organization. Fig. 1. Frequency of cases classified as tuberculosis with various scoring systems, with hierarchical and numerical outcomes condensed to a binary “tuberculosis/not tuberculosis” output , South Africa, 2001–2006 0 Tu be rc ul os is c as es (n = 1 44 5) 1400 MASA 1200 1000 800 600 400 200 n = 99 (6.9%) Migliori n = 602 (41.7%) Stoltz– Donald n = 417 (28.9%) SATVI n = 739 (51.1%) Osborne n = 1289 (89.2%) WHO– Harries n = 145 (10.0%) Fourie n = 440 (30.4%) Stegen– Toledo n = 772 (53.4%) Kibel n = 1011 (70.0%) Binary Hierarchical Numerical MASA, Medical Association of South Africa; SATVI, South African Tuberculosis Vaccine Initiative; WHO, World Health Organization. Bull World Health Organ 2010;88:312–320 | doi:10.2471/BLT.09.062893 315 Mark Hatherill et al. Screening and diagnosis of childhood tuberculosis in South Africa Research differences in tuberculosis case frequency ranged from 1.5% to 82.3%. Table 4 summarizes the observed agreement between all structured di- agnostic approaches and shows the Κ statistics for binary “tuberculosis/not tuberculosis” outcomes. For the 36 pair- wise comparisons, Κ ranged from 0.02 to 0.71 (median Κ: 0.18). Two systems based on clinical, radiological and bacteriological source data (Osborne and Kibel) generated the highest tuberculosis case frequencies, yet showed only fair agreement. Four systems – MASA, Osborne, Fourie and WHO–Harries – demonstrated poor to fair agreement with all of the structured diagnostic approaches analysed. Notably, two numerical systems – MASA and WHO–Harries– classified the fewest case episodes as tuberculosis, but showed only slight agreement. Comparison of hierarchical outcomes The distribution of diagnoses in catego- ries of ascending diagnostic certainty is illustrated for three hierarchical and two numerical-hierarchical scoring systems (Fig. 2). The distribution of the diagnostic categories assigned by the Osborne and Kibel systems was similar: a bell-shaped curve with most diagnoses grouped in the “possible” and “probable” categories. By contrast, the Stegen–Toledo and Stoltz–Donald systems yielded results with opposite distributions, with most cases in the “not”/“unlikely” or “definite” categories. Table 5 summarizes the observed agreement and weighted Κ for hierarchi- cal and numerical-hierarchical systems across categories of increasing diagnostic certainty. Hierarchical agreement was nearly perfect between SATVI and Stoltz–Donald, and substantial between Kibel and Osborne. Comparison of numerical outcomes Tuberculosis case frequency ranged from 10.0% to 70.0% across four numerical scoring systems (Kibel, Fourie, WHO– Harries and Stegen–Toledo) when set at the pre-specified threshold (Fig. 3). Rela- tive to the observed distribution of scores, two of the numerical systems (Kibel and Stegen–Toledo) used a low threshold for tuberculosis diagnosis, resulting in case frequencies of 70.0% and 53.4%, respec- tively. The other two systems (Fourie and WHO–Harries) used a relatively high diagnostic threshold, resulting in case frequencies of only 30.4% and 10.0%. Discussion The most striking finding of this study was the wide variation (6.9–89.2%) in the frequency of tuberculosis cases diag- nosed with the nine structured diagnostic systems. The fact that the differences in tuberculosis case frequency were statisti- cally significant for all but two of 36 pos- sible paired comparisons between systems suggests that the burden of childhood tuberculosis in a given population could be under- or overestimated by as much as 82%. The risk of systematic clinical error is clearly high, and excess morbidity or unnecessary treatment may result if an inappropriate diagnostic system is used for routine management. The variabil- ity in tuberculosis case frequency also underscores the importance of accurate phenotyping for interpretation of clinical trial end-points; genotypic studies, and studies of immune correlates. The second major finding is that the systems that yielded the highest and low- est tuberculosis case frequencies, namely the Osborne (89.2%) and Kibel (70.0%) and the MASA (6.9%) and WHO– Harries (10.0%) systems, demonstrated only fair or slight agreement with each other. Although the two outlier systems that generated the lowest results yielded similar tuberculosis case frequencies, the slight agreement suggests that they may be identifying different subpopulations. In this study, the variation in tubercu- losis case frequency observed when differ- ent structured diagnostic approaches were used and the relatively poor agreement between systems were more pronounced than previously reported. Edwards et al. retrospectively assessed agreement between clinical scoring systems used to diagnose tuberculosis among 91 children at a hospital in Kinshasa, Democratic Re- public of the Congo. The four approaches (Fourie, WHO provisional guidelines, Stegen–Kaplan, and Ghidey–Habte) generated tuberculosis case frequencies ranging from 87% to 96%.9,19–21 Agree- ment between systems ranged from fair (Κ: < 0.4) to moderate (Κ: 0.4–0.6).26 The reason Edwards et al. found less variation in case frequency may be that the study was hospital-based and all children had been diagnosed with tuberculosis on the original Edwards scale.18,26 We have also shown marked variation between hierarchical systems in the cer- tainty of the diagnosis of tuberculosis.13,14 The evaluation of related hierarchical approaches with similar distributions (SATVI and Stoltz–Donald) by weight- ing Κ for concordant and discordant cat- egories resulted in better agreement than for binary outcomes.8,16 Although hierar- chical and numerical systems that share key variables, such as a positive tuberculin skin test, a positive chest radiograph, and a positive sputum culture (Stegen–Toledo, Stoltz–Donald, and SATVI) showed moderate agreement, other systems with the same common variables showed less agreement and outlying case frequencies (Kibel, Osborne).8,11,13,14,16 It follows that system structure, weighting of variables and the exact order of Boolean decision- making may be as important as the constituent variables in determining the diagnostic output of each system. There are several other reasons for the observed variation in tuberculosis case frequency and the relatively poor agreement between diagnostic ap - proaches. They include differences in: (i) the purpose for which the systems were developed (as a screening tool or for definitive diagnosis; for clinical management or to obtain a trial end- point); (ii) clinical setting (community or hospital); (iii) disease severity (mild or severe tuberculosis); and (iv) re- gional prevalence of tuberculosis and/ or HIV infection (low or high). Ide- ally, for clinical trials a low-yielding diagnostic system should be used to minimize false positives at the expense of lower sensitivity.8 On the other hand, clinicians might prioritize sensitivity to avoid the potentially fatal conse- quences of underdiagnosis and delayed treatment.14,27 Therefore, approaches designed for clinical management, especially to serve as screening tools, might yield higher tuberculosis case frequencies.9,14,27 Although the SATVI trial algorithm lay in the mid-range of case frequency estimates, in the absence of a gold standard it is not possible to determine which of the nine approaches yielded the most accurate rate of tu- berculosis.8 However, the proportion of children treated for tuberculosis on clinical grounds (42.3%) was almost identical to the median tuberculosis Bull World Health Organ 2010;88:312–320 | doi:10.2471/BLT.09.062893316 Mark Hatherill et al.Screening and diagnosis of childhood tuberculosis in South Africa Research Ta bl e 3. D iff er en ce s in c as e fr eq ue nc ya y ie ld ed b y ni ne s tr uc tu re d ap pr oa ch es fo r t he d ia gn os is o f t ub er cu lo si s, S ou th A fr ic a, 2 00 1– 20 06 Sy st em M AS A M ig lio ri SA TV I Os bo rn e St ol tz –D on al d Ki be l Fo ur ie W HO –H ar rie s St eg en –T ol ed o No . ( % ) d ia g- no se d w ith tu - be rc ul os is M AS A 34 .8 (3 2. 3– 37 .3 ) 44 .2 (4 1. 7– 47 .0 ) 82 .3 (8 0. 3– 84 .3 ) 22 .0 (1 9. 8– 24 .2 ) 63 .1 (6 0. 5– 65 .7 ) 23 .5 (2 1. 0– 26 .2 ) 3. 1 (1 .3 –5 .1 ) 46 .5 (4 3. 9– 49 .2 ) 99 (6 .9 ) M ig lio ri P < 0 .0 00 1 9. 4 (6 .7 –1 2. 2) 47 .5 (4 4. 8– 50 .3 ) 12 .8 (1 0. 0– 15 .6 ) 28 .3 (2 5. 4– 31 .2 ) 11 .3 (8 .3 –1 4. 1) 31 .7 (2 8. 9– 34 .3 ) 11 .7 (9 .8 –1 3. 7) 60 2 (4 1. 7) SA TV I P < 0 .0 00 1 P < 0 .0 00 1 38 .1 (3 5. 3– 40 .8 ) 22 .2 (2 0. 1– 24 .5 ) 18 .9 (1 6. 0– 21 .7 ) 20 .7 (1 7. 5– 23 .9 ) 41 .1 (3 8. 3– 43 .9 ) 2. 3 (− 0. 3– 4. 9) 73 9 (5 1. 1) Os bo rn e P < 0 .0 00 1 P < 0 .0 00 1 P < 0 .0 00 1 60 .3 (5 7. 7– 63 .0 ) 19 .2 (1 6. 9– 21 .6 ) 58 .8 (5 6. 0– 61 .5 ) 79 .2 (7 6. 9– 81 .4 ) 35 .8 (3 3. 1– 38 .4 ) 12 89 (8 9. 2) St ol tz –D on al d P < 0 .0 00 1 P < 0 .0 00 1 P < 0 .0 00 1 P < 0 .0 00 1 41 .1 (3 8. 2– 44 .0 ) 1. 5 (− 1. 6– 4. 8) 18 .9 (1 6. 1– 21 .6 ) 24 .5 (2 2. 0– 27 .2 ) 41 7 (2 8. 9) Ki be l P < 0 .0 00 1 P < 0 .0 00 1 P < 0 .0 00 1 P < 0 .0 00 1 P < 0 .0 00 1 39 .6 (3 6. 5– 42 .5 ) 60 .0 (5 7. 3– 62 .6 ) 16 .6 (1 3. 8– 19 .3 ) 10 11 (7 0. 0) Fo ur ie P < 0 .0 00 1 P < 0 .0 00 1 P < 0 .0 00 1 P < 0 .0 00 1 P = 0 .3 44 P < 0 .0 00 1 20 .4 (1 7. 9– 22 .9 ) 23 .0 (1 9. 8– 26 .1 ) 44 0 (3 0. 4) W HO –H ar rie s P = 0 .0 00 9 P < 0 .0 00 1 P < 0 .0 00 1 P < 0 .0 00 1 P < 0 .0 00 1 P < 0 .0 00 1 P < 0 .0 00 1 43 .4 (4 0. 6– 46 .2 ) 14 5 (1 0. 0) St eg en –T ol ed o P < 0 .0 00 1 P < 0 .0 00 1 P = 0 .0 83 P < 0 .0 00 1 P < 0 .0 00 1 P < 0 .0 00 1 P < 0 .0 00 1 P < 0 .0 00 1 77 2 (5 3. 4) No . ( % ) d ia gn os ed w ith tu be rc ul os is 99 (6 .9 ) 60 2 (4 1. 7) 73 9 (5 1. 1) 12 89 (8 9. 2) 41 7 (2 8. 9) 10 11 (7 0. 0) 44 0 (3 0. 4) 14 5 (1 0. 0) 77 2 (5 3. 4) 14 45 (1 00 ) M AS A, M ed ic al A ss oc ia tio n of S ou th A fri ca ; S AT VI , S ou th A fri ca n Tu be rc ul os is V ac ci ne In iti at ive ; W HO , W or ld H ea lth O rg an iza tio n. a A bs ol ut e di ffe re nc es in p ro po rti on (d en om in at or = 14 45 ) a nd c or re sp on di ng 9 5% C Is a re a bo ve d ia go na l s pa ce s; M cN em ar ’s te st P -v al ue s fo r d iff er en ce s in p ro po rti on a re b el ow d ia go na l s pa ce s. Ta bl e 4. O bs er ve d ag re em en ta am on g ni ne s tr uc tu re d ap pr oa ch es fo r d ia gn os in g tu be rc ul os is , S ou th A fr ic a, 2 00 1– 20 06 Sy st em M AS A M ig lio ri SA TV I Os bo rn e St ol tz –D on al d Ki be l Fo ur ie W HO –H ar rie s St eg en –T ol ed o No . ( % ) d ia gn os ed w ith tu be rc ul os is M AS A 65 .2 55 .7 17 .7 78 .0 36 .8 69 .9 87 .3 53 .4 99 (6 .9 ) M ig lio ri 0. 19 72 .7 51 .1 71 .0 61 .3 68 .3 64 .4 85 .3 60 2 (4 1. 7) SA TV I 0. 13 0. 46 58 .3 77 .7 67 .3 58 .8 54 .9 76 .4 73 9 (5 1. 1) Os bo rn e 0. 02 0. 13 0. 15 37 .9 76 .9 38 .6 20 .6 61 .9 12 89 (8 9. 2) St ol tz –D on al d 0. 31 0. 38 0. 56 0. 07 53 .8 62 .6 69 .4 70 41 7 (2 8. 9) Ki be l 0. 06 0. 27 0. 34 0. 33 0. 21 52 .7 39 .2 70 .7 10 11 (7 0. 0) Fo ur ie 0. 09 0. 32 0. 18 0. 06 0. 10 0. 18 73 .8 59 .0 44 0 (3 0. 4) W HO –H ar rie s 0. 18 0. 18 0. 11 0. 02 0. 08 0. 08 0. 24 53 .4 14 5 (1 0. 0) St eg en –T ol ed o 0. 12 0. 71 0. 53 0. 19 0. 42 0. 38 0. 20 0. 12 77 2 (5 3. 4) No . ( % ) d ia gn os ed w ith tu be rc ul os is 99 (6 .9 ) 60 2 (4 1. 7) 73 9 (5 1. 1) 12 89 (8 9. 2) 41 7 (2 8. 9) 10 11 (7 0. 0) 44 0 (3 0. 4) 14 5 (1 0. 0) 77 2 (5 3. 4) 14 45 (1 00 ) Κ , k ap pa s ta tis tic ; M AS A, M ed ic al A ss oc ia tio n of S ou th A fri ca ; S AT VI , S ou th A fri ca n Tu be rc ul os is V ac ci ne In iti at ive ; W HO , W or ld H ea lth O rg an iza tio n. a O bs er ve d pe rc en ta ge a gr ee m en t f or p ai re d in di vid ua l o bs er va tio ns (n = 1 44 5) is a bo ve d ia go na l s pa ce s; Κ v al ue s ar e be lo w d ia go na l s pa ce s. 317Bull World Health Organ 2010;88:312–320 | doi:10.2471/BLT.09.062893 Mark Hatherill et al. Screening and diagnosis of childhood tuberculosis in South Africa Research case frequency across all nine diagnostic approaches (41.7%). The importance of context This study was carried out in a commu- nity in which children with suspected tuberculosis were identified early, when the disease was probably mild.8 By con- trast, the WHO–Harries system assigns the highest diagnostic weight to chronic illness, severe malnutrition and extra- pulmonary tuberculosis, all of which occur more frequently in hospitalized children. It is therefore not surprising that this approach yielded a low tuber- culosis case frequency in our context.10 Similarly, the MASA approach, which requires the presence of the complete triad of symptoms compatible with tuberculosis, as well as a positive tuber- culin skin test and a suggestive chest radiograph, is designed as a treatment guideline for hospitalized children.15 The Osborne approach, which yielded results at the upper extreme of tuberculosis case frequency, was designed in a developing country setting where the index of suspi- cion for tuberculosis is high. It functions best as a screening tool, since children with suspected or possible tubercu- losis are not necessarily treated.11,14,16 Similarly, the Kibel system is designed to guide initial treatment decisions rather than to establish a definitive diagnosis in resource-limited settings.11,27 The Fourie system, also designed as a screening tool, yielded one of the lowest tuberculosis case frequencies, which suggests that it may be unsuitable for screening in our epidemiological setting.9 Some have noted that regional HIV prevalence may affect the performance of a par- ticular diagnostic approach unless HIV infection status is incorporated.5,8,14 The confounding effect of HIV status on diagnostic decision-making is likely to be greatest in systems that emphasize the non-specific features of malnutrition.10 Edwards et al. noted that HIV-infected children scored higher on the Keith Edwards scale,18 a feature that would be common to the WHO–Harries ap- proach. Consequently, the current edi- tion of the WHO’s TB/HIV: a clinical manual no longer recommends the use of diagnostic scoring systems.10,26 Study limitations This study has several limitations. Inves- tigations were nested within a clinical trial that might not reflect clinical prac- tice in developing regions. Variables were analysed in a standardized fashion that may differ from that used in the original diagnostic systems, and we acknowledge the potential limitations of Κ scores for assessing agreement. Children were younger than 2 years (an age group in which diagnostic imprecision is highest) and the findings may not be applicable to older children with a different disease spectrum. Since the study was communi- ty-based and investigations were geared towards pulmonary tuberculosis, there may have been a bias against diagnostic approaches that included features of extra-pulmonary tuberculosis. Further- more, since all children identified by active case-finding were investigated for tuberculosis, even if they had no symp- toms, the discrepancies between clinical, symptom-based and bacteriology-based systems may have been exaggerated. Structured diagnostic approaches were selected on the basis of relevance to the sub-Saharan region. Thus, four of the nine approaches were of South African origin.8,11,15,16 We acknowledge the exis- tence of other structured approaches for diagnosing childhood tuberculosis, such as the Sant’Anna score, but they were not included in this analysis.17,28 Significance of findings The public health significance of these findings is illustrated by the marked dif- ferences in tuberculosis case frequency and the poor agreement between diag- nostic systems. Regional tuberculosis control programmes should make an informed decision to advocate a specific approach for the screening and diagnosis of childhood tuberculosis. Clearly, the Fig. 2. Frequency of tuberculosis diagnoses assigned to each category of diagnostic certainty, in order of increasing certainty of tuberculosis, with five hierarchical or hierarchical–numerical systems, South Africa, 2001–2006 0T ub er cu lo si s ca se s (n = 1 44 5) 1000 800 600 400 200 706 (48.9%) SATVI Un like ly/ no t 334 (23.1%) Po ssi ble 233 (16.1%) Pro ba ble 172 (11.9%) De fin ite 0T ub er cu lo si s ca se s (n = 1 44 5) 1000 800 600 400 200 156 (10.8%) Osborne Un like ly/ no t 603 (41.7%) Po ssi ble 507 (35.1%) Pro ba ble 179 (12.4%) De fin ite 0T ub er cu lo si s ca se s (n = 1 44 5) 1000 800 600 400 200 1028 (71.1%) Stoltz–Donald Un like ly/ no t 14 (1.0%) Po ssi ble 231 (16.0%) Pro ba ble 172 (11.9%) De fin ite 0T ub er cu lo si s ca se s (n = 1 44 5) 1000 800 600 400 200 434 (30.0%) Kibel Un like ly/ no t 457 (31.6%) Po ssi ble 382 (26.4%) Pro ba ble 172 (11.9%) De fin ite 0T ub er cu lo si s ca se s (n = 1 44 5) 1000 800 600 400 200 673 (46.6%) Stegen–Toledo Un like ly/ no t 244 (16.9%) Po ssi ble 211 (14.6%) Pro ba ble 317 (21.9%) De fin ite SATVI, South African Tuberculosis Vaccine Initiative Table 5. Observed agreementa among five hierarchical structured approaches for diagnosing tuberculosis, South Africa, 2001–2006 System SATVI Osborne Stoltz– Donald Kibel Stegen– Toledo SATVI 79.2 92.4 81.0 81.0 Osborne 0.48 72.5 85.9 77.7 Stoltz–Donald 0.80 0.40 76.5 78.9 Kibel 0.51 0.60 0.43 80.9 Stegen–Toledo 0.53 0.45 0.49 0.54 Κ, kappa statistic; SATVI, South African Tuberculosis Vaccine Initiative. a Observed percentage agreement for paired individual observations (n =1445) is above diagonal spaces; weighted Κ values are below diagonal spaces. 318 Bull World Health Organ 2010;88:312–320 | doi:10.2471/BLT.09.062893 صخلم ايقيرفأ بونج في عفترلما هراشتنا قطانم في هصيخشتو لافطلأا لس نع يرحتلل ةمظنلما بيلاسلأا ؛لافطلأا لس صيخشتل ةعستلا ةمظنلما بيلاسلأا ينب قفاوتلا سايق ضرغلا بيلاسلأا هذهب ةصخشلما لسلا تلااح ددع في تافلاتخلاا ةيمك ديدحتو .صيخشتلا نم دكأتلل ةفلتخلما تائفلا في تلااحلا عيزوت ديدحتو ،ةفلتخلما ةبرجت ءانثأ ينتنس رمع نم بركأ ًلافط 1445 نوثحابلا صىقتسا ةقيرطلا .ايقيرفأ بونج في ةيفيرلا تاعمتجلما في )2006 – 2001( ماوعلأا في حاقللا ددحو .ةيجولويبوركلماو ةيعاعشلاو ةيريسرلا تايطعلما نوثحابلا عمج ،ةعستلا ةيصيخشتلا بيلاسلأا نم لك مادختساب لسلا تلااح نوثحابلا جئاتنلا في تاعومجلما ينب قفاوتلاو تلااحلا راركت في تافلاتخلاا اوبسحو رماينكام رابتخا مادختساب )لسلاب ةباصلإا مدعو / لسلاب ةباصلإا( ةيئانثلا Cohen’s kappa اباك ينهوك لاعمو )95% ةقثلا تلاصاف عم( McNemar’s .coefficient .)41.7 :طيسولا( 89.2% لىإ 6.9% نم لسلا تلااح ددرت حوارت تادوجولما نم لقأ P لماتحلاا ةوق( تلااحلا راركت في اهب دتعي تافلاتخا كانه ناكو ىدم( ةمظنلما ةيصيخشتلا بيلاسلأا ينب ةنراقملل ًاجوز 36 نم 34 في )0.05 لىإ 0.02 نم Kappa اباك لماعم حوارتو .)82.3% – 1.5 :ةقلطلما تافلاتخلاا تلااحلل راركت لىعأ ماهل ناذللا نابولسلأا رهظأ دقو .)0.18 :طيسولا( 0.71 ناذللا نابولسلأا رهظأو ؛)0.33 :اباك لماعم( ًاديج ًاقفاوت )70.0% و 89.2%( .ًلايئض ًاقفاوت )10.0% و 6.9%( تلااحلل راركت لقأ ماهل لس نع يرحتلل ةمظنلما بيلاسلأا ينب ليئض قفاوت طقف كانه جاتنتسلاا فاشتكا ثيح نم بيلاسلأا هذه ينب يربك توافت كانهو ،هصيخشتو لافطلأا تلااحلل ضفخنم راركت جاتنإ في هباشتت يتلا ةيصيخشتلا مظنلا نإ .تلااحلا جئاتن معدت لاو .لافطلأا نم ةفلتخم ةيعرف تائف ددحت اهنأ نكملما نم ئياهنلا صيخشتلل ةمظنلما بيلاسلأل ينيتورلا يريسرلا مادختسلاا ةساردلا نوكت دق ةعفترم جئاتن لىإ يدؤت يتلا مظنلا نأ نم مغرلاب ،لافطلأا لسل .يرحتلل تاودأك ةديفم Mark Hatherill et al.Screening and diagnosis of childhood tuberculosis in South Africa Research study data do not support the routine, uncritical use of any particular diagnostic system for therapeutic decision-making. Some diagnostic approaches may in fact be best suited to specific settings. For example, a high-yielding system, such as Osborne, may be suitable as a screening tool, whereas the low-yielding WHO– Harries system may be most appropriate as a tool for diagnosing severe tuberculosis in regions with a low prevalence of HIV infection. Conclusion Although systems with a moderate case yield are less prone to extreme diagnos- tic error, the predictive value of any one system cannot be determined in the absence of a gold standard. Any struc- tured approach to estimate tuberculosis case frequency can yield biased results if used in a way that differs from that for which it was originally designed, whether for clinical care or research purposes, screening or definitive diag- nosis, mild or severe disease, or in low or high tuberculosis prevalence regions. However, in the absence of validation cohorts, there is l imited evidence that these systems would have better diagnostic accuracy in their original settings. The findings of this study should not undermine confidence in existing diagnostic methods. Instead, they should encourage innovative re- search and critical analysis in the search for improved diagnostics for childhood tuberculosis. ■ Acknowledgements We thank the staff of The South African Bacille Calmette-Guérin Trial Team for data collection; Maurice Kibel, John Bur- gess and Robert Gie for expert radiology review; Suzanne Verver for epidemio- logical support; and Lyness Matizirofa for statistical support. Funding: The study was supported by the Aeras Global TB Vaccine Foundation, a non-profit organization that aims to develop tuberculosis vaccines. Competing interests: TH and LG are current and previous full time employees of the trial sponsor. The authors have not entered into any agreements that have limited the completion of the research as planned, and they have had full control of all primary data. Fig. 3. Distribution of scores (n = 1445) obtained with different numerical scoring systems for the diagnosis of childhood tuberculosis, South Africa, 2001–2006 0 Nu m er al s co re 16 Kibel Fourie WHO–Harries Stegen–Toledo 14 12 10 8 6 4 2 Quartiles 2 and 3 Quartiles 1 and 4 Outliers Diagnostic threshold value 319Bull World Health Organ 2010;88:312–320 | doi:10.2471/BLT.09.062893 Objectif Mesurer le degré d’accord entre neuf approches structurées pour le diagnostic de la tuberculose chez l’enfant ; quantifier les différences en termes de nombres de cas de tuberculose diagnostiqués entre ces neuf approches ; et déterminer la répartition des cas dans les différentes catégories de certitude diagnostique. Méthodes Nous avons étudié 1445 enfants de moins de 2 ans appartenant à une communauté rurale d’Afrique du Sud, dans le cadre d’un essai vaccinal (2001-2006). Des données cliniques, radiologiques et microbiologiques ont été collectées prospectivement. Nous avons déterminé quel statut diagnostique (tuberculeux/non tuberculeux) était affecté par chacune des approches aux cas potentiels de tuberculose. Nous avons calculé les différences en termes de fréquence des cas et l’accord concernant la catégorie de certitude pour les résultats binaires (tuberculose/absence de tuberculose) en utilisant le test de McNemar (avec les intervalles de confiance à 95 %, IC) et le coefficient kappa de Cohen (Κ). Résultats La fréquence des cas de tuberculose se situait entre 6,9 et 89,2 % (médiane : 41,7 %). Des différences significatives sont apparues dans la fréquence des cas (p < 0,05) dans 34 des 36 comparaisons par paire entre les approches diagnostiques structurées (plage de différences absolues : 1,5-82,3 %). Le coefficient kappa variait de 0,02 à 0,71 (médiane : 0,18). Les deux systèmes donnant les plus fortes fréquences de cas (89,2 % et 70,0 % respectivement) présentaient un accord satisfaisant (Κ : 0,33) ; les deux autres systèmes, qui avaient fourni les plus faibles fréquences (6,9 % et 10,0 %, respectivement), n’étaient que faiblement en accord (Κ : 0,18). Conclusion Il n’existe qu’un faible accord entre les approches structurées du dépistage et du diagnostic de la tuberculose chez l’enfant et il apparaît entre elles une forte variabilité du rendement en cas. Les systèmes diagnostiques ayant fourni de manière similaire des fréquences de cas peu élevées pourraient identifier des sous-populations d’enfants différentes. Les résultats de cette étude ne sont pas en faveur d’un usage clinique systématique de ces approches structurées pour le diagnostic définitif des enfants tuberculeux, mais les systèmes fournissant un rendement élevé en cas pourraient constituer des outils de dépistage utiles. Resumen Sistemas estructurados de cribado y diagnóstico de la tuberculosis infantil en una región de alta prevalencia de Sudáfrica Objetivo Medir la concordancia entre nueve sistemas estructurados de diagnóstico de la tuberculosis infantil; cuantificar las diferencias en cuanto al número de casos de tuberculosis diagnosticados con los diferentes sistemas, y determinar la distribución de casos en distintas categorías de certeza diagnóstica. Métodos Se estudió a 1445 niños menores de 2 años durante un ensayo de vacunas (2001–2006) llevado a cabo en una comunidad rural de Sudáfrica. Se reunieron de forma prospectiva datos clínicos, radiológicos y microbiológicos, y se determinó si los niños sufrían o no tuberculosis usando cada una de las nueve modalidades de diagnóstico. Para calcular las diferencias en la frecuencia de casos y la concordancia de categorías para resultados binarios (tuberculosis/no tuberculosis), aplicamos la prueba de McNemar (con intervalos de confianza del 95%) y el coeficiente kappa de Cohen (Κ ). Resultados La frecuencia de casos de tuberculosis se situó entre 6,9% y 89,2% (mediana: 41,7). Se observaron diferencias significativas en la frecuencia de casos (P < 0,05) en 34 de las 36 comparaciones emparejadas entre los sistemas de diagnóstico estructurado (intervalo de diferencias absolutas: 1,5–82,3%). Kappa osciló entre 0,02 y 0,71 (mediana: 0,18). Los dos sistemas que hallaron las frecuencias de casos más altas (89,2% y 70,0%), mostraron una concordancia aceptable (Κ : 0,33); y los dos que hallaron las frecuencias de casos más bajas (6,9% y 10,0%) mostraron una concordancia baja (Κ : 0,18). Conclusión Se observa solo una baja concordancia entre los sistemas estructurados en lo relativo al cribado y diagnóstico de la tuberculosis infantil, y una alta variabilidad entre ellos en términos de detección de casos. Sistemas de diagnóstico que arrojan frecuencias de casos similarmente bajas podrían estar detectando subpoblaciones de niños diferentes. Los resultados del estudio no respaldan el uso clínico sistemático de criterios estructurados para el diagnóstico definitivo de la tuberculosis infantil, pero los sistemas que consiguen valores altos de detección pueden ser un valioso instrumento de cribado. References 1. Marais BJ, Gie RP, Hesseling AC, Schaaf HS, Lombard C, Enarson DA et al.A refined symptom-based approach to diagnose pulmonary tuberculosis in children. Pediatrics 2006;118:e1350–9. doi:10.1542/peds.2006-0519 PMID:17079536 2. A research agenda for childhood tuberculosis. Geneva: World Health Organization; 2007 (WHO/HTM/TB/2007.2381). 3. Skeiky YA, Sadoff JC. Advances in tuberculosis vaccine strategies. Nat Rev Microbiol 2006;4:469–76. doi:10.1038/nrmicro1419 PMID:16710326 4. Nicol MPDM, Wood K, Hatherill M, Workman L, Hawkridge A, Eley B et al.A comparison of T-SPOT.TB and tuberculin skin test for the evaluation of young children at high risk for tuberculosis in a community setting. Pediatrics 2009;123:38–43. doi:10.1542/peds.2008-0611 PMID:19117858 5. Hesseling AC, Schaaf HS, Gie RP, Starke JR, Beyers N. A critical review of diagnostic approaches used in the diagnosis of childhood tuberculosis. Int J Tuberc Lung Dis 2002;6:1038–45. PMID:12546110 6. Groenewald P. Boland-Overberg region annual health status report 2004. Worcester: Department of Information Management, Department of Health; 2004 . 7. Houwert KA, Borggreven PA, Schaaf HS, Nel E, Donald PR, Stolk J. Prospective evaluation of World Health Organization criteria to assist diagnosis of tuberculosis in children. Eur Respir J 1998;11:1116–20. doi:10. 1183/09031936.98.11051116 PMID:9648965 8. Hawkridge A, Hatherill M, Little F, Goetz MA, Barker L, Mahomed H et al.Efficacy of percutaneous versus intradermal BCG in the prevention of tuberculosis in South African infants: randomised trial. BMJ 2008;337:a2052. doi:10.1136/bmj.a2052 PMID:19008268 9. Fourie PB, Becker PJ, Festenstein F, Migliori GB, Alcaide J, Antunes M et al.Procedures for developing a simple scoring method based on unsophisticated criteria for screening children for tuberculosis. Int J Tuberc Lung Dis 1998;2:116–23. PMID:9562121 10. Harries A, Maher D, Graham S. TB/HIV: a clinical manual. 2nd ed. Geneva: World Health Organization; 2004. 11. Kibel M. A point system for management of childhood tuberculosis. Cape Town: Institute of Child Health, University of Cape Town; 1999. Mark Hatherill et al. Screening and diagnosis of childhood tuberculosis in South Africa Research Résumé Approches structurées pour le dépistage et le diagnostic de la tuberculose chez l’enfant dans une région d’Afrique du Sud où cette maladie est fortement prévalente 320 Bull World Health Organ 2010;88:312–320 | doi:10.2471/BLT.09.062893 Mark Hatherill et al.Screening and diagnosis of childhood tuberculosis in South Africa Research 12. Migliori GB, Borghesi A, Rossanigo P, Adriko C, Neri M, Santini S et al.Proposal of an improved score method for the diagnosis of pulmonary tuberculosis in childhood in developing countries. Tuber Lung Dis 1992;73:145–9. doi:10.1016/0962-8479(92)90148-D PMID:1421347 13. Montenegro SH, Gilman RH, Sheen P, Cama R, Caviedes L, Hopper T et al.Improved detection of Mycobacterium tuberculosis in Peruvian children by use of a heminested IS6110 polymerase chain reaction assay. Clin Infect Dis 2003;36:16–23. doi:10.1086/344900 PMID:12491196 14. Osborne CM. The challenge of diagnosing childhood tuberculosis in a developing country. Arch Dis Child 1995;72:369–74. doi:10.1136/ adc.72.4.369 PMID:7763076 15. Pinkney-Atkinson V. TB practical guidelines: managed care and quality review. Cape Town: Medical Association of South Africa; 1996. 16. Stoltz AP, Donald PR, Strebel PM, Talent JM. Criteria for the notification of childhood tuberculosis in a high-incidence area of the western Cape Province. S Afr Med J 1990;77:385–6. PMID:2330522 17. Sant’Anna CC, Orfaliais CT, March Mde F, Conde MB. Evaluation of a proposed diagnostic scoring system for pulmonary tuberculosis in Brazilian children. Int J Tuberc Lung Dis 2006;10:463–5. PMID:16602415 18. Edwards K. The diagnosis of childhood tuberculosis. P N G Med J 1987;30:169–78. PMID:3314246 19. Ghidey Y, Habte D. Tuberculosis in childhood: an analysis of 412 cases. Ethiop Med J 1983;21:161–7. PMID:6603973 20. Stegen G, Jones K, Kaplan P. Criteria for guidance in the diagnosis of tuberculosis. Pediatrics 1969;43:260–3. PMID:5304285 21. Provisional guidelines for the diagnosis and classification of the EPI target diseases for primary health care, surveillance and special studies. Geneva: World Health Organization; 1983 (EPI/GEN/83/84). 22. Marais BJ, Gie RP, Obihara CC, Hesseling AC, Schaaf HS, Beyers N. Well defined symptoms are of value in the diagnosis of childhood pulmonary tuberculosis. Arch Dis Child 2005;90:1162–5. doi:10.1136/ adc.2004.070797 PMID:16131501 23. Cundall DB. The diagnosis of pulmonary tuberculosis in malnourished Kenyan children. Ann Trop Paediatr 1986;6:249–55. PMID:2435230 24. Country profile. South Africa. In: Global tuberculosis control. WHO report 2008. Geneva: World Health Organization; 2008. Available from: http://www. who.int/globalatlas [accessed 20 November 2009]. 25. McGinn T, Wyer PC, Newman TB, Keitz S, Leipzig R, For GG. Tips for learners of evidence-based medicine: 3. Measures of observer variability (kappa statistic). CMAJ 2004;171:1369–73. PMID:15557592 26. Edwards DJ, Kitetele F, Van Rie A. Agreement between clinical scoring systems used for the diagnosis of pediatric tuberculosis in the HIV era. Int J Tuberc Lung Dis 2007;11:263–9. PMID:17352090 27. Kibel MA, Hussey G. Problems in the diagnosis of childhood tuberculosis. S Afr Med J 1990;77:379–80. PMID:2330519 28. Sant’Anna CC, Santos MA, Franco R. Diagnosis of pulmonary tuberculosis by score system in children and adolescents: a trial in a reference center in Bahia, Brazil. Braz J Infect Dis 2004;8:305–10. doi:10.1590/S1413- 86702004000400006 PMID:15565261 Corrigenda In volume 88, Number 3, March 2010: • page 169 the 4th sentence of the 5th paragraph and page 170 photo caption should read “Lee In-sook, whose mother and sister both had cancer, …” • page 170, the quotation in the final paragraph should be attributed to Yang Boon-min • page 200, the first sentence of the second paragraph should read: “Table 1 shows the estimated incidence of selected communicable diseases in 2004 in the world and in the region.” • page 201, the title for Table 1 should read: “Estimated incidence of selected communicable diseases worldwide and in the South-East Asia Region of the World Health Organization, 2004a” and the column header should read “Estimated incidence (in thousands)”