Tajikistan - Quality of child health services

HUMAN DEVELOPMENT SECTOR UNIT Central Asia Country Unit | Europe and Central Asia Region Republic of Tajikistan Quality of Child Health Services in Tajikistan THE WORLD BANK World Bank Report Number 62870-TJ ON THE COVER Family Medicine Doctor and under-five patient in Tajikistan  HUMAN DEVELOPMENT SECTOR UNIT Central Asia Country Unit Europe and Central Asia Region June, 2011 Republic of Tajikistan Quality of Child Health Services in Tajikistan THE WORLD BANK  Table of Contents Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v Acronyms, Abbreviations and Tajik Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Chapter 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Chapter 2. Key Survey Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Quality of Clinical Care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Health Systems Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Chapter 3. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Chapter 4. Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Human Development Sector Unit | Central Asia Country Unit | Europe and Central Asia Region THE WORLD BANK iii  Annexes Annex 1: Objectives of the Study and Survey Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Annex 2: Descriptive Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Annex 3: List of Surveyed PHC Facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Annex 4: IMCI Priority Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Table of Figures Figure 1: Trends in Tajikistan’s Infant Mortality Rate 1990-2006 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Figure 2: Leading Causes of Post-Neonatal Mortality in Tajikistan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Figure 3: Tuberculosis and Measles Vaccine Coverage for Children Aged 12-23 Months . . . . . . . . . . 5 Figure 4: Assessment of Danger Signs by Regions (N=300 ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Figure 5: Assessment of Weight against Growth Chart by Regions (N=300) . . . . . . . . . . . . . . . . . . . . . 9 Figure 6: Integrated Assessment: Main Tasks and the WHO Index (N=300) . . . . . . . . . . . . . . . . . . . 10 Figure 7: Rational use of Antibiotics by Region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Figure 8: Vaccines Availability in Primary Health Care Facilities (N=70) . . . . . . . . . . . . . . . . . . . . . 14 Figure 9: Availability of Basic Equipment and Supplies in the Facilities (N=70) . . . . . . . . . . . . . . . 15 Figure A1: Cases Seen by Health Care Providers’ IMCI Training Status . . . . . . . . . . . . . . . . . . . . . . . 22 List of Tables Table 1: Leading Causes of Child Mortality in Tajikistan by Region (2001-2002) . . . . . . . . . . . . . . . 4 Table 2: Comparison of Selected Indicators by Family Doctors Previous Specialization . . . . . . . . . . 13 Table A1: Sampling Frame and Survey Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Table A2: PHC Facilities Visited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Table A3: Health Care Provider Type and IMCI Training Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Table A4: Characteristics of the Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Table A5: Study PHC Facilities by Region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 iv Quality of Child Health Services in Tajikistan Acknowledgments This report was prepared by a World Bank team with support M. Atoev (Director, Center of Pediatrics and Pediatric from the Global Alliance for Vaccines Initiative (GAVI) Trust Surgery, Tajikistan), Dr. Z. Nabiev, Dr. Vakhidov (Center Fund. Wezi Msisha (Task Team Leader, Health Specialist, of Paediatrics & Pediatric Surgery, Tajikistan), and Dr. Z. ECSH1) led the preparation of the report. The team mem- Kasimova (CBHP Project Implementing Unit). bers from the World Bank included Sarvinoz Barfieva (Operations Officer, ECSH1), Gabriel Francis (Program The report has benefited greatly from the peer reviewer Assistant, ECSHD), Gulnora Kamilova (Program Assistant, comments (on both the concept note and the draft report) ECCUZ), Shoira Zukhurova (Team Assistant, ECCTJ), provided by Aparnaa Somanathan (Economist, EASHD), Ivdity Chikovani, Maya Kherkheulidze, Natia Rukhadze, and Mariam Claeson (Program Coordinator, SASHN), Salima Ketevan Chkhatarashvili (Curatio International Foundation). Kasymova (National Program Officer, WHO Tajikistan), and Editorial services were provided by Rosemarie Esber. Son Nam Nguyen (Senior Health Specialist, ECSHD). The comments and guidance received from Charles Griffin (Sr. This report would not have been possible without techni- Advisor, ECAVP) were also very useful. cal support from the WHO and UNICEF Tajikistan. We especially would like to acknowledge the support we received This task was undertaken under the guidance of Tamar from, Salima Kasymova (National Program Officer, WHO Manuelyan Atinc (former Sector Director, ECSHD), Tajikistan) and Dr. S. Kurbanov (UNICEF, Tajikistan). Mamta Murthi (Acting Sector Director, ECSHD ), Abdo Yazbeck (former Sector Manager, ECSH1), Daniel Dulitzky We also extend our thanks to Dr. S. Rhakmatuloev (Head (Sector Manager, ECSH1), Motoo Konishi (Country of Maternal & Child Health Department [MCH], Ministry Director, ECCU8), Chiara Bronchi (former Country of Health, Tajikistan), Dr. S. Rakhmatullaeva (Deputy Head Manager, ECCUTJ), and Marsha Olive (Country Manager, of MCH Department, Ministry of Health, Tajikistan), Dr. ECCUTJ). Human Development Sector Unit | Central Asia Country Unit | Europe and Central Asia Region THE WORLD BANK v  Acronyms, Abbreviations and Tajik Terms ADB Asian Development Bank NNS National Nutrition Survey ARI Acute Respiratory Illness Oblast A province CPD Continuous Professional Development ORT Oral Rehydration Therapy DPT Diphtheria Polio Tetanus PHC Primary Health Care FM Family Medicine Rayon A rural district GBAO Gorno-Badakhshan Autonomous Oblast RRP Rayons of Republican Subordination IMCI Integrated Management of Childhood Illness TLSS Tajikistan Living Standards Survey IMR Infant Mortality Rate U5MR Under Five Mortality Rate Jamoat A local self-government unit, usually a UNDP United Nations Development Program rural sub district UNICEF United Nations Children’s Fund MCH Maternal and Child Health USAID United Sates Agency for International MDG Millennium Development Goals Development MICS Multiple Indicator Cluster Survey WHO World Health Organization MOH Ministry of Health vi Quality of Child Health Services in Tajikistan Executive Summary 1. The Government of Tajikistan has identified Primary though most children were weighed at the clinic, in only 27 Heath Care (PHC), and Maternal and Child Health (MCH) percent of the cases did providers check the child’s weight as top priorities in its first Comprehensive National Health against standard growth charts. This simple step is essential to Sector Strategy (2010-2020). Poor child health outcomes reveal long-standing health problems, particularly malnutri- in Tajikistan are related to systemic health sector issues, tion. Ten tasks should routinely be performed on every child including the financing and quality of health services, as presenting at a clinic, according to the IMCI guidelines. Our well as poverty, rural residence, and access to clean water. findings show that overall, on average, only 6.9 tasks were Tajikistan’s Millennium Development Goal (MDG) target for performed, and the RRP facilities had the worst performance a two-thirds reduction in child mortality requires an infant in most instances. mortality rate (IMR) of 29.6 and the under-five mortality rate (U5MR) of 39.3 by 2015.1 The most recent estimates place 4. On accuracy of diagnosis, the findings were equally the IMR at 52 per 1,000 live births and the U5MR at 61 per troublesome. Providers appropriately referred ten sick chil- 1,000 live births.2 dren to a higher-level facility for treatment. Of the remaining 290 in the sample, only 49 percent were correctly diagnosed. 2. To reduce its high infant and child mortality rates as well Only 42 percent of children needing an antibiotic left the as develop PHC, the Government of Tajikistan introduced facility with one. While 24 percent of children not needing the Family Medicine model of practice in 2001, and the an antibiotic left the facility with one, or a prescription for Integrated Management of Childhood Diseases (IMCI) an antibiotic. Only 8 percent of children who should have strategy in 2000. This World Bank study was undertaken in received a first dose of an antibiotic at the facility during the 2010 to assess the quality of outpatient health care services visit actually received it. The treatment of children with diar- for sick children aged two months to five years at primary rhea was better. Of the children needing oral rehydration salts, health care (PHC) facilities in Tajikistan. The specific focus 84 percent of them received a first treatment at the facility, of this study is on the quality of care delivered by PHC pro- but only 22 percent with anemia received an iron supplement viders who were retrained over the last several years as Family at the facility. Medicine practitioners. The team employed the standardized WHO/UNICEF IMCI survey methodology to evaluate the 5. The study also finds that supervision of PHC workers is quality of care delivered to sick children attending outpatient irregular, and training does not seem to be performed system- facilities. The study was conducted in 19 districts in Khatlon atically to improve children’s health outcomes. For example, region, Sogd region, the Rayons of Republican Subordination in the Sogd region, no IMCI-trained nurse worked in any of (RRP), and Dushanbe City. the 20 PHC facilities surveyed, even though they play a key role in the provision of child health services. For the majority 3. The study findings which closely mirror those of the 2009 of indicators examined, no statistically significant differences WHO/UNICEF IMCI survey will be of great concern to the were found between providers retrained in the six-month Ministry of Health and should provide the impetus to take Family Medicine program and those who were not. The cur- immediate remedial actions. Of the sample 300 sick children rent Family Medicine retraining program does not seem to brought to the PHC facilities by a caretaker, only 46 percent improve the use of the IMCI guidelines for acute child health were assessed for the basic three danger signs: inability to care services. These findings indicate the need to improve the drink, vomiting, and convulsions. Providers should have overall training of Family Medicine practitioners in the basic screened 100 percent of these sick children for these seri- management of sick children following the IMCI clinical ous symptoms, according to the IMCI guidelines. Only 4 guidelines, particularly because Family Medicine practitioners percent of sick children were appropriately screened at the are often the first point of contact between sick children and RRP’s primary health centers. This shocking finding sharply the health care system. decreased the overall average. Nonetheless, the other regions scores ranged from 46 to 64 percent, still a poor showing for 6. Two main issues are highlighted by this study. Firstly, the an indicator that should be consistently 100 percent. Even quality of primary health care services provided to children is lacking in many areas, irrespective of the PHC provider’s type of training. Second, is that the family medicine and IMCI 1 UNDP (2005) Investing in Sustainable Development: Tajikistan MDG Needs Assessment. training programs and methods require further enhancement to ensure that the service quality for children improves. Some 2 UNICEF State of the World’s Children Report 2011. Human Development Sector Unit | Central Asia Country Unit | Europe and Central Asia Region THE WORLD BANK 1  of the findings of this survey may be due to the staff facing the system of routine supervision and support of primary constraints that could be relieved with additional resources in health care providers to identify problems and reinforce their all PHC facilities to ensure the availability of essential drugs, adherence to recommended clinical practice guidelines. antibiotics and vaccines in all PHC facilities. However, train- ing and incentives to provide appropriate care to sick children 8. Medium-term actions to implement include: (i) under- should be revisited urgently. The existing approach is not take an independent and external evaluation of the Family serving children or their parents well. While many determi- Medicine re-training programs to ensure adequate and rel- nants of infant and child mortality lie outside the health care evant content and training methods, with special attention system, it is vital that when a sick child comes in contact with to MCH; (ii) provide continuous skills improvement training the medical care system that it should perform well on behalf for the trainers of Family Medicine providers at the four of the child. main post-graduate training centers; (iii) improve methods of paying primary health care providers through further develop- 7. In light of these findings, several actions must be taken ment of the current per capita system, including exploring the by the Ministry of Health and its development partners to use of incentive payments to Family Medicine practitioners improve the current state of affairs. Of immediate concern are to improve quality of services, and (iv) improvement of the the following: (i) ensure the availability of essential antibiotics current system of quality assurance in primary health care and vaccine stocks in all first-level primary health care facili- facilities. ties, with a special focus on PHC facilities in the RRP; (ii) undertake a closer assessment of factors contributing to poor 9. Improving the country’s system of PHC is within the reach performance in PHC facilities in the RRP and other poor per- of the Tajikistan Ministry of Health. With careful planning forming rayons and implementing the appropriate solutions; and coordination, the MoH can successfully implement these (iii) provide intensive on the job training on correct child recommendations, which will result in much needed service growth monitoring and nutrition to reinforce the skills and improvements for sick children and their health outcomes. knowledge of primary health care workers; and (iv) improve 2 Quality of Child Health Services in Tajikistan Chapter 1. Introduction CHILD MORBIDITy & MORTALITy 1.3 Other countries with GDP levels similar to Tajikistan are making better progress towards reaching their MDG targets. 1.1 Globally, the majority of under-five mortality is attributed In the Lao People’s Democratic Republic, for example, the to a few commonly occurring and preventable childhood ill- IMR is 46 per 1,000 and in the Kyrgyz Republic and neigh- nesses. These diseases, which often occur concurrently, are boring Uzbekistan; the IMR is 32 per 1,000 live births. For pneumonia, diarrhea, malaria, measles and HIV. In many under-five mortality rates, Lao and the Kyrgyz Republic are cases, childhood diseases are worsened by under-nutrition, respectively at 59 and 37 per 1,000 live births (UNICEF, which contributes to approximately 35 percent of child 2011). Understanding the factors that promote or hinder deaths, and 11 percent of the total global disease burden.3 progress towards achieving the MDGs critical and some of 1.2 Tajikistan’s infant and child mortality rates are still quite these are discussed below. high. Current estimates place the infant mortality rate (IMR) at 52 per 1,000 live births and the under-five mortality rate TAjIKISTAN DEMOgRAPHIC (U5MR) at 61 per 1,000 live births,4 compared to an IMR of SITUATION 65 per 1,000 live births and an U5MR of 79 per 1,000 live births in 2005.5 Despite Tajikistan’s significant improvements 1.4 Poverty rates in Tajikistan have declined significantly from in MCH outcomes, to achieve its Millennium Development 72.1 percent in 2003 to 53.1 in 2007 and even further to Goal (MDG) goal of a two-thirds reduction in child mortality 47.2 percent in 2009. Even so, it remains the poorest country by 2015, Tajikistan’s task is to attain an IMR of 29.6 and an in Central Asia. Poverty is mainly a rural phenomenon, with U5MR of 39.3 by 2015.6 the rural poor accounting for 75 percent of all poor and 72 percent of the extreme poor (World Bank, 2010). Out of 3 Black et al (2008). Maternal & Child Undernutrition: Global & an estimated population of 7 million people, approximately Regional Exposures and Health Consequences. The Lancet, 371 (9068), 3 million (42%) are aged 18 and under, with the median 243-260. age of the population at just 20.7. Life expectancy at birth 4 UNICEF State of the World’s Children Report 2011. is 67.3 years. Although fertility has fallen in recent years, 5 Goskomstat. (2007). Tajikistan Multiple Indicator Cluster Survey the total fertility rate is still greater than 3 and the annual 2005 Final Report. average population growth rate is approximately 1.9 percent 6 UNDP (2005). Investing in Sustainable Development: Tajikistan MDG Needs Assessment. per year (UNICEF, 2011, UNDP, 2010). To advance overall Figure 1: Trends in Tajikistan’s Infant Mortality Rate 1990-2006 100 WHO WHOSIS: 91 90 MICS 2000: IMR = 89 80 LSMS 1999: 79 WHO WHOSIS: IMR = 75 70 MICS 2005: 65 60 50 40 WHO WHOSIS: 38 30 20 10 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 MOH Estimates HFA WHO WHOSIS MICS 2000-2005 LSMS 1999/2003/2007 Source: Multisectoral Determinants of Child Health in Tajikistan (2009). Human Development Sector Unit | Central Asia Country Unit | Europe and Central Asia Region THE WORLD BANK 3  Figure 2: Leading Causes of Post-Neonatal Mortality in Tajikistan7 2009 2010 ARI, 20% ARI, 18% Other, 36% Other, 35% Pneumonia, 19% Pneumonia, 21% Congenital Congenital Malformations, 7% Acute Diarrhea, 18% Malformations, 7% Acute Diarrhea, 19% Total Cumulative Reported Deaths = 1,535 Total Cumulative Reported Deaths = 1,604 population health, policies focused on improving child health (ARI), pneumonia, and acute diarrhea still account for more outcomes are critical. than 50 percent of reported child deaths within the first year of life, a pattern that has remained persistent over the last eight years until the present (Figure 2). CAUSES Of CHILD MORBIDITy & MORTALITy 1.6 Child morbidity and mortality patterns in Tajikistan also vary significantly by geographic area. The lowest U5MR 1.5 Preventable illnesses contribute to a considerable propor- and IMR are in the Gorno-Badakhshan Autonomous Oblast tion of all child deaths in Tajikistan. A significant proportion of neonatal deaths are due to prematurity and low birth weight, while acute infections are the leading cause of deaths 7 Data are from official medical statistics and reflect only reported child in the post-neonatal period (Table 1). Acute respiratory illness deaths. Table 1: Leading Causes of Child Mortality in Tajikistan by Region (2001-2002) Khatlon Region & Dushanbe City Sogd Region & RRP Premature birth (32%) Prematurity/low birth weight (43%) Pneumonia (21%) Birth asphyxia (33%) Congenital malformations (12%) Acute diarrhea (12%) Birth asphyxia (8.6%) Pneumonia (6%) Post-neonatal Deaths Post-neonatal Deaths Meningitis/encephalitis (20%) Acute diarrhoea (25%) Acute diarrhea (17%) Severe/moderate malnutrition (24%) Severe malnutrition (16%) Meningitis/encephalitis (16%) Pneumonia (14%) Pneumonia (16%) Severe anemia (12.6%) Severe anemia (15%) Source: UNICEF Tajikistan, 2004. *Rayons of Republican Subordination (RRP). 4 Quality of Child Health Services in Tajikistan Figure 3: Tuberculosis and Measles Vaccine Coverage for Children Aged 12-23 Months 100 90 80 70 60 50 BCG Measles 40 30 20 Total Dushanbe Other Urban Rural Sogd Khalton RRP GBAO None Primary Basic Secondary general Sec special Higher Poorest 2 Quinitile 3 Quintile 4 Quintile Richest Source: TLSS 2007. (GBAO) region (54 and 46 per 1,000 respectively), as well as age of five was 29 percent nationwide, and highest in the Dushanbe City, while the highest rates are in Khatlon region Khatlon region at 36.9 percent, 27.9 percent the in Sogd (102 and 81 per 1,000 respectively), according to the 2005 region, 22.8 percent in the RRP, 25.4 percent in GBAO, and Multiple Indicator Cluster Survey (MICS). Several factors lowest in Dushanbe at 21.8 percent (UNICEF, 2010). which vary by region, are known to increase the risk of poor 1.9 Variation in these critical factors contributes to the child health outcomes in the country including short breast- increased risk for poorer child health outcomes within the feeding duration, late vaccinations, low use of oral rehydration country. Of course, these factors alone do not entirely account therapy (ORT) for diarrheal diseases, micronutrient deficien- for all observed child morbidity and mortality, but they are cies, and low antenatal care visits. For instance, contraceptive exacerbated or indirectly caused by other broader environ- use among married women is highest in GBAO region (55%) mental factors, such as region of residence, maternal educa- and lowest in the Khatlon region (33%). Similarly, immuniza- tion and household poverty levels. tion coverage rates were lowest in the Khatlon region and the Rayons of Republican Subordination (RRP) and highest in 1.10 While child health outcomes are affected by multiple the Sogd region. factors, this study chose to assess a very specific health systems aspect, namely the quality of health services in the primary 1.7 Notably, only 41 percent of children in the RRP had been health care setting. Service quality is affected by several fac- vaccinated against measles compared to 77 percent in Sogd tors including information, infrastructure, materials, drugs, (Figure 3). In addition, the proportion of exclusively breastfed human resources, what activities are done as well as how they infants aged 0 to 3 months was 23 percent in Dushanbe and are done.8 Quality assessment is a way to measure the differ- 33 percent in Khatlon compared to 61 percent in GBAO ence between expected and actual performance to identify (TLSS 2007). The practice of early weaning combined with opportunities for improvement.9 the chronic food shortages faced by many households, par- ticularly the poorest, contribute to the relatively high levels of stunting observed in Tajik children and is also reflected in the high incidence of low birth weight babies. 8 Massoud et al (2001). A modern paradigm for improving health care quality. Quality Assurance Project. 1.8 According to the most recent National Nutrition Survey 9 Lin & Tavrow (2000). Assessing health worker performance of IMCI (NNS), the prevalence of stunting among children under the in Kenya. Quality Assurance Project Case Study. Human Development Sector Unit | Central Asia Country Unit | Europe and Central Asia Region THE WORLD BANK 5  PRIMARy HEALTH CARE IN Management of Childhood Diseases (IMCI) strategy, initially on a pilot basis in 2000. The strategy, developed by the WHO TAjIKISTAN and UNICEF, aims to improve child health and reduce child 1.11 Over the last decade, Tajikistan has been reforming its mortality due to pneumonia, diarrhea, malaria, measles and health care system, which was based on the Soviet model of under-nutrition.11 The strategy has three main components: health service provision. The Soviet system was centralized (i) in-service training of PHC workers in the delivery of effec- and focused largely on curative inpatient services provided tive and efficient preventive and curative care for the cited by highly specialized doctors, with little emphasis on pri- problems; (ii) improving family and community practices on mary health care (PHC) and preventive services. As such, the child health; and, (iii) strengthening health systems support, strengthening of PHC services, including improvements in including the supply of drugs and vaccines.12,13 The WHO financial and human resource allocation to the primary care and UNICEF recommend an eleven day in-service training level, has become the basis for the ongoing health system program for health workers on the IMCI case management reforms in the country. Key reform directions for strengthen- guidelines. This consists of seven modules that combine class- ing the management and delivery of primary health care ser- room work with hands on clinical experience.14 vices include the introduction of per capita financing mecha- nisms, building management capacity, and the introduction 1.15 In Tajikistan training of health personnel on the IMCI of the Family Medicine (FM) model of practice.10 The guidelines has been done through continuous professional Ministry of Health has introduced these reforms in selected development (CPD) workshops ranging from three to nine districts of the country on a pilot basis with the support of day duration. The guidelines are also included in the six- various international donors. Family Medicine is seen as the month FM postgraduate training program, with two days most important pillar of the PHC reform as it is expected to (16 hours) specifically dedicated to them. This training is lead to better health outcomes through efficiency gains and intended to result in improvements in the quality of care improved population health service coverage. provided to sick children in PHC facilities. 1.12 It is recognized that building an adequate cadre of FM 1.16 The IMCI Strategy components encompass the main practitioners in Tajikistan will take several years. Thus a two programmatic directions that the MOH identified in several pronged approach is being pursued to meet the country needs policy and strategic documents as critical to improving the for FM providers; (i) retraining of health personnel specialized health of children aged five and under. These documents in another area (internists, pediatricians, obstetricians and include the National Strategy on Child and Adolescent Health nurses) as Family doctors and family nurses; and (ii) prepara- (2008-2015); Action Plan on Child and Maternal Nutrition tion of new FM specialists at the undergraduate level. (2009); and more recently, Tajikistan’s first Comprehensive National Health Strategy (2010-2020), in which maternal 1.13 The FM retraining is a six month long full time post and child health and primary health care are central features. graduate program that was initiated in 2001. The program includes theoretical and clinical training covering cardiology, 1.17 The Government implemented its IMCI strategy nation- obstetrics and pediatrics among other topics. In the last few wide in 2003. Its impressive support structure included the years, the Ministry of Health and its international develop- establishment of a National IMCI Center, three regional and ment partners have undertaken various small-scale assess- more than 65 district IMCI Centers to coordinate the pro- ments of the FM retraining process and to a lesser degree gram. Yet high morbidity and mortality rates among children the practices. However, no overall assessment exists for the under the age of five due to diarrheal diseases, malnutrition, performance of Family Medicine providers in the provision of critical primary health care services, particularly those related to maternal and child health care. 11 Arifeen, Bryce, Gouws et al (2005). Quality of care for under-fives in first level health facilities in one district of Bangladesh. Bulletin of the World Health Organization; 83 (4):260-267. 1.14 The Government’s concern over the high infant and 12 Gouws, Bryce, Pariyo et al (2005). Measuring the quality of child child mortality rates prompted it to introduce the Integrated health care at first-level facilities. Social Science & Medicine; 61:613-625. 13 Bryce, Victora & the MCE-IMCI Technical Advisors (2005). Ten methodological lessons from the multi-country evaluation of integrated 10 Family medicine is the medical specialty that provides continuing, management of childhood illness. Journal of Health Policy and Planning, comprehensive health care for the individual and family. It is a specialty 20 (supplement 1): i94-i105. in breadth that integrates the biological, clinical and behavioral sciences. The scope of family medicine encompasses all ages, both sexes, each organ 14 WHO and UNICEF (1997). IMCI in service training guide. system and every disease entity. American Academy of Family Physicians Available online: http://www.who.int/child_adolescent_health/docu- (2010). ments/9241595650/en/index.html 6 Quality of Child Health Services in Tajikistan and pneumonia persist. With appropriate case management, assess how FM and other PHC practitioners use the IMCI correct clinical assessment, provision of appropriate treatment guidelines. The WHO/Unicef IMCI Health Facility survey and counseling on continued home management, and care therefore provided a systematic and well tested approach to of the child at home the majority of these deaths can be pre- examine the quality of child care using the IMCI guidelines vented. The IMCI strategy provides comprehensive guidelines are a basis for assessing the performance of FM practitioners to treat common childhood illnesses at first level facilities. As through the collection of primary data. To assist in this assess- such if these guidelines are properly followed, they should ment, the team studied the nature and quality of services result in the provision of basic quality care for sick children. received by 300 sick children who visited 70 primary health care centers in three regions of Tajikistan and Dushanbe city. 1.18 To improve the quality of health services, and ultimately Details on the study methodology and descriptive statistics are child health outcomes, policy makers, PHC managers, prac- provided in Annexes 1 to 3. titioners, training institutions, and organizations need to Human Development Sector Unit | Central Asia Country Unit | Europe and Central Asia Region THE WORLD BANK 7  Chapter 2. Key Survey Results 2.1 The main findings of the survey are presented in the assessed children for all three danger signs as IMCI guidelines subsequent sections following the main areas of focus of the require. IMCI guidelines, with particular attention paid to the perfor- 2.4 There were large variations in the distribution of the indi- mance of FM doctors. cator by regions (see Figure 4). In the RRP, only 4.3 percent of children were checked for the danger signs. QUALITy Of CLINICAL CARE 2.5 Even when family doctors and pediatricians saw the chil- 2.2 This section presents the findings of case assessment, dren, no statistically significant difference was found. Danger classification, treatment and counseling, which are organized signs were assessed by these specialists in respectively only according to the IMCI priority indicators (shown in Annex 44.0 percent and 49.3 percent of cases. When analyzed by 4). length of IMCI training, the providers trained in the longer nine-day IMCI training performed better.15 Assessment Child checked for three danger Child’s weight checked against a growth chart signs (PI.1) – 45.7% (PI.3) – 26.7% 2.3 Assessing each child for three key danger signs: inability to 2.6 Most of the children were weighed (86.7%) on the day drink, vomiting everything after each feeding and convulsions of the assessment. However, only in a very small proportion is critical to detect cases of very severe disease requiring urgent of cases (26.7%) did health providers check the child’s weight referral. The IMCI guidelines require that these key assess- against the growth chart or standard tables or formulas to ment tasks be performed on every sick child, irrespective of determine the weight-for-age, which would indicate whether the specific complaint. Health care provider should assess all the child was within appropriate guidelines. Monitoring of the three danger signs in each case of a sick child visit. This children’s weight is critical in Tajikistan because chronic assessment is the first step for early identification of patients underweight is a problem in nearly a third of the children who need urgent treatment. When assessing each danger sign aged five and under. Because malnutrition is an underlying independently, researchers found that providers assessed only cause of childhood diseases, early identification and proper 66.3 percent of children for inability to drink, 65.0 percent management are imperative. The WHO growth charts are for vomiting everything after each feeding, and 52 percent for convulsions. In less than half of all cases (45.7%), providers 15 p value < 0.05. Figure 4: Assessment of Danger Signs by Regions (N=300 ) 80 63.8 % of children observed 60 56.9 53.5 45.7 40 20 4.3 0 Dushanbe Khatlon Sogd RRP Total 8 Quality of Child Health Services in Tajikistan  Figure 5: Assessment of Weight against Growth Chart by Regions (N=30016) 100 94.2 93.8 83.1 % of observed children 80 72.5 86.7 60 42.5 40 27.9 23.1 20 10.1 26.7 0 Dushanbe Khatlon Sogd RRP Total child weighted weighted assessed against growth chart part of the physical exam in pediatrics and a powerful tool two years of age be assessed for feeding practices (including that can sensitively detect a child’s nutritional status. In the breastfeeding for children under two years old, complemen- majority of cases, providers only weighed a child but did not tary feeding, and feeding changes during acute episodes of ill- assess whether the child’s weight was age and height appropri- ness). Such assessment is very important to establish adequate ate. Weighing a child without proper assessment has no value. feeding practices and prevent development of nutritional A regional comparison shows that the lowest proportion of problems. This study shows that of the 232 children under children assessed for weight-for-age was in the RRP (Figure two years of age who were not referred to a higher level of 5). care, less than half (45.3%) were assessed for feeding practices. Children with low weight-for-age were not more likely to 2.7 Children seen by family doctors were twice as likely to receive feeding assessment than those without this condition. be weighed and have their weight checked against the growth From 40 children with low weight and not requiring referral, chart compared to children seen by pediatricians (30.2% 45.0% were assessed for feeding problems (SI.3). vs. 14.9%17 ). No difference was found in this indicator for cases seen by current and former pediatricians when con- 2.9 No statistically significant difference in adherence to pro- sidering the provider’s specialization before retraining into tocols was found between cases observed by family doctors Family Medicine). When cases were analyzed according to and those observed by pediatricians (a finding similar to other the health workers training in IMCI, it was found that no indicators). However, providers trained in IMCI were more single child was assessed weight-for-age when a health worker likely to assess children for feeding practices, and the rate was was not trained in IMCI (25). Among cases seen by IMCI- higher among cases managed by providers who attended the trained health worker (275), a third of children received seven-day IMCI course. weight-for-age assessment, with a higher proportion observed among children whose providers attended the nine-day IMCI 2.10 Assessment of respiratory rate in children with cough course. or difficulty breathing . A child’s breathing rate is one of the most sensitive and specific indicators of pneumonia. Child under two years of age assessed for feeding The IMCI guidelines recommend assessing breathing rate practices (PI.6) – 45.3% and chest retractions in every case with a cough and diffi- culty breathing. In 147 cases, caregivers complained of their 2.8 The IMCI guidelines recommend that children under children suffering from respiratory problems. Nonetheless, providers only counted the breathing rate in half of the 16 Sample sizes for regions are Dushanbe: N=65, Khatlon: N=86, Sogd: cases (53.1%). Failure to assess for this important indicator N=80, RRP: N=69; Total: N=300. can easily lead to misdiagnosis, especially in infants. IMCI- 17 p value < 0.05. Human Development Sector Unit | Central Asia Country Unit | Europe and Central Asia Region THE WORLD BANK 9  trained health workers were more likely assess the respiratory 2.12 The index of integrated assessment for Tajikistan was rate of children than those examined by non-trained staff. 6.9, indicating that on average for every child, providers Assessment of respiratory rate was positively associated with performed a mean 6.9 assessment tasks of the WHO recom- the IMCI training duration, with a statistically significant mended 10 tasks. difference.18 2.13 Of the 10 main assessment tasks, health care providers most frequently perform child weighing, assessment of vac- Index of integrated assessment (mean) cination status, and the check for the three main symptoms, (PI.5) – 6.9 while assessment of weight-for-age is the most neglected 2.11 In any sick child visit, and despite the problem and task, as shown in Figure 7 below. complaints for the visit, the provider should perform several key assessments, according to IMCI guidelines. A thorough 2.14 The highest index was found in the Khatlon region (8.0) and systematic assessment helps to identify conditions that are and the lowest in the RRP (4.5). No difference was found not reported by the caretaker. To measure the completeness of between cases observed by family doctors and those observed the assessment received by each child, an index of integrated by pediatricians. The IMCI trainings had a positive influence assessment was used. The index consists of key tasks and gives on the integrated assessment. The index was higher for equal weight to each task done (score per task done = 1). It children observed by providers trained in a nine-day IMCI is expressed as the mean of the number of tasks performed in course (7.8 out of 10 tasks) than those seen by untrained each child (out of those that should have been performed). providers or those trained for three to five or seven days (6.1, This was based on the generic index proposed by the WHO, 6.3 and 5.8 tasks out of 10 respectively). which includes ten assessment tasks. These are child checked for three danger signs (1, 2, 3); checked for the three main Classification and Treatment symptoms (4, 5, 6); child weighed (7); and weight checked Child is correctly classified (SI.5) – 49.3% against a growth chart (8); child checked for palmar pallor (9); and, for vaccination status (10). The index of integrated 2.16 Of the 290 children not requiring referral to a higher- assessment enables follow-up of improvements in care and level facility, almost half were correctly classified according to progress over time. Taking into account each of the tasks of the main symptoms of childhood illnesses. The term “correct the generic index, the higher the number of tasks performed, classification