69448 Timor-Leste Health Sector Review Meeting Health Challenges and Improving Health Outcomes Report by the World Bank Final Draft October 6, 2006 Acknowledgements: The findings, interpretations and conclusions expressed in this paper are entirely those of the authors, and do not necessarily represent the views of the World Bank, its Executive Directors, or the countries they represent. Abbreviations ........................................................................................................................................................ 4 PREFACE ............................................................................................................................................................. 5 EXECUTIVE SUMMARY .................................................................................................................................. 7 Chapter I: Health Situation Assessment.............................................................................................................. 21 1.1. Infant and Child Mortality ................................................................................................. 21 1.2. Child and Adult Morbidity................................................................................................. 25 1.3. Fertility............................................................................................................................... 30 CHAPTER II: Utilization of Health Services ..................................................................................................... 43 2.1 Levels and Trends of Health Service Utilization .......................................................... 43 2.2 Determinants of Health Service Utilization .................................................................. 53 CHAPTER III: Service Delivery in the Health Sector........................................................................................ 62 3.3. Brief Description of Public Sector Services ...................................................................... 66 3.4. Role of the Private Sector in Primary Health Care ............................................................ 68 3.5. ISSUES IN GOVERNMENT PRIMARY HEALTH CARE SERVICES ........................ 71 3.6. Issues in the Hospital Sector .............................................................................................. 74 3.7. Human Resources in the Health Sector ............................................................................. 77 Background ........................................................................................................................... 77 The size and composition of the health workforce in Timor-Leste ....................................... 78 How many health workers is enough? The challenge of health workforce planning ........... 83 Changes in the health workforce over time .......................................................................... 87 Performance, incentives, and motivation.............................................................................. 90 Financial and non-financial incentives ................................................................................ 91 CHAPTER IV: Health Sector Expenditure and Financing ................................................................................. 93 4.1. Sources of Financing for Expenditures on Health Goods and Services ....................... 93 4.2. Out-of-Pocket Expenditures by Households ................................................................. 93 4.3. Non-Governmental Organizations Expenditures and Financing .................................. 99 4.4. Public Sector Expenditure and Financing ................................................................... 100 CHAPTER V: THE CHALLENGES AND RISKS MOVING FORWARD ................................................... 113 Figure 1-1: IMR and U5MR in relation to GDP per capita, selected countries ....................................... 22 Figure 1-2: ARI and diarrhea prevalence rate among children under five, in relation to GDP per capita, selected countries ........................................................................................................................................ 27 Figure 1-3: Percent of children under-five with ARI or fever taken to health facility, .............................. 27 Figure 1-4: Percent of children under-five with diarrhea who received either ORS or RHF, .................... 28 Figure 1-5: Incidence of TB, in relation to GDP per capita, selected countries ......................................... 28 Figure 1-6: Total fertility rate in relation to GDP per capita, selected countries ........................................ 30 Figure 1-7: Total fertility rate by mothers’ education level, selected countries ......................................... 31 Figure 1-8: Contraception coverage in relation to GDP per capita, selected countries .............................. 31 Figure 1-9: Percent of currently married women and men who are not using a contraceptive method and do not intend to use in the future, by main reason ...................................................................................... 32 Figure 1-10: Trend of underweight prevalence among children under five ............................................... 33 Figure 1-11: Prevalence of stunting, underweight and wasting among children under five ....................... 33 Figure 1-12: Prevalence of stunting and underweight in relation to GDP per capita, selected countries .. 34 Figure 1-13: Timing of undernutrition ........................................................................................................ 34 Figure 1-14: Timing of stunting and underweight by maternal nutritional status ...................................... 35 -i- Figure 1-15: Prevalence of anemia among children under-five by selective background characteristics .. 37 Figure 1-16:Prevalence of anemia among children under-five by selective background characteristics ... 37 Figure 1-17: Prevalence of anemia and iron supplementation during pregnancy among women aged 15-49 years by selective background characteristics ........................................................................................... 38 Figure 1-18: Prevalence of night blindness during pregnancy and vitamin A supplementation among ..... 39 Figure 1-19: Prevalence of vitamin A deficiency, under-five mortality and vitamin A supplementation status, selected countries ............................................................................................................................. 39 Figure 1-20: Children’s feeding status and prevalence of diarrhea ............................................................ 41 Figure 1-21: Feeding during diarrhea among children <36 months ........................................................... 41 Figure 2-1: Immunization trends among children 12-23 months for selected vaccination, ........................ 44 Figure 2-2: Immunization rates among children 12-23 months and per capita GDP ................................. 45 Figure 2-3: Utilization of curative child health services in relation to GDP per capita, selected countries 46 Figure 2-4: Treatment at a medical facility for infectious diseases among children under-5 years who had a symptom in the past 2 weeks: public vs. private facility.......................................................................... 47 Figure 2-5: Anti-malaria drug for fever among children under 5 who had fever in the last 2 weeks: ........ 47 Figure 2-6: Distribution of antenatal care provider among women who had a live birth within the 5-year period before the survey for the most recent pregnancy* ........................................................................... 48 Figure 2-7: Distribution of antenatal care facilities among women who received antenatal care within the 5-year period before the survey for the most recent pregnancy* ................................................................ 49 Figure 2-8: Utilization of antenatal care services and per capita GDP ....................................................... 49 Figure 2-9: Percent of live births within 5 years before the survey by delivery place and attendant for the most recent birth* ....................................................................................................................................... 50 Figure 2-10: Utilization of delivery care and per capita GDP .................................................................... 50 Figure 2-11: Percent of population reported health complaints and percent of those who visited a health facility in the past 30 days by sex and age group........................................................................................ 51 Figure 2-12: Health facility utilization among those with health complaints ............................................. 52 Figure 2-13: Distribution of all health facilities visited by those who had health complaints in the past 30 days ............................................................................................................................................................. 52 Figure 2-14: Trends of the outpatient facility use rate and distributions of health facilities among those with health complaints in the past 30 days: 1993-2001 .............................................................................. 53 Figure 2-15: Types of the closest health facility and reasons for non-utilization by region ....................... 54 Figure 2-16: Types of the usual health facility visited by the type of the closest facility * ........................ 55 Figure 2-17: Reasons for not using any health facility among those who had health complaints in the past 30 days but did not use any service, by selected socioeconomic and regional characteristics ................... 57 Figure 2-18: Potential problems getting health care identified among women 15-49 years by.................. 57 Figure 3-1: International comparison of level of health professionals ....................................................... 84 Figure 3-2: Health worker density across countries by level of income .................................................... 84 Figure 3-3: Differences in workload across districts .................................................................................. 90 Table 1-1: MICS and DHS data on IMR and U5MR by wealth quintile .................................................... 22 Table 1-2: MICS and DHS data on IMR and U5MR by ecological zone and region ................................. 23 Table 1-3: DHS data on IMR, CMR, and UMR by demographic classification ........................................ 23 Table 1-4: MICS and DHS data on IMR and U5MR by child gender, mother’s education, and ANC/delivery assistance ............................................................................................................................. 24 Table 1-5: MICS and DHS data on U5 children morbidity by geographic and socio-economic status ..... 26 Table 1-6: Prevalence trends of selected childhood infectious diseases among children under-5 years** 26 Table 1-7: TB prevalence by biological, regional, and socio-economic status........................................... 29 Table 1-8: Prevalence of under-nutrition in children under five years ....................................................... 35 Table 2-1: Immunization coverage trends based on health card or history among children 12-23 months 44 Table 2-2: Immunization services among children 12-23 months .............................................................. 46 Table 2-3: Percent of anti-malaria drug use among children under-5 years * ............................................ 48 Table 2-4: Average time to health facilities in minutes by region .............................................................. 54 - ii - Table 2-5: Percent of households by travel time to the closest health facility ............................................ 55 Table 2-6: Usual health facility visited when a household member is ill by residential area and .............. 56 Table 3-1: Civil Servant Classifications ..................................................................................................... 80 Table 3-2: Number and share of health workers in MOH by category ....................................................... 80 Table 3-3: Number of health professionals in MOH by location of work .................................................. 81 Table 3-4: Health workforce profile by district .......................................................................................... 81 Table 3-5: The number of health facilities in Timor-Leste: past, planned, and actual ............................... 85 Table 3-6: Staffing needs and the actual workforce: A comparison ........................................................... 85 Table 4-1: Cost of Transport and Examination/Treatment for Outpatient Consultations with Usual Health Care Provider .............................................................................................................................................. 94 Table 4-2: Average Cost of Examination/Treatment and Distance Traveled by Type of Health Care Provider Consulted in the Past Twelve Months by any Members of the Household.................................. 95 Table 4-3: Average Cost of Examination/Treatment and Distance Traveled by Type of Health Care Facility Visited in the Past Twelve Months by any Members of the Household ........................................ 95 Table 4-4: Main Components of Total Household Consumption Expenditure for Timor-Leste and Several Other Low- and Middle-Income Countries (percent) ................................................................................. 96 Table 4-5: Household Expenditures on Various Health Goods and Services in the Past 30 Days and the Past 12 Months, Consumption/Expenditure Section of the LSMS Questionnaire a/ .................................. 97 Table 4-6: Household Expenditure on Medicines Without a Prescription in the Past 30 Days by Type of Seller, Health Section of the LSMS Questionnaire..................................................................................... 97 Table 4-7: Household Expenditure on Visits to Various Types of Health Care Facilities in the Past 30 Days, ........................................................................................................................................................... 98 Table 4-8: Household Expenditure on Visits to Various Types of Private Health Care Providers in the Past 30 Days, Health Section of the LSMS Questionnaire ......................................................................... 98 Table 4-9: Household Expenditure on Most Recent Hospitalizations in the Past 12 Months by Type of Hospital, Health Section of the LSMS Questionnaire c/............................................................................ 99 Box -1: Implementation of Maternal and Child Health Policies................................................................. 66 Box 3-2: World Vision Clinic in Dili.......................................................................................................... 68 Box 3-3:Changes in the Administration of the National Hospital .............................................................. 74 Box 3-4: Human resources in Timor-Leste: What we don’t know ............................................................. 79 Box 3-5: Cuban medical aid ....................................................................................................................... 82 - iii - ABBREVIATIONS ANC Antenatal care ARI Acute Respiratory Infection BMI Body Mass Index BPS Basic Package of Services CCT Clinica Café Timor CHC Community Health Center CSWs Commercial sex workers DHS Demographic and Health Survey DPT Diphteria, Pertussis and Tuberculosis EPI Expanded Program of Immunization FHI Family Health International GDP Gross Domestic Product GOTL Government of Timor-Leste HPF Health Policy Framework HSP Hospital Services Package HSV2 Human Simplex Virus (congenital herpes) HWP National Health Workforce Plan IMR Infant Mortality Rate IP In-patient IHS Institute of Health Sciences IHA Interim health Authority IDD Iodine Deficiency Disorders LSMS Living Standards Measurement Survey MICS Multi-indicators Cluster Survey MMR Maternal mortality ration MOH Ministry of Health MSM Men who have sex with men MTEF Medium-Term Expenditure Framework NDP National Development Plan NGO Non-governmental Organization OP Out-patient ORS Oral re-hydration salt SIP Sector Investment Plan SUSENAS Indonesian National Socio-Economic Household Survey TB Tuberculosis TFET Trust Fund for East Timor TFR Total Fertility Rate U5MR Under-five Mortality Rate USAID United States Agency for International Development VAD Vitamin A Deficiency WDI World Development Indicators (The World Bank) WHO World Health Organization Note: The currency unit used in this report is US dollar. -4- PREFACE Purposes and Scope of the Report This report presents an analysis of the health sector’s characteristics, main issues and constraints, while also highlighting the considerable achievements so far. In doing so, the report should be a useful reference for anybody interested in the evolution of Timor-Leste’s health sector in the period since the vote for Independence in 1999, the subsequent violence and civil strife, the UN Administration and subsequent Independence in May 2002 through 2006. The authors of the report also hope that it would be a useful input in the preparation of the Medium-Term Strategic Plan the Ministry of Health intends to formulate in cooperation with its Development Partners. The World Bank, as one of the government’s Development Partners with a keen interest in the health sector, strongly supports the formulation of such strategic plan. In reviewing the health status of the population and the utilization of health care services (Chapters I and II), the report mainly focuses on the health of under-five children and women of reproductive age, and the corresponding maternal and child care services. This is dictated by the availability of survey data. The main source of nationally-representative health survey data currently available is the 2003 Demographic and Health Survey, which (as all Demographic and Health Surveys), centers on reproductive health and the health of infants and under-five children. Obtaining nationally-representative information on the health of other segments of the population (children over five years of age, adult men, women who are past their reproductive age), and about health conditions of women of reproductive age not related to pregnancy and childbirth, would require special epidemiological studies which have not been conducted in Timor-Leste so far. Methodology This report is based on an analysis of secondary sources of data, using both simple tabulations and multivariate analysis; a qualitative study of health service delivery and utilization which was commissioned for the study; extensive discussions with government officials and others with knowledge of Timor-Leste’s health sector; and specific information provided by the Ministry of Health at the request of the team that conducted the study. Field visits were carried out during the course of the main mission in November 2005 (in addition, several team members had substantial previous experience working on Timor-Leste and/or Indonesia). The study team, comprising World Bank staff and consultants, also brought to bear its collective knowledge of the health sector in other comparable countries, in order to provide a frame of reference to various aspects of the study. [Note: in the final version of the report, we should also make reference to the discussions with the government and other stakeholders on the draft report]. Participants and Acknowledgements This report has been produced by a task team led by Ian Morris. Magnus Lindelow made a major contribution to the report, particularly its analytical content, and led a considerable amount of the dialogue with Government and other stakeholders. Other members of the task team included Dr. -5- Rui Paulo de Jesus, Dorothy Judkins, (World Bank); Alexander Edmonds, Hugo Diaz- Etchevehere, Dr. Krishna Hort, Dr.Chris Scarf ,Dr. Ina Santos, Naoko Ono (Consultants) The mission met extensively with the Minister of Health, Dr. Rui M. de Araujo, and a wide range of senior officials within the Ministry of Health (MOH) including Basilio Martins, Director, Policy and Planning; Carlos Tilman, Director, Service Delivery; Valente da Silva, Inspector for Health Ministry; Domingos da Cruz, Finance Officer; Cate Keane, Chief Financial Advisor; Dr Earjing Larson, Chief Health Advisor, Dr. Rui Calado, Senior Advisor, Ministry of Health; Robert Smyth, Health Services Delivery Advisor. The team also wishes to express its profound thanks to the many other officials who met with the team, both at head quarters level and in the districts – particularly for the willingness and patience with which they worked with the team. The report is much the richer for this assistance. The team must also thank the major development partners working in the health sector in Timor Leste including AusAID, European Community, UNFPA, UNICEF, WHO, USAID and Portugal who met with the team during the reports preparation. Finally the team thanks all those NGOs, Church Health Services, private providers other health and community stakeholders including many villages during field visits who also gave time and assisted the team. Contents and Organization of the Report In addition to the Executive Summary, the report has five chapters and six Appendices. Chapter I is an assessment of the health status of the population, with emphasis on women of reproductive age and young children. Chapter II examines the patterns of utilization of health services and its determinants. Chapter III first describes the health services available in both the public and the private components of the health sector. This is followed by a discussion of issues in the government’s primary health care services and in the hospital sector. A final section reviews human resource issues in the health sector. Chapter IV discusses the patterns and trends of health expenditure and financing. The concluding Chapter V suggests some of the directions in which the work of the Ministry of Health could evolve and develop in the medium term, in order to continue furthering the improvement of health outcomes in the country. -6- EXECUTIVE SUMMARY Health Sector Accomplishments and Remaining Disease Burden 1. The Government of Timor-Leste has accomplished a great deal in the health sector since the country’s separation from Indonesia under United Nations Administration in 1999 and subsequent Independence in May 2002. The most important accomplishments include: (i) the formulation of a set of policies for the health sector, giving a clear priority to primary health care; (ii) the rehabilitation and rationalization of health care facilities; (iii) the development and adjustment of the health workforce; (iv) the mobilization of substantial resources for the health sector, resulting in an expansion in inflation-adjusted total public expenditure on health of about 60 percent from 2000/01 to 2005/06; and (v) the development of the Ministry of Health’s (MOH) organization and its planning and implementation capacities. 2. Development Partners have made a very important contribution to what has been accomplished. Their contributions have taken the form of both funding and provision of technical assistance. Regrettably, it was not possible in this report to give sufficient credit to the large number of bilateral and multilateral agencies that have participated in the post-1999 rebuilding of the health sector. In terms of funding, the contributions of the most important Development Partners are shown in the main body of this report. But, undoubtedly, their contributions in terms of ideas and encouragement have been at least as important as their funding. 3. The Government’s efforts have been complemented by a significant private component. Health care services in many areas of the country are provided by local and international NGOs, often Church-related, and by private doctors, nurses and midwives. Based on the estimated number of health care workers in government and NGO health facilities it is estimated that NGO clinics may be providing about one-fourth of basic health service delivery. There are also many traditional healers and birth attendants providing health care services, especially in rural areas, and pharmacies and other retail shops sell medicines to the public and often provide advice as well. 4. While progress in improving the country’s health indicators has clearly been made, Timor-Leste’s population of about one million still suffers from a substantial burden of morbidity and premature mortality. The latest (2003) estimates of infant and child mortality rates indicate that their levels compare favorably with those of other countries with per capita incomes similar to Timor-Leste’s, but they are still unacceptably high—60 and 83 per thousand live births respectively. Moreover, data from a 2002 survey showed that 56 percent of children under age five had experienced some form of illness in the two weeks preceding the survey, with malaria, acute respiratory infections and diarrhea being the leading diseases in this age group. 5. Direct estimates of maternal mortality are not available. However, maternal mortality is likely to be high, because: (i) the total fertility rate, i.e. the average number of children born to a woman over her reproductive years, is one of the highest in the world [7.7 children according to the 2003 Demographic and Health Survey (DHS)]; (ii) short intervals between births are highly -7- prevalent; and (iii) relatively few births are assisted by attendants with the necessary training. Using such indirect evidence and regression analysis, WHO has estimated the maternal mortality ratio in Timor-Leste at 660 per 100,000 live births, which is similar to the levels prevailing in a number of sub-Saharan African countries. 6. Concerning adult morbidity, the data are scant. Some information is available for tuberculosis and malaria, as well as for HIV/AIDS and sexually transmitted diseases. WHO has reported a prevalence rate of tuberculosis in 2003 of 754 cases per100,000 inhabitants, or about 0.8 percent. However, in the 2003 DHS, 5.6 percent of males and 7.1 percent of females reported having at least two of the three typical symptoms of tuberculosis (persistent cough, coughing blood, weight loss) in the month prior to the survey, which suggests that the WHO figure for tuberculosis prevalence may be too low. WHO has also reported that in 2003 there were 31,819 cases of malaria confirmed by laboratory tests. The number of clinically diagnosed cases (probable cases not confirmed by laboratory tests) was much higher, at about 100,000. Both P. vivax and P. falciparum types of malaria are present. The first AIDS case was confirmed in 2001. The extent of the problem is not known with sufficient certainty due to lack of national- level HIV sentinel surveillance. But according to research by Family Health International in 2003, the overall HIV prevalence rate was a low 0.6 percent, and just three percent of female commercial sex workers in Dili were HIV-positive. Other sexually transmitted diseases were common among these workers, however: one-fourth had gonorrhea and/or chlamydia, and 60 percent had herpes. 7. The data show that there are important differentials in the health of the population according to socio-economic status, with the poor accounting for a disproportionately large share of the overall disease burden. DHS data for the ten-year period 1994-2003 indicate that the under-five mortality rate for the poorest 40 percent of households (when ranked by their wealth) was 121/1,000 live births; for the middle 40 percent, 101/1,000; and for the richest 20 percent, 86/1,000. Infant and under-five mortality are also lower for children living in the lowlands, where the majority of the urban areas are located. The 2003 DHS data also show that children were more likely to die the shorter the interval between births. As in many other countries, mothers’ education achievements were found to be strongly and inversely correlated with infant and under-five mortality. 8. A good deal of information is available on the nutrition status of the population. In the 2003 DHS sample, almost two-thirds of children under five years of age were under-weight (low weight for age), with 14 percent being severely under-weight. Moreover, 53 percent were moderately stunted (low height for age, an indicator of chronic under-nutrition) and 31 percent severely stunted. And one in ten children under the age of five was wasted (low weight for height, an indicator of acute under-nutrition). Such levels of under-nutrition are high by comparison to other countries with per capita incomes similar to Timor-Leste’s. The prevalence of both stunting and under-weight were significantly higher among children from rural areas, poorer households, and those whose mothers were less educated. The nutrition situation among adults was a little better than that among preschool children, but still posed a serious problem. The above refers to protein-calorie deficits, but there is also evidence that micronutrient deficiencies, including anemia, Vitamin A deficiency, and iodine deficiency are quite prevalent. 9. What are the causes of under-nutrition in Timorese children? One likely cause is household poverty. The 2003 Poverty Assessment estimated the incidence of poverty among -8- households by first estimating a “poverty line� and then measuring the number of households with per capita consumption below that line. Timor-Leste’s poverty line of $15.44 per capita per month was defined as the minimum expenditure needed to purchase a food basket that provided 2,100 calories per person per day and a very basic basket of non-food consumption goods. It was determined that almost 40 percent of all households in the country had levels of consumption below the poverty line at the time of the survey. This is quite a significant degree of poverty. Undoubtedly, many of these households must have at times purchased food in amounts insufficient to provide every member of the household with the minimum number of calories and protein required for adequate nutrition. It seems clear, however, that there are other causes at work, since data from DHS 2003 indicate that about 40 percent of children from the richest 20 percent of households were underweight. Other factors contributing to under-nutrition among young children are inappropriate feeding and caring practices and childhood illnesses. While according to data from DHS 2003 breastfeeding was a common practice, exclusive breastfeeding was not. By the age of five months, less than a quarter of children were exclusively breastfed, bottle feeding was common, and the prevalence of diarrhea started to increase sharply. During episodes of diarrhea only seven percent of children were given more fluids, and nearly half of them were given less, thus increasing the risk of dehydration. Many children were also not given solid foods in sufficient quantity to complement milk intake during their weaning period. 10. Inadequate dietary intake and poor health, often caused by infections, create the vicious synergistic cycle that is responsible for much of the morbidity and mortality among young children around the world. When children do not consume enough quality food, immune response is lowered, rendering them more susceptible to infectious disease. At the same time, ill children deplete their nutritional stores because of poor absorption of nutrients and the increased demands of combating disease. This vicious cycle is clearly at work in Timor-Leste. Gaps in the Utilization of Essential Health Services 11. In spite of the substantial progress achieved in developing a functional health sector, the utilization of a number of essential health services has not attained the levels that would be desirable from a public health point of view. The most recent information on immunization coverage indicates that in 2003 the percentage of children 12-23 months who were immunized against measles was about 56 percent, while the percentage who were fully immunized (i.e., BCG, polio, measles, and three doses of DPT) was only about 18 percent. These levels of immunization are substantially lower than those of other countries with per capita incomes similar to Timor-Leste’s. 12. The care of new-born children provides crucial opportunities to improve neonatal survival. WHO recommends post-natal visits at six hours, six days, and six weeks after a child is born. In Timor-Leste, however, only 12 percent of neonates born in the five years before 2003 had any postnatal care. The majority had been seen by a nurse or a midwife (61 percent), followed by a traditional birth attendant (32 percent), and a doctor (five percent). Gaps also exist concerning the utilization of services related to common conditions in under-five children: survey data indicate that in 2003 only about 63 percent and 68 percent of those children below five years of age who had acute respiratory infections and fever in the thirty days prior to the survey, respectively, were taken to medical facilities. -9- 13. In the area of maternal health, only 61 percent of pregnant women in the DHS 2003 sample had received any antenatal care from qualified health care providers, usually a nurse or midwife. Eighty-six percent of this care was provided at public facilities. In addition to how widespread the provision of antenatal care is, quality of care is an important indicator to be monitored. In DHS 2003, among those pregnant women who had received antenatal care, only 44 percent had received iron supplements; 35 percent had their blood pressure measured; 25 percent had been informed about symptoms of complications; and 51 percent had received one or more doses of tetanus immunization. With regard to delivery care, in the DHS 2003 sample only 18 percent of live births had been delivered by a skilled birth attendant (doctor, nurse, or midwife), and only 10 percent of live births had taken place at a health facility (with the rest taking place at home). As compared to other countries at similar levels of per capita income, the percentages of deliveries at a health facility and of deliveries assisted by a skilled birth attendant are low in Timor-Leste. In addition, only 0.3 percent of all live births were assisted by cesarean section, which is far lower than what international experience indicates would have been medically necessary, indicating problems in access to emergency obstetric care. The low percentage of births taking place at health facilities also results in most women being deprived of timely postpartum care, which is critical to monitor and manage maternal complications. Determinants of Health Service Utilization 14. Health service utilization is affected by various factors such as physical access to health facilities, financial constraints of clients, information, and quality of care. The 2001 Living Standards Measurement Survey (LSMS) examined the reasons for not using any health facility when health complaints were present for any member of the household in the past 30 days –i.e., these data refer to utilization of curative rather than preventive services. The most commonly reported reason for not using any health facilities was that ‘the symptom is not serious enough’ (48 percent), but this should be interpreted carefully since the actual severity of symptoms cannot be verified. About 42 percent reported access problems (distant health facility or no transportation) as a barrier. About three percent reported economic reasons such as expensive health care or transportation expenses, and four percent reported service quality related issues (health care worker unfriendly, absence of health care worker, and low quality of care). Moreover, survey data indicate that --as in many other countries-- important socioeconomic differentials in health service utilization exist in Timor-Leste. Children living in rural areas or in poor households, and those whose mothers have no formal education, are less likely to be immunized and less likely to be treated at a health care facility when ill. Similarly, women living in rural areas or in poor households, and with no education are less likely to receive preventive maternal health care. Residents in rural areas are also less likely to visit outpatient facilities than those in urban areas. 15. In addition to what can be learned from survey data, further insights on household behavior towards utilizing health care services can be gained from qualitative studies. One such study was commissioned for this review. The methodology of the study consisted of focus groups and semi-structured interviews designed to elicit information that cannot be collected as part of large, representative surveys. The study suggested that patients are open to different diagnostic and treatment approaches. Treatment options generally fell within four broad categories: self-treatment with traditional (herbal) medicine, self-treatment with pharmaceuticals, consultation with a traditional healer, and treatment by a formal sector health care provider. Complex factors affected which of these options -- or which combination -- was chosen. The - 10 - study found that there was a willingness to combine traditional cures with modern medicine. Many patients did not hesitate to visit different providers sequentially if treatment did not appear to be effective. Attitudes towards modern medicine were found to be generally positive, but the study suggested that patients may often underestimate the risks posed by their illnesses. Malaria, for example, is often viewed as not a serious illness. This informal rating of illness severity may lead patients to delay or forego medical treatment. The qualitative study generally supported the findings from survey data that distance and difficulty of transportation are major barriers to the utilization of health care facilities, though respondents indicated that willingness to travel to health facilities also depends on perception of the health risk (which as noted is often underestimated). The study further suggested that cash can be an especially significant barrier for rural communities. Since cash incomes for farming communities have substantial seasonal variations, there may be times of the year when there is no money available for transportation or health care costs. Finally, some respondents complained of rudeness and negligence on the part of health workers, along with a lack of explanation of diagnosis and treatment, and preferential treatment to some patients. Looking Forward: the Challenges Ahead 16. The discussion so far indicates that while much has been accomplished in the health sector since 1999, there is ample scope for further improving health outcomes. Infant and child mortality are still high, and so is maternal mortality. Communicable diseases, especially malaria and tuberculosis, are still major sources of morbidity and premature mortality. Nutrition deficiencies in young children are widespread. And while data on non-communicable diseases such as hypertension, cardiovascular disease and diabetes are not available, these diseases are also in all probability significant causes of morbidity and premature mortality. It is also clear that poor households bear a disproportionately high share of illness. Moreover, faced with a significant disease burden, households (and more so poor households) are not utilizing various essential health services to the extent needed, and there is also some evidence which points to deficiencies in the quality of care. 17. The final chapter of the report discusses the challenges the Government would face in the coming years in seeking to improve the health of the population, and suggests some possible approaches to addressing these challenges.1 It is first suggested that a key challenge for the MOH is to translate the existing formal policy framework into consistent action on the ground. Second, there are the challenges of motivating households to be more active participants in protecting their own health, and helping them to overcome financial constraints --the demand- side challenges. Third, there are the challenges implicit in further improving available services -- the supply-side challenges. And fourth, the MOH will have to continue its efforts to improve management, planning and budgeting, including donor coordination aspects. 1 This report focusses on the health sector. There is no doubt, however, that government activities in other sectors can have a large effect on health outcomes. Some obvious examples are water and sanitation, transport and communications, and general education – particularly of girls. Economic policies that facilitate economic growth and poverty reduction are also an important if more indirect factor in improving health outcomes. Hence improving upon its performance in these other spheres of activity is also part of the challenge the public sector faces in seeking to improve health outcomes beyond what has already been achieved, though MOH itself may only have a limited role to play in these other areas. - 11 - Health Policy Development and Implementation 18. A key challenge for the MOH is to translate its formal policy framework into consistent action on the ground. In the past several years the MOH has formulated an impressive set of policies. Some of the most important components of this policy framework include the specification of a “Basic Package of Services�, to be provided with no charges; giving priority in expenditure to these services by deliberately constraining hospital expenditures; deploying District Health Management Teams, now primarily responsible for the management of district health services and implementation of district health plans; and formulating an essential drugs list calibrated for each facility type/level of service. The Basic Package of Services, which is appropriate for the current state of development of the country, defines the minimum range of services the public can expect at each level of the Government’s primary health care facilities. A complementary Hospital Service Package is under development with the objective of setting priorities for hospital services and managing demand for non-basic services. A Private Sector Policy acknowledges that the private sector has an important role to play in complementing government health services. 19. The MOH has made a good start in implementing the formal policy framework on the ground, but some challenges remain. The main ones are:  The key health program outcomes agreed at the national level are not consistently monitored and evaluated by managers. More consistency could be attained by building on the ongoing district planning process, emphasizing the monitoring and evaluation of progress towards agreed targets. A regular forum in which district successes and problems were jointly discussed would assist in this process. It would also be necessary for MOH headquarters managers to encourage and support district managers to take corrective action where program outcomes lagged their targets.  There are frequent challenges to the ability of the MOH to “stay the course� with its core programs. It is critical that the focus of the MOH, with its limited institutional and financial capacity, not be diverted by pressures to keep adding to the list of priority programs to be implemented. The report suggests specific ways the MOH might proactively manage demand for expensive non-basic hospital services -- a potential major risk to preserving the priority on basic services.  Not all health workers have “bought into� the priority to be given to the Basic Package of Services, or fully understood/accepted their role within the current service delivery system. An approach to this problem could include intensive participatory consultation on health priorities and job delineation, appropriate pre-service and in- service training, and career planning --including attention to remuneration arrangements and other incentives.  Fourth, Development Partners sometimes pose a threat to the MOH’s ability to focus its limited implementation capacity on core programs by: (a) seeking to fund new non-core programs; and (b) insisting that their programs be managed off-budget, thus creating parallel chains of accountability. To lessen this problem, the report strongly supports the Ministry’s initiative to develop a five-year Strategic Health Plan - 12 - including a medium-term public expenditure framework. Such plan could form the basis for the formulation of Development Partner programs which increasingly used government systems and accountability patterns (see section on expenditure priorities below). Demand Side Challenges 20. It has already been noted that many Timorese are not making use of available essential health services. The discussion under Determinants of Health Utilization indicated that two important factors preventing greater use of those services are inadequate household-level information and financial constraints--both of which could be considered “demand-side factors�. There are significant opportunities for government interventions to: (a) encourage households to adopt better health practices and better understand the risks associated with various health conditions; and (b) help them to overcome financial barriers to increased utilization. 21. Making Households More Active and Better Informed Participants through Health Education. In Timor-Leste, as in many other countries, the Government has tended to see households mostly as passive recipients of health services. But experience from around the world indicates that when households become more active participants in looking after the health of their own members, health outcomes improve. The impact of this factor is stronger the better informed the households are about common health conditions, the self-help measures that they can take, and the health services available within their reach. The evidence suggests that many households are not adequately informed when it comes to health. The deficiencies observed in the care and feeding of young children and the tendency to underestimate the risks posed by some medical conditions such as malaria or childbirth are cases in point. The widespread reliance on traditional healers is another aspect of the gaps in households’ knowledge about health and health care services. Hence the report suggests that the MOH give a higher priority to educating households and changing household-level behavior. Some of the messages to be given could focus on actions the households themselves can take to prevent disease, such as good basic hygiene (e.g. regular hand-washing) and child feeding and care practices. Another set of messages could address the beliefs and practices of households in relation to various health conditions and the related utilization of health services ((e.g. malaria, ARI and its association with indoor air pollution). Different options should be explored for delivering the messages, such as mass media, the use of community health workers, and an expanded role of NGOs and other civil society groups. The potential of health education in schools should also be explored. These options would have to be piloted and evaluated to ensure messages are positively affecting household behavior. 22. Addressing financial constraints. Respondents in both the DHS 2003 and the qualitative study pointed to financial constraints as a reason for not using the available services. This finding suggests that the Government may wish to explore instituting some sort of demand-side financing scheme. For example, a number of countries have been experimenting with conditional cash transfers, whereby the Government pays a subsidy to households against a commitment to engage in certain behaviors, such as to make use of certain under-utilized essential health services (e.g., immunization, antenatal care). There are complex design issues that would need to be addressed, such as the degree and mechanics of targeting (if the subsidy were not a universal one to which all households are entitled). The experiences of other low- and middle-income countries which have started demand-side financing schemes should be studied - 13 - prior to piloting any such scheme in Timor-Leste. Major lessons with conditional cash transfers include the need for strong administrative capacity, clear and simple rules, and an adequate health system to supply services in response to demand --a failure on the supply side would bring the program into disrepute. Supply-Side Challenges 23. Three main supply-side challenges confront the MOH in its drive to expand the utilization of essential health services. First, there is the challenge of improving the technical quality of services. Second, while about 80 percent of the population lives within two hours of a health facility (a major achievement), a significant proportion of the population faces access issues. Third, there are several important human resources related challenges, including ensuring that qualified health workers are deployed where needed; that they treat patients with respect; that the skill mix in each facility is appropriate; that the role of the rapidly expanding cadre of doctors is clearly defined; and that the structure of incentives for health care workers is conducive to keeping workers motivated and committed to their jobs. 24. Improving Technical Quality. There is some evidence that the technical quality of certain essential health services does not meet minimum internationally accepted standards, thus detracting from the health benefits patients obtain when they do use the services. A case in point, as already noted, is antenatal care. More information is needed to determine the extent of the problem. However, now that most of the basic health care infrastructure is in place, the time seems right for MOH to start a concerted effort to develop a system to monitor and upgrade the quality of health care services, in both its technical and non-technical aspects, starting with services in the public sector. Spelling out the details of a suitable quality assurance program exceeds the scope of this report, but work on such a program is suggested in this report as one of the desirable directions for the development of the Ministry’s work. Enhanced in-service training and emphasizing the role of health facility managers in helping to ensure technical quality may be a useful starting point. The increase in the number of doctors in the system and decisions about how they could best be used should provide part of the solution to these quality issues. 25. Improving Access. Difficulty of access --distance and difficulty in securing transportation-- has often been mentioned by survey respondents as reasons for not seeking services from the formal health care sector, especially by those respondents from rural areas. Nevertheless, DHS 2003 showed that, nationwide, 81.7 percent of all households were within two hours’ travel time from the nearest health facility, with the remaining households being further away. The difficult terrain and scattered distribution of the population, often residing in small villages, presents a challenge for ensuring cost-effective access. How best to improve access further is a problem that needs to be analyzed locally. Additional health care facilities are one of the options, but there is a trade-off between ease of access in terms of distance and efficiency in terms of facility utilization rates and staff workloads. Beyond a point it would be difficult to justify adding more facilities. Hence the option of establishing new facilities needs to be considered together with three other options to facilitate access: (i) increased outreach activities from existing facilities; (ii) piloting and evaluating the proposed community health worker program; and (iii) taking measures to improve transportation and/or reduce transportation costs for those seeking care. Furthermore, in those cases where establishing new health care facilities appeared to be the best alternative, the MOH should consider the option of contracting - 14 - out services to private providers, as opposed to creating additional in-house capacity. A clear distinction should be made between the desirability of public financing of a program and whether government should implement the program itself. Contracting out would generally provide the MOH with more flexibility in the deployment of resources and a greater scope for imposing discipline and demanding results. The MOH is to be commended for its agreement to enter a partnership with the largest health sector NGO, Café Timor, involving government financing of drugs in return for Café Timor continuing to serve non-cooperative members. Options to enter similar agreements with other NGOs could be explored and may also assist in improving the quality of services provides. 26. The lack of an extensive and functioning road network is a challenge in many areas. However, a large proportion of the population, including mothers and children, regularly visit towns that function as regional and district markets to both buy and sell food and other goods, and to attend church services. If the MOH began providing some health services in these locations, either in outreach mode or in facilities provided by the community, accessibility to services could be greatly increased. This would require some changes in how staff is deployed, which could be built into district and sub-district plans in consultation with local authorities and communities. 27. Human Resources Issues. The MOH must ensure that qualified health workers are deployed where needed. However, the MOH is facing difficulties in filling positions in remote health care facilities. Deployment is currently application based, and mandatory deployment to specific posts is not used. Changing to a system of mandatory deployment, however, may lead to greater absenteeism. A better strategy may be to recruit staff originally from the local communities where there are vacant posts. The MOH has already started using this approach for recruitment to midwifery training. Other strategies could include special incentives such as establishing fixed three-year assignments in areas with staff shortages and offering accelerated access to promotions, cash bonuses and/or access to training for those who take up these assignments. 28. It is also critical for increasing utilization of services that health workers’ treatment of patients is friendly and respectful. As noted, there is evidence that this is not always the case. Survey data do not clarify whether this is a problem which is widespread in both government and private facilities, or it is mostly a problem in public sector facilities. An indication that it may be the latter comes from the qualitative study commissioned for this report. One of the findings of the study was that government health workers do not see that an effective performance monitoring system and incentive system is in place in the public sector, unlike what they perceive to be the case in NGO health care facilities. This is an important area of work for the MOH, which should aim to more closely monitor staff performance and applying rewards and sanctions in line with performance. 29. In defining the staff complements to be deployed at each type of facility (skill mix) in the context of formulating the overall health policy framework, the MOH took into account: (a) the numbers and composition of health workers thought likely to be available over the medium term (e.g., it recognized that there was likely to be a significant shortage of doctors, midwives and technical staff); and (b) the services to be made available by each type of facility. It was also recognized that it would be possible to gradually increase the number of midwives through domestic training. It was a reasonable assumption at the time that financial resource constraints - 15 - and the long period required to train doctors would severely constrain the MOH’s ability to increase their numbers significantly in the short- to medium term. However, there have been two significant changes since the health policy framework was formulated: (a) the recruitment of 300 Cuban doctors and the decision to train over 600 Timorese doctors in Cuba; and (b) significantly increased government revenues from oil/gas extraction. These developments pose important challenges to the management of the skill mix and opportunities to improve the quality of services. 30. The challenges arising from the deployment of large numbers of doctors, particularly expatriate doctors, should not be underestimated. What will be the role of foreign doctors relative to Timorese health facility managers in overseeing facilities and programs? Which patients will be seen by doctors and thus what are to be the new roles of nurses, midwives and other paramedical staff in service delivery? Given the current relatively low staff workloads in health facilities, will the numbers of paramedical staff deployed need to fall to make way for doctors? Are the functions of health facilities to change as a result of doctors’ deployment? Will all this have implications for the design and equipment of facilities, or the range of pharmaceuticals consumed at each facility level? Clearly, these are all challenges that the MOH will need to address in the coming years. It is also important that MOH continue with its effort to train and deploy more midwives throughout the country to meet the demand for reproductive health services. 31. In the qualitative survey, a number of staff argued that financial and other incentives were needed to ensure willingness to serve in remote rural areas. Currently there are no provisions for special payments or bonuses for this purpose in the Statute of Civil Service. The expansion of newly trained doctors in the coming years, many of whom will be expected to work at district and sub-district level, is likely to intensify the deployment challenge. Discussion of staff incentive issues, however, should be much wider than deployment issues. Staff is the most critical input for the health system, but unlike inputs such as facility infrastructure and drugs, people have legitimate needs and preferences. Non-material incentives such as inclusiveness in the important decisions –on health priorities, the objectives of health programs, the role of different staff cadres in the health system, etc-- can positively affect staff commitment and productivity. In Timor-Leste, it has been a challenge for staff to move from a very specialized cadre structure under the former Indonesian administration to one where staff is expected to perform multiple tasks under the current health policy framework. 32. The salary scale which the MOH can offer is set on a government-wide basis and has not been changed for the last six years. The scale is also very restrictive, with scant possibility for most staff to gain promotion. Over time this situation will become less and less tenable, and the MOH may wish to consider taking the initiative in making proposals to the Government for civil service-wide changes, including: (a) introducing salary points within each grade that could be climbed over time based on performance; (b) a career path for staff to be able to reach higher grades with relevant training, experience and performance achievements; and (c) reducing the salary compression between the highest and lowest grade points. Management, Planning and Budgeting 33. The report recognizes the very significant achievements by the MOH in developing institutional capacity starting from a dire situation. In addition to loosing upwards of 85 percent - 16 - of its health infrastructure in the violence at separation from Indonesia, the country lost some 80 percent of its senior and middle managers in the MOH. Most of the new staff at MOH headquarters and hospital and health facility managers had to learn new jobs for which they had no previous experience. Further, there were less than 15 Timorese doctors in the whole country. Nowadays, MOH has a capable senior management team overseeing the health system. A District Health Management Team manages the services in each district, and is responsible for preparing district health plans and for reporting on program outcomes and expenditures on a quarterly basis. Hospitals have been given increased autonomy to manage resources within agreed activity plans and an increasingly autonomous medical store ensures timely supplies of pharmaceuticals and other supplies. All-in-all, it has been a stellar performance. Yet, institutional capacity perhaps remains the binding constraint. There are clear challenges in transforming policy priorities into effective services on the ground. A few general principles may be of help in achieving this goal. 34. First, the MOH should maintain the centrality of annual district plans as a basis for focusing program priorities and activities according to local needs. These plans should set realistic targets to be achieved and be adequately funded (by both Government and Development Partners) through government systems, making sure that managers fully understand what their budgets are and what they are expected to achieve with those budgets. Second, implementation of plans (including annual hospital plans) should be regularly monitored and evaluated by facility and district health managers and by central MOH staff. Problems, successes and failures should be identified and discussed with the active support of MOH headquarters in order to learn from mistakes and determine what works best. Third, accountability for program results and resource use should be increasingly devolved to districts and facilities. The ongoing efforts to establish accounts at the district and sub-district/facility level should be applauded as a decisive move in the right direction. Financial discipline is of course paramount and it is critical staff are appropriately trained and supported by central MOH staff. Fourth, the MOH needs to further define the appropriate roles for its headquarters staff and how best to support the district (and hospital) planning and management processes. In recent years there seems to have been a clearer role for the Policy and Planning Division than it has been the case for the technical divisions. There has been some tension at MOH headquarters between supporting districts (and hospitals) in implementing their programs and taking over responsibility for implementing those programs. There are a number of reasons for this tension, including donors lobbying the MOH to organize vertical programs separate from the core MOH programs. The planned organizational and management review should provide a good basis to work through these issues. Expenditure Priorities 35. A central theme of this report, as discussed, has been that while much has been achieved there remains significant scope for further improvements in health outcomes. Infant and child mortality remains unacceptably high as does maternal mortality. Communicable diseases, especially malaria and tuberculosis, are still major sources of morbidity and premature death. Nutrition deficiencies and upper respiratory tract infections in young children remain widespread. It is also probable that non-communicable diseases are emerging as a threat to population health. It is also evident that poor households bear a disproportionate share of the burden of disease and that demand for health services by households is less than optimal – in part due to deficiencies in the quality of care, lack of information on the seriousness of specific illnesses and problems with access, particularly due to poverty and costs of getting to health - 17 - providers. Within the government systems framework (planning, budgeting and monitoring of programs) outlined above the MOH would strongly benefit from establishing a medium term program as outlined below. 36. Thus a question facing the MOH is how to set priorities for expenditure on health given the major health problems facing the country and the critical importance of the capacity constraint and the challenge of transforming policy priorities into effective services on the ground. Given the limited capacity of the public sector to deliver services, the central aim should be to prioritize delivery of services, particularly at the district and facility level, and to reduce the government’s direct responsibility for service delivery and put in place more viable options. One approach to setting expenditure priorities is to systematically work through the following six elements of a health strategy: 1. Focus on high-priority health outcomes using the most cost-effective strategies, with particular emphasis on targeting the poor. There is agreement that the high-priority health outcomes include reducing infant, child and maternal morbidity and mortality, controlling infectious diseases (tuberculosis, malaria vaccine-preventable diseases) and encouraging healthy population behaviors and increasing the use (demand) for quality basic health services including by reducing the access costs for the poor. Improvements in health outcomes will also be achieved by improved implementation of cross sectoral priorities (water supply and sanitation, transport and communications, education, particularly of girls, and control of indoor air pollution). In order to achieve these MOH should: a Identify technical options for achieving the desired health outcomes and ways in which options might be packaged for delivery. This might include programs to extend vaccination coverage, targeted health promotion (e.g. nutrition awareness, weaning practices and recognition of health conditions and their seriousness), local level based treatment of pneumonia in children to reduce child mortality, efforts to ensure basic steps are followed with anti-natal counseling (e.g. iron supplementation, measuring blood pressure, counseling on recognizing signs of potential pregnancy complications), STD treatment, attending to etc. b Set priorities for public expenditure using public finance criteria. Three major criteria are applicable. One, if benefits (or costs) from the intervention accrue to more than the individual from its consumption there is a case on public fiancé criteria for public funds to be used (e.g., the benefits to vaccination spill over to the whole population so long as 80-90 percent of the population are covered, cure of TB results in benefits to the whole population by preventing the further spread of the disease. Two, the absence of perfect information means that individuals do not have the information required to have appropriate health related decisions means that health promotion designed to effect behavior change is justified on public finance criteria. Three, on equity grounds it is appropriate for public subsidies to be provided to ensure obtain services. - 18 - c Commit to monitoring and evaluating progress towards the agreed health targets. This involves ensuring the Health Information System (including disease surveillance) is sustained and further developed and that appropriately structured population based surveys (e.g., demographic household surveys, quantative and qualative health seeking behavior surveys) are conducted on a timely basis and used for evidence based monitoring and evaluation of programs. 2. Public-Private Sector Interfaces. There is a need to explicitly consider the potential of the whole sector – public and private – to contribute to health outcomes. Thus it is important to: a Focus the public sector on those public goods (1b above) it can effectively deliver. It is critical that MOH focuses on those services that cannot be provided by the private sector and that its institutional capacity can deliver. b Promote appropriate involvement of the private sector in the delivery of private goods such as ambulatory care. c Strengthen collaboration between the public and private sectors in service delivery. This provides a major opportunity for MOH to tap the capacities in the private sector to assist in meeting overall health objectives (e.g. further development of the relationship with CAFÉ Timor; increased cooperation with church health facilities (e.g. access to pharmaceuticals and basic health equipment to improve the quality of services in church facilities). 3. Prepare an Expenditure and Implementation Plan for the whole of the publicly funded part of the health sector (covering all government and donor-financed programs) based on the priorities and activities identified in the preceding steps and consistent with available resources. 4. Create Fiscal Space. This can be achieved in three ways: (i) achieve savings in the sector; (ii) making more efficient use of existing funds; and (iii) increasing sectoral allocations. 5. Prepare a Financing Plan. This needs to be prepared once MOH has decided what programs to emphasize and implement and having prepared and expenditure and implementation plan consistent with the overall resource envelop available to the MOH. This will require ensuring that the government and major development partners indicate what resources they can make available for the health sector over the medium term. 6. Develop Consistent Accountability Framework and Guarantee the Integrity of Government Processes. It is important that as the Government develops its combined resources budget that increasing effort is made to ensure accountabilities for development partner financed programs are as close as possible to those financed by Government. It is important that the MOH and the managers of its constituent parts (divisions, districts and facilities) know their total budget for the year and are responsible for accounting for the achievements and expenditures on the programs for which they are responsible. To - 19 - ensure development partners can work in this way with Government it is critical that every effort is made by the Government (and MOH) to guarantee the integrity of government processes. Knowledge Gaps 37. This report has brought a considerable amount of new information together and consolidated a range of pre-existing information on the health system. The DHS, in particular, has made a major contribution to the knowledge available for health planning. The report has also been written prior to the availability of the full results of the new census which will provide important additional information for health planning and management. The MOH is also working to significantly improve its health information and management system. Nevertheless there remain a number of key knowledge gaps a number of which have been identified in this report. These include:.  Limited knowledge of both the communicable and non-communicable disease burden in Timor Leste arising from diseases not covered by the DHS;  Need for improved knowledge of health seeking behaviors -- both qualitative and quantative – to address the need to improve the demand for health services; including improved understanding of possible barriers to access including costs of access.  Limited systematic knowledge of provider behavior and concerns including: (i) attitudes of health professionals to clients; (ii) commitment to national policy objectives; (iii) technical quality of advice and services provided; and (iv) concerns about career paths, material and non-material incentives, postings and workloads.  Limited details of development partner expenditures, their program focus, input categories of expenditure, and whether recurrent or development expenditures; and the scale of incremental recurrent costs arising to the Government’s recurrent budget from development partner expenditures. Next Steps 38. One feature that has been prominent in the development of the country’s health sector since 1999 has been the excellent working relationship between the MOH and its Development Partners, including the World Bank. In this spirit, it is suggested here that if MOH management were to find sufficient merit in the report’s analysis and suggestions, it could convene a workshop to discuss the report with Development Partners and other key stakeholders. The workshop could be a useful input in the preparation of the medium-term Strategic Health Plan envisaged by MOH. It could also further develop the six point program outlined above. - 20 - CHAPTER I: HEALTH SITUATION ASSESSMENT 1.1. INFANT AND CHILD MORTALITY 1. After the Referendum in 1999, two comprehensive assessments of people’s health, nutrition, and demographic status have been conducted in Timor-Leste: the Demographic and Health Survey (TL 2003 DHS) and the Multiple Indicator Cluster Survey (MICS 2002). These two surveys are still the best sources to get most reliable information on infant and child mortality in Timor-Leste. The results from MICS 2002, however, need to be treated with caution, especially for the estimates of infant and child mortality rates. Given the methodological limitations of indirect estimation methods, 2 it may be concluded that DHS 2003 is the best available source for estimating IMR and U5MR at this point. The 2004 Census will make it possible to re-calculate IMR and U5MR using indirect methods. Trends 2. The overall IMR and U5MR from MICS 2002 were 88 and 125 per 1,000 live births respectively, whereas for DHS 2003 they were 60 and 83 per 1,000 live births. 3 DHS data showed a significant decline of both IMR and U5MR by 50 percent in the last 15 years preceding the survey. Timor-Leste performs relatively well in terms of child mortality when compared to other developing countries of similar per capita income (Figure 1-1). Some countries such as Indonesia and Vietnam showed a superior outcome for child mortality for their income level, but the achievement of Timor-Leste is noteworthy because the country regained its political stability and independence just a few years ago. 2 MICS 2002 employed “indirect estimation methods� to calculate infant mortality rate (IMR) and under-five mortality rate (U5MR). Indirect methods use information on survival status of children to specific age cohorts of mothers. Meanwhile, DHS 2003 estimated the mortality rates using ‘direct reports of survivorship of children by their mothers (“direct estimation methods�). The indirect methods can utilize data that are commonly collected in census and many general surveys such as the number of children ever born and the number living to women and age. Biases and other potential sources of error exist in both methods. 3 The results from MICS 2002 presented here referred to four-year preceding the survey, i.e., 1999-2002, while mortality rates in DHS 2003 were calculated for five-year period preceding the survey, i.e., 1999-2003. - 21 - 140 250 120 200 100 MZTZ CD MZ UG U5MR (DHS) UG 150 IMR (DHS) 80 TZ BD CD 60 TL 100 BD 40 TL ID ID 50 20 VN VN 0 0 0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500 GDP per capita (current USD) GDP per capita (current USD) Figure 1-1: IMR and U5MR in relation to GDP per capita, selected countries IMR U5MR Legend: BD – Bangladesh, CD – Cambodia, ID – Indonesia, MZ – Mozambique, TZ – Tanzania, VN – Vietnam, UG – Uganda Source: DHS 1999-2003 and World Development Indicators (WDI) Economic differentials 3. It is clearly seen from the results from both MICS 2002 and DHS 2003 show that child mortality rates decline as the households’ wealth level increases (Table 1-1). 4 MICS data indicated a much higher prevalence of child mortality for the poorest households. The difference in the level of mortality between the poorest and richest households is much greater in U5MR than IMR. Table 1-1: MICS and DHS data on IMR and U5MR by wealth quintile IMR U5MR Wealth quintile MICS DHS MICS DHS Lowest 108 158 89 121 Second 103 148 Middle 88 125 78 101 Fourth 73 102 Highest 65 73 89 86 Pop. Average 88 82 125 107 Note: Data period for MICS 1999-2002, DHS 1994-2003. DHS assigned bottom 40 percent households as "poorest", next 40 percent as "middle", and the last top 20 percent as "richest". Source: TL DHS 2003 and MICS 2002 Geographical differentials 4. Living in “lowlands� appears to be a factor to lower IMR and U5MR, according to both DHS and MICS data. The “lowlands�, in fact, includes the majority of urban areas in Timor- Leste, therefore, it could be safely assumed that children living in urban areas are less likely to die in their childhood as compared to children in rural areas. Better access to basic needs in lowlands and urban areas such as safe water and health services would be the main factor to explain the differences in childhood mortality. 4 The data period of the data from DHS 2003 used in Table 1-1 was calculated for ten-year period between 1994 and 2003. Meanwhile, as Note 2 suggested, the data on Figure 1-1 referred to five-year period preceding the survey, i.e., 1999-2003. - 22 - 5. DHS and MICS data showed conflicting patterns of IMR and U5MR when stratified by regions. In DHS 2003, the lowest IMR and U5MR were observed in the rural East area, whereas MICS 2002 found that the rural West region had the lowest rates (Table 1-2). Both DHS and MICS adapted the same broad classification of geographical regions (East, Central, and West). Table 1-2: MICS and DHS data on IMR and U5MR by ecological zone and region IMR U5MR MICS DHS MICS DHS Ecological zones Highlands 106 106 153 130 Lowlands 80 77 113 102 Region* Urban 65 72 89 86 Rural 95 136 East 93 77 133 103 Central 94 90 134 117 West 75 88 106 122 Note: Data period for MICS (1999-2002); DHS (1999-2003) Geographical classification (East/Central/West): Both MICS and DHS used the same regional classification, but DHS excluded urban areas of the East/Center/West regions and included them into the “urban� region. Meanwhile, the East/Center/West classification which MICS used did not subtract urban areas. Source: MICS 2002 and TL DHS 2003 Demographic differentials 6. The results from both MICS and DHS indicate that under-five mortality rates for males are higher than those for females (Table 1-3). This difference in U5MR by gender is mainly attributed to the sharp disparity in IMR between males and females. Table 1-3: DHS data on IMR, CMR, and UMR by demographic classification IMR CMR U5MR IMR CMR U5MR Child's sex Previous birth intervals Male 90 23 113 <2 years 112 30 141 Female 75 27 102 2 years 71 26 97 3 years 54 11 65 Mother's age at birth 4+ years 29 18 46 <20 49 19 69 Birth size Small or 20-29 84 21 105 58 n/a n/a very small Average or 30-39 85 24 109 47 n/a n/a larger 40-49 75 30 105 Don't know 86 n/a n/a Birth order 1 97 21 118 2-3 79 21 100 4-6 77 34 111 7+ 87 17 104 Source: TL DHS 2003 - 23 - 7. The data confirmed that first and high-order births are associated with higher IMR and U5MR (Table 1-3). The DHS results could not demonstrate a significant association between the mortality rate of the first born and the mothers’ age at child birth. Teenage pregnancies are usually associated with premature and low-birth-weight newborns, however, in the case of Timor-Leste, newborns from very young mothers had the lowest probability for dying before their first and fifth birthdays. Too small sample size (0.9 percent of the total births) for mothers under 20 years may account for this result. 8. The birth interval, instead, showed a clear association with mortality rates – the shorter the birth interval, the higher the child’s probability to die. Children born with less than two-year intervals had four and three times higher IMR and U5MR, compared to children born with more than four-year intervals. 9. Birth weight is known to be one of the factors affecting child survival, but measuring baby’s weight at delivery is not practiced widely in Timor-Leste. Thus, DHS 2003 attempted to make mothers describe the size of their babies (and hence implied weight) using their own judgment, as “very small�, “small�, “average� or “large�. Using this method, DHS found that small babies had a 44 percent higher IMR, especially neonatal mortality, compared to average- sized babies (Table 1-3). Other factors influencing child mortality 10. Mother’s education level was strongly and inversely correlated with IMR and U5MR (Table 1-4). Children from mothers who received both antenatal care (ANC) and delivery assistance had the lowest mortality rates, while children of mothers with neither ANC nor delivery assistance had the highest mortality rates (Table 1-4). But, it is noteworthy that the children from mothers who had delivery assistance but no ANC had similar probability of dying before their first birthday than those from mothers with neither delivery assistance nor ANC. This result is probably due to the fact that most delivery assistance in Timor-Leste is provided by non-trained persons such as family members. ANC, on the other hand, seems to be effective in reducing the risk for child mortality. Table 1-4: MICS and DHS data on IMR and U5MR by child gender, mother’s education, and ANC/delivery assistance IMR U5MR MICS DHS MICS DHS Child's sex Male 99 90 142 113 Female 77 75 108 102 Mother's education None 100 89 145 119 Some primary 79 106 89 126 Completed primary 77 93 Secondary+ 62 69 85 81 Antenatal care (ANC) /delivery assistance Both ANC and assistance n/a 20 n/a n/a ANC only n/a 27 n/a n/a Assistance only n/a 63 n/a n/a Neither ANC nor Assistance n/a 65 n/a n/a Source: MICS 2002, TL DHS 2003 - 24 - 1.2. CHILD AND ADULT MORBIDITY 11. Reliable information sources on adult morbidity currently lacks in Timor-Leste, as a nationwide epidemiologic surveillance system has not been yet established. Thus, DHS 2003 and MICS 2002 are also the main sources of epidemiological information at present. DHS conducted in other countries does not contain the information on morbidity, however, the DHS 2003 steering committee made a special request to include some questions on adult morbidity, given the lack of morbidity data at national level. Child morbidity 12. MICS 2002 indicated that about 56 percent of children under age five experienced some form of illness in the two weeks preceding the survey. Malaria seems to be the most common disease among these children aged less than five years. MICS indicated that 26.7 percent of children under five had a high fever, which is a typical symptom of Malaria, in the last two weeks prior to the survey (Table 1-5). The corresponding figure from the DHS was 29.8 percent (see Table 1-5). ARI is the second leading cause of child morbidity – according to both MICS and DHS about 14 percent of children under five had experienced ARI symptoms in the two weeks preceding the survey. The prevalence of diarrhea among children is also high, though in this case there was a large difference between the two surveys; MICS reported 25.4 percent (2002), whereas DHS reported only 10.1 percent (2003) (Table 1-6). 13. Economic differentials. Surprisingly, the result from DHS showed almost no significant difference in child morbidity among households with different levels of wealth (Table 1-5). MICS, on the other hand, showed a clear decline in morbidity as households got richer. 14. Geographical differentials. The DHS data indicated that children from the rural West region had the highest risk of getting ARI, malaria, and diarrhea, while children in the rural East region enjoyed the lowest probability to get these infectious diseases, which is even lower than those of the urban area (Table 1-5). Meanwhile, the MICS data presented the highest risk of contracting ARI and diarrhea in the West Region and the lowest risk in the urban area. 15. Demographic differentials. The data from DHS 2003 and MICS 2002 clearly demonstrate that children less than one year old, especially those aged 0-6 months, had a greater resistance to ARI, Malaria, and diarrhea than older children (Table 1-5). This is probably due to the fact that breastfeeding contributes to enhancing immunity. Children’s gender does not seem to influence morbidity although female children had a slightly lower risk of getting sick. 16. Other factors. The results from MICS showed a clear decline of the probability of ARI, malaria and diarrhea as the mother’s education level increases, but this was not the case with the DHS (Table 1-5). - 25 - Table 1-5: MICS and DHS data on U5 children morbidity by geographic and socio-economic status All Figures in Percentages: Illness ARI Malaria Diarrhea MICS MICS DHS MICS DHS* MICS DHS Total 56.3 14.1 14.0 26.7 29.8 25.4 10.1 Gender Male 57.2 15.0 14.5 29.6 29.9 26.5 11.1 Female 55.4 13.2 13.6 25.8 29.7 24.3 9.2 Child's age (month) 0-6 49.0 13.0 12.8 18.9 19.8 19.5 7.4 7-11 18.0 37.4 15.1 12 - 23 64.2 15.5 16.1 34.4 38.1 33.0 13.4 24 - 35 57.3 16.9 13.8 31.1 30.9 26.8 10.7 36 - 47 58.2 14.4 12.6 31.4 26.7 25.8 7.4 48 - 59 53.3 10.1 11.1 25.4 22.4 21.7 6.2 Ecological zone Highland 56.6 15.3 12.2 27.9 26.9 26.3 9.9 Lowland 56.1 13.5 14.4 28.0 30.4 24.9 10.2 Region** Urban 50.7 7.4 12.4 23.4 29.5 22.7 10.5 Rural 58.0 16.1 29.3 26.2 East 55.8 15.8 22.6 26.7 32.4 18.8 11.8 Central 58.1 10.9 12.5 29.6 30.2 27.9 9.5 West 52.6 19.5 6.4 25.8 25.3 28.9 8.5 Mother's education None 57.2 15.9 13.7 30.1 27.5 26.6 9.4 Some primary 14.2 31.8 10.7 57.4 15.0 29.1 25.5 Completed primary 15.8 33.5 12.4 Secondary 53.5 9.8 14.0 22.7 31.1 22.9 10.3 Household Wealth index Lowest 61.8 16.9 33.4 29.1 15.3 30.3 9.3 Second 55.5 14.6 31.1 24.1 Middle 58.1 18.0 26.8 26.0 13.1 29.5 10.8 Fourth 54.3 14.7 28.5 27.1 Highest 50.6 5.8 13.0 19.3 29.2 20.3 10.7 Note: *The figures here are described as “children under five with fever� because DHS did not separate “malaria� and just a fever. ** Geographical classification (East/Central/West): Both MICS and DHS used the same regional classification, but DHS subtracted urban areas of the East/Center/West regions and included them into the “urban� region. Meanwhile, the East/Center/West classification Source: MICS 2002, TL DHS 2003 Table 1-6: Prevalence trends of selected childhood infectious diseases among children under-5 years** Year ARI Fever Diarrhea N* Source 1991 5.4 104 Indonesian DHS 1991 Final Report 1994 5.7 15.2 6.9 115 Indonesian DHS 1994 Final Report 1997 5.7 12.1 6.0 113 Indonesian DHS 1997 Final Report 2002 14.1 27.0 25.4 4452 TL MICS 2002 Final Report*** 2003 14.0 29.8 10.1 5225 TL DHS 2003 Final Report*** Note: * Un-weighted number of children ** Prevalence during the 2-week period before the survey ***Timing of data collection is August-September for MICS 2002 and April-August for DHS 2003. 17. As Figure 1-2 shows, the prevalence of ARI and diarrhea in Timor-Leste is relatively low when compared to other developing countries of similar per capita income. - 26 - Figure 1-2: ARI and diarrhea prevalence rate among children under five, in relation to GDP per capita, selected countries ARI Diarrhea 30 45 Diarrhea prevalence (% of children under ARI prevalence (% of children under five) 40 25 35 30 20 UG five) (DHS) CD 25 (DHS) 15 20 UG MZ CD BD TZ 15 10 ID TL TL TZ 10 MZ 5 BD 5 ID 0 0 0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500 GDP per capita (current USD) GDP per capita (current USD) Legend: BD – Bangladesh, CD – Cambodia, ID – Indonesia, MZ – Mozambique, TZ – Tanzania, VN – Vietnam, UG – Uganda Source: DHS 1997-2003 and WDI Prevention and treatment of child illness 18. Using a pesticide-impregnated bed net is a main strategy to prevent malaria infection in Timor-Leste. About 50 percent of children under age five slept under a bed net, according to the MICS 2002. The frequency of use of bed nets is strongly correlated with mothers’ education level and ecological zones (low- and highland). However, only 8 percent of children of those using bed nets slept under an insecticide-impregnated bed net. 19. Overall, 24 percent of children with symptoms of ARI and/or fever were taken to a health facility. A child from lowlands, urban areas, a wealthier family, and/or an educated mother is more likely to be taken to a health facility when he/she gets an ARI or fever. The level of service utilization by children with ARI or fever is very low when compared to other countries with similar income level, as shown in the following graph (Figure 1-3). Figure 1-3: Percent of children under-five with ARI or fever taken to health facility, in relation to GDP per capita, selected countries 90 % of children with ARI or fever taken to a 80 70 health facility (DHS) UG 60 ID 50 MZ 40 CD 30 BD 20 TL 10 0 0 500 1000 1500 2000 2500 GDP per capita (current USD) Legend: BD – Bangladesh, CD – Cambodia, ID – Indonesia, MZ – Mozambique, TZ – Tanzania, VN – Vietnam, UG – Uganda Source: DHS 1997-2003 and WDI 20. Concerning the treatment of diarrhea, there is a very high level of knowledge and practice of oral re-hydration therapy (ORT) among mothers (73 percent) (Figure 1-4). The data showed - 27 - that a younger, educated, and/or wealthier mother is more likely to apply her knowledge of ORT to children with diarrhea. However, about 25 percent of children with diarrhea were given no treatment. Figure 1-4: Percent of children under-five with diarrhea who received either ORS or RHF, in relation to GDP per capita, selected countries 80 % of children with diarrhea who received BD TL 70 60 either ORS or RHF (DHS) 50 TZ MZ ID 40 UG 30 20 CD 10 0 0 500 1000 1500 2000 2500 GDP per capita (current USD) Legend: BD – Bangladesh, CD – Cambodia, ID – Indonesia, MZ – Mozambique, TZ – Tanzania, VN – Vietnam, UG – Uganda Source: DHS 1997-2003 and WDI Adult morbidity 21. Tuberculosis. WHO has reported an incidence of tuberculosis of 556 cases per 100,000 people, and a prevalence rate of 754 cases per 100,000 people, or about 0.75 percent in 2003.5 the DHS, however, presented a very different picture; with 5.6 percent of males and 7.1 percent of females reporting to have at least two of three typical symptoms of TB (persistent cough, cough blood, weight loss) during the month prior to the survey. According to the DHS data, males and females with the lowest body mass index (BMI) status, from the poorest households, and/or from the rural East region had a high risk of being transmitted to TB. Figure 1-5: Incidence of TB, in relation to GDP per capita, selected countries 800 Incidence of tuberculosis (per 100,000 700 600 TL people) (WDI) 500 CD MZ 400 UG TZ 300 ID 200 BD VN 100 0 0 500 1000 1500 2000 2500 GDP per capita (current USD) Legend: BD – Bangladesh, CD – Cambodia, ID – Indonesia, MZ – Mozambique, TZ – Tanzania, VN – Vietnam, UG – Uganda Source: WDI 5 WHO, 2005, WHO Report on Global Tuberculosis Control: Surveillance, Planning, Financing: ANNEX 2. The figures here take account of all cases including HIV+ cases. - 28 - Table 1-7: TB prevalence by biological, regional, and socio-economic status Female Male BMI status <16.0 11.9 11.2 16.0-16.99 10.0 8.7 17.0-18.49 7.3 7.7 18.5-24.99 6.5 4.7 >25.0 0.6 2.4 Region Urban 4.2 5.0 Rural east 9.9 10.2 Rural central 7.4 4.2 Rural west 6.2 3.7 Ecological zones Highland 9.5 3.6 Lowland 6.6 6.1 Household wealth index Poorest 40 percent 9.1 6.3 Middle 20 percent 6.7 5.2 Richest 40 percent 3.3 5.0 Total (TB prevalence) 7.1 5.6 Note: DHS 2003 estimated TB prevalence from the survey respondents who reported to have at least two of three typical symptoms of TB (persistent cough, cough blood, weight loss) during the month prior to the survey. Source: TL DHS 2003 22. Malaria. WHO reported that in the year of 2003 there were 31,819 malaria cases confirmed by laboratory, of those 409 cases were severe (impatient or hospitalized), and eight malaria reported deaths in 2003, however, probable or clinically diagnosed malaria cases were estimated to be 100,000 and 100 deaths.6 Two types of malaria mainly exist in Timor-Leste: P. Vivax and P. falciparum. The reported number of cases has fluctuated in recent years – the peak was 2001 with 63,440 cases. 23. HIV/AIDS. The first AIDS case was confirmed at Dili Hospital in December 2001. The extent of the impact of HIV/AIDS in Timor-Leste is not known yet due to lack of national-level HIV sentinel surveillance.7 But, according to the research by Family Health International (FHI)8 in 2003, the overall HIV prevalence rate appeared to be very low (0.6 percent).9 3 percent of 6 WHO, 2005, Roll Back Malaria Monitoring and Evaluation Report. The Report pointed out limited capacity in diagnosing malaria cases and data management, and underreporting from districts in Timor-Leste. 7 There are currently 10 diagnosed HIV+ cases in the entire country (Background document, 2005). 8 The survey by FHI is the first quantitative survey of HIV, other STIs and risk behavior in Timor-Leste, however, the potential selection bias should be considered in this survey. Participation from high-risk groups to this survey was voluntary. Voluntarism in epidemiological research may lead to selection bias either by attracting the healthiest individuals (those more conscious of the potential risks) or by attracting those who are already complaining symptoms supposed to be related with the study area of interest. The direction of this bias over the study results is unpredictable and may jeopardize internal validity of the observed associations (Background document, 2005). 9 It is important to consider that although this survey had enrolled a at risk population, the observed prevalence rate was below the observed prevalence in low risk population in other countries. In Brazil, for example, the sero-prevalence rate among pregnant women, blood donors, or others with no known risk factors is 2.7 times higher (1.6 percent). Among at risk population (sex workers and clients, STI patients, or others with known risk factors) the prevalence rate is of 42.0 percent (Background document, 2005). - 29 - female commercial sex workers (CSWs) in Dili and 1 percent of men who have sex with other men (MSMs) were HIV positive. One quarter of CSWs in Dili had gonorrhea and/or Chlamydia, and 60 percent were infected with HSV2 (Herpes Simplex Virus; congenital herpes). Among taxi drivers and men who have sex with men, 29 percent had HSV2. 24. DHS 2003 showed that the majority of population in Timor-Leste had very limited knowledge of HIV/AIDS – 70 percent of men and 80 percent of women had never heard of HIV/AIDS. MICS 2002 also indicated that about 85 percent of women between 15-49 years of age had never heard of HIV/AIDS. Only 15 percent of men knew that there was a way to avoid HIV infection. The level of knowledge of HIV/AIDS increase as the level of education and wealth level increase. About one-half of all people in the urban areas had heard of HIV/AIDS, while only 6 percent of women and 18 percent of men in the Rural West region knew about HIV/AIDS. 1.3. FERTILITY 25. MICS 2002 counted births occurring in the last 12 months (2001-2002) and from these data a total fertility rate (TFR) of 7.6 children per women was estimated. DHS 2003 estimated TFR by asking about births, abortions, stillbirths, and miscarriages occurred in the 36 months prior to the survey for women aged between 15 and 49 years. DHS found that Timorese women in the period of 2000 to 2003 were bearing 7.7 children over their reproductive lives, which would be one of the highest levels of fertility in the world. 26. Figure 1-6Surprisingly, the TFR seems to be increasing; the most recent TFR estimate, for the period April 2002 to March 2003, was 8.3 births per women. Figure 1-6: Total fertility rate in relation to GDP per capita, selected countries 9 8 TL 7 UG 6 MZ TZ TFR (DHS) 5 4 CD BD 3 ID 2 VN 1 0 0 500 1000 1500 2000 2500 GDP per capita (current USD) Legend: BD – Bangladesh, CD – Cambodia, ID – Indonesia, MZ – Mozambique, TZ – Tanzania, VN – Vietnam, UG – Uganda Source: DHS 1997-2003 and WDI 27. It is noteworthy that there was virtually no difference in TFR across women’s education levels. Women with primary education had 8.6 children while women with no education had less than 8.0. Usually, a higher education level tends to lower the number of children that women bear, according to international experience (Figure 1-7). Likewise, the difference in TFR by the level of wealth was small. TFR in urban areas was slightly lower than in rural areas (7.4 and 7.7-8.0, respectively). - 30 - Figure 1-7: Total fertility rate by mothers’ education level, selected countries 9 8.5 7.9 8 No education Primary Secondary or higher 7.5 7 6 5.3 4.8 5.0 5 TFR 4.3 4.1 3.8 4 3.2 3.3 3.1 2.7 3 2.7 2.6 2.4 2.5 2.5 2.2 2.0 2 1.7 1 0 Nepal Cambodia Philippines Bangladesh Indonesia Vietnam Timor-Leste Source: DHS 1999-2003 28. The median age of women at first birth appears to be declining, unlike the case in other developing nations. About half of the women aged 45-49 years at the time of the survey had their first baby after 25 years of age. Meanwhile, only 25 percent women who were 35-39 years of age at the time of the survey had their first baby after their 25th birthday. 29. Contraceptive coverage. The DHS 2003 data indicated that about 10 percent of all ever- married women aged 15-49 years were current users of contraception, and an additional 10 percent were no longer using contraception but had done so in the past. Thus, about 80 percent of ever-married women aged 15-49 years had never used contraception. MICS 2002 also confirmed the low contraceptive coverage – only eight percent of all women who were currently married or living with a partner used contraceptive methods. The socioeconomic status of users correlates positively with contraceptive use: a higher contraceptive use rate prevails among educated and wealthier women and those from urban areas. Injectable contraceptives were the most popular choice of method, although it requires medical assistance every three months and can cause serious side effects. The level of contraceptive coverage in Timor-Leste is extremely low in comparison with other countries at a similar income level, such as Vietnam and Bangladesh (Figure 1-8). Figure 1-8: Contraception coverage in relation to GDP per capita, selected countries 90 % of married women currently using of any 80 VN 70 contraceptive method (DHS) 60 ID BD 50 40 30 MZ 20 TZ CD UG 10 TL 0 0 500 1000 1500 2000 2500 GDP per capita (current USD) Legend: BD – Bangladesh, CD – Cambodia, ID – Indonesia, MZ – Mozambique, TZ – Tanzania, VN – Vietnam, UG – Uganda Source: DHS 1997-2003 and WDI - 31 - 30. Knowledge of contraceptive methods among Timorese was remarkably low – 60 percent of ever-married women and 70 percent of ever-married men could not identify any contraceptive methods. The knowledge of contraceptive methods increased with household wealth and educational levels. Respondents from urban areas and/or aged between 20 and 44 years also have a higher knowledge level of contraceptive methods. Women who could identify at least one contraceptive method were likely to identify public health facilities as the source of supply. Women from wealthier and more educated households in urban areas tended to procure their contraceptives from private sources as well as from public health facilities. 31. The reasons not to use contraceptives among currently married people who were not using a contraceptive method were slightly different between women and men. A fear of side effects was one of the biggest motives not to take contraceptives for women, while one-quarter of men simply did not know any method of contraceptives. Similar percentages of women and men selected “fertility-related reason (desire to have as many children as possible)� and “opposition to use� by spouse as the reasons not to use contraceptives. Figure 1-9: Percent of currently married women and men who are not using a contraceptive method and do not intend to use in the future, by main reason 100% 4.5 4.5 90% 10.4 25 80% 23.2 70% 60% 19.9 50% 29 40% 29.6 30% Not specified Lack of knowledge 20% Fertility-related reason 33 Opposition to use 10% 21 A fear of side effects 0% Women Men Source: TL DHS 2003 32. Figure 1-6 Direct estimates of maternal mortality are not available for Timor-Leste. However, maternal mortality is likely to be high, because: (i) the total fertility rate --the average number of children born to a woman over her reproductive years-- is one of the highest in the world (7.7 children according to the 2003 DHS); (ii) short birth intervals between births are highly prevalent; and (iii) relatively few births are assisted by attendants with the necessary training. Using such indirect evidence and regression analysis, WHO has estimated the maternal mortality ratio (MMR) in Timor-Leste at 660 per 100,000 live births, which is similar to MMR levels in Laos, Haiti, and a number of countries in Africa. However, regression-based estimates of the MMR are subject to a large degree of uncertainty.10 10 MMR estimates in 2000 were developed by WHO, UNICEF and UNFPA (http://www.reliefweb.int/library/documents/2003/who-saf-22oct.pdf). The data are estimated with a regression model using information on fertility, birth attendants, and HIV prevalence. The data shown here have been collected in various years and adjusted to a common 1995 base year. - 32 - 33. 1.4. Nutrition 34. While there was a considerable improvement of the nutritional situation during the 1990s,11 two national surveys in 2002 and 2003 confirmed that prevalence of under-nutrition among children under five years is one of the highest in the world. Between 1989 and 1999, the prevalence of underweight among children under five was almost halved from 63 percent to 32 percent. However, in the 2003 DHS, almost two-thirds of children in this age bracket were under-weight (low weight for age) (Figure 1-10), with 14 percent being severely under-weight (Figure 1-11). Moreover, 53 percent were moderately stunted (low height for age, an indicator of chronic under-nutrition) and 31 percent severely stunted. And one in ten children under the age of five was wasted (low weight for height, an indicator of acute under-nutrition). Most children became under-nourished by the first half of their second year. Malnourished children were not confined to those born of under-nourished mothers (i.e., height below 145 cm and BMI below 18.5), indicating some additional socio-economic and behavioral factors at work. Figure 1-10: Trend of underweight prevalence among children under five 70 Underweight prevalence (%) Moderate underw eight 60 Severe underw eight 50 40 30 20 10 0 1985 1990 1995 2000 2005 Source: SUSENAS (1989, 1992, 1999); MICS (2002); DHS (2003) Figure 1-11: Prevalence of stunting, underweight and wasting among children under five 60 50 Moderate Severe Prevalence (%) 40 52.7 30 45.7 20 30.5 10 13.1 13.9 0 3.2 Stunting Underw eight Wasting Source: Authors calculation: data from TL DHS 2003 35. International comparison showed that the level of malnutrition among children under five in terms of stunting and underweight was high when compared to other countries with similar GDP per capita level (Figure 1-12). 11 SUSENAS (1989, 1992, 1999) - 33 - Figure 1-12: Prevalence of stunting and underweight in relation to GDP per capita, selected countries Stunting Underweight 60 60 % of children stunted (-2 SD) (DHS) 50 % of children underweight (-2 SD) 50 TL BD TL BD 40 MZ CD TZ CD 40 UG (DHS) 30 30 TZ 20 MZ 20 UG 10 10 0 0 500 1000 1500 2000 2500 0 GDP per capita (current USD) 0 500 1000 1500 2000 2500 GDP per capita (current USD) Legend: BD – Bangladesh, CD – Cambodia, ID – Indonesia, MZ – Mozambique, TZ – Tanzania, VN – Vietnam, UG – Uganda Source: DHS 1997-2003 and WDI 36. Figure 1-13 below shows that most children become undernourished by the first half of the second year of life: mean Z scores of height and weight among Timorese children were above zero up to two months, but gradually decreased to less than -2 between 12-17 months and stayed relatively constant thereafter. Figure 1-13: Timing of under-nutrition 2 Height for age 1 Weight for age Weight for height Mean Z scores 0 -1 -2 -3 -4 0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 Age (mo) Source: Authors calculation: data from TL DHS 2003 37. Figure 1-14 below shows the timing of stunting and underweight by maternal nutritional status. Even though nutritional status of children born to undernourished mothers (i.e., those with height under 145 cm. and BMI under 18.5) was likely to deteriorate further than for other children, the patterns observed were similar. - 34 - Figure 1-14: Timing of stunting and underweight by maternal nutritional status 4 6 Mean height for age z score Mean weight for age z score Maternal height>=145 cm Maternal BMI>=18.5 2 Maternal height<145 cm 4 Maternal BMI<18.5 0 2 -2 0 -4 -2 -6 -4 0 6 12 18 24 30 36 42 48 54 60 0 6 12 18 24 30 36 42 48 54 60 Age (m o) Age (m o) Source: Authors calculation: data from TL DHS 2003 Note: SD varies significantly due to small sample sizes in each age group, especially for mean height for age z score of the children born to stunted mother. 38. The prevalence of both stunting and under-weight were significantly higher among children from rural areas, poorer households, and those whose mothers were less educated. For example, prevalence of stunting was above 70 percent among children born in the highlands of the rural West, as compared to 44 percent among urban children (Table 1-8). More than 60 percent of children born into the poorest 20 percent of households in the rural West were stunted, while less than 40 percent of children born in the richest 20 percent of households in urban areas were stunted. Table 1-8: Prevalence of under-nutrition in children under five years Stunting Underweight Wasting Number 1 of Background characteristics Moderate Severe Moderate Severe Moderate Severe children2 Sex Boys 54.6 32.6 46.3 14.2 13.5 3.3 2355 Girls 50.8 28.4 45.0 13.5 12.8 3.1 2447 Age (months) <6 20.1 8.2 2.7 0.5 8.8 3.6 541 6-11 31.8 17.8 27.1 7.3 11.5 2.2 578 12-17 53.7 32.2 53.5 17.4 18.9 5.0 589 18-23 64.0 39.8 54.5 17.7 17.9 5.7 457 24-29 57.0 33.3 61.0 20.8 15.0 3.1 611 30-35 65.5 37.1 55.2 16.5 10.2 2.3 421 36-41 67.3 41.0 55.7 16.9 11.9 1.8 514 42-47 62.0 35.7 50.7 14.5 10.2 2.5 381 48-53 57.6 32.5 53.9 13.5 12.8 3.0 444 54-59 62.4 35.6 50.3 14.9 11.7 1.7 267 Size at birth Larger than average 47.2 25.0 39.3 9.8 11.1 2.7 887 Average 54.4 31.9 47.2 13.7 13.0 3.2 2860 Smaller than average 55.4 34.4 48.8 19.3 16.2 3.8 744 Birth order 1 55.1 30.2 46.7 14.7 14.6 3.7 673 2-3 52.9 30.1 46.8 12.6 12.8 2.8 1581 4-5 51.0 30.7 44.6 14.2 13.9 3.2 1302 >=6 52.9 31.0 44.8 14.6 12.0 3.3 1245 Maternal height Ht>=145 cm 50.9 29.1 44.2 12.7 13.0 3.2 4145 Ht<145 cm 63.4 38.9 55.2 21.1 14.1 3.0 639 Maternal BMI BMI>=18.5 50.0 28.2 40.5 11.4 11.0 3.0 2741 BMI<18.5 55.3 32.8 51.7 17.7 17.9 3.8 1438 Maternal None 55.7 33.9 49.4 16.4 13.9 3.9 2178 education Primary or less 52.3 30.0 43.7 12.6 11.9 2.6 1331 - 35 - Secondary or more 47.8 25.2 41.4 11.0 13.2 2.6 1290 Region Urban 44.5 21.3 41.4 11.5 13.9 2.3 1156 Rural West 59.8 37.8 56.3 19.3 18.3 5.6 815 Rural Central 53.4 32.0 44.9 14.6 12.6 2.9 1601 Rural East 54.6 32.4 43.7 11.6 9.6 2.7 1230 Ecological zones Highlands 59.0 37.1 48.5 17.7 12.3 3.2 824 Lowlands 51.4 29.1 45.1 13.1 13.3 3.2 3978 Household food None in last week 51.8 30.5 46.3 14.0 14.7 3.6 2396 shortages Any in last week 53.5 30.5 45.0 13.7 11.6 2.8 2405 Wealth index Poorest 20% 58.8 37.1 49.0 17.4 13.1 3.8 998 Poorer 20% 56.9 35.5 45.5 14.8 13.2 2.9 1027 Middle 20% 54.5 30.9 49.3 14.9 13.5 3.6 976 Richer 20% 50.0 27.9 44.3 12.4 12.2 3.2 921 Richest 20% 41.4 19.4 39.6 9.1 13.7 2.3 879 Total 52.7 30.5 45.7 13.9 13.1 3.2 4802 Note: 1 Moderate under-nutrition includes severe under-nutrition: 2 Total population weighted * Bold: significant difference (p<0.05) in the prevalence of stunting, underweight, wasting by background characteristics Source: Authors calculation: data from DHS 2003 Prevalence of under-nutrition among adults 39. The nutrition situation among adults is a little better that that among preschool children, but still poses a serious problem. More than a third of non-pregnant women aged 15-49 years (35 percent) and a quarter of men aged 15-54 years were chronically undernourished (BMI<18.5). Fourteen percent of women were also shorter than 145 cm, at which level the risk of pregnancy and delivery complications increases. Under-nutrition among adults was more prevalent among the underprivileged – those from the poorest households and the rural West, the less educated, and the older. Interestingly, food insecurity, measured by “food shortage in the last week,� was positively correlated with only women’s nutritional status, possibly indicating inequality of intra-household food distribution. Micronutrient deficiency – Anemia 40. Timorese children are not only affected by protein-energy under-nutrition, but also suffer from micronutrient deficiency disorders, notably anemia, even though the prevalence is much lower than expected at the level of protein-energy under-nutrition. According to DHS 2003, 29 percent of children under the age of five were anemic. Prevalence of anemia among children was significantly correlated with that among mothers: while only a quarter of children with non- anemic mothers were anemic, almost 60 percent of children, whose mother were severely anemic, were also anemic. Boys, children under the age of two, and those living in rural West and rural East also had significantly higher anemia rates. 41. Unlike the prevalence of protein-energy under-nutrition, that of anemia was not lower among children from more educated mothers or from better-off households. Also, there was little variation in the prevalence of anemia by nutritional status of children: whether a child had any protein-energy under-nutrition or not, the prevalence of anemia was about the same at 30 percent (Figure 1-15). When other factors were controlled for, however, the likelihood of being anemic among undernourished children increased significantly: moderately underweight children were 29 percent more likely to be anemic and severely underweight children were 41 percent more likely to be anemic. Children who had anemic mother were almost twice more likely to be - 36 - anemic than children who had mother without anemia, when other background characteristics were controlled. Figure 1-15: Prevalence of anemia among children under-five by selective background characteristics 60 severe moderate mild 45 Prevalence (%) 30 15 0 Rural C. Rural E. Mild None None Urban Moderate Rural W. Girls <6 6-11 12-17 18-23 24-29 30-35 36-41 42-47 48-53 54-59 Severe Boys Any Maternal anemia Sex Age (mo) Region PEM Source: Authors calculation: data from TL DHS 2003 42. Anemia was also common among women aged Figure 1-16:Prevalence of anemia among 15-49 years, affecting nearly a third of them. The children under-five by selective prevalence of anemia did not vary substantially by the background characteristics status of pregnancy, but the pregnant women had much 100 Proportion of women (%) higher rates of moderate and severe anemia, while the majority of non-pregnant with anemia suffered from 80 mild form (Figure 1-16). Overall, 29 percent of 60 pregnant women and 31 percent of non-pregnant women had anemia 12 with about 15 percent and 2 40 percent of pregnant women having moderate and severe anemia, while 8 percent and less than one 20 percent of non-pregnant women having moderate and 0 severe anemia, respectively. The prevalence was Non-preg Pregnant significantly higher in the lowlands and since most severe moderate mild None rural East and Urban areas are lowlands, the prevalence of anemia was also high in those areas. In Source: Authors calculation: data from TL DHS 2003 addition, women from the households experiencing food shortage had higher prevalence of anemia as well. However, nutritional status measured by height and BMI was not significantly correlated with women’s anemia status. There was also little difference in the proportion of anemic women either by the level of household wealth or maternal education level. After controlling other factors, women in the lowlands were still 80 percent more likely to be anemic and those in the rural West were about a third less likely to be anemic. 43. Unfortunately, only about four percent of women who gave birth in the five years preceding the survey, according to DHS 2003, took the recommended doses of iron supplementation; a further 44 percent of them received some supplementation. Iron supplementation was particularly low among women with no education (37 percent), those from the poorest 20 percent of households (36 percent), and those living in highlands (32 percent), while more than 60 percent of women from the richest 20 percent of households and urban areas, as well as those with secondary education received iron during their pregnancy (Figure 1-17). 12 The cut off points for mild anemia were Hb <110 g/L for pregnant women and Hb <120 g/L for non-pregnant women. Hb <90g/L and <70g/L were used for moderate and severe anemia for both pregnant and non-pregnant women. - 37 - Figure 1-17: Prevalence of anemia and iron supplementation during pregnancy among women aged 15-49 years by selective background characteristics 70 Proportion of women (%) 60 Anemia Iron supplementation 50 40 30 20 10 0 Rural C. Rural E. Primary None None Middle Urban Rural W. Poorest Richest Poorer Richer Highlands 1 2-3 4-5 >=6 >average 30 percent have a mid upper arm circumference measurement of < 21 cm�, there is no recommendation in favor of supplementary feeding for children under 5 or school children. - 72 - administration assistants make up the remaining staff. Paramedical workers such as pharmacists and laboratory technicians are rarely found. Midwives are in particularly short supply, and in the districts visited, were limited to working in the CHCs. The midwifery diploma for nurses at the NIHS is particularly suitable to address this problem. 32. The absence of key staff resulted in provision of services that did not reach the standards set down in national policy. For example, the absence of midwives in the health posts limited the ability of the health posts to provide antenatal care and delivery care, as required. With only one staff being able to operate mobile health posts, there was insufficient time to provide the full range of services, particularly health education and immunization. 33. Training. Although district health managers are supposed to have an input in identifying training needs, in practice training is determined at central level, and district managers have only a role in the selection of participants. Selection generally follows seniority, once the location and category of workers to be trained have been determined, leaving little opportunity for using training as a reward to encourage good performers. 34. Field visits for this report indicated that post-training supervision tends to be neglected. The IMCI trainers made a post-training follow up visit, but there was no ongoing clinical supervision provided. The same was true of the clinical nurse trainees. The district “focal points� are only able to provide management supervision. The increased availability of doctors through the Cuban cooperation agreement provides an opportunity to address these clinical supervision needs. Human resource issues are discussed in detail in section 3.7 below. 35. Health facility / infrastructure issues. During the field visits for this report, district staff noted that there was a need for the provision of radio communication at the health posts, to enable nurses to consult with clinicians in the CHC or district hospitals. Most of the health posts also lacked motor- cycles, which limited the ability of the health post staff to communicate with CHCs, or to undertake mobile health posts. This constraint affects the ability of the health post to provide immunization, in accordance with the service standards, as vaccines need to be brought from the refrigerator located at the nearest CHC. Shortages of both staff and transportation also limit the frequency with which mobile posts can be visited (MOH policy standard is two visits per week). In CHCs where there are only two nurses, the CHC can only service two mobile health posts if they require two visits a week. However, some CHCs service three or more mobile posts, which they can only visit on a lesser frequency. 36. One issue raised in the districts visited was the lack of suitable space for the TB laboratory (which supports the TB control program) in the Level 2 CHCs, as this is an activity not previously included in the BPS standards for Level 2 CHC. Moreover, in two of the districts visited (Covalima and Bobanaro), all of the CHCs are at Level 2, and do not have facilities or staff for inpatient care. Thus all patients requiring longer observation or even simple inpatient care have to be referred to higher level facilities. Because patients are often unwilling to be referred, or there are difficulties obtaining ambulance transport, patients are sometimes kept overnight although facilities are not appropriate. 37. In all three districts visited, the demand for the use of the ambulance was reported to be high, and particularly in Covalima and Bobanaro, the ambulance role in referral to the national - 73 - hospital in Dili competed with its availability for transport of patients within the district. In both cases only one ambulance was available for the district. 38. As can be seen from these examples, the development of policies/strategies for each program, and the identification of coverage targets, may lead to a demand for increased facilities (e.g., construction of additional health posts) or staff (midwife coordinator, specialized mental health workers) beyond the original BPS planned levels. It is important that further development of staffing or facilities, if required, be managed in a planned way, with decisions made against appropriate criteria and data on utilization. 3.6. ISSUES IN THE HOSPITAL SECTOR 39. The MOH has drafted a Hospital Services Package (HSP), which would complement the Basic Services Package. The HSP aims to set out in detail what services are offered by each of the six hospitals and how they relate to the overall health system. More specifically, the HSP will define: (i) the key service streams that will form the hospital services package; (ii) the levels of hospital services and approach to the referral system; and (iii) the institutional and management structure and mechanisms through which the HSP is delivered. 40. Service Streams. Four main service streams are identified in the draft HSP: (a) obstetrics/gynecology; (b) neonatal/perinatal/pediatric care; (c) internal medicine; and (d) trauma/surgery. They are meant to provide higher-level support for complicated cases referred from the primary (district) health services. 41. Levels of Service. For each of the service streams, the HSP will set out what should be dealt with: (a) at the primary care level (aid posts and community health centers); (b) referral hospitals in the district; and (iii) second level of referral at the National Hospital Guido Valadares in Dili. The National Hospital has ceased to attend to outpatients with the exception of emergencies and referral cases. 42. Management of Hospitals. Recent legislation provides for the establishment of hospital management boards (Concejo de Administracion). The details of how the management boards are to work and how they will be phased in is yet to be formulated. A beginning has been made at the National Hospital (see Box 3.3). 43. The MOH has also drafted instruments for annual activity plans which will form an important basis for monitoring by MOH of hospital activities. Box 3-3:Changes in the Administration of the National Hospital Subsequent to the Estatuto Hospitalar decree of May 31, 2005, the day-to-day management of the National Hospital is now the responsibility of a Management Board (Concejo de Administracion) composed of a General Director, an Administrator, the Clinical Director and the Director of Nursing. A Controller appointed by the Ministry of Planning and Finance oversees the hospital’s financial operations and accounts. The National Hospital is now authorized to prepare its annual budget, which has to be approved by the Minister of Health and the Minister of Planning and Finance. Once the budget is approved, the Management Board has full authority to execute it. Moreover, the Management Board is authorized to draw the hospital’s internal regulations without approval from the MOH. The hospital now also enjoys more freedom in terms of recruitment. Before the decree, when the hospital needed to recruit for a vacant position, they would relay the request to the MOH, which would advertise for the post, conduct interviews, select the person to occupy the post, and enter a contract with him/her. Now the hospital management can carry out these recruitment steps directly. The hospital prepared an Annual Action Plan - 74 - for 2005/06 for the first time, and will continue this exercise from now on. Every month the hospital prepares a report which is submitted to MOH. The hospital can now initiate purchase orders directly (previously they had to request the MOH to initiate the orders), but procurement process continues to be conducted by the Central Procurement Unit at the Ministry of Planning and Finance for all contracts in excess of US$1,000. The Management Board does not have authority to make changes to the mix of services offered by the hospital, but it can (and does) make suggestions to that effect to the Minister of Health. More generally, the hospital is now able to formulate its own strategic plan for the short, medium and long-run which is then subject to the Minister of Health’s approval. While the changes in the hospital management’s authority have been rather modest so far, it is likely that if everything goes well greater administrative and financial authority will be granted gradually. Source: Mission discussions. 44. Hospital Services Package and its Operationalization. The development of the HSP will be an important milestone. It will be important that all stakeholders including hospital managers fully discuss and understand the HSP and how it will be rolled out. The move to increase the autonomy of hospitals is an interesting development which if implemented in a gradual manner as capacities are developed could bring significant improvements. Important Constraints and Challenges for the Hospital Sector 45. There are serious gaps in the management capacity at the hospital level, and as yet few systems have been established to support management processes. Important management decisions are centralized and local managers have very limited authority. 46. The National Hospital faces high demand for its 244 beds and outpatient services, and this demand has been growing with the addition of new clinical skills. Baucau Hospital’s 114 beds have an approximately 70% occupancy rate, though the average length of hospital stay is considerably longer than in the National Hospital. The four 24 bed community hospitals in Oecussi, Maliana, Maubisse and Suai have lower levels of demand. 47. As yet no mechanisms have been established to assess the quality of care provided in the hospital system. There is anecdotal evidence of service quality being very variable and at times unacceptably poor. Systems of clinical accountability are yet to be developed and processes of clinical supervision and support are weak. 48. The development of the HSP and moves to autonomy provides an opportunity and a necessity to address these issues. There are several areas of priority for hospital management development in Timor-Leste. Most of these areas are receiving attention from hospital managers and the MOH in conjunction with Development Partners, though action to date has not been coordinated or given a particular priority. The priority areas relate to service delivery limits and demand management; ensuring quality of care; budgetary control; and efficient use of resources. 49. Service delivery limits and demand management. The basic package of services will determine level of services to be provided. The referral services to be provided at Bacau and Dili hospital need to be defined detailing the points at which referral will be undertaken. This will require sampling of hospital activities and disease coding of medical records, as well as collection of reliable data on individual clinic activities for outpatient clinics. From this process MOH and the hospitals will have a basis for determining annual activity plans that will relate to budgets and agreed activities. As noted above, the MOH has already drafted instruments for the preparation of annual activity plans. The plans now need to be prepared. - 75 - 50. Ensuring quality of care. Management plays a central role in ensuring the quality of care delivered in hospitals. The care itself is delivered by the professional clinical staff, but is strongly impacted by the management process. Some of the quality-influencing aspects of hospital management are ensuring the reliable availability of resources, bolstering the morale and pride of staff, ensuring the efficient transit of patients through the processes of the hospital, ensuring the maintenance of equipment, and keeping an open communication with the public. 51. One potential approach to improving quality of care is the development of standardized diagnostic and treatment guidelines for common conditions. The development of clinical pathways is one such approach, where agreement among clinical staff is negotiated, resulting in one or on occasions two or three “standard� approaches to the diagnostic workup and treatment of particular conditions. Standard medical records for the condition are then prepared, and these records require only a signature to verify the following of the standard treatment, or a more detailed report where clinical or other circumstances require that the standard approach not be followed. The standardized approach and an efficient medical record system allow for objective review of the processes of care and assist management in making continuing improvements. 52. Budgetary control. Delegation of authority to hospital managers has been very limited in Timor-Leste. Staffing controls are centralized, and expenditure on goods and services is tightly controlled at senior levels of the MOH and the Ministry of Planning and Finance. The recent creation of the Management Board for the National Hospital is an important step in the decentralization of authority within the hospital sector. 53. Now that institutional-level budget information is available, senior managers and finance officers in the hospitals –initially at Dili National Hospital—need to be trained on the new reporting system, its capacity and limitations and how reports should be interpreted. 54. The MOH must decide what level of accountability it will require of hospital managers concerning budgetary performance, including providing answers to the following:  Who will review the reports?  Who will follow up with hospital managers on deviations from expected spending patterns?  What remedial actions might be instituted where indicated?  What support will be given to hospitals to introduce agreed changes in spending patterns?  Will comparative data be provided to hospitals? What use should be made of the data?  Who will be accountable for budget performance in hospitals? How will this be achieved? 55. With routine reports on spending by individual hospitals, annual budgets can be expected to be more focused and to more accurately reflect proposed expenditure patterns. The development of these more realistic budgets will also require that hospitals accurately record and report activity levels in sufficient detail to permit reasonable cost estimates. 56. Efficient use of resources. A number of steps could be taken to begin to assess relative efficiency and where appropriate to reward hospitals for outstanding performance, while assisting others to improve their efficiency. These include: - 76 -  Developing a small set of performance indicators that relate outputs to resources used, in a way that permits comparison of the hospitals. Managers in the hospitals and other relevant MOH managers should be involved in the development of these indicators.  Developing systems to routinely and reliably gather data required to produce these performance indicators.  Providing management training on activities that can improve hospital performance.  Developing incentive systems that reward improving efficiency (for both institutions and managers).  Sharing achievements with the community.  Adapting and expanding the range of performance indicators as information systems develop and management confidence increases.  Introducing benchmarking data from similar size hospitals from other relevant countries. 3.7. HUMAN RESOURCES IN THE HEALTH SECTOR Background 57. Human resources comprise an essential input into the delivery of health services. However, unlike other health inputs, such as facility infrastructure and drugs, doctors and nurses have legitimate needs and preferences. This means that the size and composition of the health workforce cannot simply be determined by government, but depends, at least in part, on the active choices of current or prospective health professionals. Similarly, the skills and motivations of health workers cannot be taken for granted, but reflect how prospective health workers are selected to the profession, and the incentive and support environment they face in the workplace. These features of human resources in the health sector result in difficult policy and management challenges. A failure to effectively address these challenges risks resulting in health system problems which, by now, are all too familiar around the world: internal and external brain drain, unequal geographical distribution of health workers, inadequate skills and inappropriate skill mix, low morale, etc. With these issues in mind, the purpose of this section is to provide an overview of the human resource situation in the health sector in Timor-Leste, and to outline key challenges for the future. 58. Following the disturbances in 1999, the NDP and the eventual HPF espoused a health system model based on broad access to publicly-financed and delivered primary care, with essential referral care being provided by regional and central hospitals. The new model represented a considerable departure from the Indonesian model, both in terms of infrastructure requirements and human resource needs (numbers and skill mix). In this context, both the NDP and HPF clearly identified a number of human resource challenges, including dealing with the departure of a large number of health professionals and managers, the absorption of low-skilled workers, low morale in the health workforce, the needs and demands of the private sector; re- alignment of skill-mix, and the development of capacity for local training. 59. Considerable progress has been made in improving the human resource situation in recent years. A clear budget framework for the health workforce has been defined. Within this framework, the MOH has gradually adjusted and restructured the workforce through a process of training, re-training, and filling of vacancies. With the Statute of Civil Service (Law No. 8/2004), - 77 - Organic Structure of MOH (Government Decree No. 5/2003), and the Practice of Health Professions (Government Decree Law No. 14/2004), the basic legislative and regulatory framework has been established. Looking forward, the MOH, with support from UNDP, has developed a National Health Workforce Plan (HWP), covering the period 2005-2015. The HWP has important strengths. It presents a general overview of the current workforce, and sets out a scenario that would gradually expand the number of doctors to a doctor-physician ratio of 1:8000, and the number of nurses and midwives through new 2- and 4-year education programs starting in 2007.58 60. Yet, while there has been much progress, the MOH, and the GOTL more broadly, faces important human resource challenges in the health sector. First, recent developments in the health workforce—in particular the arrival of Cuban doctors to Timor-Leste, and a dramatic expansion of the actual and planned number of Timorese students in medical school—make the relatively conservative plans set out in the HWP obsolete. At the same time, new and expanding oil revenues has opened prospects for a considerable expansion of the overall budget envelope over the medium and long term. These developments raise important questions about the long- term size and composition of the health workforce in Timor-Leste. 61. In part, these questions concern the sustainability and cost-effectiveness of different workforce scenarios. But human resource policy is not only about numbers. An important limitation of the HWP is some of the assumptions that underpin the projections. Similarly to workforce planning exercises in many countries, the HWP assumes that the health profession remains attractive to potential applicants, that attrition from the public sector, as well as from the health sector in general, remains low, that trained health professionals can be effectively deployed where they are needed, and that health workers are highly motivated and productive. Reliance on assumptions is inevitable in any planning exercise. Yet, attrition, deployment, and performance are clearly important policy issues in their own right, and trends in workforce will depend in important ways on policy decisions relating to these issues. The size and composition of the health workforce in Timor-Leste An overview of the current situation 62. The ideal starting point for any health system human resource assessment is an overview of stocks and flows of the health workforce as a whole—considering both employed and unemployed health workers, and health workers in both the public and private sectors. Unfortunately, comprehensive health workforce data are difficult to come by in Timor, as is the case in many other countries (see Box 3-4). 58 The HWP was based on the staffing situation in May 2005, and basic assumptions about the future demand for health care and the link between health care demand and staffing requirements. Approval of the HWP by government is pending. Needless to say, the HWP was prepared in a context of considerable uncertainty, and the need for regular updating was explicitly acknowledged. Key areas of uncertainty include fiscal outlook, population trends, future dimension of and compensation in the civil service, developments in the private sector, etc. Moreover, many policy decisions that would impact on any workforce planning exercise—e.g. concerning the hospital service package—are also pending. - 78 - 63. Due to data limitations, much of this human resource assessment refers only to the MOH workforce, although issues of attrition and the links between the public and private sector are discussed. At the time when the HWP was prepared in May 2005, the MOH employed 1,526 people in permanent positions, and 364 people on contract basis. In what follows, we use some of the tables contained in the HWP. However, in order to get a more disaggregated picture of the MOH workforce, we also turned to the raw data. As these data were received from the MOH in March 2005—slightly earlier than the preparation of the HWP, there are some discrepancies between the two sources.59 Box 3-4: Human resources in Timor-Leste: What we don’t know While the MOH in Timor-Leste maintains a detailed data base about public sector health workers, considerably less is known about the health workforce more broadly (see Table below). Efforts are currently under way within the MOH to get a better handle on who is working in the private sector, but there are still important information gaps. Moreover, although it is said that there are many unemployed health workers in Timor, no hard data are currently available. Table: The current number of health professionals in Timor-Leste Workforce plan Other sources Ministry of Health 1,390 (inc.admin) 1,477 (inc.admin) Private sector About 200 190 NGO clinics N/A N/A For-profit clinics N/A N/A Other private providers N/A 31 (CCT) Unemployed or working in other sector N/A N/A Timorese health professionals working overseas N/A N/A Note: Only permanent private professionals are included in this table (part-time and volunteer professionals are not included). Source: Figures for CCT are as of 2003, from the annual report. Beyond the stocks, information is needed on inflows to and outflows from the health workforce. Some data are available on training of new health workers, but there is less information on death and retirement, and about ‘discouraged’ health workers that decide to seek alternative employment in other sectors. It is inherently difficult to get good data on many of these issues. However, considering the importance of human resources to the sector, there is a strong case for investing more effort in generating and maintaining good data on the overall health workforce. 64. MOH employees are classified into seven levels, according to their specialty and experience (Table 3-1). The majority of the health workers categorized under level 3 and 4 are midwives and nurses. Medical doctors, district managers, and the senior staff at MOH are categorized at level 5 or above. Although the majority (82 percent) of the MOH workforce is civil servants, there are also a considerable number of contract workers. This includes both health professionals and support staff. Some districts (e.g. Ainaro) have hired a considerable number of nurses and midwives as contract staff. The proportion of contract staff is also higher in Dili relative to other districts, in part due to the relatively heavy reliance on contract nurses and doctors in the National Hospital. Overall, however, little is known about the precise composition of contract workers. 59 The data in HWP indicates MOH staffing of 1,890 (1,526 established and 364 contract positions) including 40 medical doctors, 274 midwives, and 820 nurses. In contrast, the March 2005 data base includes 1719 staff, (1,415 established positions and 303 contract staff). Foreign medical doctors and five Timorese doctors working for administration at MOH are not included in the total number of doctors. - 79 - Table 3-1: Civil Servant Classifications Civil MOH positions applicable to the level60 % Female Salary Total Share of Share of servant level Number permanent contracte level (monthly) of health health d health workers workers workers (%) (%) Level 1 Logistics, support services 25.0% US$85 160 0% 100% Level 2 Assistant nurse 21.7% US$100 198 87.4% 12.6% Level 3 Midwife, nurse 49.1% US$123 850 89.2% 10.8% Level 4 Senior midwife/nurse 28.7% US$155 335 95.8% 4.2% Level 5 Doctor (contract), CHC director, Nurse coord. 13.9% US$201 137 95.6% 4.4% Level 6 Medical doctor, District Manager 23.5% US$266 35 82.9% 17.1% Level 7 Inspector, Permanent Secretary 0% US$361 2 100% 0% 36.3% 1717 82.4% 17.6% Note: Minister and vice minister are not included into any level. Source: MOH personnel database as of March 2005. 65. Looking at the composition of the MOH workforce by profession, there are 17 Timorese medical doctors hired by MOH and this figure does not include five medical doctors currently working for administration (directors and minister). Table 3-2: Number and share of health workers in MOH by category Share of Share of Total Number permanent contracted of health health workers health workers workers (%) (%) Administration 248 83.1% 16.9% Medical doctor 17 58.8% 41.2% Midwife 255 96.9% 3.1% Nurse 763 94.4% 5.6% Dentistry 27 70.4% 29.6% Health Assistant 31 90.3% 9.7% Laboratory 33 60.6% 39.4% Other technician 103 67.0% 33.0% Logistics/support 241 39.8% 60.2% Total 1718 82.4% 17.6% Note: 1 observation missing Source: MOH personnel database as of March 2005. 66. Just over 45 percent of the MOH staff work in health centers of health posts, and another 9.3 percent in district services. Hospitals, meanwhile, account for almost 35 percent of the MOH workforce, while the remaining 10 percent work in various central administrative units. 60 A medical doctor soon after finishing education as a general practitioner is usually hired at level 5, and after a year practice, the person is promoted to level 6. An assistant nurse is paid at level 2, whereas a nurse’s salary starts at level 3. Salary for a midwife, who has a nursing education plus one-year midwifery education, starts at level 3. The salary level is the same between nurse and midwife. - 80 - Table 3-3: Number of health professionals in MOH by location of work Share of health % Female workers (%) Blood Bank 0.5% 33.3% CNFPC 1.9% 24.2% District Service 9.3% 8.1% Health Center 32.9% 40.3% Health Post 13.6% 42.2% Ministry 5.4% 19.6% Mission comisao pastoral 0.6% 90.9% National Hospital 19.4% 43.7% National Laboratory 1.4% 25.0% Referral Hospital 15.0% 35.5% Total 100% 36.4% Note: 7 observations missing Source: MOH personnel database as of March 2005. 67. There are considerable disparities across districts in the size and composition of the health workforce. The two largest cities in Timor-Leste in terms of population, Dili and Baucau, have the highest density in health professionals, which is 355.2 and 221.9 per 100,000 population, respectively, whereas most of other districts have between 120 and 190 health workers per 100,000 population (Table 3-4). Ermera is a clear outlier, with only 72 health workers per 100,000 population, 40 percent of the national average. Table 3-4: Health workforce profile by district Workers per Level 1-2 Level 3-4 Level 5+ % Female Mean age Permanent Contracted 100,000 Ermera 71.7 17.6% 71.6% 10.8% 43.2% 36.8 87.8% 12.2% Liquica 105.3 20.7% 70.7% 8.6% 29.3% 36.1 81.0% 19.0% Ainaro 124.9 20.9% 70.2% 9.0% 32.8% 36.3 73.1% 26.9% Bobonaro 137.2 19.5% 70.8% 9.7% 31.0% 36.7 85.0% 15.0% Oecusse 140.1 24.4% 65.9% 9.8% 39.0% 33.2 81.7% 18.3% Aileu 146.4 11.1% 77.8% 11.1% 37.0% 34.0 92.6% 7.4% Lautem 153.2 23.9% 69.3% 6.8% 28.4% 41.7 87.5% 12.5% Viqueque 156.5 20.2% 73.1% 6.7% 23.1% 37.1 82.7% 17.3% Covalima 169.8 24.2% 65.3% 10.5% 33.7% 35.9 88.4% 11.6% Manatuto 180.9 20.0% 70.0% 10.0% 35.0% 36.8 86.3% 13.8% Manufahi 197.0 25.0% 67.1% 7.9% 38.2% 36.3 88.2% 11.8% Baucau 221.9 22.0% 72.0% 6.0% 36.2% 36.8 89.2% 10.8% Dili 355.2 20.1% 66.3% 13.3% 40.8% 35.8 75.8% 24.2% Total 185.9 20.8% 68.9% 10.1% 36.2% 36.1 82.4% 17.6% Note: health workers here include administrative staff as well as health professionals. Source: MOH personnel database as of March 2005. Recent developments 68. The most important development since the preparation of the HWP is the new collaboration between Timor-Leste and Cuba in the health sector, which involves temporary posting of Cuban doctors in Timor-Leste (see Box 3-5), and training of Timorese medical - 81 - students in Cuban medical schools.61 In the first stage, the inflow of Cuban doctors increased the total number of doctors to 189 (approximately 1:4,893 population), including 55 Timorese doctors, 99 Cuban doctors, and 35 of other nationalities.62 More recently, a new collaboration agreement has further increased the number of Cuban doctors to a total of 302, which will take the overall number of doctors in Timor-Leste to 392 (approximately 1:2,359 population). In addition to Cuban doctors serving in Timor-Leste, Cuba is also supporting the medical education of Timorese students. In the short to medium term, this involves sending a considerable number of students for medical training in Cuba—according to current plans, a total of 628 students will participate in this program. In addition, a decision has been made to establish a medical faculty in Timor-Leste. Box 3-5: Cuban medical aid Cuban ‘medical aid’ started in 1963, when a medical brigade was sent to Algeria. Since this modest beginning, Cuba has sent nearly 51,000 health workers (mostly physicians) to 91 African, Asian, and Latin American countries. Today there are about 4,000 Cuban physicians in more than 50 countries, including, for example, 500 in South Africa and nearly 2,000 in Haiti and Central America [8]. Similar forms of assistance also exist in education and, to a lesser degree, in technical areas such as agronomy and agriculture. The medical aid program takes two forms: (i) ‘compensated collaboration’, whereby receiving nations pay the doctors’ salary (50% of which accrue to the Cuban government), transport, and accommodation; and, (ii) assistance programs (Programa Integral de Salud), under which the receiving countries only pay for transport, accommodation, and a small stipend. In both cases, the families of the overseas Cuban doctors continue to receive to receive the usual salary—around USD20 a month. Most doctors under the compensated collaboration scheme are specialists, while it is mostly general practitioners that go under the assistance program. Cuban doctors tend to have a good reputation in the receiving countries, reflecting their general willingness to work in rural areas, ability to work effectively in poorly resourced environments, and good relations with patients. For many countries, they are also an inexpensive way of getting needed expertise. On the other hand, critics have contended that doctors’ qualifications are not up to international standards, that the Cuban medical presence is part of a larger plan to bring socialism to these countries, and that the programs simply comprise a way for the Cuban government to generate hard currency. It has also bee argued that participation by Cuban doctors often is not voluntary, and some receiving countries have seen ‘defections’ [9, 10]. While complaints about qualifications often seem unwarranted, there may be some merit to the argument that Cuban medical aid seeks to support the establishment of a particular health system model [11]. Considering the impressive health outcomes in Cuba, some aspects of the model may seem worth replicating. The Cuban health system emphasizes primary care and community health, in which ‘family doctors, who live in the community they serve, form the cornerstone. On the other hand, the doctor-intensive approach to health care delivery may not be cost- effective outside planned-economy contexts, where doctors are selected not only on the basis of technical skills, but also on the basis of commitment to ‘revolutionary ideas’ [12, 13]. 61 At the time of preparing the HWP, there were 53 foreign doctors of varied nationalities working in Timor-Leste, about half of which were specialist. There are also various programs of foreign doctors visiting Timor-Leste for shorter periods. Prior to receiving support from Cuba, a number of Timorese doctors had been sent abroad—mainly to the Philippines, Indonesia, and Portugal—for advanced training. 62 Information provided to the authors by MOH staff in November 2005 and January 2006. - 82 - A second strand of the medical aid program is medical training of foreign students from developing nations. There is now a new medical school in Havana especially for foreign students. Students receive free tuition, room and board, while their home governments must pay for travel and stipend. In some countries Cuba has also supported the establishment of medical schools (e.g. Ethiopia, Yemen, Gambia), and has helped staffed these with doctors and professors. The idea in both of the above cases is to gradually educate enough indigenous doctors who are committed to serving their countries and people’, who will then go back home and work in the public sector where many Cuban doctors are currently completing missions. 69. Less has happened in the area of nursing, midwifery, and allied health professions. NCHET/IHS have trained/upgraded 30 midwifes and 10 laboratory technicians and will continue this effort next year. The institute has also run a series of courses to upgrade nurses to nurse practitioner level. Yet, the strategic direction of the IHS remains uncertain (in part due to the recent decisions concerning foreign doctors and medical education), and IHS facilities and capacity remains limited. How many health workers is enough? The challenge of health workforce planning 70. All countries have to grapple with the question of whether the size and composition of the health workforce is adequate and appropriate. In Timor-Leste, this question was given much attention when the new health system was established.. Recent developments—early experiences in establishing the health system, the development of the HWP, new fiscal scenarios, and the decisions regarding the collaboration with Cuba—has brought the question to the fore again. 71. The simplest way to approach this question is to focus on health worker to population ratios—by identifying target ratios, and by comparing the size and composition of the health workforce with other countries. The HWP was based on a target of increasing the ratio of doctors to population to 1:8,000 from a level of 1:12,242.63 72. From the perspective of international comparison, Timor-Leste does not fare particularly badly. Among a narrow set of comparator countries (Figure 3-1), Timor-Leste has the highest density for both midwives and nurses in 2004, and the number of medical doctors even does not demonstrate a critical shortage compared to other selected countries. 63 The latter figure is based on a total number of doctors of 77, including 17 in the MOH, 7 in the private sector, and 53 foreign doctors. - 83 - Figure 3-1: International comparison of level of health professionals After receiving 302 Cuba doctors Doctors Midwives Nurses GDP per capita 2003 90 1,400 80 1,200 70 1,000 60 per 100,000 50 800 US$ 40 600 30 400 20 200 10 0 0 am G a . os e de a e Is di qu st si N or er ne P bo N e on bi -L om V t do am am om ie or e C ap V In m C oz ol C Ti S M Source: Data for Timor- Leste is from the MOH, and international data is from the WHO Global Atlas of the Health Workforce (http://www.who.int/globalatlas/dataQuery/default.asp). GDP per capita is 2003 and shown in constant 2000 US$, retrieved from World Development Indicators. Note: The definition of the “health professionals� means that persons who are currently working in the place. Thus, it is assumed that non-native doctors are also included in the figure (WHO Global Atlas methodology). 73. However, comparisons with a broader set of countries (Figure 3-2) illustrate some of the difficulties associated with this type of analysis: the wide variation in health worker density across countries at the same level of income raises the question of what should be considered an appropriate target density, and the positive relationship between density and per capita income make it clear that, in any event, targets should be seen as relative rather than absolute. Figure 3-2: Health worker density across countries by level of income Doctor per 100,000 population Nurse per 100,000 population 400 1,500 Georgia Nurses, density in 100,000 population Azerbaijan Armenia 350 Physician, density in 100,000 1,200 300 Kyrgyz Republic Mongolia 250 population Tajikistan 900 Azerbaijan 200 Kyrgyz Republic Nicaragua 150 Cuba 600 Philippines Armenia Tajikistan Philippines Cuba Georgia 100 Mongolia Lao PDR 300 50 Madagascar Timor-after Lao PDR Nicaragua Indonesia Mozambique Timor Madagascar Indonesia Mozambique 0 0 Timor 0 200 400 600 800 1,000 1,200 1,400 1,600 1,800 2,000 0 200 400 600 800 1,000 1,200 1,400 1,600 1,800 2,000 GDP per capital (current US$) GDP per capita (current US$) Source: Data for Timor- Leste is from the MOH, and international data is from the WHO Global Atlas of the Health Workforce (http://www.who.int/globalatlas/dataQuery/default.asp). GDP per capita is 2003 and shown in constant 2000 US$, retrieved from World Development Indicators. 74. Although appealing in its simplicity, the focus on health worker to population ratios has serious limitations—simply ensuring that adequate numbers of health professionals are trained is typically not enough to achieve the government’s objectives. In particular, crude staffing ratios do not adequately take into account actual health care needs, the range of possible interventions - 84 - to meet these needs, the different technologies or approaches for delivering services, and the impact of incentives and motivation of health workers on both productivity and quality. 75. A more meaningful approach to thinking about workforce needs is to link it to the planning of health services—infrastructure, delivery models, intervention priorities, the role and function of different health professions, etc.—which in turn should be rooted in a fiscal framework and an assessment of health needs. This was essentially the initial planning exercise undertaken in Timor-Leste by the IHA and immediately after independence. As discussed earlier, an assessment was made that, considering fiscal and other constraints, the health needs of the Timorese were best met through a more limited health facility network. Indeed, relative to 1999, the planned facility infrastructure entailed a substantial scaling down of all facility levels. However, as can be seen from Table 3-5, the number of facilities is currently substantially higher than was planned, largely due to an expansion of the number of health posts relative to the initial Post-independence plan. Table 3-5: The number of health facilities in Timor-Leste: past, planned, and actual Central Referral Health Centers Health Total hospital hospital CS4 CS3 CS2 posts Pre-1999 1 11 21 46 309 388 August 2000 1 5 17 52 146 221 Post-independence plan 1 4 5 64 85 159 Current 1 5 5 4 55 191 261 Note: actual facility data based on MOH facility data base maintained by the pharmaceutical department; obtained Oct. 30 2005; historical and planned infrastructure based on table 2 in "Initial Steps in Rebuilding the Health Sector in East Timor" by Tulloch, J. et al. (2003) 76. The MOH has also adopted a Basic Package of Services Policy, which defines the set of interventions that each type of facility is intended to provide, and the staffing requirements to deliver these services (Table 3-6). Taking into account both the actual facility network, and the staffing norms implied by the BPS, it is straightforward to calculate the implied (MOH) staffing need. Table 3-6: Staffing needs and the actual workforce: A comparison Staffing need Cent Ref. Tot. Tot. Central MOH Total (Basic package standards) hospital hospital district hospital agency Doctor NA NA 9-14 NA NA NA NA Nurse NA NA 461 NA NA NA NA Midwife NA NA 324 NA NA NA NA Allied and other health NA NA 46 NA NA NA NA Admin/support NA NA NA NA NA NA NA Actual workforce Cent Ref. Tot. Tot. Central MOH Total (Based on Table 5 in WFP) hospital hospital district hospital agency Doctor 14 5 21(?) 19 40 Nurse 155 122 540 277 817 Midwife 30 45 199 75 274 Allied and other health 20 18 65 38 31 134 Admin/support 75 30 153 105 15 86 359 Contract 34 55 250 89 10 15 364 Total 328 275 1228 603 56 101 1988 - 85 - Note: Staffing needs based on number of facilities (64 CS and 191 HP) and standards in Basic Package of Services; in actual staffing data, it was not possible to categorize staff in MOH and central agencies, as well as contract staff, by qualification. 77. Although simplistic, this analysis raises a number of issues: a) Considering the staffing standards defined by the BPS, the actual number of health professionals at district level is more than adequate for most categories, with midwives being the notable exception. The MOH is very aware of the need to expand the number of midwives in the health system, and some efforts are under way to this effect. b) Due to the expansion of health facility infrastructure, the required number of health professionals, in particular nurses and midwives, is considerably higher than originally intended. This highlights the need for careful thinking about further infrastructure expansion. c) The actual number of doctors (before arrival of the new cohort of Cuban doctors) at district level is higher than needed (as defined in BPS). With the arrival of Cuban doctors, there are now plans to have a doctor in each sub-district, and it is envisioned that many of the returning Timorese medical students will also work at district and sub-district level. While the expansion of the number of doctors in the health system can bring important benefits, it is associated with a number of risks. For one thing, the trend raises questions about cost-effectiveness.64 Are doctors needed at district and sub-district level to ensure that the Basic Package of services is delivered with a high level of quality? What are the costs associated with staffing rural positions with doctors? Do any improvements in quality justify the additional costs? There are no easy answers to these questions, but the proposed heavy reliance on doctors in primary care deserves some scrutiny. International experience also suggests that decisions about staffing and skill-mix at any point in time will shape the development of the future health system—through regulatory development, changes in the service package, and the expansion of capital and recurrent spending to match the capacities and preferences of physicians. Some of these developments may improve health outcomes, but the question is at what cost, and whether the cost is sustainable.65 d) There is a considerable number of administrative and contract staff at district level. Little is known about what specific skills and functions these categories of staff have, and there are no staffing norms to guide an assessment of whether the staffing levels are reasonable or not. 64 Indeed, in the interest of containing costs, many countries are substituting nurses for physicians in primary care. There is only limited evidence on how changes in the skill mix impact on quality and costs, most of which comes from developed countries [14, 15]. The available evidence suggests that, for defined clinical activities (focus has often been on primary care or management of specific types of cases), patient outcomes are as good for nurses as for doctors, that there are no systematic differences in prescriptions or testing practices, that patient satisfaction is often higher for nurses., … Cost savings have however often not been realized due to the fact that nurses tend to spend longer with patients. This evidence cannot be straightforwardly generalized, however. Evidence is not strong… Particular care must be given to the specific clinical tasks and activities that are considered, the training and skills of nurses and doctors in particular contexts, supporting organizational and institutional context, etc. 65 Robert Evans, a Canadian health economist notes that “…manpower policy, however sophisticated, cannot be developed independently of the structure of the delivery system. If changes in type, quantity, and mix of personnel are desired, then they must somehow be fitted in, either to the present or to some other modified form of care provision… Present manpower training decisions determine future production choices over as long as a generation, by the ‘human capital’ they create.� [16] - 86 - e) A considerable number of health professionals work in hospitals, central agencies, or the MOH. There is currently no clear basis for assessing whether the current number of staff and their skill mix is appropriate, or for determining future staffing needs. Changes in the health workforce over time 78. So far, we have focused on the overall MOH workforce. This approach clearly has important limitations. First, in order to achieve any desired workforce targets, changes in the workforce over time, both inflows (training, international migration, etc.) and outflows (death, retirement, alternative employment of trained health professionals, international migration, etc.) must be taken into account. Second, in most health systems, part of the health workforce is employed by the private sector. This is also true in Timor-Leste. This is an important consideration because the mix and location of services in the private sector may not reflect government priorities, such that although the overall workforce is adequate, health needs are not effectively met. For example, the private sector may be concentrated in urban areas, leaving rural areas underserved. This raises questions about the role and relative size of the private sector, and the role of government policy in this regard. These issues are the focus of this section. Training 79. The HWP set out detailed training scenarios for different health professions. However, as noted, these scenarios have not been reflected in developments on the ground. The Institute for Health Sciences (IHS) has recently been established and been given a clear mandate for both basic and in-service training of nurses and allied health professions. IHS continues some of the training programs initiated under the IHA, but there has been a lack of clarity about future directions. Training scenarios for physicians, in turn, will depend critically on whether current commitments to send Timorese students to Cuba, and to establish a medical faculty at the University are followed through. Attrition 80. While training of new health professionals comprises an inflow into the health workforce, medium- and long-term trends in the workforce also depend on outflow. This, in turn, is a function of health professionals dying, retiring, seeking alternative employment, or emigrating to another country. Similarly to workforce planning exercises in many countries, the HWP in Timor-Leste makes very general assumptions about attrition. Given the dearth of information on these issues—in Timor and internationally—this is a reasonable approach. However, it risks resulting in a neglect of how government policy and other factors can influence attrition over time. The big questions in this regard concern labor market choices by health workers and international migration. Currently, it seems that the health profession remains attractive. Moreover, as a result of the scaling down of the health sector by the IHA and the initial post- independence MOH, there is a surplus of health professionals. 66 Over time, however, this situation may change, depending on trends in civil service remuneration, job satisfaction in the health sector, alternative employment opportunities, etc. 66 As noted above, the extent to which to which there are actually large numbers of unemployed health professionals is unclear. - 87 - 81. Similarly, international migration of health workers currently does not seem to be an important concern. However, in part, this may reflect the limited number of health workers with transferable clinical and language skills. Of course, with the training of an expanded number of doctors, who moreover will acquire foreign language skills, international migration may become a growing challenge. There are no simple approaches to tackling attrition. It is however important for the MOH to continually strengthen information about these issues so as to inform policy.67 Public-private choices 82. The private sector can, and often is, an effective partner to the government in delivering health services. However, although the NDP and HPF promote a ‘mixed’ health system model, it is envisioned that the MOH will remain responsible for financing and delivering the majority of health services in Timor-Leste for the foreseeable future. In order to do so effectively, the public sector needs to be able to attract and retain the right numbers and types of health professionals. Currently, this does not seem to be a problem, although data on the private sector is limited. The qualitative study that was undertaken as part of the HSR suggests some reasons why this is the case. Study respondents in both the public and private sectors often indicated a strong preference for the public sector—rooted in the perception that the public sector offers greater security and a retirement pension. In the words of some of the participants in the focus group discussions: I think I prefer to work in the government sector than the private sector because in the private sector your contract can be terminated any time. But in the government sector you will work until you reach 50 or 60 years old then you will retire. (Rural health worker) Personally, I would choose the government sector, because if you work for the government you will get a retirement benefit. In the private sector, as long as you are healthy you can work but if you are sick you will have to stop. (Rural private sector health worker) 83. Indeed, it seems that many health workers in the private sector are there because they did not pass the entrance exam to the public sector, or because they could not find employment in the public sector for some other reason. 84. Some public sector workers also expressed pride in being a civil servant in the newly independent government. Thus, it seems that some form of “intrinsic motivation�—often attributed to NGO and faith-based services—also exists in the public sector. I was working in the private sector then there was a recruitment from the government sector and I started to apply because I felt that I was called to do something for this nation. In the end I was selected so I moved to the government sector. (Urban CHC worker) 85. Nevertheless, the private sector is expanding and is likely to continue to expand in coming years with the channeling of considerable amounts of external financing to the NGO sector. Some health workers in the qualitative study were aware of higher NGO salaries and better benefits (such as transportation) and viewed them as a potential lure to the private sector. 67 Since recruitment process started in 2002, approximately 50 health professionals have left the civil service, reflecting both death, health problems, private sector jobs, and migration. More detailed and systematic tracking is needed. - 88 - Personally, as a health worker I have the responsibility to do what I can do for the nation. But if the government can not look after us I think people might abandon their job to work in the private sector. (Rural CHC worker) 86. Moreover, some respondents in the qualitative study complained about nepotism and corruption in both hiring and transfers. Due to the nature of qualitative work, we do not know how wide spread this perception is, or whether it is well-founded. Nonetheless, insofar as it is a problem, it can gradually undermine the appeal and public sector. 87. Although anecdotal, these accounts suggest that tensions between the public and private sectors may grow over time. Considering that doctors tend to have different expectations and aspirations than health workers with lower qualifications, these challenges are likely to intensify when the number of doctors in the Timorese health system expands. In this regard, it is also important to note that the Statute of Civil Service does not proscribe civil servants from holding a second job on a part time basis. At the moment, there is however no systematic information about how widespread dual job-holding is. Deployment 88. A final issue that merits attention is the deployment of the MOH workforce. The pursuit of equity in access to health care requires that at least some health workers are effectively deployed to rural and remote health facilities. The primary questions in this regard are what types of health workers should be deployed where, and what model of service delivery is most appropriate. Whatever model is adopted, however, all health systems face a difficult challenge in actually deploying health workers where they are needed. Countries rely on a range of different approaches to meet this challenge, including contract bonding tied to financing of training, rural bonuses, recruitment from rural areas, etc. 89. There are indications that deployment is already a challenge in Timor-Leste. For example, the MOH is facing difficulties in filling positions in rural and remote facilities.68 The focus groups discussions with health workers provided some insight into why these difficulties arise—health workers in the qualitative study expressed frustration about both security, personal and professional isolation, and a lack of access to training. The MOH is currently relying on recruitment from local communities, and is finding that these recruits are willing to return to work in their communities.69 However, many of the respondents in the qualitative study argued that financial incentives were needed, and that more should be done to ensure that rural health workers were not passed over in training. Indeed, rural or hardship incentives was a feature of the health system under Indonesian occupation. However, currently there are no provisions for special payments or bonuses for this purpose in the Statute of Civil Service. Clearly, the expansion of newly trained doctors in the coming years, many of whom are expected to work at district and sub-district level—is likely to intensify this challenge. 68 Although the Statute of Civil Service states that individuals in permanent positions should be “…always ready to be assigned anywhere on the national territory or overseas� (Article 14), MOH deployment is currently application based, and non-voluntary deployment to specific posts is not used. 69 It should be noted that CCT has a policy of not posting nurses and midwives near their families, apparently so that family obligations will not contribute to absenteeism. - 89 - Performance, incentives, and motivation 90. What do we know about the performance of health professionals? Health worker performance has many dimensions and is difficult to measure. Due to data limitations, we focus here on some crude measures of productivity, and non-technical quality.70 Productivity 91. There is considerable variation in the number of outpatient visits per health professional across districts in Timor-Leste. In part, this reflects staffing patterns: districts which have higher ratios of health workers to population tend to have fewer OP visits per professional. But there are also other issues at play. For one thing, we should take into account inpatient activity. Indeed, lower OP visits per professional in part reflects higher IP activity (Figure 3-3). But even taking this into account, there are likely to be substantial productivity or workload differences across districts. The average workload was 1,547 OP visits per MOH professional in 2004, while it was 1,725 patient visits per staff among non-government health care service providers. Figure 3-3: Differences in workload across districts 60 3,500 Inpatient days per health professional Oecussi Dili OP visits per health professional 3,000 50 Ermera 2,500 Viqueque Aileu 40 Liquiça 2,000 Covalima Lautem Manatuto 30 Ainaro Total Manufahi Ainaro 1,500 Lautem Total Covalima Bobonaro Dili 20 Manufahi Oecussi Baucau Bobonaro 1,000 Manatuto 10 Viqueque Baucau 500 Ermera Aileu Liquiça 0 0 0 50 100 150 200 250 300 350 400 0 50 100 150 200 250 300 350 400 health professionals per 100,000 OP vists per health professional Non-technical quality: staff attitudes and client orientation 92. There is only limited information on non-technical quality in Timor-Leste. Although on a small-scale, many patients in the qualitative study complained of rudeness and negligence in their interaction with health workers, along with a lack of explanation of both diagnoses and treatments, and preferential treatment of some patients. For example, one patient in an urban area had the following complaint: If we say we don’t understand something the nurses scold us or tell us to go clean up our ears so that we can hear better. We are scared to ask questions. (Urban client) 93. Given the scale of the qualitative work, is not clear how widespread this problem is. It is also not clear why the problem arises. For example, some health workers argued that problems 70 The important omission is, of course, technical quality—i.e. clinical knowledge and consistency of diagnostic and treatment procedures with ‘best practice’. Information on technical quality is normally collected through verbal or written tests of clinical knowledge, discussion of standardized cases (clinical vignettes), observation in clinical setting, or record reviews. To our knowledge, no studies of technical quality have been undertaken in Timor-Leste since independence. - 90 - reflect a lack of training, heavy workload, and other factors related to the work environment. Whatever the source of the problem, it clearly merits further attention. Financial and non-financial incentives 94. Financial incentives are clearly an important determinant of both labor market choices and performance. The salary of health professionals in Timor-Leste is based on the civil service pay scale (see Table 3-1). Although this salary scale has not changed since July 2000, this may change as a result of the new fiscal space that the government is currently enjoying. 95. An obvious comparison with the civil service pay-scale is the salary level in the private sector. Comparison between the salary scale for health personnel working at CCT clinics and the one for those working for the government, clearly indicates that CCT health personnel earn significantly higher pay than the government employees. For instance, a medical doctor in CCT is paid $400 on average, but the government salary for a medical doctor ranges only from US$201 to US$266. Catholic Church related organizations compensate doctors at similar level to the MOH doctors, but medical doctors at a clinic owned by international NGOs are paid about twice as high as the MOH doctors. A general nurse in CCT earns US$217 per month, while a nurse at level 3 is paid US$123 and even a senior nurse with level 4 earns only US$155. Both Catholic Church related organizations and NGO clinics pay nurses and midwives at similar level to the government salary scale (see Appendix 4 for more detailed data). 96. In addition to monthly salary, health professionals, especially nurses and midwives working for NGOs may qualify to receive a bonus in a certain occasion. CCT pays about $183 on average per bonus per health personnel, but the bonus at other organizations for nurses and midwives is just $25 per personnel (see Appendix 4). 97. Labor market choices and performance are also influenced by opportunities for salary increases and promotion. In the Statute of Civil Service, promotion of civil servants is defined as to be ‘based on merit criteria and subject to a specific competitive application’ (Article 22). There are however no within-grade salary differentiation. In practice, this means that the only route to a salary increase or promotion is by applying to a vacant position at a higher grade. The qualitative work provides some evidence of frustration with salary level, both in relation to private sector, and in relation to other civil servants who are seen to work less, and exposed to less risk. Now the salaries of the government workers is not enough [compared to the Indonesian era]. If people don’t earn enough they will have to do some other business. After they finish work they might sell drugs. We can not stop people from doing this because they don’t have money for living. Now there are also a lot of people who have left East Timor to work, for example, in Ireland, Japan, and Malaysia. (Urban worker) 98. Motivation in the public sector may also be undermined a feeling that hiring, promotion, transfers to urban areas, access to training opportunities, and other rewards are not allocated in a way that reflects performance—some of the focus group participants complained of nepotism. There is no official way to transfer for the time being. But some people do this through “KKN.� If you know someone who has a high position in the Ministry they can transfer you to a place of your choice. - 91 - One of my friends did it, she moved from [rural subdistrict] to Dili because she knows someone. (Urban worker) 99. Overall, there is a strong case for stronger performance monitoring in the public sector, with associated rewards and sanctions (not necessarily pecuniary). Currently, government workers do not see that an effective performance monitoring system and incentive system is in place. In the CHC, those who come to supervise us are from the WHO… they do not directly reprimand us, but just observe what work we do. Then they will just put a check or cross on their supervision sheet. They do not tell us whether or not we do things properly… We do not know where are the weaknesses because they do not tell us. (Rural worker) 100. This can be contrasted to the private sector. Since we are private the rules are very strict. If you come 10 minutes late, tomorrow you will get called to the office to talk about it. If you continue three times then you’ll be out. The discipline in working is also controlled tightly. Once again since we are working with private NGO we have to follow what the NGO ask us to do. If not, even though you have a one year contract it can be decreased into 6 months contract. (CCT worker) - 92 - CHAPTER IV: HEALTH SECTOR EXPENDITURE AND FINANCING 4.1. SOURCES OF FINANCING FOR EXPENDITURES ON HEALTH GOODS AND SERVICES 1. The main sources of financing for expenditures on health goods and services (henceforth “health expenditures�) in Timor-Leste are:  (i) Households;  (ii) Non-Governmental Organizations (NGOs);71  (iii) The Government’s Budget; and  (iv) Development Partners’ Grants. 2. Health expenditures financed from (iii) and (iv) are referred to in this report as public sector health expenditures. On the other hand, those financed from (i) and (ii) are private health expenditures. 4.2. OUT-OF-POCKET EXPENDITURES BY HOUSEHOLDS 3. Data on health expenditures by households were collected in the 2003 Demographic and Health Survey (DHS), as well as in the 2001 Timor-Leste Living Standard Measurement Survey (LSMS). No data on household expenditure were collected in the 2002 Multiple Indicator Cluster Survey (MICS).72 Household Health Expenditure According to the 2003 DHS 4. The DHS collected information on household expenditure per outpatient consultation with: (i) their usual health care provider; (ii) the health care facility closest to the household; (iii) by type of health care provider; and (iv) by type of health care facility. These data are summarized in Table 4-1- Table 4-3.73 5. Table 4-1 shows that the mean cost of examination and treatment per consultation with the household’s usual provider was $0.22, while the mean cost of transport associated with the consultation was $0.18, for a total mean cost per consultation of $0.40. Many people (patients and their attendants) had walked to their consultations, thus incurring no out-of-pocket expenditure.74 The data are broken by wealth quintile and by region. The breakdown by wealth quintile shows the richest quintile spending a much larger amount per examination and treatment than other quintiles, but otherwise the pattern is erratic. In the breakdown by region, the highest mean cost of examination and treatment is for urban areas. 71 Including church-run health facilities and Café Timor clinics. 72 Some expenditures by NGOs are financed by the Government and/or development partner grants and are public expenditures managed by NGOs. 74 The opportunity cost of the time spent in transportation and at the health care provider was not estimated in the survey. - 93 - Table 4-1: Cost of Transport and Examination/Treatment for Outpatient Consultations with Usual Health Care Provider Mean cost Mean cost of Mean cost of Number of of transport transport excluding examination households c/ ($) b/ those consultations and treatment where only walking ($) b/ was used as means of transport ($) b/ Region Urban 0.33 0.66 0.47 887 Rural East 0.15 0.57 0.20 981 Rural Central 0.09 0.58 0.05 1265 Rural West 0.20 0.80 0.25 671 Household Wealth Index Poorest 0.10 0.66 0.18 752 Poorer 0.15 0.51 0.10 786 Middle 0.16 0.90 0.22 772 Richer 0.19 0.72 0.09 751 Richest 0.30 0.56 0.53 743 All households (weighted average) 0.18 0.64 0.22 3804 Note: The categories of "usual health care provider" listed in the survey's questionnaire were as follows: traditional healer, midwife, nurse, doctor, and other. Fifty-seven percent of households in the sample identified "nurse" as their usual health care provider; 15.3% "doctor"; 14.2% "midwife"; 0.9% "traditional healer"; 0.7% "other"; and 12% indicated that they consult no provider when someone in the household is sick. b/ Per consultation. c/ Not included are 517 households that reported not usually seeking a consultation for ill household members. The total sample size was 4,321 households. Source: 2003 DHS, page 246. 6. Table 4-2 shows the average cost of examination and treatment (and distance traveled) by type of health care provider. The table indicates that there are striking differences in cost between different types of providers, with pharmacist ($35.98) and “doctor in private clinic� ($33.61) being the highest-cost providers, followed by traditional healers ($18.01) and midwives in private clinics ($10.73). The cost of consulting government providers (rows fourth, sixth and eighth) were very low by comparison to the various private providers. Because the questionnaire did not ask how many consultations a given household had carried out for any given type of provider, it is not possible to calculate a weighted average consultation cost. - 94 - Table 4-2: Average Cost of Examination/Treatment and Distance Traveled by Type of Health Care Provider Consulted in the Past Twelve Months by any Members of the Household Average cost Average Number of of examination distance households and treatment traveled to use that consulted ($) a/ the services this provider (km) Type of Provider 1 Traditional Healer (dukun) 18.01 2.98 318 2 Traditional Birth Attendant 4.49 1.22 177 3 Pharmacist 35.98 10.57 106 4 Midwife in Government Facility 0.24 5.80 906 5 Midwife in Private Clinic 10.73 5.46 295 6 Nurse/Paramedic in Government Health Clinic 0.43 6.77 2253 7 Nurse/Paramedic in Private Clinic 3.36 5.81 470 8 Doctor in Government Health Facility 0.02 6.97 680 9 Doctor in Private Clinic 33.61 6.07 202 10 Dentist 0.00 5.13 40 Total number of households reporting having consulted one of the above types 5447 of providers in past twelve months b/ Note: a/ Per consultation. Transport costs are not included. b/ The total number of reporting households (5,447) exceeds the sample size (4,321) because some households consulted more than one type of provider in the past twelve months. Source: World Bank calculations from 2003 Timor-Leste Demographic and Health Survey data 7. Table 4-3 presents information by type of health care facility instead of type of provider. There is a striking difference in cost between private clinics and government facilities. As with Table 4.2, it is not possible to calculate a weighted average for all households. Table 4-3: Average Cost of Examination/Treatment and Distance Traveled by Type of Health Care Facility Visited in the Past Twelve Months by any Members of the Household Average cost of Average Number of examination and distance households that treatment ($) a/ traveled to visited this type use the of health care services (km) facility Type of Health Care Facility Mobile Clinic 0.08 2.99 149 Government Sub Community Health Center 0.37 4.94 1039 Government Community Health Center 0.04 7.31 1626 Private Clinic 8.87 5.19 733 Government Hospital 0.02 7.11 1095 Total number of households reporting having consulted one of the above types of providers in past twelve months b/ 4642 a/ Per consultation. Source: World Bank calculations from 2003 Timor-Leste Demographic and Health Survey data. - 95 - Household Health Expenditure According to the 2001 Timor-Leste Living Standard Measurement Survey 8. The Timor-Leste LSMS collected information on household expenditure on food items, rent, and other non-food items; health expenditure was part of the latter. For all non-food items, the survey collected information on expenditures only, i.e., no information on quantities was collected. No separate report of the results of the Timor-Leste LSMS has been published. Instead, many of the results were used in the preparation of the Timor-Leste Poverty in a New Nation: Analysis for Action report. 9. Information on household health expenditures was collected not just in the Expenditures/Consumption module of the survey but also in the Health module. The Poverty in a New Nation report used the health expenditure estimates from the Expenditures/Consumption module. Household health expenditure per capita per month estimated from this module was $0.17, as it can be seen from Table 4-5. Table 4.4 shows that household expenditure on health represented about one percent of total household consumption expenditure. This percentage would be on the low side by international standards, as suggested by the comparisons with other countries in Table 4-4.75 Table 4-4: Main Components of Total Household Consumption Expenditure for Timor-Leste and Several Other Low- and Middle-Income Countries (percent) Timor Leste Vietnam Nepal Kyrgyz Ecuador S. Africa Panama Brazil 2001 1992-93 1996 1996 1994-95 1993 1997 1996-97 Food 63 51 64 45 50 30 46 28 Non-Food Items Education 1 3 3 2 8 3 8 6 Health 1 6 3 1 2 1 5 Other Non-Foods 8 21 13 19 21 40 37 21 Consumer Durables 13 1 4 5 5 Housing 27 8 15 30 16 25 3 40 Total 100 100 100 100 100 100 100 100 GNP per capita (US$) a/ 470 170 210 550 1,280 2,980 3,080 4,400 Note: a/ GNP per capita corresponds to the same year of the survey, except for Panama (1996). Source: Timor-Leste Poverty in a New Nation: Analysis for Action, Volume 2, page 29. The estimates for Timor-Leste are from the 2001 LSMS. 10. Table 4-5 also shows that about one-third of household expenditure on health goods and services was in the form of medicines purchased for self-treatment, and an additional one-fourth was in the form of payments to traditional healers and birth attendants. Table 4-6, prepared with data from the Health module of the LSMS questionnaire, shows that the most popular source of medicines purchased without a prescription were vendors/kiosk owners – i.e., shops not specialized in selling medicines. 75 Total monthly household per capita consumption expenditure (food and non-food) estimated from the LSMS was about $24.23. - 96 - Table 4-5: Household Expenditures on Various Health Goods and Services in the Past 30 Days and the Past 12 Months, Consumption/Expenditure Section of the LSMS Questionnaire a/ Average per household amount spent in past 30 days ($) b/ Type of Expenditure Payments to Public Hospital/Community Health Center/Clinic 0.12 Payments to Church or Private Hospital/Community Health Center/Clinic 0.15 Payments to Mobile Clinics 0.06 Payments to Midwife/Private Nurse 0.07 Payments to Traditional Healer/Traditional Birth Attendant 0.27 Medicines Purchased for Self-Treatment 0.33 Other Health Expenditures (Contraceptives, Vitamins, etc.) 0.01 Total Per-Household Health Expenditure, $ 1.01 Per capita, $ 0.17 Note: The number of households in the sample is 1,800. a/ The Timor-Leste Living Standards Measurement Survey (LSMS) was a nationally representative survey of 1,800 households covering about one percent of all households. The average household size in the sample was about 6 persons (Poverty Assessment, Volume II, page 62). b/ Excluding any items purchased for processing or resale as business. Source: World Bank calculations from 2001 LSMS data. Table 4-6: Household Expenditure on Medicines Without a Prescription in the Past 30 Days by Type of Seller, Health Section of the LSMS Questionnaire [1] [2] Average expenditure of Number of households a/ per household ($) Type of Seller of Medicines Unweighted Weighted Pharmacy 3.52 43 2858 Drug Store 2.89 91 5261 Village Medicine Man 1.79 44 4623 Vendor/Kiosk Owner 1.86 220 19622 Ambulant Vendor 1.50 1 123 Other 2.58 66 7146 Note: a/ Excluding households which reported not having bought any medicines without a prescription in the past 30 days. Source: World Bank calculations from 2001 LSMS data. 11. Table 4-7 from the Health module of the LSMS, showing expenditures on visits to health facilities, can be compared with Table 4-3. While the lists of types of health care facilities are not identical, there is considerable overlap. For overlapping categories, there are striking differences in the average cost per outpatient visit (excluding transportation) between the two tables. For example, the costs are much higher in Table 4-7 for government hospitals ($1.65 versus $0.02), community health centers ($0.21 versus $0.04) and mobile clinics ($0.34 versus $0.08). However, for private clinics, the average cost is much higher in Table 4-3 ($8.87 versus $1.09). More information would be needed in order to assess which set of figures is more realistic. - 97 - Table 4-7: Household Expenditure on Visits to Various Types of Health Care Facilities in the Past 30 Days, from the Health Section of the LSMS Questionnaire [1] [2] [3] [4] [5] Cost per visit Transportation Total cost Number of Number of excluding cost per visit per visit households c/ visits per transportation b/ ($) household ($) a/ (past 30 days) c/ Type of Health Care Unweig Weig Facility [1] + [2] hted hted Public Hospital 1.65 1.21 2.87 210 15165 2.40 Community Health Center 0.21 0.16 0.37 227 25287 1.94 Public Health Post 0.16 0.29 0.45 14 1431 1.49 Mobile Clinic 0.34 0.04 0.39 53 6838 1.82 Private Hospital/Health Clinic 1.09 0.67 1.76 127 7894 2.24 Church Clinic 1.06 0.46 1.52 83 6980 2.47 Note: a/ Including any medicines prescribed during this visit even if purchased elsewhere. b/ Includes payments in either money or kind. It also includes the transportation costs of anyone who accompanied the patient in his visit. It does not include the opportunity cost of time spent traveling to/from the facility or at the facility itself. c/ Excluding households which reported no visits to health care facilities in the past 30 days. Source: World Bank calculations from 2001 LSMS data. 12. Similar major discrepancies arise when comparing Table 4.2 (from DHS) and Table 4.8 (from the Health module of the LSMS). For example, the average cost of a visit to a private doctor (excluding transportation) according to Table 4.8 was $3.66, but according to Table 4-3 it was $33.61. Again, more information would be needed to clarify these discrepancies. Table 4-8: Household Expenditure on Visits to Various Types of Private Health Care Providers in the Past 30 Days, Health Section of the LSMS Questionnaire [1] [2] [3] [4] [5] Cost per visit Transportation Total Number of households c/ Number of excluding cost per visit cost visits per transportation b/ per household (US$) a/ visit (past 30 (US$) days) c/ Type of private provider [1] + [2] Unweighted Weighted Private Doctor 3.66 1.60 5.26 55 4575 2.79 Private Nurse 1.27 0.39 1.66 57 5154 0.96 Midwife 0.31 0.33 0.64 23 3044 1.70 Other 0.72 0.01 0.73 11 1053 2.10 Note: a/ Including any medicines prescribed during this visit even if purchased elsewhere. b/ Includes payments in either money or kind. It also includes the transportation costs of anyone who accompanied the patient in his visit. It does not include the opportunity cost of time spent traveling to/from the facility or at the facility itself. c/ Excluding households which reported no visits to private health care providers in the past 30 days. Source: World Bank calculations from 2001 LSMS data. 13. The above information refers in all cases to outpatient care. Table 4-9 from the Health module of the LSMS presents some interesting data on the cost of inpatient care. The cost per hospital stay (excluding transportation) was the highest in private hospitals ($34.42), followed by Church hospitals ($19.31) and public hospitals ($8.92). It may also be noted that while the average length of stay was roughly similar for public and private hospitals at 13.25 and 11.54 - 98 - days respectively, it was much shorter for Church hospitals at 3.17 days. This suggests that the case mix was very different in the latter as compared to the first two although practice on protocols for the same ailments may also vary. Table 4-9: Household Expenditure on Most Recent Hospitalizations in the Past 12 Months by Type of Hospital, Health Section of the LSMS Questionnaire c/ [1] [2] [3] [4] [5] Number of Cost per Transportation Total cost Number of households d/ days spent hospital stay, cost per per hospital in hospital excluding hospital stay stay (average transportation (average per (average per (average per household) ($) per household) household) ($) b/ household) a/ ($) Type of Hospital [2] + [3] Unweighted Weighted Public Hospital 13.25 8.92 8.43 17.36 91 6968 Private Hospital 11.54 34.42 13.67 48.09 8 403 Church Hospital 3.17 19.31 5.81 25.12 3 199 Other Note: a/ Including payments in money or in kind, and also including any medicines prescribed during this stay, even if purchased elsewhere. b/ Includes payments in either money or kind. It also includes the transportation costs of anyone who accompanied the patient to/from the hospital. It does not include the opportunity cost of time spent traveling to/from the hospital or at the hospital itself. c/ I.e., for each member in a household who reported having been hospitalized at least once in the past 12 months, the data in the table refers to the most recent of those hospitalizations. d/ Excluding households which reported no hospitalizations in the past 12 months. Source: World Bank calculations from 2001 LSMS data. 4.3. NON-GOVERNMENTAL ORGANIZATIONS EXPENDITURES AND FINANCING 14. Health expenditures are also incurred by a number of NGOs. As already noted in the previous chapter, the largest of these NGOs is Clinic Café Timor, a primary health care program within the Cooperativa Cafe Timor, the coffee farmers’ cooperative. Clinic Café Timor currently has an annual operations budget of about $450,000. The Clinic does not charge user fees. Its funding comes from coffee trade profits, private donations, and in recent years USAID. Funding from USAID has amounted to about $150,000 per year until the current year 2005/06, but it is being phased out and will cease after 2007/08. Since USAID’s grants to Clinic Cafe Timor would also be counted as part of the financing from Development Partners, the Clinic’s net contribution to health sector financing would be about $300,000. 15. After 2007/08, the MOH has agreed to provide a subsidy to Clinic Café Timor to cover the costs of drugs, which would permit the Clinic to continue serving patients who do not belong to the coffee farmers’ cooperative. This subsidy would approximately compensate for the loss of USAID funding ($150,000 per year). 16. In addition to Clinic Café Timor, there are a number of other NGOs active in the health sector. These NGOs finance their health service provision through a combination of user fees, donations (mostly from abroad), and grants by development partners. The user fees portion of their revenues would also be included in households’ out-of-pocket expenditures, and development partner grants would similarly be included under donor programs discussed below, so the NGOs’ net contribution to health sector financing would be the value of donations. - 99 - 4.4. PUBLIC SECTOR EXPENDITURE AND FINANCING Trends in Total, On-budget and Off-budget expenditures76 17. As noted above, public sector health expenditures include those funded from the government’s own budget, i.e., the Ministry of Health’s budget, and those funded by bilateral and multilateral aid agencies in the form of off-budget grants. The trends in these expenditures from 2000/01 to 2006/07, and government projections for 2007/08 to 2009/10, are shown in Table 4.10. 18. It can be seen from that table that until 2005/06 off-budget expenditure accounted for a larger share of total expenditure than the MOH’s budget. In 2005/06, of an estimated total public expenditure on health of $44.27 million, $27.38 million (or about 62 percent) was off-budget expenditure and $16.88 million was expenditure budgeted in the MOH’s budget77. However, projections for 2006/07 indicate that the shares of MOH and off-budget expenditure in total public expenditure on health would be approximately equal. Based on current information, it appears that off-budget expenditure peaked in 2005/06 and would subsequently decline. Projections of off-budget expenditure are based on donor commitments as of the time of preparation of the 2006/07 budget. Details of the contributions of different donors are given in Table 4.11. 76 Timor-Leste’s fiscal year is from July 1 to June 30. 77 The estimated actual 2005/06 expenditure from the MOH budget was somewhat lower than US$16.88 million, i.e. US$14.12 million. - 100 - Table 4-10: Timor-Leste: Public Expenditure on Health, All Sources of Funding, 2000/01 to 2009-10 a/ (in US$ million at current prices) 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2005/06 2006/07 2007/08 2008/09 2009/10 Budget Budget Actual Actual Actual Actual Actual Estimated Projected Projected Projected Estimates Estimates Actual 1. Expenditures in Government Budget, 5.09 5.34 7.78 8.85 9.49 16.88 14.12 25.72 21.94 21.17 21.06 MOH 1.1. Recurrent Expenses 4.09 4.76 6.70 7.68 8.88 12.71 10.14 15.63 16.90 17.78 18.08 (i) Salary and Wages 2.73 3.03 3.03 3.85 3.69 3.80 3.91 (ii) Goods and Services 5.69 9.14 6.56 10.49 12.43 13.03 13.22 (iii) Minor Capital Expenses 0.46 0.55 0.55 1.29 0.78 0.95 0.95 1.2. Capital Development Expenses 1.00 0.58 1.08 1.17 0.61 4.17 3.98 10.09 5.03 3.39 2.97 (Infrastructure) 2. Expenditures Outside the Central Government Budget (Confirmed 15.21 16.71 13.32 15.85 18.78 27.38 27.38 26.00 12.52 9.82 5.68 Funding Only) b/ (i) Recurrent Expenses 14.36 14.23 11.20 13.85 16.05 22.86 22.86 20.10 8.44 5.74 1.95 (ii) Capital Expenses 0.85 2.48 2.12 2.00 2.74 4.53 4.53 5.90 4.08 4.08 3.73 3. Total Public Expenditure on Health 20.30 22.05 21.10 24.70 28.27 44.27 41.50 51.72 34.46 30.99 26.74 [(1) + (2)] (i) Recurrent Expenses 18.45 18.99 17.90 21.53 24.93 35.57 33.00 35.73 25.34 23.52 20.03 (ii) Capital Expenses 1.85 3.06 3.20 3.17 3.35 8.70 8.50 15.99 9.11 7.47 6.70 4. Government-Wide Combined Sources 432.00 463.00 379.00 305.00 195.00 247.40 247.40 452.30 319.60 288.30 270.2 Budget Expenditure 4.a. On-budget Government-wide 52.0 55.0 73.0 82.0 87.0 142.3 142.3 315.90 265.40 263.80 259.9 Expenditure 5. Total Public Expenditure on Health as 4.7 4.8 5.6 8.1 14.5 17.9 16.8 11.4 10.8 10.7 9.9 a Percentage of Government-Wide Combined Sources Budget Expenditure 5.a. MOH Expenditure as a Percentage of Government-wide on-budget 9.8 9.7 10.7 10.8 10.9 11.9 9.9 8.1 8.3 8.0 8.1 Expenditure 6. Annual rate of inflation (US$-based 3.0 0.0 10.0 4.0 4.0 2.5 2.5 6.0 2.5 2.5 CPI for Dili), % 7. Total Public Expenditure on Health at 20.30 22.05 19.18 21.59 23.00 34.91 32.73 38.5 24.6 21.8 2000/01 constant prices - 101 - Note: a/ Excludes health expenditure included in the budgets of government agencies other than the Ministry of Health . b/ "Projectized" expenditures funded by grants from bilateral and multilateral agencies. Sources: For 1999/00 to 2003/04, MOH, "Health Care: Priorities and Proposed Sector Investment Program", April 2005, page 43. For 2004/05 to 2009/10, The Democratic Republic of Timor-Leste 2005/06 and 2006/07 Budget Documents, Ministry of Planning and Finance. Inflation (CPI) projections for 2006/07 and 2007/08 are from IMF. Table 4-11: External Assistance for Timor-Leste's Health Sector, 2005/06 to 2009/10 (US$ 000 in current prices) Program Donor 2005/06 2006/07 2007/08 2008/09 2009/ 10 Policy, Planning, Research and Management - Health Sector Rehabilitation-Administration EU 116 121 - Health Sector Rehabilitation-Policy and Management EU 172 178 - Health Sector SIP Implementation EU 163 489 489 489 - Evidence and Information Policy WHO 616 - Health sector advisor WHO 105 180 180 180 75 Sub-total 1,172 969 669 669 75 Medical Education and Training - Community health care trainees Japan 141 - Masters’ degree in public health Japan 18 13 - Health Rehabilitation II-Training health workers - Health Rehabilitation II Supplement- Training health EU 281 297 workers - Health sector SIP human resource development EU 150 449 449 449 Sub-total EU 120 359 359 359 559 819 808 808 449 Communicable Diseases - HIV/AIDS II 15 - National strategic plan on HIV/AIDS UNDP 5 - HIV/AIDS prevention and control UNDP 700 - Malaria control program USA 1,800 1,500 750 - Dengue prevention USA 200 200 - Malaria control USA Global 766 766 - TB Control Fund Global 968 - HIV/AIDS I Fund 120 - Immunization Plus I UNICEF 396 - HIV/AIDS II UNICEF 150 Sub-total 5,119 2,466 750 Primary Health Care -Community health and empowerment Australia 137 - Health sector program initiative Australia 206 - Family and community health WHO 304 - Non-communicable disease and mental health WHO 63 - Expansion of health care program Japan 100 - Health sector SIP implementation service delivery - 102 - - Support for community health programs EU 128 385 385 385 - Public Health III USA 2,000 2,000 2,000 - Public Health V Thailand 4 - Rehabilitation of health clinics Thailand 215 Sub-total Ireland 41 2,993 2,592 2,385 385 Maternal Health - Maternal health care USA 375 375 - Improving maternal health Japan 440 Sub-total 815 375 Child Health - Child health USA 2,250 2,000 - Immunization and child health USA 1,000 1,000 1,200 1,250 Sub-total 3,250 3,000 1,200 1,250 Health and Nutrition - Nutrition Program WFP 4,400 4,400 Sub-total 4,400 4,400 Hospitals and Specialized Services - Health Sector Rehabilitation-hospitals EU 3,771 3,959 - Health Sector Rehabilitation II Supplement-hospitals - Health sector SIP capital fund EU 1,244 3,733 3,733 - Specialists service project EU 116 347 347 347 3,733 - Laboratory diagnosis of resistant microbes Australia 598 364 Sub-total Japan 14 4,498 5,914 4,080 4,080 3,733 Pharmaceuticals and Supplies - Health technology and pharmaceuticals WHO 305 - AMI assistance Dili Portugal 1,442 1,515 - Health Sector Rehabilitation II-supplies EU 1,950 2,047 - Health Sector Rehabilitation II Supplement-supplies - Health sector SIP pharmaceutical support EU 476 1,428 1,428 Sub-total EU 97 292 292 292 1,428 3,794 4,330 1,720 1,720 1,428 Health Promotion and Education - Integrated health and nutrition education program Australia 69 46 - Integrated health and nutrition education program EU 330 - Health education and promotion Emera Japan 83 185 Sub-total 482 231 Health, Unidentified - Health sector SIP implementation TA EU 301 902 902 902 Sub-total 301 902 902 902 Grand Total 27,382 25,998 12,514 9,814 5,684 Source: Democratic Republic of Timor-Leste, General Budget of the State 2006/07, Vol. I, pp. 110-11. - 103 - 19. Table 4.10 shows that MOH expenditure increased from $5.09 million in 2000/01 to $14.12 million in 2005/06 (estimated actual), and a further large increase to $25.72 million is budgeted for 2006/07. The large recent increase in on-budget health expenditure mirrors a large increase in overall, government-wide on-budget expenditure. As shown in Table 4-12, such expenditure rose from $87 million in 2004/05 to $142.3 million in 2005/06, and a further increase to $315.9 million is expected for 2006/07. These increases have been made possible by rising petroleum revenues. The trends in the overall Combined Sources government budget are shown in Table 4.12. 20. Table 4-10 also indicates that total public sector expenditure on health (on-budget plus off-budget) rose steadily from $20.30 million in 2000/01 to about $42 million in 2005/06, and based on current information it is expected to rise further to about $52 million in 2006/07. When corrected for inflation (as measured by changes in the CPI for Dili), total public expenditure on health at constant 2000/01 prices increased from $20.30 million in 2000/01 to $32.73 million in 2005/06, or by about 61 percent, and it would further increase by about 18 percent in 2006/07 relative to 2005/06. 21. The same table also shows government-wide combined sources expenditure figures, and health’s share of that total. Health’s share rose steadily from 4.7 percent in 2000/01 to 16.8 percent in 2005/06, but is projected to decline thereafter. On the other hand, the share of MOH in total (government-wide) on-budget expenditure fluctuated within a fairly narrow range of 9.7% to 10.9% between 2000/01 and 2005/06 and was estimated at 9.9% in 2005/06, but would decline to a range of 8.1-8.3% in the subsequent four years (Table 4-10). 22. In absolute terms, MOH expenditure is projected to increase very sharply between 2005/06 and 2006/07, from $14.12 million to $25.72 million, or an increase of about 82%, although actually implementing such a large expenditure increase in such a short time may tax the capacity of MOH.78 After 2006/07 MOH expenditure is expected to decline, due to lower projections of capital expenditure, as the ongoing upgrading of hospital infrastructure is completed. 23. The rise in the international price of oil in 2005 and 2006 and expanded production levels has led the government to sharply increase its estimate of what constitutes a sustainable level of on-budget expenditure. This is the annual level of expenditure the country could afford after making provision for the targeted capitalization of the Petroleum Fund, which has been designed along the lines of the Norwegian model. At the time of the preparation of the 2005/06 budget, the Government had estimated that the sustainable level of annual on-budget expenditure was about $150 million in 2005/06, rising to about $170 million in 2008/09. 79 However, the Government now estimates that the sustainable 2006/07 level of on-budget expenditure would be about $322 million, which is actually a bit higher than the actual amount budgeted for 2006/07 of $315.9 million. The figure of $322 million is the sum of $39.1 million of estimated revenue from 78 It is notable that in spite of such a large increase in the MOH’s budget its share in total government on-budget expenditure would decline between 2005/06 and 2006/07, as total government expenditure is expected to increase a whopping 122% (from $142.3 million to $315.9 million). 79 Ministry of Planning and Finance, Draft Budget Background Paper, April 2005, page7. - 104 - domestic taxes and fees plus $283 million of sustainable petroleum income (i.e., the 2006/07 sustainable level of withdrawal from the Petroleum Fund).80 24. The Government is projecting the sustainable level of annual on-budget expenditure to rise further to about $352 million by 2009/1081, comprising $48.9 million in domestic taxes and fees plus $303 million of sustainable withdrawal from the Petroleum Fund. In determining the latter amount, the Government has conservatively assumed that the price of oil would decline from their present high levels to $44 per barrel by 2010, in line with the price projections of the International Energy Agency, and it has only taken into account the expected yield from oil fields currently in production. Economic Classification of Public Health Expenditure82 25. Detailed data on the economic classification of MOH expenditure is shown in Table 4.13, for the years 2004/05 to 2006/07. Several points may be noted. The share of Capital Expenditure has risen from 6.1% in 2004/05 to 28.2% in 2005/06 and further to 39.2% in 2006/07. The escalation reflects the push to upgrade the GuidoValadares National Hospital in Dili and the referral hospitals, a project that would be completed in 2006/07. Payments to labor are reflected not only in the Salaries and Wages category, but also under Professional Services. Some ancillary services, notably catering and cleaning in hospitals, have been outsourced and the correspondent expenditures are reflected under Other Miscellaneous Services. 26. The table also highlights the sources of the large increase in Recurrent Expenditure between 2005/06 and 2006/07. Most of the increase is accounted for by: (i) Salaries (+$800,000), reflecting actual and expected recruiting by MOH; (ii) Training and Workshops (+$1,007,000), including $376,000 for scholarships for medical students studying in Cuba; (iii) Vehicle Operation (+$1,180,000), reflecting a more realistic provision than in previous years as well as an increase in the number of vehicles (due to donations of UN vehicles); (iv) Fuel for Generators (+$350,000), reflecting expenses related to the new/upgraded hospitals and generators provided to housing for Cuban doctors; (v) Professional Services (+$597,000), on account of a larger number of Cuban doctors working in Timor-Leste; and (vi) “Minor Capital� (+$767,000), reflecting renovations to new buildings, and solar panels and generators for CHCs and Health Posts. 80 See General Budget of the State 2006/07, Budget Document No. 1, pp. 5 and 8. 81 Comprising $48.9 million of domestic taxes and fees and $303 million of sustainable withdrawal from the Petroleum Fund. 82 The “economic classification� of government expenditure (also sometimes referred to as the “object classification�) is the classification of expenditure into types of inputs purchased. - 105 - Table 4-12: Combined Sources Government Budget, 2000/01 to 2009/10 (US$ Million at Current Prices) 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 Est. Est. Est. Est. Est. Budget Projected Projected Projected actual actual actual actual actual est. 1. Government Revenue 28 36 54 65 284 466.2 728.9 1,023.6 1,117 1,086.1 1.1. Domestic Taxes and Fees 13.1 10.8 29.5 31.8 243 36.7 39.1 42.9 46.1 48.9 1.2. Petroleum Revenue 14.1 20.5 19.3 27.5 33 418.8 683.3 972.4 1,062.1 1,027.6 1.3. Autonomous Agency Revenues 1 4 5 6 8 10.7 6.5 8.3 8.8 9.6 2. Direct Budget Support from Development Partners 33 35.7 31 10.3 10.3 10 0 0 2a. Sub-total: Government Budget Revenues [(1) + (2)] 28 36 87 100.7 315 476.5 739.2 1,033.6 1,117 1,086.1 3. Other Funding from Development Partners a/ 380 408 306 223 108 105.1 136.4 54.3 24.5 10.3 4. Total Combined Sources Budget Revenues [(1) + (2) + (3)] 408 444 393 323.7 423 581.6 875.6 1,087.9 1,141.5 1,096.4 5. Government Budget Expenditure 52 55 73 82 87 142.3 315.9 265.4 263.8 259.8 Of which Recurrent Expenditure 43 46 62 67 73 103.6 196.2 177.1 184.3 195.7 Of which Capital Expenditure 9 9 11 15 14 38.7 119.7 88.3 79.5 64.1 6. Expenditure Outside the Government Budget b/ 380 408 306 223 108 105.1 136.4 54.3 24.5 10.4 7. Total Combined Sources Budget Expenditure [(5) + (6)] 432 463 379 305 195 247.4 452.3 319.7 288.3 270.2 8. Fiscal Balance [(4) - (7)] -24 -19 14 18.7 228 334.2 423.3 768.2 853.2 826.2 Note: a/ Grants from bilateral and multilateral Development Partners for the funding of public sector activities outside the Government Budget. b/ "Projectized" expenditure funded by grants from bilateral and multilateral agencies. Sources: 2006/07 Budget Document, Ministry of Planning and Finance. IMF Staff Report for the 2004 Article IV Consultation, June 2004, for budget figures before 2004/05. - 106 - Table 4-13: Economic Classification of MOH Expenditure, 2004/05 to 2006/07 (US$ 000 at Current Prices) Categories of Expenditure 2004/05 2005/06 2006/07 (Economic Classification) Est. Actual Est. Actual Budget est. 1. Salaries and Wages 3,013 (30.7%) 3,025 (21.4%) 3,851 (15.0%) 1.1. Salaries 2,957 (30.2%) 2,963 (21.0%) 3,763 (14.6%) 1.2. Overtime/Allowances 56 (0.6%) 62 (0.4%) 89 (0.3%) 2. Goods and Services 5,714 (58.3%) 6,560 (46.4%) 10494 (40.8%) 2.1. Local Travel 62 (0.6%) 109 (0.8%) (40.8%) 145 (0.6%) 2.2. Overseas Travel 0 0 0 2.3. Training and Workshops 176 (1.8%) 522 (3.7%) 1,529 (5.9%) 2.4. Utilities 245 (2.5%) 244 (1.7%) 271 (1.1%) 2.5. Rental of Property 10 (0.1%) 17 (0.1%) 100 (0.4%) 2.6. Vehicle Operation, Fuel 230 (2.3%) 307 (2.2%) 1,186 (4.6%) 2.7. Vehicle Operation, Maintenance 298 (3.0%) 372 (2.6%) 673 (2.6%) 2.8. Vehicle Rental, Insurance and Other Services 0 0 0 2.9. Office Stationery and Supplies 170 (1.7%) 158 (1.1%) 252 (1.0%) 2.10. Pharmaceuticals and Other Supplies 982 (10%) 942 (6.7%) 1,161 (4.5%) 2.11. Fuel for Generators 107 (1.1%) 168 (1.2%) 518 (2.0%) 2.12. Maintenance of Equipment and Buildings 307 (3.1%) 232 (1.6%) 292 (1.1%) 2.13. Other Operational Expenses 111 (1.1%) 125 (0.9%) 161 (0.6%) 2.14. Professional Services 1,859 (19.0%) 2,057 (14.6%) 2,654 (10.3%) 2.15. Translation Services 18 (0.2%) 0 28 (0.1%) 2.16. Other Misc. Services (catering, cleaning, etc.) 1,139 (11.6%) 1,307 (9.3%) 1,281 (5.0%) 2.17. Grants/Subsidies 0 0 243 (0.9%) 3. Equipment, Furniture and Vehicles (Minor Capital�) 479 (4.9%) 522 (3.7%) 1,289 (5.0%) 3.1. Security Equipment 6 (0.1%) 7 (0.1%) 28 (0.1%) 3.2. Information Technology Equipment 32 (0.3%) 0 105 (0.4%) 3.3. Communications Equipment 64 (0.7%) 120 (0.8%) 99 (0.4%) 3.4. Generators 35 (0.4%) 32 (0.2%) 438 (1.7%) 3.5. Water Equipment 38 (0.4%) 44 (0.3%) 84 (0.3%) 3.6. Office Equipment 74 (0.8%) 76 (0.5%) 116 (0.5%) 3.7. Other Miscellaneous Equipment 130 (1.3%) 273 (1.9%) 239 (0.9%) 3.8. Furniture and Fittings 100 (1.0%) 0 93 (0.4%) 3.9. Purchase of Vehicles 0 0 87 (0.3%) Sub-Total Recurrent (1+2+3) 9,206 (93.9%) 10138 (71.8%) 15634 (60.8%) 4. Capital Development (Infrastructure) 600 (6.1%) 3988 (28.2%) 10,087 (39.2%) 9,806 14126 25,721 Total Ministry of Health Expenditure (100%) (100%) (100%) Source: The Democratic Republic of Timor-Leste 2005/06 Budget Document, Volume 2, page 190, for 2004/05 expenditure figures; MOH staff for 2005/06 and 2006/07 figures. 27. Concerning off-budget health expenditure, similarly disaggregated data by heads and sub- heads of the economic classification do not seem to be available from published material. A broad classification between “Recurrent� and “Capital� expenditures is shown in Table 4.10. It can be seen that most of the off-budget expenditure has been in the form of recurrent expenditure - e.g., about 85 percent in 2004/05. - 107 - Classification of Ministry of Health Expenditure by Program (Functional Classification of Expenditure) 28. Starting with the 2004/05 fiscal year, the MOH’s budget has been organized into 24 programs, namely: the Office of the Minister of Health and that of the Vice-Minister; Central Services; the National Hospital Guido Valadares at Dili; the five hospitals at Bacau, Maliana, Maubisse, Oe-cussi, and Suai; the Institute of Health Science; the National Laboratory; and the 13 District Health Services. Table 4.14 presents a summarized version of the distribution of MOH’s budget by program for the years 2004/05 to 2006/07. Table 4-14: Ministry of Health Expenditure on Hospitals, District Health Services, and Other Programs, 2004/05 to 2006/07 (US$ 000 at Current Prices) 2004/05 2005/06 2006/07 % of Budget % of Budget % of Budget total est. total est. total Program est. 1. Hospitals a/ 3,797 38.7 7,447 52.7 13,120 51.0 1.1. Recurrent 3,797 41.2 3,947 38.9 5,030 32.2 1.2. Capital Development - 0 3,500 87.8 8,090 80.2 2. District Health Services b/ 4,608 47.0 4,870 34.5 8,865 34.5 1.1. Recurrent 4,008 43.5 4,433 43.7 6,960 44.5 1.2. Capital Development 600 100.0 437 10.9 1,905 18.9 3. Other Programs c/ 1,401 14.3 1,809 12.8 3,736 14.5 1.1. Recurrent 1,401 15.2 1,757 17.3 3,640 23.3 1.2. Capital Development - 0 52 1.3 95 0.9 4. Total MOH Expenditure 9,806 100 14,126 100.0 25,721 100.0 4.1. Recurrent 9,206 100 10,138 100.0 15,630 100.0 4.2. Capital Development 600 100 3,988 100.0 10,090 100.0 Note: a/ Includes the National Hospital Guido Valadares and the five referral hospitals at Bacau, Maliana, Maubisse, Oe-cussi, and Suai. b/ Includes all 13 districts. c/ Includes the Office of the Minister of Health, the Office of the Vice-Minister of Health, Central Services, the Institute of Health Science, and the National Laboratory. Source: MOH budget documents. 108 29. There is a standing agreement between the government and donors that the share of hospital recurrent expenditure would not exceed 40 percent of total MOH recurrent expenditure. Table 4.14 shows that for 2004/05 the share of hospital recurrent expenditure was slightly above the cap, at 41.2 percent, but it has since fallen and would be well within the limit in 2006/07 (at 32.3% as per budget estimates). 30. Data on the functional classification of off-budget health expenditure are available for 2004/05 from the Health SIP report. 83 Table 4.15 presents these data together with the corresponding on-budget expenditure figures. The functional classification in Table 4.15 is the same as in Table 4.14, namely, including the three broad categories of “Hospitals�, “District Health Services�, and “Other Programs�. Table 4-15: Functional Classification of Public Sector Health Expenditure, 2004/05 (US$ 000 at Current Prices) 2004/05 Program Budget % of total Estimates 1. Hospitals a/ 4,778 16.7 1.1. MOH Budget 3,797 13.3 1.2. Ongoing Donor Programs (off-budget) 981 3.4 2. District Health Services b/ 20,714 72.5 2.1. MOH Budget 4,608 16.1 2.2 Ongoing Donor Programs (off-budget) b/ 16,106 56.3 3. Other Programs c/ 3,096 10.8 3.1. MOH Budget 1,401 4.9 3.2. Ongoing Donor Programs (off-budget) 1,695 5.9 4. Total MOH Expenditure 28,588 100 4.1. MOH Budget 9,806 34.3 4.2. Ongoing Donor Programs (off-budget) 18,782 65.7 Note: a/ Includes the National Hospital Guido Valadares and the five referral hospitals at Bacau, Maliana, Maubisse, Oe-cussi, and Suai. b/ Includes all 13 districts. c/ Includes the Office of the Minister of Health, the Office of the Vice-Minister of Health, Central Services, the Institute of Health Science, and the National Laboratory. Sources: For MOH Budget figures, Table 4.15; for Ongoing Donor Programs, from Health Care Priorities and Proposed Sector Investment Program, MOH, April 2005. 31. Table 4.15 shows that in 2004/05 the bulk of off-budget expenditure was in the “District Health Services� category ($16.106 million out of a total of $18.782 million), with the rest mostly pertaining to “Other Programs� ($1.695 million), and a small share of expenditures on “Hospitals� ($0.981 million). Of the total public sector health expenditure (on-budget plus off- budget) of $28.588 million, 72.5 percent was for the “District Health Services�, 16.7 percent for “Hospitals�, and 10.8 percent for “Other Programs�—a distribution that clearly shows the commitment to primary health care. Such commitment continues, although Table 4-14 shows a jump in the share of Hospitals within the MOH budget from 38.7 percent in 2004/05 to 52.7 percent in 2005/06 and 51.0 percent in 2006/07, reflecting the peak of expenditure for the hospital upgrading effort. 83 MOH, Health Care Priorities and Proposed Sector Investment Program, April 2005. The data in this report are not however presented according to the new detailed classification of the MOH’s budget (Table 3.6). 109 Per Capita MOH Expenditure by District, 2005/06 32. Table 4.16 below presents data on expenditure by the District Health Services on a per capita basis, including in addition to allocations in the MOH budget off-budget allocations for drugs funded from the TFET for 2006/07. The range is from a low of $4.9 in Dili to a high of $19.0 in Manatuto, or a factor of about four. If only recurrent expenditure is considered, the range is about the same --from $3.8 in Dili to $15.3 in Manatuto. 33. But while Dili ranks low in terms of per capita expenditure by the District Health Services, it is clear from Table 4.16 that from a broader perspective Dili district fares quite well: it ranks highest among all districts in terms of the number of government health professionals per 100,000 population (355) as well as in terms of the number of annual per capita outpatient visits to MOH facilities (3.0). These indicators undoubtedly reflect the presence in Dili of the National Hospital, which until recently was open to all outpatients without the need for a referral. 84 34. Table 4.16 also shows, for each district, expenditure per CHC/HP (i.e., district expenditure divided by the number of CHCs and health posts). It can be seen that the inter- district dispersion is lower here as compared with per capita expenditure: expenditure ranges from a low of $21,800 in Bobonaro to a high of $41,200 in Dili district, or a factor of about 1.9. Thus, as in many other countries, the geographical distribution of health expenditure is more closely related to the number of facilities than to population. 35. An additional complication in assessing inter-district equity in the distribution of public expenditure on health is the fact that, as already discussed, a large part of public expenditure on health has been off-budget. Table 4.16 does include the off-budget expenditure on drugs funded from TFET, but other categories of off-budget expenditures by the District Health Services are not included in the table due to lack of detailed information. 36. Finally, it should also be remembered that the figures in Table 4.16 are limited to the public sector. For a clearer picture of differences across districts in access to health services, the activities of NGOs and other (qualified) health service providers would also have to be taken into account. 84 In addition to Dili, the population of several other districts benefit from the presence of hospitals: Ainaro (Maubisse Hospital, just re-starting operations in 2006/07 after upgrading), Baucau (Baucau Hospital), Bobonaro (Maliana Hospital), Covalima (Suai Hospital), and Oecusse (Oecusse Hospital). However, Table 4.16 shows that such benefits are not reflected in either above- average numbers of per capita outpatient visits to MOH facilities in these districts or above-average number of government health professionals per 100,000 population (with the exception of Baucau). 110 Table 4-16: District Health Services: Per Capita Expenditure by District, 2006/07 2006/07 District Per capita Govt. Annual per Expendi- Budget populat expenditure health capita ture per est. ($000) ion, est. ($) profession- outpatient CHC/HP c/ a/ als per visits to ($000) d/ District Health Service 100,000 MOH b/ Aileu 639 38,365 16.6 146 facilities b/ 2.6 (14) -Recurrent 471 12.3 33.6 -Capital 168 Ainaro 609 55,774 10.9 125 1.4 (19) -Recurrent 504 9.0 26.5 -Capital 105 Baucau 846 108,754 7.8 222 2.0 (28) -Recurrent 654 6.0 23.4 -Capital 192 Bobonaro 772 85,680 9.0 137 1.3 (24) -Recurrent 524 6.1 21.8 -Capital 248 Covalima 868 58,179 14.9 170 2.1 (17) -Recurrent 514 8.8 30.2 -Capital 354 Dili 851 174,488 4.9 355 3.0 (16) -Recurrent 659 3.8 41.2 -Capital 192 Ermera 856 107,296 8.0 72 1.5 (21) -Recurrent 678 6.3 32.3 -Capital 178 Lautem 877 59,751 13.0 153 1.7 (24) -Recurrent 643 10.0 26.8 -Capital 234 Liquicia 589 57,260 10.3 105 1.5 (17) -Recurrent 486 7.7 28.6 -Capital 103 Manatuto 882 40,124 19.0 197 2.4 (22) -Recurrent 647 15.3 29.4 -Capital 235 Manufahi 714 46,004 13.2 181 2.2 (19) -Recurrent 587 11.2 30.9 -Capital 127 Oecusse 676 68,840 8.4 140 1.0 (12) -Recurrent 484 6.2 40.3 -Capital 192 Viqueque 1,007 69,091 11.7 157 2.8 (22) -Recurrent 776 9.5 35.3 -Capital 231 Total All Districts 10,329 9.0 186 2.0 (255) -Recurrent 8,424 7.2 33.0 -Capital Development 1,905 (Infrastructure) Total Population Timor-Leste 969,606 Notes: a/ District population figures are MOH estimates based on the 2004 Census. b/ Data correspond to 2004/05. c/ MOH budget plus TFET- funded expenditure on drugs for the District Health Services. d/ The numbers in parenthesis are the numbers of health posts plus CHCs. 111 Making Efficient Use of Expanding Budgets 37. It was noted earlier in this chapter that, because of increases in the price of oil and expanded production levels, there has been a large increase in the estimated sustainable on- budget total government expenditure and by implication of the sustainable MOH budget as well. This upward trend in the sustainable level of government expenditure may continue as new oil fields are brought under production. The expansion of its resource envelope presents a challenge to the MOH to spend the additional resources efficiently. 38. In this context, the MOH’s initiative to formulate a Medium-Term Strategic Plan is a welcome development. It is critical that the Medium-Term Strategic Plan include a Medium- Term Expenditure Framework (MTEF) with a realistic estimated resource envelope (over the period of the Plan), so that trade-offs among various programs and/or categories of expenditure can be assessed. 39. The MTEF should encompass not only the projected expenditures to be funded from the MOH budget, but also the projected off-budget expenditures. As already noted, while off-budget public sector expenditures on health appear to have peaked, they are still substantial and likely to remain so for at least the medium term. This need to incorporate off-budget expenditures in the MTEF implies that the construction of the MTEF has to be a real cooperative effort between the MOH and its Development Partners. In this sense, construction of the MTEF would also serve the purpose of facilitating aid coordination. 40. The construction of the MTEF could also pave the way towards a unified public sector (annual) budget for health, which would integrate all expenditures and all sources of funding. The current fragmentation of the budgeting of public sector expenditures in the health sector is an impediment to an adequate assessment of various aspects of equity and efficiency in those expenditures --such as the assessment of inequities in expenditure across districts and the examination of the balance among various categories of expenditure (salaries, drugs, etc.) in the economic classification of expenditures. 112 CHAPTER V: THE CHALLENGES AND RISKS MOVING FORWARD Introduction 1. The discussion in Chapter I on the health status of the population of Timor-Leste clearly indicates that, while progress has been made, there is ample scope for improving health outcomes in the country. Infant and child mortality and maternal mortality rates remain unacceptably high. An extremely high Total Fertility Rate (TFR) and the consequent high population growth rate coupled with extensive poor nutrition exacerbate these mortality rates. Communicable diseases, especially malaria and tuberculosis, are still major sources of morbidity and premature mortality. Data on common non-communicable diseases such as hypertension, cardiovascular disease and diabetes are not available, but these diseases are also in all probability emerging causes of morbidity and premature mortality. HIV/AIDS does not appear to be a major problem at this time but, given the experience with this disease in other countries in the region, continued vigilance is indicated. 2. This final chapter of the report suggests some broad strategies the Government may wish to explore in order to effect further improvements in health outcomes—i.e. to further reduce the burden of morbidity and premature mortality across the entire population – but particularly for mothers and children. The intention is not to provide a detailed blueprint for what the Government in general and MOH in particular ought to do in the coming years, but rather to point to some of the directions in which the health-related activities of the public sector could evolve and support private and community action. 3. In this context it is first suggested that a key challenge for MOH is to translate the formal policy framework into consistent and sound action on the ground. Second, it identifies demand side issues including encouragement of households to be more active participants in their own health by improving knowledge and improved health seeking behavior. Opportunities to improve incentives for seeking health care are also discussed. Third, options to improve utilization of health care services through supply-side measures are proposed – including suggestions on how quality of service may be improved, barriers to access to health care facilities may reduced and how multi-dimensional aspects of human resource issues may be addressed. Fourth, options to improve management, planning and budgeting including donor coordination are discussed. Finally, the importance of cross-sector interventions, knowledge gaps and how the report may be used are discussed. Health Policy Development and Implementation 4. A key challenge for the MOH is to translate the formal policy framework into consistent and sound action on the ground and to avoid a number of risks by staying the course and building capacity to implement priority program. 5. In the relatively short time since the troubled separation of Timor Leste from Indonesia the MOH has formulated an impressive set of health policies and a sound vision for the future development of the health sector. The Health Policy Framework defines the mission of the MOH and sets out principles to guide the development of Government health services. 113 In this regard, as discussed, MOH has decided a selected set of “basic� services be developed and given priority over less essential services; expenditure priority has been given to basic health services by deliberately constraining hospital expenditures; District Health Management Teams are now primarily responsible for the management of district health services and implementation of district health plans; priority has been given to access to basic and specific emergency services – particularly emergency obstetric care; basic services will be free; an essential drugs, list calibrated for each facility type/level of service would underpin public sector drug supply policy, an autonomous medical supply system would be developed and private pharmaceutical services would be licensed. The Basic Package of Services defines the minimum range of services the public could expect at each level of Government health facility. The report, as discussed below, confirms the appropriateness of the basic package. A complementary Hospital Service Package is under development with the objective of setting priorities for hospital services and managing demand for non-basic services and the Private Sector Policy establishes that the private sector should compliment and supplement Government services. 6. The report makes suggestions on what should be given priority in the provision through the Governments’ health system. The most important criteria for deciding priorities on what services to deliver is their relative impact on the populations’ health status (or in the case of emerging diseases such as HIV/AIDS on their potential impact on health status without appropriate health services). Thus priorities for the basic package should be selected based on the magnitude of the problem (both its severity and prevalence); the ability of know interventions to redress the health problem and both the economic feasibility of the intervention (i.e. its cost- effectiveness) and administrative feasibility. 7. The following four challenges and risks to the capacity of MOH to consistently implement its proposed core program have been identified in the report.  First, the laudable focus on key program outcomes agreed at both the national and district level are not consistently and aggressively monitored and evaluated by national MOH and district authorities. This could be done by building on the district planning process and emphasizing the monitoring and evaluation of progress towards agreed targets in a manner which leads to improved understanding of what works and does not work in Timor Leste conditions and across districts. A regular forum in which district successes and problems are jointly discussed would assist this process. Decisive support from MOH headquarters to support district managers to take corrective action where program outcomes are lagging should become a core role of headquarters. This would help create a learning environment clearly linked to on the ground experience in achieving desired outcomes.  Second, there are many challenges to the capacity of MOH to consistently “stay the course� with its core programs as reflected in the policy framework (which are known to reduce the disease burden). It is critical that the focus of MOH, with is limited institutional and financial capacity, is not diverted by pressures to keep adding to the list of priority programs to be implemented until targets and objectives are being consistently met. The report also suggests specific ways the MOH might proactively 114 manage demand for expensive hospital services (non-basic services) – a potential major risk to sustaining expenditure priorities on core public health programs.  Third, health workers need to accept (“buy into�) the priority being given to the basic package of services and critically each health worker cadre needs to fully understand and accept their core role in service delivery. To this end there is a specific need to consistently reinforce the collective understanding of the revised roles of each of the health cadres in the new “multi-skilled� paradigm as skill mixes change, staff numbers expand and strategic priorities of the basic package policy are implemented. This will be enhanced by participatory consultation on health priorities and job delineation, appropriate pre-service and in-service training, and career planning -- including attention to remuneration arrangements and incentives (e.g., access to refresher training, remote area allowances and incentives and housing allowances/policies). The recent decision to radically expand the supply of doctors means that a range of current policies concerning the role of each cadre in service delivery (which essentially was developed on the assumption of a non-doctor model of primary health care delivery) need to be modified.  Fourth, development partners pose a threat to MOH capacity to sustain its limited implementation capacity on core programs by: (a) making recommendations to adopt new programs not included in the core program; and (b) by insisting large parts of their programs are managed off-budget and thus creating incoherent and parallel chains of accountability (see below). To this end, the report strongly supports the efforts of the MOH to develop a five year Health Strategic Plan within a medium term public expenditure framework and which can form the basis for the design of development partner programs which increasingly should use government systems, accountabilities and budgeting framework. Demand Side Health System Issues 8. There are significant opportunities for interventions designed to improve health outcomes by: (a) systematically encouraging households to be more active participants in looking after their own health by encouraging more healthy behaviors (including improved health service seeking behaviors for some specific conditions through health information and education programs); and (b) providing incentives to overcome barriers to increased demand for health services. The key challenge is to develop a strong population based health education and healthy behavior change program sensitive to the culture, belief systems and languages of Timor Leste. The key risks include an inadequately designed health information and education program and inadequate attention to the design details and institutional capacity development required to implement demand side incentive programs. 9. Making Households More Active and Better Informed Participants through Health Education There has been a tendency in Timor-Leste, as in many other countries, to view households mostly as passive recipients of health services. But experience from around the world indicates that having households that are more active participants in looking after the health of their own members is an important factor in improving health outcomes. The impact of this factor is stronger the better informed the households are about common health conditions, 115 the interventions that households themselves can carry out when faced with those conditions, and the services available from health care facilities within their reach. Thus while health-related behaviors and practices by households are currently problematic, they are amenable to change through health education and promotion interventions. As discussed in Chapter II, the evidence in Timor-Leste suggests that many households are not adequately informed when it comes to health. The deficiencies observed in the feeding of young children are a case in point. The alarmingly low and limited knowledge about HIV/AIDS is another. It was also noted that there is evidence from the qualitative study carried out for this review that many clients may underestimate certain health threats, such as malaria and those conditions related to pregnancy and childbirth, which in turn tends to result in under-utilization of some basic health services. The widespread reliance on traditional healers is another aspect of the gaps in households’ knowledge about health -- the qualitative study found that households move fluidly among traditional and modern providers.85 10. Many of the micro-policies emphasize the importance of health education and promotion at community level. However, the report raises questions about whether this emphasis gets reflected in practice. There is hence a need to ensure effective health education and promotion activities at community level. There are no simple answers to how this can be achieved. Different options should be explored, including strengthened outreach, appropriate health messages during service consultations, community health workers (compensated and volunteers), expanded role of NGOs and other civil society groups. The potential role of health education in schools and road safety and driver education should not be underestimated. These options will have to be piloted and evaluated, so that the implementation of key community level interventions can be strengthened over time. 11. In line with the above, it is suggested that the MOH give a higher priority to reaching households with key health-related messages. Some of these messages could focus on actions the households themselves can take to prevent disease from affecting their members, such as good basic hygiene (e.g. regular hand-washing) and child feeding practices. The importance of expanding the level of knowledge about HIV/AIDS and how it can be prevented remains a key and urgent challenge because of the potential impact the disease can have on the population. Another set of messages could address the beliefs and practices of households in relation to various health conditions and the related utilization of health care services. Taken together, both sets of messages would aim at turning households into better informed and more proactive participants in all matters related to the health of their own members. 12. Expanding the Utilization of Basic Health Care Services Through Demand-Side Incentive Measures. As discussed, inadequate knowledge at the household level would seem to play a key role in households’ failure to seek services from qualified providers when faced with certain health conditions. It was in part because of this finding that it was suggested above that the MOH place a greater emphasis on reaching out to households with health education. But many respondents in both the DHS 2003 and the qualitative study pointed to financial constraints as another reason for not using the available services – particularly for the poor.86 85 DHS 2003 did not inquire about use of the services of traditional health care practitioners. This matter came up frequently, however, during the interviews and focus groups for the qualitative study. 86 Government primary health care facilities do not charge fees, but there are other costs associated with visiting the facilities, such as transportation costs, the cost of medicines when they have to be purchased elsewhere, and lost earnings. Moreover, for 116 13. The latter finding suggests that one of the new strategies the Government/MOH may wish to explore is instituting some sort of demand-side financing scheme. For example, a number of countries have been experimenting with conditional cash transfers, whereby the Government pays a subsidy to households against the households’ obligation to engage in certain behaviors, such as sending their children to school –or, in the case of health, to make use of certain basic health services which are currently under-utilized (e.g., immunization, anti-natal visits and child growth monitoring and counseling). There are complex design issues that would need to be addressed in introducing a scheme of this nature, such as the degree and mechanics of targeting (if the subsidy were not a universal one to which all households are entitled). The experiences of other low- and middle-income countries which have experimented with demand- side financing schemes should be studied prior to piloting any such scheme in Timor-Leste. Major lessons with Conditional Cash Transfers include the need for strong and transparent administrative capacity, clear and simple rules, and an adequate health system to supply services in response to demand – a failure on the supply side would bring the mechanism/program into disrepute. We would recommend piloting and evaluating small scale interventions to learn from experience prior to scaling up such programs. Expanding the Utilization of Basic Health Care Services through Supply-Side Measures 14. Three main supply-side challenges confront the MOH as it attempts to expand the utilization of basic health care services. First there is the challenge of improving the technical quality of services provided. Second, while about 80 percent of the population lives within two hours of a health facility (a major achievement), a significant proportion of the population faces access issues – importantly a key additional challenge is that the cost of getting to facilities, particularly for the poor, appears to be a significant barrier. Third, there is a challenge of ensuring adequate staff are deployed and are present at service delivery points; that treatment of patients by staff is friendly and respectful; that the skill mix in each facility is appropriate and that role of the now rapidly cadre of doctors is fully agreed – in particular that their role vis-a - vis health facility managers (typically a [male] nurse) is determined and agreed. Expanding utilization of and access to core services, given the documented under-utilization of these services by the poor, would be central plank to any poverty reduction strategy – a key objective of the National Development Strategy and the Health Policy Framework. 15. Improving the Technical Quality of Health Care Services. A key challenge to improving the utilization of services is to improve the technical quality of services provided. There is significant evidence that the technical quality of some basic health services does not meet minimum internationally accepted standards, thus detracting from the health benefits patients obtain when they do use the services. A case in point is antenatal care. As noted in Chapter II, of all the women who received antenatal care in 2003, only 44 percent received prenatal iron supplements; just 35 percent had their blood pressure measured; and a very low 25 percent were informed about signs of complications (DHS 2003). 16. More information would be required in order to determine the extent of the problem. However, now that most of the basic health care infrastructure is in place, the time seems right some households the nearest government facility may be too far, while the fees of private clinics closer to home may be out of reach financially. The available data on household expenditures on health goods and services was reviewed in Chapter IV. 117 for MOH to start a concerted effort to develop a system to monitor and upgrade the quality of health care services, in both its technical and non-technical aspects, starting with services in the public sector. Spelling out the details of a suitable quality assurance program for Timor-Leste exceeds the scope of this report, but work on such a program is suggested here as one of the desirable directions of change/development in MOH’s work. In-service training and emphasizing the role of health facility managers (and doctors) in helping ensure technical quality may be a useful starting point. The dramatic increase of the number of doctors in the system and decisions of how they could best be used should provide part of the solution to these quality issues. 17. The discussion in Chapters I and II also indicates that basic health services are not reaching large segments of the relevant population groups. The discussion mainly centered on maternal and child health care services, for which nationally representative survey data are available. It is likely, however, that a similar situation prevails with regard to other basic services such as those relating to major infectious and non-communicable diseases. In this regard the reported, but unsystematically documented scale of TB, is clearly a major challenge for the MOH (WHO estimates an incidence of 556/100,000 population) – most of which is not treated. Coupled with the low level of knowledge about HIV/AIDS and its prevention, given the susceptibility of TB carriers to HIV, it is clear that treatment of TB and prevention of HIV are a higher priority than at present. 18. In maternal and child health care, recapitulating the discussion in Chapters I and II, specific data exist to document the under-utilization of the following categories of basic health services, which are critical for the improvement of health outcomes in women and children:  Childhood immunization coverage (at less than 60% Chapter II, paras. 2-3), which if improved could also be used to expand vitamin A supplementation in under-5 children which is reported at only one third (Chapter I, para 44).  Post-natal care (a very low 12% neonate care coverage Chapter II, para. 4).  Treatment of acute respiratory infections, fever and diarrhea in under-5 children (about 35% don’t seek treatment Chapter II, paras. 5-6).  Malaria prevention and treatment -- Use of insecticide-impregnated bed nets by under-5 children (about 50% children sleep under bed-nets but only 8% of these bed- nets are treated Chapter I, para. 18) and only 18% of those with fever take appropriate presumptive anti-malarial drugs (Chapter II para 7).  Family planning services for better spacing, which would contribute to reducing both infant mortality and maternal mortality (Only 10% women 15-49 use contraceptives Chapter I, paras. 8, 29, and 32).  Antenatal care (40% of women receive no antenatal care Chapter II, paras. 8-9), including provision of iron supplementation (one-third of women 15 – 49 years anemic Chapter I, paras. 41-42). 118  Attendance at deliveries by skilled attendants (only 18% are attended by trained staff Chapter II, para. 10).  Cesarean sections and other emergency maternity care have low recorded rates (Chapter II, para. 11).  Post-partum care (almost 90% mothers had no post-partum care Chapter II, para. 12).  Nutrition education (less than 25% of children exclusively breast fed by 5 months of age Chapter I, paras. 50-51). 19. Moreover, a further challenge for the MOH, is that the analysis in this report has shown that under-utilization of many of these essential services is more widespread among poor households which is relatively large share of the total in Timor Leste. Thus expanding utilization of these services needs to be an important aspect of any comprehensive poverty alleviation drive in Timor-Leste. 20. Access to Health Care Facilities: Distance/Travel Time and Service Delivery Design. The terrain and scattered distribution of the population, often in small villages, provides a challenge for ensuring cost-effective access and service delivery design87. On the supply side, difficulty of access -- distance to the nearest health care facility and difficulty in securing transportation -- was often mentioned by respondents as reasons for not seeking services from the formal health care sector, especially by those respondents from rural areas. Nevertheless, DHS data (2003) showed that, nationwide, 81.7% of all households were within two hours’ travel time from the nearest health facility in that year, with the remaining 18.3% being further away in terms of travel time (see Table 2.5, Chapter II). 88 Region-wise, the percentage of households outside the two hours’ travel time range was 6.8% in Urban Areas, 20.9% in Rural Central, 22.2% in Rural East, and 24.2% in Rural West. 21. How best to improve access further is a problem that needs to be analyzed locally, and should be an important consideration in the preparation of district (and sub-district) plans. Increasing the number of health care facilities is one of the options, but there is a trade-off between ease of access in terms of distance and efficiency in terms of facility utilization rates and staff workloads. Beyond a point it would be difficult to justify adding more facilities. Hence the option of establishing new facilities needs to be considered together with three other options to facilitate access: (i) increased outreach activities from existing health facilities; (ii) piloting and evaluating the proposed community health worker program; and (iii) measures to improve transportation and/or reduce transportation costs for those seeking care. 87 There is antidotal evidence that many Timorese are moving back to their ancestral homes and away from centralized villages encouraged by the Indonesian administration which if extensive will add to the scattered nature of the population living in areas without easy access to services. The recent conflicts have led to very significant population movements – while a significant proportion of these recently displaced groups will eventually return to their homes a significant number may not. This will need to be monitored. 88 This means that households could reach the nearest health care facility within two hours by using the fastest available means of transportation, or alternatively by walking if no other means of transportation was available. “Facility� means either a government or a private facility. 119 22. Furthermore, in those cases where establishing new health care facilities appear to be the best alternative, the MOH should consider the option of contracting out services to private providers who would establish and run the facilities, as opposed to creating additional in-house capacity for service provision. Contracting out would generally provide MOH with more flexibility in the deployment of resources and a greater scope for imposing discipline and demanding results. 23. The lack of an extensive and functioning road network, as noted, is a challenge in many areas. On the other hand a very large proportion of the population, including mothers and children, regularly visit regional and district markets to both buy and sell food and other supplies (and to attend church services). If the MOH began providing key services (immunization, anti- natal and post natal screening, child health clinics etc.), even in out-reach mode without permanent facilities, or in facilities provided by the community, accessibility to key services could be dramatically increased and just as importantly the costs of accessing services could be reduced. This would require some changes in how staff deployment is managed – including days worked, when and where. Strategies for this should be built into sub-district and district plans in consultation with local authorities and communities. 24. Human Resources Issues: Provider Behavior, Skill Mix, Staff Numbers and Incentives. All services and their quality critically depend on health professions, their skills, commitment and accountability to clients, their work load and the incentives – financial and non-financial -- they face. A key risk to the improvement of health outcomes in Timor Leste is that MOH, and Government more generally, do not develop strategies that redress some of the important staff challenges currently becoming evident. 25. Provider behavior and how to overcome problems with the non-technical quality of services remains an import challenge for the MOH. The problem is multi-dimensional. Staff not being present at the facilities could either be a problem of recruitment/deployment, or of absenteeism. The former would tend to be reflected in long periods of time when the facilities are not staffed, or are only partially staffed. The latter may be reflected in sporadic availability of the staff. Clients, particularly those traveling long distances and/or at significant expense to themselves, need to know services will be available when they attend facilities. 26. It was noted in Chapter III that MOH is facing difficulties in filling positions in rural and remote health care facilities (para. 89). MOH deployment is currently application based, and non-voluntary deployment to specific posts is not used. Changing to a system of mandatory deployment, however, may lead to greater absenteeism. A better strategy may be to recruit staff originally from the local communities where there are vacant posts. The MOH has already started doing this for recruitment to midwife training. Other strategies could include special incentives such as establishing designated three year posts in areas with staff shortages and offering accelerated access to promotions and/or access to training for those who take up these posts for the stipulated period; or financial incentives for work in remote areas. 26. The DHS data does not clarify whether the problem of unfriendly or disrespectful treatment of patients by staff is a problem which is widespread among both public sector and private facilities, or it is mostly a problem in public sector facilities. An indication that it may be the latter comes from the qualitative study commissioned for this report. One of the findings of the study was that MOH workers do not see that an effective performance monitoring system and 120 incentive system is in place in the public sector, unlike what they perceive to be the case in NGO health care facilities. Hence it would seem important that MOH aimed at improving staff incentives to be consistently in attendance at their posts and to do a proper job, including treating patients with respect, by more closely monitoring staff performance and applying rewards and sanctions in line with performance. 28. Skill mix. As discussed in Chapter 3, the MOH policies on health functions to be undertaken at each type of facility and the mix of staff to be allocated to each facility type has been determined by: (a) a consideration of the skill mix of staff thought likely to be available over the medium term (e.g., it recognized that there was likely to be a significant shortage of doctors, midwives and technical staff); and (b) the services, as discussed, to be made available to the population. It was also recognized that it would be possible to gradually and cost-effectively increase the number of midwives through enhanced domestic training to meet the need to fill funded vacancies of this key cadre. It was a reasonable assumption when the policy was developed that severe financial resource constraints and the long term overseas training period and cost to train doctors (and to a lesser extend key technical staff), would likely severely constrained capacity to increase their numbers significantly in the short – medium term. 29. Two significant changes since the health policy framework was established: (a) the recruitment of large number (300) of Cuban doctors and the decision to train over 600 Timorese doctors in Cuba; and (b) the possibility of significantly increased fiscal space, including for health, arising from additional new gas revenues for government domestic financing, provide key challenges to the both the health policy framework as specified and to the management of the skill mix within the health sector. 30. From a health program perspective it is important that MOH continues with its effort to deploy adequate numbers of midwives throughout districts to meet the effective demand for reproductive health programs given their importance to desired health outcomes. This will require enhanced training (numbers and quality) taking into account attrition rates as well as efforts to deal with deployment problems discussed above. 31. The challenges (and opportunities) arising from the deployment of large numbers of doctors, particularly expatriate doctors, should not be underestimated. What will be the role of (foreign) doctors relative to (Timorese) health facility managers in overseeing facilities and programs? Which patients will be seen by doctors and thus what are to be the new roles of nurses, midwives and other Para-medical staff in service delivery and how do they relate to the doctor(s)? Are the functions of health facilities to change as a result of doctor deployment? Will this have implications for the design and equipment of facilities? What implications will this have for the range and demand for pharmaceuticals at each facility level? Given the current relatively low staff workloads in health facilities (with current staff numbers) will total numbers of Para-medicals deployed need to fall to make way for doctors? Do foreign doctors follow Timorese protocols and if not to what extent is this a problem? How does this impact on the roles and capacities of Timorese staff to support service delivery? How can MOH best induct both foreign doctors and foreign trained Timorese doctors into the public health system and ensure national protocols are followed? 32. Staff numbers and their deployment. Decisions on the skill mix needed to deliver services at each level of the health system, the number of facilities, and documentation of staff 121 workloads should influence total staff numbers required in the public sector. As discussed in Chapter 3, Timor is doing relatively well for a low income country in health staff numbers per 1,000 population, including for Timorese doctors per 1,000 population (i.e., excluding foreign doctors and those Timorese Doctors now under training in Cuba). This has resulted from the considerable expansion of the number of facilities staffed over recent years (even though midwife numbers have not kept up with demand) to ensure access objectives were met. It is important to carefully monitor staff numbers, their work loads (particularly as doctor numbers expand) and as planned efforts to expand the demand for health services realize results. 33. Current deployment of staff reveals considerable disparities relative to population numbers – both across districts, and cities relative to rural areas. Improved health staff information and use of recent census results would enable district level disparities to be analyzed. On the other hand staff distribution needs to reflect: (a) health needs (demand); (b) the range of possible interventions to be dealt with within a district/ sub-district city; (c) the different technologies or approaches for delivering services (for example in remote and mountainous regions); and (d) the impact of incentives and motivation of health workers on both productivity and quality. Nevertheless, there are indications that MOH faces a major deployment challenge in filling positions in rural and remote facilities. The qualitative study of health workers suggests there is frustration about security, personal and professional isolation, and a lack of access to training. MOH, as discussed, is relying on recruitment from local communities with some success. The question is can this be relied upon alone and in the longer run and what other incentives might be needed. 34. Staff incentives. Many staff argued, as documented in the qualitative survey, that financial incentives were needed to ensure willingness to serve in remote or rural areas and to ensure that rural staff were not passed over in access to training opportunities. Currently there is no provision for special payments or bonuses for this purpose in the Statute of Civil Service. Clearly, the expansion of newly trained doctors in the coming years, many of whom will be expected to work at district and sub-district level, is likely to intensify this challenge. 35. Discussion of staff incentive issues should be much wider than deployment issues and in fact are a theme, at least implicitly, running through out the above discussion of challenges and policy options. Staff are a critical input to the health system but unlike inputs such as facility infrastructure and drugs have legitimate needs and preferences. Understanding these and keeping in mind program objectives of the health system can greatly improve the option to create incentives for improved behavior in terms of both the quality and amount of services delivered by staff. In a professional environment such as a health system, or at a health facility level, non- material incentives such as inclusiveness in the decisions on health priorities, the objectives for health programs and on the role of staff cadres in the health system as discussed can positively affect staff commitment and productivity. It has been a challenge for many staff to move from a very specialized cadre structure under the former Indonesian administration to one where staff are expected to do multiple tasks under the current health policy framework. Understanding the issues felt by staff would assist in improved design on non-material incentives including transparency and principled selection of staff for in-service training and selection for vacant positions (particularly in urban areas) and participation in planning, managing and monitoring and evaluating health system performance. 122 36. An important staff incentive issue is career planning incentives and possibilities for recognition of good service over time and in remote areas (as discussed above). While the salary scale which MOH is able to offer staff is determined on a Government wide basis it has not been altered for 6 years for inflation and the seven point salary scale is very restrictive with very little possibility for the majority of staff to gain promotion. Over time this will become less and less acceptable and it is appropriate for MOH to consider options for proposals to the Government as a whole for: (a) salary points within each scale that could be climed over time based on staff performance; (b) a career path (backed by real access to training to upgrade skills where appropriate) for staff to be able to reach higher grades with relevant training, experience and performance achievements; and (c) reducing the compression of the spread of salary points between the highest and lowest grade points to enable greater differentiation in the grading of positions according to level of responsibility and work load. These trends seem to have happened in the private and non-government health sectors – but given the labor market structure in Timor Leste it is to be expected that private and non-government sector would set wage rates on a cost-plus basis in order to attract staff from Government. These proposals would need to be negotiated with Government by MOH within the context of a Government wide wages policy consistent with the fiscal framework. Management, Planning and Budgeting 37. This report has recognized the very significant focus on and achievements by MOH in developing institutional capacity to manage a functional health system from scratch. Timor Leste, in addition to loosing upwards of 85 percent of its health infrastructure in the violence at separation from Indonesia lost 80 percent of its senior and middle management – i.e., the vast majority of central MOH staffers, hospital and health facility managers had to learn new jobs for which they had no previous experience. Further, there were less than 15 Timorese doctors in the whole country to form the leadership together with experienced nurses and to plan and manage services. 38. Today, MOH has a senior management team responsible for oversight of the health system, accountable to a Minister, National Council of Ministers and Parliament for health outcomes and use of budget resources. Management of district health programs is by a District Health Management Team, which is responsible for preparing district health plans, associated budgets and reporting on health outcomes and expenditures on a quarterly basis. Hospitals are being given increased autonomy to manage resources within agreed activity plans and an increasingly autonomous medical store ensures timely supplies of pharmaceuticals (based on an essential drugs list) and medical supplies. Efforts have also been made to contract out ancillary services (grass cutting, laundry and meals for hospitals). At the same time, as discussed, MOH has developed an impressive health policy framework and managed a major facility reconstruction program. All-in-all, a stellar performance. Yet, capacity remains perhaps the binding constraint to improved health outcomes in Timor Leste. The question or key challenge is: How can MOH improve its capacity and give priority focus to interventions which will have the greatest impact in improving health outcomes? There are clear challenges in transforming policy priorities into effective services and programs on the ground, raising questions about management mechanisms, incentives, and service delivery models. 39. First, the MOH should sustain the centrality of districts and annual district plans (and annual hospital plans) which are regularly monitored and evaluated as a basis for focusing 123 program priorities and activities according to local needs to achieve key outcomes. These plans should focus on outcomes and realistic targets to be achieved and be resourced (by both government and development partner funds) through government systems in a manner which ensures managers have budgets to manage and fully understand what their budget is to achieve the desired outcomes. 40. Second, implementation of plans should be regularly monitored and evaluated by facility and district health managers and by central MOH. As discussed, experiences and problems, successes and failures should be identified and discussed across districts with the active support of central MOH in order to learn from mistakes and determine what works best 41. Third, accountabilities for programs and resources to implement programs should be increasingly devolved to districts and facilities. Efforts to establish revolving or imprest accounts at the district and sub-district/facility level should be applauded as a decisive move in the right direction. Financial discipline is of course paramount and it is critical staff are appropriately trained and supported by central MOH. One challenge arising from the devolution of responsibility for purchasing and cash management is to ensure clear rules exist on what can be procured locally and what should, for efficiency reasons be procured centrally. Resources available at each facility level and their expenditures should be publicly displayed at each facility so communities know what the money is being spent on and can monitor what is happening. 42. Fourth, the MOH needs to further define the appropriate role of central MOH and how best to organize to support the district (and hospital) planning and management processes. At present there seems to have been a clearer role for the policy and planning division of the MOH in organizing the policy and planning processes which have been the hallmark of the successes achieved over the past 5-6 years. The role of the technical divisions has been less clear in part because it has taken time to finalize the health policy framework (albeit with their input). One part of this confusion has been the tension between setting out to support districts (and hospitals) to implement programs and to become responsible for implementing programs. There are a number of reasons for this tension including the need to marshal resources for programs from donors (or donors going to the MOH to organize vertical programs separate from the core MOH program). Clarity on the program priorities to be implemented and a firm commitment to fully resource district and hospital plans provide a basis for agreement on the core role of central MOH. The planned organizational and management review as part of the National Strategic Plan provides a sound basis to work through these issues. This review should also consider the need for and how to create appropriate incentives at the respective levels. 43. Completion of the National Strategic Plan, in consultation with national stakeholders and development partners, provides a sound basis on which MOH can engage donors and encourage commitment to an agreed program by both Government and development partners. This could form the basis for agreement of a health sector wide approach (SWAp) for donor support to the sector and an agreement for development partners to increasingly use government systems. Multiple chains of accountabilities at each level of the health system which arise from multiple donor projects are not conducive to clear management accountabilities and often create or encourage a fragile and fractures bureaucracy. An increasing use of government systems by development partners, as in a SWAp, to fund health programs would enable streamlined accountabilities and build capacities within MOH. 124 44. The use of government systems does not imply that the MOH has to implement programs. A clear distinction should be made between the desirability of public financing (by government or development partners) and whether government undertakes the program itself. The MOH is to be commended for the extent to which it has tapped non-government institutions and staff to assist it with program delivery. Some 17 percent of staff are contracted and an increasing number of NGOs are being encouraged to support the implementation of health programs. Agreements on a partnership with Café Timor involving government financing of CCT clinic pharmaceuticals in return for CCT continuing to serve non cooperative members are sound policy. Options to do this exist with other agencies including Church health services which currently have very fragmented and inefficient pharmaceutical purchasing and sourcing processes. Options also exist (or could be explored) to contract out service delivery in some rural and remote areas currently underserved and where duplicating current arrangements are: (a) either not cost effective; or (b) current incentive arrangements mean that staff are not willing to go to the area. Expenditure Priorities 45. One approach to development of a medium term expenditure program is to develop the six elements outlined below. The starting point is the central theme of this report, as discussed, that while much has been achieved there remains significant scope for further improvements in health outcomes. Infant and child mortality remains unacceptably high as does maternal mortality. Communicable diseases, especially malaria and tuberculosis, are still major sources of morbidity and premature death. Nutrition deficiencies and upper respiratory tract infections in young children remain widespread. It is also probable that non-communicable diseases are emerging as a threat to population health. It is also evident that poor households bear a disproportionate share of the burden of disease and that demand for health services by households is less than optimal – in part due to deficiencies in the quality of care, lack of information on the seriousness of specific illnesses and problems with access, particularly due to poverty and costs of getting to health providers. Within the government systems framework (planning, budgeting and monitoring of programs) outlined above the MOH would strongly benefit from establishing a medium term program as outlined below. 46. Thus a question facing the MOH is how to set priorities for expenditure on health given the major health problems facing the country and the critical importance of the capacity constraint and the challenge of transforming policy priorities into effective services on the ground. Given the limited capacity of the public sector to deliver services, the central aim should be to prioritize delivery of services, particularly at the district and facility level, and to reduce the government’s direct responsibility for service delivery and put in place more viable options. One approach to setting expenditure priorities is to systematically work through the following six elements of a health strategy: 1. Focus on high-priority health outcomes using the most cost-effective strategies, with particular emphasis on targeting the poor. There is agreement that the high-priority health outcomes include reducing infant, child and maternal morbidity and mortality, controlling infectious diseases (tuberculosis, malaria vaccine-preventable diseases) and encouraging healthy population behaviors and increasing the use (demand) for quality basic health services including by reducing the 125 access costs for the poor. Improvements in health outcomes will also be achieved by improved implementation of cross sectoral priorities (water supply and sanitation, transport and communications, education, particularly of girls, and control of indoor air pollution). In order to achieve these MOH should: a Identify technical options for achieving the desired health outcomes and ways in which options might be packaged for delivery. This might include programs to extend vaccination coverage, targeted health promotion (e.g. nutrition awareness, weaning practices and recognition of health conditions and their seriousness), local level based treatment of pneumonia in children to reduce child mortality, efforts to ensure basic steps are followed with anti-natal counseling (e.g. iron supplementation, measuring blood pressure, counseling on recognizing signs of potential pregnancy complications), STD treatment, attending to etc. b Set priorities for public expenditure using public finance criteria. Three major criteria are applicable. One, if benefits (or costs) from the intervention accrue to more than the individual from its consumption there is a case on public fiancé criteria for public funds to be used (e.g., the benefits to vaccination spill over to the whole population so long as 80-90 percent of the population are covered, cure of TB results in benefits to the whole population by preventing the further spread of the disease. Two, the absence of perfect information means that individuals do not have the information required to have appropriate health related decisions means that health promotion designed to effect behavior change is justified on public finance criteria. Three, on equity grounds it is appropriate for public subsidies to be provided to ensure obtain services. c Commit to monitoring and evaluating progress towards the agreed health targets. This involves ensuring the Health Information System (including disease surveillance) is sustained and further developed and that appropriately structured population based surveys (e.g., demographic household surveys, quantative and qualative health seeking behavior surveys) are conducted on a timely basis and used for evidence based monitoring and evaluation of programs. 2. Public-Private Sector Interfaces. There is a need to explicitly consider the potential of the whole sector – public and private – to contribute to health outcomes. Thus it is important to: a Focus the public sector on those public goods (1b above) it can effectively deliver. It is critical that MOH focuses on those services that cannot be provided by the private sector and that its institutional capacity can deliver. b Promote appropriate involvement of the private sector in the delivery of private goods such as ambulatory care. c Strengthen collaboration between the public and private sectors in service delivery. This provides a major opportunity for MOH to tap the capacities in the 126 private sector to assist in meeting overall health objectives (e.g. further development of the relationship with CAFÉ Timor; increased cooperation with church health facilities (e.g. access to pharmaceuticals and basic health equipment to improve the quality of services in church facilities). 3. Prepare an Expenditure and Implementation Plan for the whole of the publicly funded part of the health sector (covering all government and donor-financed programs) based on the priorities and activities identified in the preceding steps and consistent with available resources. 4. Create Fiscal Space. This can be achieved in three ways: (i) achieve savings in the sector; (ii) making more efficient use of existing funds; and (iii) increasing sectoral allocations. 5. Prepare a Financing Plan. This needs to be prepared once MOH has decided what programs to emphasize and implement and having prepared and expenditure and implementation plan consistent with the overall resource envelop available to the MOH. This will require ensuring that the government and major development partners indicate what resources they can make available for the health sector over the medium term. 6. Develop Consistent Accountability Framework and Guarantee the Integrity of Government Processes. It is important that as the Government develops its combined resources budget that increasing effort is made to ensure accountabilities for development partner financed programs are as close as possible to those financed by Government. It is important that the MOH and the managers of its constituent parts (divisions, districts and facilities) know their total budget for the year and are responsible for accounting for the achievements and expenditures on the programs for which they are responsible. To ensure development partners can work in this way with Government it is critical that every effort is made by the Government (and MOH) to guarantee the integrity of government processes. Cross-Sector Interventions 47. This report has focused on the health sector. There is no doubt, however, that government activities in other sectors can have a large effect on health outcomes. Some obvious examples are water and sanitation, transport and communications, and general education. Economic policies that facilitate economic growth and poverty reduction are also an important if more indirect factor in improving health outcomes. Hence improving upon its performance in these other spheres of activity is also part of the challenge the public sector faces in seeking to improve health outcomes beyond what has already been achieved, though MOH itself may only have a limited role to play in these other areas. Knowledge Gaps 48. This report has brought a considerable amount of new information together and consolidated a range of pre-existing information on the health system. The DHS, in particular, has made a major contribution to the knowledge available for health planning. The report has 127 also been written prior to the availability of the full results of the new census which will provide important additional information for health planning and management. The MOH is also working to significantly improve its health information and management system. Nevertheless there remain a number of key knowledge gaps a number of which have been identified in this report. These include:.  Limited knowledge of both the communicable and non-communicable disease burden in Timor Leste arising from diseases not covered by the DHS;  Need for improved knowledge of health seeking behaviors -- both qualitative and quantative – to address the need to improve the demand for health services; including improved understanding of possible barriers to access including costs of access.  Limited systematic knowledge of provider behavior and concerns including: (i) attitudes of health professionals to clients; (ii) commitment to national policy objectives; (iii) technical quality of advice and services provided; and (iv) concerns about career paths, material and non-material incentives, postings and workloads.  Limited details of development partner expenditures, their program focus, input categories of expenditure, and whether recurrent or development expenditures; and the scale of incremental recurrent costs arising to the Government’s recurrent budget from development partner expenditures. Next Steps 49. One feature that has been prominent in the development of the country’s health sector since early 2000has been the excellent working relationship between the MOH and a number of Donor Partners, including the World Bank. In this spirit, it is suggested here that if MOH’s senior managers found sufficient merit in the report’s analysis and suggestions, they could convene a Workshop to discuss the report with their Donor Partners and other key stakeholders. Such a Workshop could help the MOH to decide on their priorities for change and development in the coming years: in what ways should MOH’s work evolve in the next five years, given what is known about the health sector at this time? An exercise of this nature would also fit in with the Medium-Term Strategic Plan envisaged by MOH. 128 BIBLIOGRAPHY Asowa-Omorodion, F. I. 1997. Women’s perceptions of the complications of pregnancy and childbirth in two Esan communities, Edo State, Nigeria. Social Science & Medicine, 44(12), 1817-1824. Atkinson, S., & Farias, M. 1994. Perceptions of risk during pregnancy amongst urban women in northeast Brazil. Social Science & Medicine, 41(11), 1577-1586. Buse, Kent and Waxman, Amalia. “Public-Private health partnerships: a strategy for WHO.� Bulletin of the World Health Organization, 2001, 79: 748-754 Chapman, R. 2003. Endangering safe motherhood in Mozambique: prenatal care as pregnancy risk. Social Science & Medicine 57, 355-374. Diamond, Nancy. 2004. USAID Timor-Leste Gender Assessment. DevTech Systems, Inc. East Timor National Mental Health Project. 2003. Ministry of Health, Democratic Republic of East Timor. East Timor National Development Plan, May 2002. Foster SD. 1991. Pricing, distribution, and use of antimalarial drugs. Bulletin of the World Health Organization 1991;69:349-63 Health Alliance International, Strengthening Maternal and Newborn Care in Timor-Leste, Focus Group Discussions with Midwives, December 2004. International Monetary Fund. Staff Report for the 2004 Article IV Consultation, Democratic Republic of Timor-Leste, June 2004. International Monetary Fund. Staff Report for the 2005 Article IV Consultation, Democratic Republic of Timor-Leste, May 2005. Jefferds, M. D. 2002. “Concepts of iron deficiency anemia and public health measures in rural Costa Rica, Social Science & Medicine, 55(7):1143-1156. Jirojwong, S. 1996. Health beliefs and the use of antenatal care among pregnant women in Southern Thailand. In P. Rice & L. Manderson (eds.), Maternity and reproductive health in Asian societies (pp. 61-82). Amsterdam: Harwood Academic Press. Kamat, V.R. 2001. “Private practitioners and their role in the resurgence of malaria in Mubai and Navi Mumbai, India: serving the affected or aiding an epidemic?� Social Science & Medicine. V52i6:885(25) Kleinman A., Eisenberg L., and Good B. 1978. Clinical lessons from anthropological and cross- cultural research. Annls Intern. Med. 88, 251-258. 129 Krueger, R. A. 1994. Focus groups: a practical guide for applied research (2nd ed.). Thousand Oaks: Sage. Kumar, S. 2000. “Indian health organizations call for control of private sector health care.� The Lancet. V356i9247:2077 Lindelow, M., Serneels, P. and Lemma, T. 2004. The Performance of Health Workers in Ethiopia: Results from Qualitative Research. (Draft paper.) World Bank. Livermore, C. 2002. Tuur Ahi: Childbirth and child death in Aileu, East Timor. Mills, A, Brugha, R, Hanson, K, McPake, B. 2002. What can be done about the private health sector in low-income countries? Bulletin of the World Health Organization, April 2002 v80i4:325 Ministry of Health and National Statistics Office, Timor-Leste, and ACIL Australia Pty Ltd, University of Newcastle and the Australian National University. 2004. Timor-Leste 2003 Demographic and Health Survey. Dili, Timor-Leste: Ministry of Health. Ministry of Health of Timor-Leste, Health Policy Framework, June 2002. Ministry of Health of Timor-Leste, Basic Package of Health Services Policy, March 2004. Ministry of Health of Timor-Leste, Private Sector Policy, December 2003. Ministry of Health of Timor-Leste, National Workforce Plan for the Democratic Republic of Timor-Leste, May 2005. Ministry of Health of Timor-Leste, Micro-Policy Documents: Reproductive and Health Strategy (September 2004); Integrated Management of Childhood Illnesses (IMCI, June 2004); National Nutrition Strategy (July 2004); National Immunization Strategy (July 2004); National Strategy for the Elimination of Lymphatic Filariasis and Control of Intestinal Parasitic Infections and Yaws (June 2004); National Mosquito-Borne Disease Control Strategy (in draft, September 2005); National Tuberculosis Strategy (July 2004); National Mental Health Strategy (November 2004); and National Oral Health Strategy (June 2004). Ministry of Planning and Finance of Timor-Leste. Overview of Sector Investment Programs, April 2005, Volumes I and II. Ministry of Health, Budget Document 2005-06. Mize, L. 2004. Maternal and neonatal health in East Timor: Findings from the USAID situational analysis January 11-30, 2004 and recommendations for possible program interventions. Morgan, D. 1998. The Focus Group Guidebook. Thousand Oaks: Sage. 130 Muela, S.H., Ribera, J.M., Mushi, A.K. & Tanner, M. 2002. “Medical syncretism with reference to malaria in a Tanzanian community. Social Science & Medicine, 55(3): 403-413. Nichter, M. 1996. Pharmaceuticals, the commodification of health, and the health care-medicine use transition. In M & M Nichter (eds), Anthropology and International Health. Australia: Gordon & Breach. Pfeiffer, J. 2003. “International NGOs and primary health care in Mozambique: the need for a new model of collaboration.� Social Science & Medicine 56 725-738. Reich, M. 2002. Public-Private Partnerships for Public Health. Cambridge: Harvard Univ. Press. Republica Democratica de Timor-Leste, The Road Map (for the Implementation of NDP Programs), April 2003. Republica Democratica de Timor-Leste, Intersectorial Action for Health and Wellbeing: A Framework for Action, December 2004. Sesia, P. M. 1997. Women come here on their own when they need to: Prenatal care, authoritative knowledge and maternal health in Oaxaca. In R.E. Davis & C.F. Sargent (eds.), Childbirth and authoritative knowledge: Cross cultural perspectives (pp. 397-420). Berkeley, CA: University of California Press. Soares, A., Martins, J., et al. 2004. Knowledge, attitude and practice on malaria in Timor-Leste. University da Paz, Dili & Menzies School of Health Research, Darwin. UNICEF, Dili and Insan Hitawasana Sejahtera, Jakarta, Indonesia. The Democratic Republic of Timor-Leste, Combined Sources Budget 2003-04, June 2003. The Democratic Republic of Timor-Leste, Combined Sources Budget 2004-05, May 2004. Tulloch, J. et al. 2003. Initial steps in rebuilding the health sector in East Timor. Washington: National Academies Press. UN Millennium Project Task Force on Child Health and Maternal Health :“Who’s got the power? Transforming health systems for women and children�. UN Millennium Project 2005. Coming to Grips with Malaria in the New Millennium. Task Force on HIV/AIDS, Malaria, TB, and Access to Essential Medicines, Working Group on Malaria. See www.unmillenniumproject.org/documents/. UN Millennium Project 2005. Investing in Strategies to Reverse the Global Incidence of TB. Task Force on HIV/AIDS, Malaria, TB, and Access to Essential Medicines, Working Group on TB. UNICEF. 2003. Multiple Indicator Cluster Survey –2002, Republic of Timor-Leste. 131 Van Schoor, V. 2003. Sitting Fire: a community-based study of home births and postpartum care in East Timor. University of Amsterdam Faculty of Social and Behavioral Sciences, Medical Anthropology Unit, August. Widdus, Roy. “Public-private partnerships for health: their main targets, their diversity, and their future directions. Bulletin of the World Health Organization, 2001, 79: 713-720 World Bank. 2004. World Development Report: Health and Nutrition Services. World Bank. Timor-Leste Poverty Assessment. Timor-Leste: Poverty in a New Nation-Analysis for Action, May 2003, Volumes I and II. World Bank. Background Paper for the Timor-Leste and Development Partners Meeting, April 2005. World Bank, Aide Memoir of Joint Donor Mission, Trust Fund for East Timor, December 2005. World Bank, Implementation Completion Report, Timor-Leste Health Sector Rehabilitation and Development Project, June 2005. World Bank, East Timor Public Expenditure Review, 2004. World Health Organization. 2004. WHO country cooperation strategy 2004-2008, Democratic Republic of Timor-Leste. WHO regional office for South-East Asia: New Delhi. World Health Organization, Eliminating Lymphatic Filariasis, Strategy Page. See http:www.who.int/ctd/filariasis/Elimination/strategy.html. World Health Organization, De-worming for Health and Development, Report of the Third Global Meeting of the Partners for Parasite Control, November 2004. World Health Organization Country Cooperation Strategy 2004-2008, Democratic Republic of East Timor, February 2004. Yach, D. 1992. The use and value of qualitative methods in health research in developing countries. Social Science & Medicine, 35(4), 603-612. 132