AFRICA REGION HUMAN DEVELOPMENT WORKING PAPERS SERIES THE WORLD BANK Improving HedIth for the Poor in Mozdmbique The fight Continues 23789 February 2002 :7 IL Other Titles in This Series Dynamic Risk Management and the Poor-Developing a Social Protection Strategy for Africa Engaging with Adults-The Case for Increased Support to Adult Basic Education in Sub-Saharan Africa Inclure les adultes - Pour un appui A I'6ducation de base des adultes en Afrique subsaharienne Enhancing Human Development in the HIPC/PRSP Context-Progress in the Africa Region during 2000 Early Childhood Development in Africa-Can We Do More for Less? A Look at the Impact and Implications of Preschools in Cape Verde and Guinea Le d6veloppement de la petite enfance en Afrique - Peut-on faire plus avec moins? Impact et implications des centres pr6scolaires au Cap-Vert et en Guin6e AIDS, Poverty Reduction and Debt Relief-A Toolkit for Mainstreaming HIV/AIDS Programs into Development Instruments Systemic Shocks and Social Protection-Role and Effectiveness of Public Works Programs Social Protection of Africa's Orphans and Vulnerable Children-Issues and Good Practice Program Options Can Africa Reach the International Targets for Human Development? An Assessment of Progress towards the Targets of the 1998 Second Tokyo International Conference on African Development (TICAD 11) Education and Training in Madagascar-Towards a Policy Agenda for Economic Growth and Poverty Reduction A Summary Education et Formation A Madagascar - Vers une politique nouvelle pour la croissance 6conomique et la r6duction de la pauvret6 Un r6sum6 Issues in Child Labor in Africa Community Support for Basic Education in Sub-Saharan Africa Le syst6me 6ducatif mauritanien - El6ments d'analyse pour instruire des politiques nouvelles Rapid Guidelines for Integrating Health, Nutrition, and Population Issues into the Poverty Reduction Strategies of Low-Income Countries Int6grer les questions de sant6, de nutrition et de population aux strat6gies de r6duction de la pauvret6 dans les pays A faibles revenus : quelques directives rapides Deux 6tudes pour la scolarisation primaire universelle dans les pays du Sahel en 2015 Africa Region Human Development Working Paper Series Improving Health for the Poor in Mozambique The Fight Continues Shiyan Chao Kees Kostermans Africa Region The World Bank ii AFRICA REGION HUMAN DEVELOPMENT WORKING PAPER SERIES @ February 2002 Human Development Sector Africa Region The World Bank The views expressed herein are those of the authors and do not necessarily reflect the opinions or policies of the World Bank or any of its affiliated organizations. Cover photo by Kees Kostermans. Cover design by Tomoko Hirata. IMPROVING HEALTH FOR THE POOR IN MOZAMBIQUE Contents Foreword v Abstract vii 1 Introduction 1 2 Health of the People 3 Health status 3 Health status of the poor 5 3 Health Sector Performance 6 Health service provision 6 Health service outputs 7 4 Health Expenditure and Financing 10 National health expenditures 10 Health financing policy 11 5 Equity in Health Care 14 Distribution of resources 14 Equity in access to health care 15 IV AFRICA REGION HUMAN DEVELOPMENT WORKING PAPER SERIES 6 Making the Health Sector More Pro-Poor 20 Health sector development in the PRSP context 20 Mounting an effective health sector response 21 Serving the poor more effectively 22 Improving overall service delivery 24 References 26 IMPROVING HEALTH FOR THE POOR IN MOZAMBIQUE V Foreword S ince the 1992 Peace Agreement, Mozam- The Government is fully aware of the key role bique has made excellent progress in recov- that the health sector can play in lifting people out ering from its war-torn past. In the health of poverty and increasing their productivity. How- sector, the Government has been rebuilding its net- ever, the poor will not be reached simply through work of health services under the ambitious Health increased expenditures and increased coverage. Sector Recovery Program. This program has pro- Experience in many African countries shows that vided a framework for the country's partnership the poor often benefit much less than the nonpoor with most of the bilateral and multi-lateral develop- from government health care expenditures. As this ment agencies active in the health sector. More report demonstrates, Mozambique is no exception recently, Mozambique's health sector has benefited in this regard. An effective poverty reduction plan from debt relief under the Heavily Indebted Poor must therefore also show specifically how the Countries (HIPC) Initiative and the Enhanced HIPC health sector can serve the poor. Over the past few Initiative. These events, in addition to a rapidly years, the Ministry of Health and its development growing economy, have led to a considerable partners have worked together to prepare a new increase in resources for the health sector. health sector strategy. The results have been used to Despite the favorable developments, enormous inform the preparation of the country's PRSP, and challenges remain. Mozambique is still one of the will be used to guide thinking on health sector poorest countries in the world, with around 70 per- development and implementation in the next cent of the population living below the poverty line. decade. The Government has prepared a Poverty Reduction This report attempts to summarize key aspects of Strategy Paper (PRSP), entitled the Action Plan for the knowledge base upon which the health sector Reduction of Absolute Poverty (2001-2005), as part strategy was built. Utilizing existing studies and of a concerted national effort to articulate a strategic data, it documents how health sector development, vision and action plan to improve the welfare of the debt relief, and poverty reduction strategies can poor. The plan identifies six priority areas: educa- work together to produce substantial and sustain- tion, health, agriculture and rural development, able progress in the health sector. The report is basic infrastructure, good governance and improved therefore best seen as a piece of work-in-progress macroeconomic and financial management. intended to capture and institutionalize the current VI AFRICA REGION HUMAN DEVELOPMENT WORKING PAPER SERIES state of knowledge on health sector issues in Mozambique. Our hope is that its publication would facilitate sharing of our evolving under- standing of the link between health sector develop- ment and poverty reduction, as well as prepare the way for further documentation of this important link, as the country's health sector strategy is implemented in the broader PRSP context. Ok Pannenborg Sector Leader for Health, Nutrition and Population Human Development Department Africa Region IMPROVING HEALTH FOR THE POOR IN MOZAMBIQUE VII Abstract T his study describes the development of the The study builds upon the existing studies on health sector over the last decade. The sector health and consolidates the sector knowledge. has made significant progress in terms of Based on the analysis, the study makes various rec- increasing coverage of services. However, health ommendations on how the health sector reforms remains a major concern in the area of poverty can be made more pro-poor by focusing on certain reduction. Huge inequalities exist with regards to interventions, by targeting certain areas and popu- resource allocation, deployment of staff and avail- lation groups, by designing new delivery models ability of services among various geographic areas, that would bring the services closer to the popula- between the urban and rural population, and tion, and by improving financial management to between the poor and the non-poor. serve the poor more effectively.  Improving Health for the Poor in Mozambique The Fight Continues Africa Region Human Development Working Paper Series  1 Introduction M ozambique is one of the poorest countries ment, Mozambique falls well below the averages of in the world, with a population of 17.3 mil- Sub-Saharan Africa. lion and a per capita income of US$230 Since the Peace Accord in October, 1992, peace, (Table 1). In the Human Development Index economic liberalization and extensive external aid Mozambique ranked 169 out of 174 countries together have contributed to a significant economic (UNDP, 1999). The country became independent in recovery and steady economic growth. Per capita 1975. A continuous civil war between 1976 and 1992 GDP grew on average by six percent during 1992- left it with weak infrastructure and poor human 99, and inflation was under control. However, the capital. Poverty is pervasive, with as much as 66 economic growth and revenue generation have not percent of the population falling below the poverty marched in pace with growing investment. The line. In many aspects of socioeconomic develop- dependence on external aid has increased drasti- cally. External funds finance about 50 percent of the total government expenditure. Major sources of Table I financing for the social sector are from donors. Key socioeconomic indicators of Mozambique Mozambique is one of the African countries that Indicator Latest Year have benefited from the Heavily Indebted Poor Countries Initiative (HIPC). In June 1999, the Exec- Economic indicators utive Boards of the World Bank and the Interna- Population 17.3 million 1999 tional Monetary Fund agreed that Mozambique had GNP US$3.9 billion 1999 GNP per capita US$ 230 1999 Povety idicaorsbenefit from the debt relief. It was also agreed that Poverty indicators Poverty headcount' 66.1% 1997 Mozambique would get additional assistance under Poverty gap2 27.3% 1997 the Enhanced HIPC Initiative. The total debt relief Nutritionto Mozambique from both original and enhanced Nutrtiontindators199 HIPC was about US$2 billion in net present value Percent stunted 43.22% 1997 Percent wasted 6.42% 1997 (about US$4.3 billion in current value). Social sec- Percent underweight 23.98% 1997 tors are the major beneficiary of the debt relief. The Sources: Economic indicators from World Bank Data Base and the rest from expenditure for the social sector is projected to "Understanding Poverty and Well-Being in Mozambique: the First National Assessment," increase from US$158 million in 1999 to US$203 mil- Ministry of Planning and Finance, Government of Mozambique, 1998. lion in 2002. A critical element that links to HIPC 1. The percentage of the population in households with consumption per capita less than the poverty tine. and country's future economic development is the 2. The mean distance below the poverty line, i.e., a measure of the depth of poverty, not development of Poverty Reduction Strategy Paper just its incidence. ecnmclbrlzto1ndetnieetra i 2 AFRICA REGION HUMAN DEVELOPMENT WORKING PAPER SERIES (PRSP). The PRSP will outline a strategic frame- ton on health in Mozambique and links between work and action plan for poverty reduction. The health and poverty. It tries to summarize the current debt relief would certainly help to reach the targets knowledge on the health of the population, particu- under the poverty reduction strategies. larly the poor, the health system's performance, and Poverty can be defined in many ways. The limi- the health sector's policies. The main sources of tation on people's abilities and opportunities to information for this paper are the health sector enjoy long and healthy lives is one way to measure expenditure review and the poverty assessment, poverty. Poverty has strong impact on people's Health Sector Strategic Plan (2001-2005), the most health, in turn, ill-health can put people into recent information available on health and poverty. poverty. Health improvement is one of the key Section 2 of this paper provides information on paths to poverty reduction. To develop a solid strat- health status. Section 3 summarizes the recent per- egy framework, one has to understand health and formance in the health sector. Section 4 assesses its links to poverty. The government of Mozam- equity in access and use of health care. Section 5 bique is preparing its PRSP called Action Plan for reviews health expenditure and financing sources. the Reduction of Absolute Poverty, 2001-2005, The final section discusses major issues and policies (PARPA). As a part of the preparation for PARPA, related to health and poverty reduction. this paper intends to provide background informa- 2 Health of the People prolonged civil war only ended in 1992. The population; and (c) develop the Ministry of Health's war led economic hardship and severe (MOH) technical and managerial capacity for plan- destruction of social infrastructure, includ- ning, implementing and evaluating health care and ing damage to the health care system. After the war, support services. The Health Sector Strategy Plan Mozambique inherited a very weak health care defines the health sector's contribution to poverty system that was urban-biased because of large reduction through interventions: health care provi- destruction in rural areas. Inequalities of distribu- sion; strengthening individuals and communities; tion of health facilities existed not only among and health advocacy (Council of Ministers, 2001). provinces, but also within provinces. The state has played a predominant role in providing health serv- Health Status ices. The private sector in health was abolished during 70s and 80s and only experienced a rapid The health status of Mozambican people is among growth in mid 90s after the new legislation permit- the poorest in the world. Basic health indicators are ting private practice (Law 26/91 and Decree 9/92). worse than the average for Sub-Saharan Africa. The Even now, the private sector is still operating disease pattern remains pre-transitional, that is, mainly in the capital area. The public sector has mainly infectious and parasitic diseases, diarrhea, been one of the best-supported sectors in terms of acute respiratory infection, measles, malaria, and both government and external foreign resources . tuberculosis and child malnutrition. Three basic Donor aid in the health sector continues to be exten- health status indicators, infant mortality, under-five sive. The government intends to expand health child mortality, and maternal mortality, are among services, both curative and preventive, to the entire the highest in the world (Table 2). country through the National Health Service. The spread of HIV infection has increased at an The health policy goal of the government is for alarming speed in recent years. According to all Mozambicans to have access to quality health UNAIDS, HIV prevalence among male STD clinic care. The national health policy sets out the follow- patients tested in Maputo increased from 3 to 20 ing objectives for the health sector: (a) reduce mor- percent from 1987 to 1996. Among female STD clinic tality, morbidity and suffering , especially among patients tested, HIV prevalence increased from 5 high risk groups such as women, children and all percent to 8 percent in 1997. Outside of Maputo, those displaced due to the war and natural disas- HIV prevalence among male STD clinic patients ters; (b) keep primary health care as the basis for the tested was 37 percent in 1998 and 26 percent among provision of good quality and sustainable health female STD clinic patients in 1997. The overall esti- care and make it accessible to the majority of the mated HIV prevalence rate was 13.2 percent among 3 4 AFRICA REGION HUMAN DEVELOPMENT WORKING PAPER SERIES Table 2 Key Indicators of Health Status (Years indicated in footnote) Indicador Mozambique Malawi Zambia Zimbabwe SSA Life expectancy 44 39 40 44 52 Infant mortality 115 134 80 69 102 Under-5 child mortality 219 234 197 74 170 Total fertility rate 6.2 6.7 5.5 3.8 5.3 Adult HIV prevalence (%) 13.2 15.9 19.95 25 8.57 Maternal mortality rate 1500 620 940 153 690 Low birth weight (%) 20 20 13 14 Sources: All data comes from Mozambique's Health Sector Strategic Plan, 2001-2005 except HIV prevalence rate for SSAwhich comes from UNAIDS 2000b. adults in 1999 (UNAIDS, 2000a), but increased to rates for infants and children under five fell signifi- 16.1 percent in 2000 according to the UN and MOH. cantly over the longer interval of 1970-1997, as The very latest estimates, however, based on better shown in Figure 1. surveillance methods show a prevalence of 12 per- The growth monitoring program also shows a cent. The number of AIDS-related deaths is slight improvement in nutritional status among expected to rise from 118,000 in 1998 to approxi- children: between 1996 and 1999, the proportion of mately 400,000 in 2002 (UNDP, 2000). children with poor growth has decreased from 10.5 Malnutrition is prevalent, particularly among percent to 8.9 percent. children. Data indicated that about 30-40 percent of Improved performance in many areas sets the children surveyed suffered from chronic malnutri- stage for major improvements in health outcomes. tion (stunted growth) while six percent of children But this is true only if the threat of AIDS is taken had acute malnutrition, indicated by wasting. Nutri- seriously and addressed rapidly and adequately. tional problems directly aggravate other health prob- The AIDS epidemic is relatively young in Mozam- lems and increase the overall burden of diseases. bique compared with neighboring countries, but it Some improvement in health status has been is expanding fast and its impact on the society and made since 1992, when the peace agreement was families is increasing drastically. According to the signed, and the economy has been growing at 8-14 latest MOH estimates (2000), adult HIV prevalence percent per year. Some health outcomes have is now around 16 percent. All countries surround- shown a positive trend. For example, mortality ing Mozambique have very high adult prevalence Figure 1 Mortality rates for infants (under 1 year) and children under five (per 1000 live births), 1970-1997 300- 250- 200- 150.-- Infant 150rcent to 8.9Under 5 100* - - _ _ _ _ _ _ _ _ _ _ _ _ 50- 1970-75 1980-85 1992-97 Source: The World BankIsMF estimates IMPROVING HEALTH FOR THE POOR IN MOZAMBIQUE 5 rates (between 20 and 35 percent), and these rates diseases and have much higher mortality rates than are already resulting in economic losses and sub- non-poor. An analysis of the Demographic and stantial reductions in life expectancy. There is little Health Survey (DHS) 1997 data shows that children doubt that Mozambique has to brace itself for the from poor households are more likely to be mal- tsunami. The epidemic was first concentrated along nourished and have a much higher mortality than the main transport corridors but is now spreading children from non-poor households. The data also widely to other sectors and areas. AIDS is capable of show that the poor are more likely to report illness, reversing all improvements gained in health, but less likely to get treatment (more discussion in absorbing a large proportion of the health budget section on equity in access to health care). and increasing the burden on the fragile health Infant mortality and under-five mortality are system, as seen in other neighboring countries. closely related to the poverty. Children from poor provinces are more likely to die. Figure 2 indicates Health status of the poor that the correlation between under-five mortality and the poverty status of the provinces. The The poor are in worse health than the non-poor provinces with more poverty have higher child (Table 3). Generally, they bear a higher burden of mortality rates. Clearly the poor need more health care. The fol- lowing section assesses whether the health care Table 3 system has addressed the needs of the poor and Health indicators by poverty status, whether health services have reached the poor. Mozambique, 1997 National Non- Figure 2 Indicator Average Poor poor Child mortality and poverty Infant mortality rate (IMR) 147.4 187.7 94.7 Under-five mortality rate (U5MR) 218.7 277.5 144.6 T0 Children stunted (%) 35.9 47.8 21.8 50 0 Children underweight - moderate (%) 26.1 36.9 14.3 'a a) 30_ - --- - Children underweight - severe (%) 9.1 15.3 4.5 20 Low mother's BMI (%) 10.9 17.2 4.2 10 - 0 Age specific fertility rate (15-19 years) 171.0 191.0 126.0 0 50 100 150 200 250 300 350 Sources: Instituto Nacional de Estatistica and Macro International, 1997 and Gwatkin et al., 2000. Sources: Ministry of Planning and Finance at al., 1998 and oHS, 1997. 3 Health Sector Performance V arious indicators show that performance of there is also the evidence that existing facilities are the health sector in Mozambique has more productive. improved. Since 1992, the health sector has Table 4 shows that great inequality exists among developed significantly. The system is mixed with provinces. Zambezia Province seems to be one of public, private, and quasi public/private institu- the most underserved. However, the figures must tions. The public sector still plays a dominant role in be interpreted with caution. A low number of health care provision while the private sector and people per facility, such as in Niassa, does not nec- NGO facilities are growing fast. The private sector essarily mean that health services are more accessi- consists of non-profit and profit institutions. ble. The low population density and great distance Attempts to outlaw traditional practice in the 1970s to facilities also affect access. Data on differences in and 80s were unsuccessful. The traditional sector health service utilization, presented later in Table 6, operates in parallel with the government health may be more revealing about inequities among the services. The public/private mix in health service provinces in access to services. The Central Hospi- provision has been increasing. tals have a coverage area reaching beyond the bor- ders of the province where they are located, while Health service provision the Maputo Central Hospitals, a general and a psy- chiatric facility, serve - at least officially - the whole The public sector expanded after the end of the war nation. in 1992. Under the umbrella of the Health Sector The management system under the NHS is still Recovery Program, the government started to rather centralized. The resources are allocated from rebuild the health sector, restructuring the National the Ministry of Health to provincial directorates of Health System (NHS) and improving service deliv- health, and to district directorates of health. ery. More than 400 health care facilities have been Provinces have been programming their annual rehabilitated or newly built. Many health posts activities based on an analysis of past performance were upgraded to health centers with maternity and needs. More recently, the provinces and dis- facilities and provide both curative and preventive tricts have embarked on more substantive planning services. Table 4 indicates the distribution of the for health service delivery. New planning methods health infrastructure by province. Staff figures have been piloted in several provinces. follow similar patterns. The public sector provides regular and "special" Service provision has increased dramatically services. The regular services are offered in public since 1992, partly due to new health facilities, but facilities for very low prices or free of charge. Spe- cial clinics are attached to larger government hospi- 6 IMPROVING HEALTH FOR THE POOR IN MOZAMBIQUE 7 Table 4 MOH/NHS health infrastructure network, 1997 1997 Pop. Central Provincial or Rural Health Health Province (x 1,000) Hospital General Hospital Hospital Center Post Niassa 764 - 1 (764)* 1 (764) 15 (51) 90 (9) Cabo Delgado 1,284 - 1(1,284) 3(428) 43(30) 37(35) Nampula 3,065 1 1 (3,065) 4 (766) 45 (68) 106 (30) Zambezia 3,202 - 1(3,202) 3 (1067) 24 (133) 138 (23) Tete 1,149 - 1(1,149) 3(383) 30(38) 50(23) Manila 975 - 1(975) - 14(70) 61(16) Sofala 1,380 1 - 4 (345) 17(81) 99(14) Inhambane 1,112 - 1(1,112) 2(556) 47(24) 26(43) Gaza 1,034 - 1(1,034) 4 (259) 11(94) 69(15) Maputo Province 809 - 1 (809) 1 (809) 14 (58) 43 (19) Maputo City 966 2 3(322) - 16 (60) 17 (57) Total 15,740 4 12(1,311) 25(630) 276 (57) 736 (21) * Number of facilities and (number of people (x1000) served per facility). Source: Management Sciences for Health, 1999. tals. These special clinics are highly subsidized Mozambique has a better developed information operations catering to the highest socioeconomic system than many Sub-Saharan countries with com- class. They have their own revenue-generating parable levels of per-capita income. One indicator capacity and thus are able to provide better quality developed and used in the health planning is the service. They have their own independent account- care unit." Calculated on the basis of the time ing system and are a major magnet for specialists to spent on the service, the care unit gives a weight to remain in the public sector. each of the five major health services that together The private sector is limited mainly to the large account for the vast majority of service outputs: cities. Private individual and group practices have vaccinations, outpatient consultations, MCH con- increased in some areas. International and national sultations, deliveries, and hospital bed days.' The NGOs run some non-profit health facilities and also care unit provides a measure of service output and provide funds directly to the NHS. service utilization. Care units produced per health worker, a measure of efficiency, increased from 6005 Health service outputs to 6744, an improvement of 11 percent from 1993 to 1999. Some health outcome indicators also show an Between 1993 and 1999, service outputs of health improvement such as nutrition indicators and mor- centers and hospitals have increased by 50 percent. tality rates. Service units for polio vaccinations increased by 210 One needs to be cautious interpreting these fig- percent, and out-patient consultation units grew by ures because the increase in outputs is certainly 107 percent, while MCH consultations, institutional partly due to improved reporting. The overall coy- delivery and bed occupancy service units increased erage of the national health system is still limited. It by 44 percent, 47 percent and 20 percent respec- is conservatively estimated that about 50 percent of tively. The large increase in vaccinations is due population have access to basic preventive and cur- mainly to the polio eradication campaigns. ative health services, meaning that they live within 8 AFRICA REGION HUMAN DEVELOPMENT WORKING PAPER SERIES Table 5 Performance of the health sector, 1993-1999 1993 1994 1995 1996 1997 1998 1999 Utilization/coverage No. of consultations/habitant 0.36 0.37 0.41 0.46 0.57 0.66 0.77 Percent deliveries by trained staff 26 29 28 30 35 36 37 Antenatal care coverage (%) 57 63 65 73 90 90 94 Postpartum care coverage (percent) 22 26 28 31 37 41 44 Percent of children received DPT 3rd doses 45 55 57 59 73 80 81 Anti-tetanus vaccination coverage (percent) 60 65 66 67 80 89 90 Care units per habitant (percent) 2.37 2.43 2.47 2.62 3.20 3.18 3.26 Care units/staff 6005 6713 6078 6310 6524 6685 6744 Difference in coverage between DPT Ist and 3rd doses (%) 25 26 23 23 21 18 17 Health status Intra partum mortality (per 1000) 0.15 0.44 0.42 0.35 0.31 0.33 0.29 Maternal mortality ratio (per 100,000 deliveries, in the clinic) 2.34 2.28 1.84 2.02 1.81 1.58 1.54 Percent children stunted (Z<-2) 12.8 10.8 11.1 10.5 9.7 9.6 8.9 Percent low birth weight (<2500 gr.) 13.4 13.5 13.5 12.6 12.1 12.3 12.2 Source: Mozambique Ministry of Health Annual Reports, 1993-99. 10 kilometers of a facility. The DHS, 1997 data give be explained partly by the low population density the following figures for service coverage: about in those provinces, while Maputo City clearly has a 44.2 percent of deliveries were assisted by a health relative oversupply of workers and relative overuse professional and 47.3 percent of children aged 12-23 of services. The low level of service units per inhab- months had received full immunization (MOH, itant in the two northern provinces shows that the 1999). services still have low availability probably because Service outputs have increased, but are still dis- of distance. The solution therefore seems to be to tributed unequally. Maputo City has much better increase basic-level multi-purpose health cadres in indicators for outpatient and inpatient visits per relatively small facilities. In Nampula, Tete, and person and thus higher total of care units per capita. Zambezia provinces, the issues are different: low However, in terms of care units per health provider utilization rates combined with high outputs per Maputo City ranks lowest, partly because of a con- health worker. These provinces would need more centration of health professionals there. staff to expand service coverage. The best perform- The rather low efficiency of health workers in the ing province seems to be Gaza, with a high produc- northern provinces Niassa and Cabo Delgado can tivity of staff and good use of services by the popu- IMPROVING HEALTH FOR THE POOR IN MOZAMBIQUE 9 Table 6 Care units per person by province, 1997 Outpatient MCH Care Units Per Total Province visit Inpatient Deliveries Contacts Immunization Health Provider Care Units Niassa 0.639 0.138 0.023 0.551 0.800 6,462 3.113 Cabo Delgado 0.415 0.114 0.011 0.293 0.722 5,328 2.220 Nampula 0.412 0.124 0.013 0.424 0.752 7,619 2.486 Zambezia 0.392 0.089 0.010 0.297 0.707 7,811 1.956 Tete 0.574 0.165 0.015 0.356 0.735 6,589 2.953 Manica 0.599 0.166 0.019 0.472 0.786 7,294 3.194 Sofala 0.736 0.297 0.016 0.312 0.687 6,934 4.256 Inhambane 0.608 0.172 0.017 0.565 0.802 5,742 3.330 Gaza 0.715 0.279 0.023 0.646 1.046 8,317 4.673 Maputo Province 0.806 0.203 0.015 0.515 0.734 8,251 3.692 Maputo City 1.352 0.616 0.039 0.589 0.850 4,614 8.378 National 0.583 0.185 0.016 0.421 0.766 6,815 3.247 Source: Management Sciences for Health, 1999. lation (this province suffered most from the flood in 2000). Comparing these data with similar data from 1. Care Units (CU) were calculated as follows: vaccination = other years, one can detect clear increases in health 0.5 CU; MCH contact = 1 CU; outpatient = 1 CU; inpatient = 9 CU; and delivery 12 CU. UNDP, 1999, Mozambique: worker efficiency and in utilization of services. National Human Development Report. 4 Health Expenditure and Financing National health expenditures such as Malawi and Ghana. This level falls short of the US$12.00 standard established under the World In an effort to better understand the health ex- Development Report, 1993, and the US$9.24 stan- penditure, the MOH constructed a national dard under Better Health in Africa, 1994. health account (NHA) using 1997 data. Useful Table 7 indicates there are four major sources of information was collected on health expenditure funds for health financing: government treasury, and financing. However, the available data are external donors, employers, and households. Exter- incomplete and cannot be sufficiently disaggre- nal aid financed more than 50 percent of the total gated into a "sources and uses" matrix. The infor- health expenditures and the government took 22 mation on health expenditures and financing for percent of the share. Even at high levels of poverty, 1997 is presented in Tables 7 and 8. The information households spent almost as much as the govern- includes both public and private sector health ment did on health care (19 percent) (see Figure 3). expenditures. Funds from each source were channeled through According to the NHA estimate, the health sector financing agencies that either provide or purchase spent about US$140 million 1997. This is equivalent health services. Much of the health expenditure is to US$8.84 per capita and is comparable to the level channeled through the MOH (54 percent). In fact, of health spending in many low-income countries, health has become one of the larger sectors sup- Table 7 Health expenditures by financing sources and financing agents, 1997 (in million U.S. dollars) Financing Agents Sources of funds MOH Other Ministries Employers NGOs Households Total % of Total Treasury 28.5 2.4 ee s- 30.9 22.0 Donors 46.8 - - 26.4 -73.2 52.0 Employers - - 9.2 - - 9.2 7.0 Households - 1.8 - - 24.9 26.7 19.0 Total 75.3 4.2 9.2 26.4 24.9 140.0 100.0 Percent of total 54.0 3.0 7.0 19.0 18.0 100.0 Sources: Management Sciences for Health, 1999. 10 IMPROVING HEALTH FOR THE POOR IN MOZAMBIQUE 11 Table 8 Health expenditures by financing agents and providers, 1997 (in million U.S. dollars) Financing Agents Health providers MOH Other Ministries Employers NGOs Households Total % of Total MOH/NHS 75.3 - 3.8 24.4 1.5 105.0 75.0 MOH/SC n.a. - - - 5.2* 5.2* 4.0 Other min. n.a. 4.2 - - - 4.2 3.0 NGOs - - - 2.0 - 2.0 1.5 For-profit providers - - 3.0 - 4.5 7.5 5.0 Employers - - 2.2 - - 2.2 1.5 Traditional medicine - - - - 9.9 9.9 7.0 Communal pharmacies - - 0.2 - 3.8 4.0 3.0 Providers abroad - - n.a. - n.a. n.a. - Total 75.3 4.2 9.2 26.4 24.9 140.0 100.0 Percentage of total 54.0 3.0 7.0 19.0 18.0 100.0 Source: Management Sciences for Health, 1999. *MSH estimations. ported by the government and by donors. NGOs No reliable data exist on health expenditures by handle 19 percent of the health expenditures. levels of care. However, the rough estimates idi- As health service providers, the MOH/HS and cate that government and donors spent US$2.42 per special MOH clinics consume a substantial share of capita on primary and secondary care and US$4.89 resources (79 percent). MOH facilities receive per capita on all levels of care in 1997 (MSH, 1999). resources from a variety of financing agents: the government itself, employers, NGOs, and house- Health financing policy holds. Traditional medicine, also a major service provider, accounted for 7 percent of the health The current financing policy is based on the princi- expenditures. Private for-profit providers took 5 pie that all Mozambicans should have access to percent of the total resources. Very limited services quality care at an equitable price. The NHS in were provided by employers (1.5 percent). Mozambique has received substantial contributions from the international community. Donor funds contributed more than 50 percent of total health Figure 3 financing. Major multilateral and bilateral agencies Sources of health financing, 1997 active in Mozambique include UNICEF, WHO, UNTPA, the African Development Bank, the Islamic 19% 22%Development Bank, the World Bank, and coopera- tion agencies from the European Union, USA, Ire- land, the United Kingdom, Denmark, the Nethere- lands, Italy, Norway, Switzerland, and Spain. External support is provided as direct budgetary support, program support, project support or tech- nical assistance. Overall, donors financed about 60 52% percent of the national budget through budget sup- 0 Govt. 11 External Lc Employers 0 Households port. A portion of project aid to the health sector 12 AFRICA REGION HUMAN DEVELOPMENT WORKING PAPER SERIES Table 9 Sources and uses of MOHINHS health expenditures in 1997 (in million US dollars) Government Items MPF/Treasury Debt Relief Fee Revenues Subtotal Gov't Donors Total Personnel emoluments 11.713 0.004 0.089 11.806 9.684 21.490 Drugs and med. eqpt. - 1.500 - 1.500 16.844 18.344 Other recur. costs 10.179 1.702 0.803 12.684 11.065 23.749 Investment expenses 2.567 - - 2.567 9.295 11.862 Total expenses 24.459 3.206 0.891 28.557 46.889 75.446 Share (percentage) 32.42 4.25 1.18 37.85 62.15 100.0 Source of basic data: Management Sciences for Health, 1999. assumed the form of earmarked budget support. accounted for 4.25 percent of the total MOH expen- Earmarked budget support began in 1990 and has diture and mainly financed non-salary recurrent increased significantly. On average, at least US$ expenditures. 5million has been provided to the health sector Funds for personnel emoluments are split annually as earmarked budget support (Pavignani between MPF/ Treasury (54.5 percent) and donors and Durao 1999). (45.1 percent). Drugs are funded almost solely by Out-of-pocket expenditures are one of the major donors (91.8 percent) and debt relief (8.2 percent). sources of health financing. User fees for curative There is hardly any Treasury funding for drugs. outpatient services in the public sector were first Funding for other recurrent costs comes from a introduced in 1997. The fees were set low initially variety of sources: MPF /Treasury 42.9 percent; debt and increased later through changing the law. Hos- relief, 7.2 percent; fees, 3.4 percent; and donors, 46.6 pitals were also allowed to charge in-patients and percent. foreigners and to charge for special services in 1994 Using government tax revenues and donor (Medical Care Development International, 2000). budgetary support, MOH/NHS provides global Fee revenues from inpatient and outpatient facili- budgets to provincial district health offices, which in ties are small (around 3 percent of total government turn allocate these to the district health offices resources for health) but appear to be growing. below them. Some donors also provide off-budget Since 1997, the HIPC debt relief initiative has support directly to provincial health offices. These provided a new source of health financing funds. off-budget items pose particular difficulty to the HIPC has meant an increased allocation to the Ministry of Planning and Finance (MPF), as they are health sector. Table 8 indicates the debt relief not recorded in the Triennial Public Investment Pro- gram (and in the annual development budget). Thus, their execution rate is often unknown. Some Figure 4 donors also channel "off-budget" funds to NGOs, Health recurrent budget, 1993-1999 which then provide support to MOH/NHS. Cur- rent planning is not activity-based (budget linked to 600 objectives or outputs) although MOH has shown 500 - - interest in moving in this direction. At present, the 4 00 - - -- - -- - - - 3 00----------------- state accounting system cannot provide an appro- 300 --- ---- - 0 200 -- - ----_ priate picture of activities and outputs as a function 0 -a of budget allocation or spending. 93 94 95 96 97 98 99 Under HIPC and later the Enhanced HIPC Initia- tive (eHIPC), the government also aimed to increase Nominal Value E - Conatant Value current health expenditures annually, both in real IMPROVING HEALTH FOR THE POOR IN MOZAMBIQUE 13 terms and as a share of total current expenditure in ment's recurrent budget for health increased from line with medium-term expenditure framework. 91.6 billion to 193.8 billion Mt in constant value. Current expenditures on health, as a proportion of Under eHIPC, the government of Mozambique's total current expenditures of the government, budget for the social sectors increases by about US$ evolved as follows: 1998 - 9.1 percent; 1999 - 10.2 40-50 million per annum. Since 1997, the Debt relief percent; and 2000 - 13.4 percent. Under HIPC the has contributed about 110 billion Mt to the health government continues to show its commitment to sector. the sector. Over the period 1993-1999, the govern- 5 Equity in Health Care Distribution of resources em part of the country for all inputs. Many of the health services delivered in Maputo City have, mn Inequality persists in health resource distribution theory a national function -the Central Hospital of among provinces. Maputo City has the advan- Maputo is the highest referral facility in the country tage over all the provinces in human and physi- and is a major training ground for many levels of cal capital as well as in financial resources. Zam- health workers. In practice the services are mainly b6zia Province presents the worst situation in both delivered to the local urban population. human and physical resources (Table 10). In general, Inequality in fund allocation among provinces the northern provinces are worse off than the south- was substantial. Table 11 indicates that Maputo City Table 10 Distribution of health resources by province, 1999 Maternity beds Beds/a 000 per 1000 women Drug exp. Population per Funds per Province population in reproductive age per person* (MZM) technical staff person (MZM) Maputo City 2.36 2.37 51,546 578 176,911 Sofala 1.24 1.54 19,361 1,478 67,270 Gaza 1.05 1.64 15,759 2,020 44,985 Maputo Pry. 1.05 1.61 12,476 1,942 51,470 Inhambane 0.99 1.62 15,553 1,991 53,036 Niassa 0.89 1.03 9,288 1,833 48,816 Manica 0.80 1.11 16,576 2,115 56,318 Tete 0.78 0.94 18,039 2,263 63,035 Nampula 0.77 0.77 6,497 2,814 14,877 Cabo Delgado 0.59 0.71 9,897 2,431 34,151 Zamb6zia 0.48 0.68 7,836 3,351 24,665 National Average 0.89 1.12 14,134 1,955 47,461 Sources: MOH, 1999. Includes the drugs distributed by the center to the provinces. 1US$ = MZM 12,000. Funds per person refer to the total recurrent costs (state budget and external funding), and does not include funds for central institutions. 14 IMPROVING HEALTH FOR THE POOR IN MOZAMBIQUE 15 Table 11 Funds per capita available in the capital city and the least favored province (expressed in MZM) Ratio of Zambezia's funding to Maputo's Maputo City Zambezia Province MaputolZambezia Year State External Total State External Total State External Total 1994 5763 1753 7516 449 559 1008 12.8 3.1 7.5 1995 15350 1955 17305 1415 3007 4422 10.8 0.7 3.9 1996 26851 7294 34145 1816 6902 8718 14.8 1.1 3.9 1997 27306 14463 41769 1797 3807 5604 15.2 3.8 7.5 1998 38482 25261 63743 2849 3854 6703 13.5 6.6 9.5 1999 49376 12862 62238 4608 4637 9245 10.7 2.8 6.7 Source: MOH, 2000. received from the government more than ten times not improved over the years. Limited absorptive the funds per capita than Zamb6zia did. External capacity is an important issue that needs to be funds tended to be distributed to disadvantaged addressed. Effort need to be made to improve areas; however the allocation did not fully correct absorptive capacity to break a vicious cycle: the less the inequality. Overall, Maputo City still got about the poor provinces are able to spend, the less funds seven times more per capita than Zamb6zia did. they receive. The progress in narrowing the funding gap is slow, and inequality in resource allocation persists over Equity in access to health care the years. The inequality in fund allocation clearly results from unequal distribution of other resources As indicated earlier, the poor have much worse such health professionals and infrastructure, which health status than the non-poor. In addition to can change only slowly over time. socioeconomic status, inequality in access to health An inequality also existed in levels of care. While care contributes to their inferior health status. primary level facilities furnished 37 percent of the Household survey data show large differences in activity outputs measured in health care units, they health behavior and health access between the poor only received 22 percent of the government and non-poor. resources. On the other hand, the three central hos- Although the poor generally suffer more from ill- pitals only produced 15 percent of the services ness than the non-poor, perceived needs for care and (measured by service units) but received 37 percent actual care-seeking behavior do not necessarily of these funds. Even considering that the central reflect their real needs as defined by clinical condi- hospitals provide a more sophisticated level of care tions. Household survey data show that the poor are and supply services that benefit the other levels, less likely to report illness than the non-poor, and such as training, the fund allocation to them still among those who reported ill, the poor are less likely seems relatively generous. seek care than the non-poor (Christy and Ferrara The extent of use of the available funds varies 1999; Cabral, 1999). In general, younger age groups, significantly among provinces (Table 12). In 1999, female population, and people living in northern the use ranged from 29 percent to 102 percent for provinces are more likely to report illnesses. state funds and 53 percent to 99 percent for external Table 13 shows the poverty characteristics of funds. In general, Maputo province, Maputo City those who reported illness and received care during and Cabo Delgado use funds better. Zamb6zia as the month prior to the survey. The poor are defined the least served province also has the lowest level of as people living below the poverty line (estimated spending the available resources. The situation has in 1996 to be at 3,941 Mt, 4,520 Mt, 6,934 Mt and 16 AFRICA REGION HUMAN DEVELOPMENT WORKING PAPER SERIES Table 12 Budget execution in percentage by province for 1997-99 Actual expenditure/budget (percent) Province Government External funds Total 1997 1998 1999 1997 1998 1999 1997 1998 1999 Niassa 77 96 95 65 82 67 68 86 79 Cabo Delgado 86 93 102 88 83 82 87 82 87 Nampula 68 88 - 88 91 - 77 84 - Zambezia 97 72 75 87 56 58 90 65 65 Tete 54 60 29 89 90 94 76 78 78 Manica 98 100 88 89 94 64 94 96 71 Sofala 94 96 78 98 96 99 88 95 85 Inhambane 79 90 93 76 107 59 77 92 73 Gaza 92 68 95 93 69 53 92 68 78 Maputo 100 83 88 100 98 99 98 90 90 Maputo City 90 97 95 76 99 94 80 95 85 H.C. Maputo 98 51 - 93 45 Country 86 80 86 88 85 82 Sources: MOH, 2000. 13,323 Mt respectively for the north, center, south, urban areas certainly have advantage over the rest and Maputo City) while the ultra-poor are those of the population. For all groups, children had living at less than 60 percent of the poverty line. higher use of services than the rest of the popula- Table 13 shows that the likelihood of receiving tion, suggesting that households make greater treatment increases for those living in urban areas efforts to treat children. (74 percent rather than 57 percent). Differences Overall, about 40 percent of population who between poor and non-poor are small in rural areas, reported illness did not seek for care. For the rural while they are significant in urban areas. Among population, the percentage is slightly higher. The those who reported illness, urban residents are two main reasons for not seeking care are distance more likely to obtain treatment. The non-poor in and lack of money (Table 14). More than for the rest Table 13 Characteristics of those reporting illness and receiving treatment Rural Urban Total Ultra Non- Ultra Non- Ultra Non- poor* Poor poor All poor Poor poor All poor Poor Poor All Percentage of those who reported illness and received treatment 56 56 60 57 64 70 81 74 58 59 64 61 Percentage of children 0-5 yr. who reported illness and received treatment 66 64 73 67 72 77 84 80 67 67 75 70 Source: MPFU 1998. *Ultra poor is defined as 60 percent of the reference poverty line, which was US$170 consumption per person per year in 1996-97 survey. IMPROVING HEALTH FOR THE POOR IN MOZAMBIQUE 17 Table 14 Reasons for not seeking health care at a facility Percent Estimated Percent poor Reason of total total pop. Median age Percent rural or ultra poor Percent still ill Facility too far 38 256,872 25 99 35 45 Lack money 35 240,293 23 78 37 42 Illness not severe 8 50,906 19 86 22 25 Lack of drugs 6 39,344 24 96 15 50 No confidence 0.4 3,065 40 85 11 41 Other 12 80,488 23 80 27 42 Total 100 682,869 24 88 34 42 Source: Christy and Ferrara, 1999. of the population, the health care seeking behavior Table 16, which shows the accessibility of health of the poor is affected by lack of confidence in the professionals by poverty status indicates that the system, lack of drugs at the facilities, and the per- non-poor have better access to doctors, health cen- ception that their illness is not severe. The rural ters and pharmacies. population is over-represented for all reasons, but The following data from the 1997 DHS indicate especially so when distance and lack of drugs are that the poor benefit less from health services than given as a reason for not seeking care. the non-poor. The data were stratified for urban and In a large but thinly populated country like rural expenditure quintiles. Table 17 shows health Mozambique, distances play an important role in service statistics for a wide variety of elements, such access to health care , as the following table indi- as vaccination, treatments for common diseases, cates. Unfortunately median distances are not avail- antenatal care, delivery attendance, and use of con- able for the various groups. The mean distance, traception. The data on knowledge of HIV trans- which doesn't characterize a non-normal distribu- mission are also presented for the various groups. tion well, does not show big differences between The numbers speak for themselves: large differ- the poor and non-poor. ences are recorded in service coverage for the poor TableTable 16 non-poorshaveebetterloccessrtotdoctors,ehealthecen serviestance for r lousehos toselectePercentage of rural population with specified that the poor benefit health services in their village, 1997 Service Ultra-poor Poor Non-poor All Health service Ultra poor Poor Non-poor All Doctor 47 47 43 46 Doctor 1.3 1.6 3.2 2.1 Traditional practitioner 1 1 2 1.5 Midwife 23 22 19 21 Traditional healer 94.7 94.3 92.3 93.7 Health post 19 19 17 19 Nurse 14.3 15.3 20.4 16.8 Health center 31 30 26 29 Midwife 19.2 19.9 19.2 19.7 Pharmacy 31 29 25 28 Health post 18.4 18.5 20.5 19.1 Market 17 17 15 16 Primary school 4 4 Health center 3.9 4.6 6.4 5.1 Public transport 18 17 15 16 Pharmacy 3.7 4.1 6.0 4.6 Source: Ministry of Planning and Finance eta., 1998 (the sample from LSM 1996-97). Source: Ministry of Planning and Finance et al., 1998. 18 AFRICA REGION HUMAN DEVELOPMENT WORKING PAPER SERIES Table 17 Analysis of DHS 1997 data on health behavior and access by expenditure quintiles Urban Rural Indicator Definition Expenditure quintiles Expenditure quintiles 1 2 3 4 5 1 2 3 4 5 Immunization coverage (Children age 12-23 months by (percent): vaccination card or mother's report): Measles * * 91.7 92.7 33.0 38.2 40.8 69.7 95.5 DPT3 * * * 93.7 93.1 32.2 50.6 45.2 70.4 95.5 All * * * 83.8 84.1 19.7 27.6 30.3 61.9 88.2 None * * * 1.6 1.3 36.1 28.3 28.4 7.1 0.9 Medical treatment of illnesses (percent): Treatment of diarrhea Prevalence Percent ill in the preceding two weeks * * 32.5 37.7 25.8 20.9 26.0 18.2 12.1 5.1 ORT use ORS, RHF, or increased liquids * * (82.1) 66.0 83.3 42.4 52.5 68.4 73.0 (84.0) Seen medically Brought to a health facility if ill * (67.9) 32.8 52.8 25.2 30.1 30.4 37.8 (67.7) Percent seen in a public facility Among those medically treated * * (67.9) 32.1 51.3 25.2 30.1 23.6 37.8 (67.7) Treatment of acute respiratory infection Prevalence Percent ill in the preceding two weeks * * 23.1 14.9 15.7 11.7 11.6 9.3 7.7 16.7 Seen medically Brought to a health facility if ill * * * (73.7) 63.8 17.3 32.4 45.2 39.7 (16.3) Antenatal care visits (percent): To a medically trained person Doctor, nurse, or nurse-midwife * * 87.6 94.4 97.9 46.6 67.1 61.9 87.5 99.1 To a doctor * * 1.2 1.4 12.2 0.1 2.0 0.3 0.9 0.6 To a nurse or trained midwife Nurses and nurse-midwives * * 86.4 93.1 85.6 46.5 65.1 61.6 86.6 98.5 2+ visits * * 82.3 81.8 82.4 36.8 62.7 54.4 75.2 74.9 Delivery attendance (percent): By a medically trained person Doctor, nurse, or nurse-midwife * * 71.2 81.6 83.8 18.1 34.0 27.1 55.0 78.8 By a doctor * * 0.6 3.3 10.1 0.2 1.0 0.4 2.3 1.5 By a nurse or trained midwife * * 70.6 78.4 73.7 17.9 33.0 26.7 52.7 77.3 Percent in a public facility * 71.2 81.6 83.5 17.0 31.8 26.6 56.0 78.5 Percent in a private facility * * 0.0 0.0 0.1 0.0 0.0 0.3 0.1 0.0 Percent at home * * 28.8 17.2 14.7 82.4 67.5 71.1 42.1 20.9 Use of modern (Currently married persons using a contraception (percent): modern method): Females * (1.7) 9.9 9.1 21.8 0.9 1.8 2.4 3.8 6.1 Males * * (13.6) 6.4 21.6 0.6 1.1 4.1 6.8 17.8 Knowledge of HIV transmission (percent) Females * (28.2) 23.2 29.0 44.8 26.8 36.2 30.4 30.7 32.4 Males * * (64.3) 63.7 69.3 40.8 48.1 36.8 44.0 77.2 Source: Gwatkin D. et al., 2000. IMPROVING HEALTH FOR THE POOR IN MOZAMBIQUE 19 and the non-poor. The rural people are worse off for 37 percent and 75 percent). Use of contraceptives is almost all services. For example, vaccination cover- low among all groups but extremely low among the age differs threefold (33 percent for the lowest quin- poor, both male and female. tile and 95 percent for the highest). ORT use is twice Overall use modem contraception is very low. as high among the highest quintile as it is among Almost none of the poor use any modern methods the lowest, that is, 84 percent vs. 42 percent. Of the at all. Knowledge about HIV transmission is also far very poor, only 46 percent attend antenatal care, too low, particularly for women. Differences in while 99 percent of the highest quintile does so (for knowledge levels among quintiles are smaller for more than two visits, these figures are respectively females than for males. 6 Making the Health Sector More Pro-Poor Health sector development in the PRSP context progress has been made in the areas of privatiza- tion, public enterprise reform, and fiscal reform. T he 1997 poverty assessment indicates that Social service reforms, as one of the key strate- almost 70 percent of the population live in gies for poverty reduction, were implemented in absolute poverty. Rural areas, where more both the education and health sectors. The follow- than 80 percent of the poor live, bear the greatest ing section focuses on the health sector reform. burden in terms of poverty incidence, depth, and severity. Based on the poverty assessment, the gov- Health sector development ernment developed an Action Plan for Reduction of Absolute Poverty or PARPA (2000-2004). The Health sector reform started after the war in poverty strategies outlined under the plan are response to the need to rebuild the health system (a) generating rapid and sustainable growth; that was damaged by the war. The objectives of the (b) investing in human capital through improved Health Sector Recovery Program (HSRP) launched delivery and quality of social services; and in 1995 were to increase access to and quality of (c) developing a program including safety nets that services by rehabilitating and adding to the net- fosters the social and economic integration of the work of first-level care facilities, and rural hospitals, most vulnerable groups. The new PARPA (2001- and by providing adequate staffing, drugs, and 2005) continues to emphasize the importance of supplies. The program, with an original cost of rapid and broad-based growth through creating a US$355 million for six years, has been supported by favorable climate for investment and productivity government budgets (33 percent) and external aid and promoting human development. Health has funds (67 percent), including a World Bank sector been identified as one of the six fundamental areas investment credit of almost US$100 million. In its for action because the health sector plays a key role five years of implementation, the HSRP has made in directly improving the well-being of the poor good progress in rebuilding the health infra- while it also contributes to economic growth. structure. Health coverage has been improving So far, Mozambique has performed well to meet steadily. the targets set in the poverty action plan. The over- The program originally aimed mainly at restorat- all growth of the economy has been strong and ing physical infrastructures destroyed during the inflation has been kept low, although the floods of war, but soon it started to redefine sector priorities 2000 have caused setbacks. The growth has been and readjust imbalances in resources allocation and broad-based, with agriculture, industry and serv- inequity in access to care. The government and ices growing more than seven percent. Substantial donors recognized that the health sector was 20 IMPROVING HEALTH FOR THE POOR IN MOZAMBIQUE 21 moving from a phase of recovery from the civil war Mounting an effective health sector response to a more forward-looking phase of improving the health system and services. The emergency man- The government's poverty program for the health agement approach would not meet the require- sector is targeted towards basic service delivery, ments of the sector's development. There was a with a vision that the poor will benefit from these need for policy reform in the sector. Under this services. However, without other specific targeting reform, the MOH stated its mission was to promote mechanisms in place, there is no guarantee that the and preserve the health of the people of Mozam- services will actually reach the poor. Health sector bique, and to promote and deliver services of good policies and services have to pay specific attention quality in a sustainable way, making them available to the poor and their health needs. A new health to all Mozambicans with equity and efficiency. The expenditure review will be carried out. This will mission statement was guided by the following provide more information to adjust policies and principles: efficiency and equity, flexibility and interventions for more pro-poor health outcomes. diversification, development of partnerships and community participation, transparency and Major health issues accountability, and integration and coordination. The main objectives of the health center are now to The morbidity and mortality of Mozambique's pop- increase the availability of good quality services ulation has a pre-transitional epidemiologic pattern, and to improve efficiency and equity. in which infectious diseases dominate over degener- Since the end of the war, the health sector in ative diseases. As indicated above, health in Mozam- Mozambique has attracted substantial contribu- bique is extremely poor, particularly for rural popu- tions from the international donor community. lation. Poor health can put people into poverty and Donor funds helped and continue to help in filling can keep them in poverty The poverty reduction in the financing gap for health. However, donor strategies therefore have to address the disease involvement often resulted in fragmentation and burden imposed on the nation and on families. inefficiency in resource management. The HSRP Diseases such as malaria and tuberculosis are was an important step towards more coordinated major causes of illness and death in the country. The sector financing. Even if it did not meet all the fights against such diseases are the health sector's requirements of sector-wide approach (SWAp), priorities. The "Roll Back Malaria" campaign pro- which provides an instrument for common plan- vides a window of opportunity to scale up the ning and managing of both government and donor national response, as does the Stop TB Initiative. TB funds. The sector is now moving towards the SWAp needs extra attention in the context of increasing with a broad sector policy framework and coordi- HIV prevalence, since the disease is now often nated resource management. Formulation of a new symptomatic for AIDS. Strategic Plan was part of the SWAp process. The Improving nutrition is another important ele- government and the large majority of the external ment in a poverty strategy for the health sector. The partners in the health sector signed a Code of Con- poor suffer disproportionately from malnutrition. duct in 2000, which defined the rules cooperation Furthermore, the attributable risk of malnutrition to between MOH and external partners. A new Health common morbidity and mortality far outweighs the Sector Strategic Plan (2001-2005) was approved by attributable risk of any other health condition. The the Council of Ministers in April 2001. health sector can play a crucial role in nutritional The financing strategy of the sector will focus on education and in making micronutrients such as (a) increasing the overall resource envelope; (b) Vitamin A and iron available to the people who improving the efficiency and execution of funds need them. School deworming programs also available through improved management of would have a large impact on the educational resources; (c) making resource allocation criteria attainment of otherwise anemic children. and methods more transparent and equitable; and The HIV/AIDS epidemic adds huge weight to (d) overhauling the user fee system. the national burden of disease. The impact of AIDS 22 AFRICA REGION HUMAN DEVELOPMENT WORKING PAPER SERIES on poverty at the national and household levels will ventions against AIDS have to be multi-sectoral. be enormous in the coming years. About 11,000 The role of the health sector, however, is critical. cases were registered by the end of 1998' however, Major health services or disease priorities that the registration probably captured less than 10 per- are explicitly listed under the government's poverty cent of all cases. Indications show that the adult program are: HIV rate is currently climbing rapidly and is now estimated at 12 percent. The prevalence rate in Polio eradication, elimination of neonatal Mozambique could soon become as high as in its tetanus, eradication of leprosy as a public health neighboring countries such as Malawi, Zambia, and problem, reduction of the incidence of common Zimbabwe unless drastic actions are taken immedi- diseases, such as HIV/AIDS, tuberculosis, ately to contain the epidemic. malaria, and childhood diseases that cause high HIV / AIDS poses a very serious threat to the morbidity and mortality; development of Mozambique, but its impact has not been fully recognized. It is an undeclared - Improvement of the nutritional status of the pop- national emergency. Denial and stigma associated ulation, especially children, including prevention with AIDS are still common in Mozambique of micro-nutrient deficiencies; although the impact of HIV / AIDS on society, com- munities, and households is becoming obvious. 9 Increased access to obstetric services; Only relatively recently has the government taken more aggressive steps to fight the HIV/AIDS epi- * Reduction in incidence of preventable diseases demic. In September 1999, the government adopted through (a) vaccination of children 0-23 months, a National Strategic Plan to Fight STDs / HIV / AIDS school-age children, and women of childbearing over the period 2000-2002. Given Mozambique's age, (b)expanding the coverage of the target limited implementation capacity, the plan focuses groups, and (c) introduction of Hepatitis B vac- realistically on population groups that are espe- cine; cially vulnerable to HIV/AIDS. It aims to provide essential prevention and care interventions to at Reduction of oral/ dental problems in school-age least 1,600,000 people with irregular sex partners children and adolescents, in strict cooperation and 15,000 people living with HIV/AIDS. The gov- with the education sector. ernment has also recognized that fighting HIV/AIDS requires a national response involving Serving the poor more effectively all sectors. It planned to create two coordinating bodies for a multisectoral response at the central The poverty study showed clearly that the main fac- level: an Inter-ministerial Committee for AIDS tors limiting access to health services by the poor (which will have general oversight responsibilities are distance and cost. and involve representation from eight ministries) and a National AIDS Commission (which will have Reducing geographic inequality in more direct national management responsibilities). access to care The implementation of the AIDS strategy requires not only resources at the central level but substan- The above analysis show that health resources are tial effort from communities. still unequally distributed among provinces and AIDS needs to be dealt with as one of the top pri- that the poor have to travel longer distances to any orities in the poverty reduction strategy. The AIDS health facilities and health personnel. The distribu- epidemic has to be addressed not only as a health tion of health staff does not necessarily match the issue, but also as a development issue and a poverty needs of the population. The distance issue is more reduction issue. Since AIDS affects every aspect of acute in the four northern provinces (Niassa, Cabo the society and can most often only be prevented Delgado, Nampula and Zamb6zia). Service statis- long before people get to a health facility, the inter- tics, especially the low utilization rates in those IMPROVING HEALTH FOR THE POOR IN MOZAMBIQUE 23 provinces, indicate the need for extra efforts to system does not generate substantial revenues. For reduce access barriers in these provinces. example, a consultation fee is about US$0.09 in With regard to distance, the system has to design 1997. The user fee system began in 1977. During the ways by which the services can be brought closer to 1980s, it was merely symbolic for cost recovery pur- communities. This is not necessarily a matter of poses. In 1996, it recovered 2.7 percent of the gov- building more health facilities. Other ways to ement's recurrent health spending. Nevertheless, ensure that services are delivered to under-served the household survey data show that people pay for areas include developing outreach services and services, and costs become barrier to health care, deploying of community-based health workers. In particularly for the poor. One of the major issues is accordance with the government's policy, a package the illegal charge to patients, which became perva- of basic services has to be provided to the whole sive in the 1990s. These illegal charges are a multi- population. Even after years of effort, by 1997, it ple of the official fees, but because of their very was estimated that only about 50 percent of the nature, little systematic information exists about population had access to most basic services. It is them. obvious that most of the people excluded from the When the user fee system was introduced, there system are the poor. Inclusion is expensive and was a sense of social justice, reflected by a long list requires specific geographic targeting. of exemptions. The exemptions clearly indicated Resource allocation in the past has not helped that those who could pay should do so while those much in reducing inequality among provinces. The who could not pay should not be penalized. The disadvantaged provinces received far fewer funds exemption list includes certain types of services, than better-off provinces. For example, per capita such as preventive care or STD care, and so on, and government expenditure on health in Maputo City certain categories of people, such as children under is US$2.8, while Zambezia only gets US$0.60 (1998). five, the elderly, the poor, and so on. In the reality, The government is taking steps to correct inequality the system is complex and rarely functions. There in access. At the beginning FY 2000, the Ministry of are no clear guidelines defining exemption cate- Health allocated its resources to the provinces in a gories or giving instructions on how to collect fees more equitable way. New budget allocations are and how to use the funds collected. The categories based less on historical patterns and more on popu- that are difficult to define, such as the poor, do not lation size and density. Given the debt relief, posi- get exempted. And even if the people get exempted tive economic development in the country, and from official fees, in order to actually get services, increasing proportion of the budget going to the they have to pay unofficial charges. The objectives health sector, the Ministry of Health is in a unique of the user fee system also evolved over time. The position to improve equity among provinces with- system was designed as a mechanism to generate out having to decrease any provincial allocation. revenue rather than as a measure to raise awareness Similarly, MOH can provide a larger proportion of of the value of health services and encourage social the budget to basic services without having to justice and the better use of services. As one of the decrease the budget for the tertiary and quaternary indictors under HIPC debt relief, the collection of level care. user fees has to be increased to 10 percent of the government's recurrent expenditure (IMF, 1997). Reducing inequality in financial Fee collection in recent years has increased from access to care US$0.7 million in 1996 to 1.7 million in 1998. At the same time the government's recurrent budget has Cost of services is another barrier preventing the increased significantly, so the fee collection target of poor from seeking care. Lack of money was the 10 percent is difficult to achieve. So far, user fees number one reason for the poor for not using serv- contribute to about 5 percent of the government's ices when they are ill. The user fee system has seri- recurrent budget. ous deficiencies. The official fees for health services The Expenditure Review (MSH 1999) analyzed are set relatively low in Mozambique, and the the user fee system and concluded that the fee col- 24 AFRICA REGION HUMAN DEVELOPMENT WORKING PAPER SERIES lection system is highly inefficient, abuses public investing more on facilities and human resources in resources, dissatisfies patients, and thwarts the underserved areas. achievements of public sector goals. It suggested The equity index the health sector uses in that the government define the objectives of the Mozambique is a very useful instrument to com- user fee system, set up realistic targets for fee col- pare the use of health services among population lection, simplify the fee structure, ensure the fees groups. Use of the index could be widened to meas- collected to be used as intended, improve financial ure inequities within provinces or districts, or management of the fee system, and regulate special between urban and rural populations. The current services. So far, no systematic assessment of illegal equity index already provides a focus on equity charges has been published. The issue of illegal well beyond what health systems in many other charges has to be addressed before considering any countries have achieved, but the index still may increase of fees to reach the revenue generation hide many inequities because it is based on aver- target. The system is clearly hurting the poor and ages. Therefore it seems important to use the small- increasing inequity. Given that the majority of pop- est possible unit of analysis, such as a district or ulation is poor, the user fee system will certainly not even areas within a district. generate substantial amount of funds for the health sector nor improve equity. The country needs to Improving overall service delivery explore other financing options that would promote equity and risk sharing. Efficiency in the use of funds and in getting value To improve both physical and financial access to for money are areas that need substantial improve- health care by the poor, the government will have to ment. focus more on resource allocation, service delivery, and the user fee system. The strategies need not be Increasing efficiency limited to improvement of supply. They can also involve changing the demand for services and Service outputs are increasing, but these increases making them more affordable. have yet to translate into better health outcomes. The strategies for improving supply and access Even though socioeconomic factors other than do not necessarily require building more physical health sector factors contribute to the poor health, infrastructures. One of the proposed solutions to one may still question the quality and efficiency of improving access is to move health services out of health services. Very limited data exist to allow an facilities and become more community-based. By adequate assessment of the efficiency of service pro- giving local communities more say in the services, vision in Mozambique. Under the expenditure the health care workers will need to become more review (MHS 1999), cost per service unit was used responsive to the actual needs of the people and to to measure efficiency in service provision. Signifi- treat patients in a humane way. Illegal charges can cant variations in cost per service unit were found only be controlled at the local level, where health among provinces. Maputo City has the highest cost workers become accountable to the communities per service unit, where a relative oversupply of they serve. Such a community-based approach fits workers may contribute to the service costs. A wide well with the decentralization policy of the govern- range of inefficiency indicates poor quality. How- ment. The health system needs to reach out to ever, without control of quality of services and case people, particularly to the poor. mix, it is difficult to draw conclusions about effi- The above analysis also suggests that targeting ciency from cost-per-service unit alone. A more strategies need to focus on geographic targeting to comprehensive study on quality and efficiency was help the poorer provinces catch up in terms of carried out in three provinces (Gaza, Niassa, and access and better use of resources. The health Zambezia) with analysis of various indicators for system needs to increase its inclusion of the popu- quality of antenatal care and outpatient consulta- lation, particularly the poor population that is usu- tion. The results also show the variations among ally left out of the system. This can be done by facilities and among the three provinces. Some facil- IMPROVING HEALTH FOR THE POOR IN MOZAMBIQUE 25 ities definitely use resources better than others do. explains the problem. Complex financing proce- Higher expenditures do not necessarily result in dures from the donor side and poor resource plan- better quality of services or overall sector perform- ning and management also contribute to under-use ance. or inefficient use of resources. An effort has been made to address those issues through system devel- Enhancing budget execution opment and program financing. Finally, to improve accountability and prudent The low ratio of actual expenditures versus allo- use of scarce resources, it is critical to increase cated budget indicates poor resource management involvement of the local population in the manage- and limited absorptive capacity. Even when the ment of those resources. While the center provides sector is clearly under-funded, funds often remain guidance on standards and targeting, decentraliz- unspent and have to be reprogrammed to the fol- ing resource management may increase efficiency lowing year. Poor absorptive capacity only partly and effectiveness. 26 AFRICA REGION HUMAN DEVELOPMENT WORKING PAPER SERIES References Cabral Z., 1999. A Study of Access to Basic Education and Ministry of Planning and Finance, Government of Health in Mozambique, Final Report: Community Mozambique, Eduardo Mondlane University, and Level, Oxfam, Mozambique. International Food Policy Research Institute, 1998. Christie F. and P. Ferrara, March 1999. "Health Expendi- "Understanding Poverty and Well-Being in Mozam- ture Review in Mozambique: The Social Context," bique: The First National Assessment (1996-97)." Management Sciences for Health. Mozambique National Health Services, Ministry of Council of Ministers, 2001. "Health Sector Strategic Plan Health, 1999 and 2000. "Informaqo Estatfstica (2001-2005-2010)." Republic of Mozambique. SumAria." Gwatkin D. R., S. Rutstein, K. Johnson, R. Pande, and A. Pavignani, E. and J. Durao, 1999. "Managing external Wagstaff, 2000. "Socio-Economic Differences in resources in Mozambique: building new aid relation- Health, Nutrition, and Population in Mozambique." ships on shifting sands?" Health Policy and Planning: HNP/Poverty Thematic Group. The World Bank. 14(3):243-253. Washington, D.C. Republic of. Mozambique, 1999. "National Report on Institito Nacional de Estatistica and Macro International Social Development, .1995-1998." Inc., 1997. "Moqambique, Inqukrito Demogrifico e de Republic of Mozambique, 2001. "Action Plan for the Saide." Reduction of Absolute Poverty (2001-2005)," 4th draft. International Monetary Fund and The International UNAIDS and World Health Organization, 2000a. Epi- Development Association, 2000. "Decision Point Doc- demiological Fact Sheets on HIV/AIDS and Sexually ument for the Enhanced Heavily Indebted Poor Coun- Transmitted Infections - Mozambique, Malawi, tries (HIPC) Initiative." Zambia and Zimbabwe. Mahon, J., and C. Schwabe, 2000. "Health Service Costs UNAIDS and World Health Organization, 2000b. AIDS and Cost Recovery in Cuamba." Epidemic Update: December 2000. Management Sciences for Health, 1999. "Health Expendi- UNDP, 1999. "Mozambique: Economic Growth and ture in Mozambique: An Analysis of Major Policy Human Development: Progress, Obstacles and Chal- Issues." lenges." National Human Development Report. Medical Care Development International. March 2000. UNDP, 2000. "Mozambique: National Human Develop- "Health Service Costs and Cost Recovery in Cuamba ment Report." District, Mozambique," Ministry of Health, Government of Mozambique, 1993-99. Annual Reports.  THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. Telephone: 202 477 1234 Facsimile: 202 477 6391 Internet: www.worldbank.org E-mail: afrhdseries@worldbank.org Since the 1992 Peace Agreement, Mozambique has made excellent progress to- wards recovering from its war-torn past, including in the health sector. However, enormous challenges remain. Mozambique is still one of the poorest countries in the world and large inequalities exist in the allocation of health resources, de- ployment of staff and availability of services across geographic areas and be- tween the poor and non-poor. Over the past few years, the Ministry of Health and its partners have worked together to prepare a new health sector strategy to guide the long-term devel- opment strategy for the sector. The results have been used to inform the prepa- ration of the country's Poverty Reduction Strategy Paper (PRSP), as part of a concerted national effort to articulate Mozambique's strategic plan to improve the welfare of the poor. Improving Health for the Poor in Mozambique: The Fight Continues attempts to summarize key aspects of the knowledge base upon which the health sector strategy was built. The study makes recommendations on how health sector re- forms can be made more pro-poor by targeting certain areas and population groups, by designing new delivery models, and by improving financial manage- ment. It is hoped that this work-in-progress will facilitate sharing of our evolving understanding of the link between health sector development and poverty re- duction as well as prepare the way for further documentation of this important link, as Mozambique's health sector strategy is implemented in the broader PRSP context.