UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 MODULEIII(B) EFFICIENCY, EFFECTIVENESS, AND EQUITY IN HEALTH SPENDING UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. © 2023 International Bank for Reconstruction and Development/International Development Association or The World Bank Group 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. The World Bank encourages dissemination of its knowledge, so this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; FAX: 202-522-2422; e-mail: pubrights@worldbank.org. Photo credits: Brian Amanya, 2023. Photo credit for the cover photo: Nurse attending to a patient, Kashozi HC3, Sheema District (Amanya Brian, 2022) Design/Layout: Artfield Graphics Printed in Uganda by Artfield Graphics Additional material relating to this report can be found on the World Bank Uganda website (www.worldbank.org/uganda). 2 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. Health worker drawing vaccine during the covid mass vaccination at Ministry of Health, Kampala (Rachel Mabala, 2021) UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 3 Table of Contents AABBREVIATIONS.............................................................................................................................................. V ACKNOWLEDGEMENTS....................................................................................................................................VII EXECUTIVE SUMMARY....................................................................................................................................VIII 1. INTRODUCTION....................................................................................................................................... 1 1.1 Public Expenditure Reviews in the Health Sector in Uganda...................................................................... 1 1.2 Specific Objectives of the Analyses in the Health Sector............................................................................ 2 1.3 Methodology and Data Sources................................................................................................................ 3 1.4 Key Features of the Health System in Uganda........................................................................................... 3 1.5 Health Status in Uganda........................................................................................................................... 5 2. LEVEL AND COMPOSITION OF TOTAL CURRENT HEALTH EXPENDITURE................................................ 7 2.1 Overview of Total Current Health Expenditure............................................................................................ 7 2.2 Government Health Expenditure................................................................................................................ 8 2.3 External Development Partners (Donor) Expenditure on Health................................................................ 10 2.4 Households Expenditure on Health.......................................................................................................... 11 2.5 Alignment of the overall spending on health to the disease burden......................................................... 12 3. COMPOSITION OF PUBLIC EXPENDITURE ON HEALTH.......................................................................... 13 3.1 Overview................................................................................................................................................ 13 3.2 Public Expenditure on Health at Sub-Regions........................................................................................ 13 3.3 Public Expenditure on Health by Level of Healthcare............................................................................... 14 3.4 Distribution of Public Expenditure on Health at Public Health Facilities by Facility Type............................ 15 3.5 Health Expenditure by Key Health Systems Inputs.................................................................................. 16. I UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 4. EFFICIENCY AND EFFECTIVENESS OF SPENDING IN THE HEALTH SECTOR......................................... 19 4.1 Benchmarking Uganda with Peer Countries............................................................................................ 19 4.2 Budget Performance............................................................................................................................... 20 4.3 Availability and Productivity of the Health Workforce............................................................................... 21 4.4 Supply and Availability of Drugs and Medical Products........................................................................... 24 4.5 Findings From Existing Studies on Efficiency in the Health Sector in Uganda........................................... 25 4.6 Detailed Review of the Health System at Local Government Level........................................................... 25 4.6.1 Planning, Budgeting and Financial Management..................................................................................... 26 4.6.2 Health Resources Management and Functionality of Health Facilities...................................................... 26 4.6.3 Drugs and Medical Supplies................................................................................................................... 27 4.6.4 Recommendations - Health System at Local Government Level............................................................. 29 5. EQUITY ANALYSES................................................................................................................................. 31 5.1 Introduction............................................................................................................................................ 31 5.2 Benefit Incidence Analysis...................................................................................................................... 31 5.2.1 Trends in Benefit Incidence..................................................................................................................... 32 5.2.2 Trends in Benefits Received Compared to Needs.................................................................................... 39 5.3 Catastrophic Health Spending and Impoverishment due to Health Spending........................................... 40 5.3.1 Trends in Incidence of Catastrophic Health Spending.............................................................................. 41 5.3.2 Incidence of Household Impoverishment due to OOP Spending on Health............................................... 45 REFERENCES.................................................................................................................................................................. 50 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. II List of Figures Figure 1.1 Trends in Ownership of Health Facilities in Uganda.................................................................................... 14 Figure 2.1 Trends in Total CHE in Uganda................................................................................................................... 18 Figure 2.2 Trends in General Government Health Expenditure – Nominal Terms......................................................... 19 Figure 2.3 Regional Comparison of GGHE-D – 2020/21............................................................................................ 19 Figure 2.4 Nominal vs Real General Government Health Expenditure – Domestic....................................................... 19 Figure 2.5 Trends in Total External (Donor) Expenditure on Health – Nominal Terms................................................... 20 Figure 2.6 Trends in On- and Off-Budget Donor Funding.......................................................................................... 21 Figure 2.7 Composition of Households Spending on Health........................................................................................ 21 Figure 2.8 Top 10 Causes of Death and Disabilities in 2019 and percent change 2009–2019, all ages...................... 22 Figure 2.9 Total CHE by Diseases and Conditions, 2016/17–2020/21........................................................................ 22 Figure 3.1 Growth in Per Capita Expenditure on Health at Sub-Regions..................................................................... 23 Figure 3.2 Public Expenditure on Health at Public Health Facilities............................................................................. 26 Figure 3.3 Breakdown of Expenditure by Key Health Systems Inputs......................................................................... 26 Figure 3.4 Health wage bill as a share of the total public expenditure on health......................................................... 27 Figure 3.5 Adequacy of Public Spending on Drugs and Medical Commodities............................................................ 27 Figure 4.1 Producing quality services with available funding – 2020/21.................................................................... 29 Figure 4.2 Translating available services into better health outcomes – 2020/21....................................................... 30 Figure 4.3 Health Sector Budget Performance........................................................................................................... 30 Figure 4.4 Trends in training outputs and staff in-post.............................................................................................. 31 Figure 4.5 Distribution of Skilled Health Providers by Sub-Region in Uganda - 2022................................................ 32 Figure 4.6 Productivity of Skilled Health Providers by Sub-Region, 2022................................................................... 33 Figure 4.7 Effectiveness of Maternal Healthcare by Sub-Region, 2022...................................................................... 33 Figure 4.8 Expenditure on Drug and Medical Commodities: 2016/17–2020/21.......................................................... 34 Figure 5.1 Per Capita Public Spending and Under-Five Mortality Rate....................................................................... 41 Figure 5.2 Trends in Benefit Incidence: Public Health Centers.................................................................................... 42 Figure 5.3 Trends in Benefit Incidence: Public Hospitals............................................................................................. 43 Figure 5.4 Trends in Benefit Incidence: Private Health Centers................................................................................... 43 Figure 5.5 Trends in Benefit Incidence: Private Hospitals............................................................................................ 44 Figure 5.6 Trends in Benefit Incidence in Urban and Rural Areas: Public Health Centers............................................. 44 Figure 5.7 Trends in Benefit Incidence in Urban and Rural Areas: Public Hospitals...................................................... 45 Figure 5.8 Trends in Benefit Incidence in Urban and Rural Areas: Private Health Centers............................................ 45 III UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. Figure 5.9 Trends in Benefit Incidence in Urban and Rural Areas: Private Hospitals.................................................... 46 Figure 5.10 Trends in Total Benefits for Outpatient Services: 2012/13.......................................................................... 47 Figure 5.11 Trends in Total Benefits for Outpatient Services: 2016/17.......................................................................... 47 Figure 5.12 Trends in Total Benefits for Outpatient Services: 2019/20.......................................................................... 47 Figure 5.13 Trends in Total Benefits for Inpatient Services: 2012/13............................................................................ 48 Figure 5.14 Trends in Total Benefits for Inpatient Services: 2016/17............................................................................ 49 Figure 5.15 Trends in Total Benefits for Inpatient Services: 2019/20............................................................................ 49 Figure 5.16 Comparing Total Benefits with Total Needs: 2012/13–2019/20.................................................................. 50 Figure 5.17 Trends in the Incidence of CatHE - 10 Percent Threshold.......................................................................... 51 Figure 5.18 Incidence of CatHE by Socioeconomic Quintiles - 10 Percent Threshold................................................... 51 Figure 5.19 Incidence of CatHE by Poor and Non-Poor: 10 Percent Threshold............................................................. 52 Figure 5.20 Incidence of CatHE by Rural and Urban Setting - 10 Percent Threshold.................................................... 52 Figure 5.21 Incidence of CatHE by Region - 10 Percent Threshold.............................................................................. 53 Figure 5.22 Incidence of CatHE by sub-region in 2019/20: 10 Percent Threshold........................................................ 53 Figure 5.23 Incidence of Impoverishment due to OOP Spending on Health................................................................... 55 Figure 5.24 Incidence of Impoverishment due to OOP Spending on Health - Rural vs Urban........................................ 57 Figure 5.25 Incidence of Impoverishment Due to OOP Spending on Health by Socioeconomic Groups.......................... 58 Figure 5.26 Incidence of impoverishment due to OOP Spending on Health by Region................................................... 58 List of Tables Table 1.1 Summary of Key Issues and Recommendations from this Public Expenditure Review.................................. 6 Table 1.2 Implementation of Recommendations from Previous PERs in the Health Sector......................................... 11 Table 1.3 Health Facilities in Uganda – 2022/23....................................................................................................... 14 Table 1.4 UHC Index Scores: Uganda vs Benchmark Countries................................................................................. 15 Table 1.5 Key demographic and health indicators..................................................................................................... 16 Table 2.1 Key Health Finance Indicators for Uganda – 2020/21................................................................................ 17 Table 2.2 Current Health Expenditure and Per Capita Expenditure............................................................................ 17 Table 3.1 Growth in Public Expenditure on Health by Level of Healthcare: 2020/2021 vs 2018/19............................ 25 Table 3.2 Trends in Public Expenditure at Public Health Facilities by Facility Type..................................................... 25 Table 5.1 Factors Influencing CatHE - 10 Percent Threshold.................................................................................... 54 Table 5.2 Impoverishment at Uganda’s National Poverty Line (2005/6 - 2019/20)................................................... 56 Table 5.3 Impoverishment at International Poverty Lines (2005/6 - 2019/20).......................................................... 56 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. IV Abbreviations CatHE Catastrophic health expenditure CHE Current Health Expenditure GDP Gross Domestic Product GGHE-D General Government Health Expenditures from Domestic Sources GGHE General Government Health Expenditure GoU Government of Uganda HAQ Health Access and Quality Index HCDP Human Capital Development Program HHFA Harmonized Health Facility Assessment HRH Human Resources for Health IFMIS Integrated Financial Management Information System MMR Maternal Mortality Rate MoFPED Ministry of Finance, Planning and Economic Development MoH Ministry of Health MTEF Medium Term Expenditure Framework NGO Non-Governmental Organisation NMS National Medical Stores OOP Out-of-Pocket PER Public Expenditure Review PFP Private-for-Profit PHC Primary Health Care PNFP Private-not-for-Profit RBF Results-Based Financing SDG Sustainable Development Goals UBOS Uganda Bureau of Statistics UgIFT Uganda Intergovernmental Fiscal Transfer UGX Uganda Shilling UHC Universal Health Coverage UNHS Uganda National Health Survey URMCHIP Uganda Reproductive, Maternal and Child Health Services Improvement Project WHO World Health Organization V UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. Covid vaccination center in Kampala (Rachel Mabala, 2022) UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. VI Acknowledgements T he Uganda Public Expenditure Review 2022–23 was undertaken as a programmatic, analytical, and advisory task executed jointly between the Government of Uganda and World Bank staff and consultants. At the World Bank, the overall task was led by Rachel K. Sebudde (Senior Economist, EAEM1), while the Government was led by Mr Charles Byaruhanga (Advisor, Budget). Specifically for Module III(B), the work was led by the Health, Nutrition, and Population (HNP) Global Practice of the World Bank Group. The team was comprised of Collins Chansa (Task Team Lead and Senior Health Economist), Rogers Ayiko (Senior Health Specialist), Ezrah Trevor Rwakinanga (Consultant), William Kiganda (Consultant), Elizabeth Mabel Asege Ekochu (Health Specialist), Rachel Sebudde (Senior Economist), Julia Mensah (Senior Health Specialist), and Brendan Michael Hayes (Senior Health Specialist). Great collaboration and leadership was provided by the Permanent Secretary and Secretary to Treasury, Mr. Ramathan Ggoobi, Dr. Diana Atwine, the Permanent Secretary (Ministry of Health) and Mr. Ishmael Magona (Ag Director Budget), with the core team working closely with the World Bank led by Dr. Sarah Byakika, and comprising Dr. Tom Aliti, Mr. Richard Kabagambe-Tureebe, Mr. Nimrod Agasha (Economist and Health Desk Officer, Ministry of Finance Planning and Economic Development), and Mr Charles Byaruhanga (Advisor Budget), Ministry of Finance, Planning and Economic Development. The team appreciates the overall guidance of Vivek Suri (Practice Manager, EA1M1) during the first phase of the work; Philip Schuler (Lead economist/Ag. Practice Manager); Abha Prasad (Practice Manager AE1M1); and specific health sector guidance from Francisca Ayodeji Akala (Practice Manager); Anne Margreth Bakilana (Program Leader); Marek Hanush (Program Leader, AFCE2); Amparo Elena Gordillo- Tobar (Senior Health Specialist); Rosemary Mukami (Country Manager, AFMUG); and Keith Hansen (Country Director, AFCE2). The team would like to thank Esther Ampumuza, Pearl Namanya, and Catherine Joy Ajiku Obitre Gama for providing logistical and administrative support throughout the process. We are also grateful to Virginia Larby for the editorial support. We are grateful for the good partnership we enjoyed with the Ministry of Health and Ministry of Finance, Planning and Economic Development, and for the access they gave to most of the data used in this report – the preliminary findings of which were shared and discussed extensively with them. The contents of this report were discussed and validated in various PER focused-consultative meetings and the Ministry of Health Strategic Planning, Financing, and Development Technical Working Group. VII UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. Executive Summary Key Findings Health Outcomes in Uganda Over the past two decades, Uganda has made significant progress towards the attainment of Universal Health Coverage (UHC). The UHC service coverage index score for Uganda increased from 22 in 2000 to 49 in 2021 which implies that investments in the health sector in Uganda are bearing fruit. Consequently, this has contributed to improved health and nutrition outcomes. For example, between 2011 and 2022, the Under-5 Mortality Rate reduced from 90 to 52 deaths per 1,000 live births, while the Maternal Mortality Ratio reduced from 438 to 189 deaths per 100,000 live births over the same period. The prevalence of stunting among children below the age of 5 years also reduced from 33 percent to 26 percent between 2011 and 2022. Despite these improvements, fertility levels among the adolescents and women in general, are higher in Uganda than the averages for low-income countries. This contributes to the high teenage pregnancy rate in Uganda which is estimated at 24 percent among women aged 15-19. In addition, the percentage of children aged 12-23 months with all basic vaccinations estimated at 54 percent in 2022 is very low. Furthermore, Uganda is vulnerable to disease outbreaks given its location and faces a growing incidence and prevalence of non- communicable diseases, for which its health system is inadequately prepared for. Service availability and readiness at health facilities are still very low in Uganda. According to the 2022 Harmonized Health Facility Assessment, the general service readiness index is 0.59. This means that only 60 percent of the facilities are ready to provide services. These gaps are symptomatic of a poorly financed or managed health system. Thus, Uganda requires additional investments in the health sector to improve the quality of health service delivery. In light of these issues, Uganda has to do much more to achieve the health-related Sustainable Development Goals (SDGs). At the sub-national level, where local government manages the delivery of health services, the challenges are profound. The governance and management functions at district level are weakened by inadequate funding and staffing and weak capacities in the management of health facilities, especially medicines and financial management. Adequacy of Total Spending on Health in Uganda The overall level of spending on health in Uganda is inadequate. At US$50.5 per capita, Uganda’s total current health expenditure (CHE) per capita is significantly less than the average spending of US$75.6 per capita in aspirational peer countries, and the recommended US$86 per capita in low-income countries to provide basic health services. Households and external development partners contribute 85 percent of the total CHE in Uganda. In 2020/21, expenditure by external development partners was the highest at 55 percent of total CHE, followed by households at 30 percent, government at 13 percent, and corporations (employers) at 1.8 percent. Households’ contribution to the total CHE is mostly through out-of-pocket (OOP) payments on medicines (49 percent). This high level of OOP spending on health is one of the leading causes of catastrophic expenditure on health, forgone care, and impoverishment. There is low prioritization of health by the Government of Uganda (GoU). General government health expenditure from domestic sources (GGHE-D) as a share of the total government expenditure (TGE) has declined from 6.5 percent in 2014/15 to 3.9 percent in 2020/21. Declining GGHE-D as a share of TGE implies that the GoU is not adequately prioritizing funding to the health sector. Consequently, funding of health services in Uganda is dependent on households and external development partners who provide 85 percent of the total CHE. Uganda’s GGHE-D per capita estimated at US$6.8 in 2020/21 is also below the average level of spending in peer countries. When the GGHE-D is adjusted for inflation, the results show that the value of GGHE-D was almost half the nominal value over the period 2014/15– 2020/21. External development partners are still the largest source of funding to the health sector in Uganda. In nominal terms, external funding to the health sector was relatively the same over the period 2014/15–2018/19 but increased substantially by 56 and 26 percent in 2019/20 and 2020/21, respectively. However, 84 percent of the total external spending in the health sector over the period 2014/15–2020/21 was off-budget. Off-budget funding reduces government’s flexibility in resource allocation, its ability to re-prioritize funding to emergent needs, and is a missed UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. VIII opportunity to improve the public financial management system in the country. The large share of external funding and off-budget support poses a huge risk of making health financing unsustainable. In general, health expenditures from the primary sources of financing in Uganda are aligned to the disease burden. Focusing health expenditures on the leading causes of mortality and disability has enabled Uganda to substantially reduce the prevalence of its most common diseases and conditions. Composition of Public Expenditure on Health There are variations in the level and growth in per capita public expenditure on health across the sub-regions in the country. This suggests that the existing formula for allocating financial resources from the center to the local government level has not been fully effective at adjusting the variations across the sub-regions. However, variations in funding by sub-regions can also be affected by differences in the number and type/size of the health facilities, and the number and mixture of health workers’ skills in each sub-region. Uganda focuses more on funding higher-level health care than primary health care (PHC). Government- owned regional, specialized, and national referral hospitals get 83 percent of all public spending at government health facilities, even though they only make up one percent of all government health facilities. On the other hand, district- level government health facilities receive only 17 percent of all public spending at government health facilities despite making up 99 percent of all government health facilities. While PHC services can be provided at all levels of the health system, the problem then is that higher-level facilities are not being used as intended. Therefore, the government has to examine the functionality of health facilities in the country and rationalize expenditures at different levels. Spending on salaries and wages as a share of the public expenditure on health in Uganda has been growing over the years and is within the historical spending patterns in Eastern Africa. Uganda’s health workforce wage bill as a share of the total public expenditure on health has increased from 24 percent in 2016/17 to 37 percent in 2020/21. However, the health wage bill in the health sector in Uganda is still below the recommended range of 45 to 60 percent for low-income countries in Africa. Despite the increase in the volume of public expenditure on drugs and medical commodities in nominal terms, the level of spending in Uganda is still significantly below expected norms. At 29 percent, the share of expenditure on drugs and medical commodities is lower than the African regional average of 33 percent (Bennett et al 1997). Secondly, the per capita spending on drugs and medical commodities – estimated at US$3.2 in 2020/21 in Uganda – is significantly below the recommendation by the Lancet’s Commission on Essential Medicines for low-income countries to finance a basic package of 201 essential medicines at US$13 to US$25 per capita. Inadequate spending on drugs and medical commodities in Uganda is one of the factors contributing to the low availability of tracer drugs and commodities at health facilities in the country. Only 46 percent of the health facilities had over 95 percent of the tracer drugs and medical commodities in 2019/20 and 2020/21. There are some discrepancies between spending on drugs and medical products and their availability due to lack of a system to link the electronic procurement system at the National Medical Stores (NMS) to the Integrated Financial Management Information System (IFMIS) at the Ministry of Finance, Planning and Economic Development (MoFPED). Thus, multi-year framework contracts for drugs and medical products are signed outside the IFMIS. Because contract management is entirely out of the system, the IFMIS internal budgetary controls do not apply as each contract has to be checked manually to ensure compliance with available budgetary allotments. Efficiency and Effectiveness of Spending in the Health Sector Government budget performance in the health sector over the period 2016/17–2020/21 was fairly good. On average, about 97 percent of the budgeted funds were released, 97 percent of which were used. However, given the need for additional funding in the health sector, it is expected that 100 percent of the budgeted funds are released and used each year. Despite having a lower total CHE per capita, Uganda’s 2020/21 score on the health access and quality (HAQ) index was comparatively the same as the score for its aspirational peers, Rwanda, and Kenya. This suggests that Uganda is more efficient than its aspirational peers and some countries in East Africa (Rwanda and Kenya) at using its limited resources to produce quality health services. However, this analysis doesn’t take into consideration differences in the cost-of-service provision. Considering that health service provision in Uganda is generally less expensive, this IX UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. could be the reason why Uganda is relatively more efficient than peer countries. For example, salaries and wages for health workers in Uganda are much lower than in Kenya and Rwanda. Uganda is also more effective at using the existing health services to produce better maternal health outcomes than aspirational peer countries, Congo Republic, Kenya, Sudan, and Cameroon. However, there are substantial disparities in maternal mortality rates within the country, particularly across the sub-regions. This is explained below. Some sub-regions are more efficient or productive in translating resources into outputs than others. Sub- regions like Karamoja and Kigezi have relatively lower institutional deliveries than other sub-regions even though they have more skilled health providers. This could be explained by other limitations to service delivery such as physical and economic access, and social acceptability of the services provided. Countrywide, maternal deaths at public facilities are higher in sub-regions which have higher numbers of skilled health providers. The only exceptions are Bukedi, Tooro, and Kigezi sub-regions. These results are very concerning and suggest that there are huge implementation gaps including poor quality of maternal healthcare services in almost all the sub-regions in Uganda. Despite an increase in the health workforce in the public sector from 59,105 in 2016/17 to 64,808 in 2019/20, the number of skilled health providers in Uganda is critically low and the distribution is skewed towards some sub-regions. In 2022, the highest number of skilled health providers per 10,000 population was in Kampala at 13.6 which was more than twice the national level (Uganda) estimate of 5.9. This means that none of the sub-regions – nor Uganda as a whole – had met the required standard of 23 skilled health providers per 10,000 population. However, the data only includes skilled health providers in the public sector, suggesting that the density of these providers could increase if private sector data is included. Nevertheless, the national target of 23 skilled health providers per 10,000 population will still be unachieved given that the government is the largest employer in the health sector. Furthermore, given Uganda’s desire to reach the health-related SDGs, the SDG target of 44.5 physicians, nurses, and midwives per 10,000 people is a long way off. Recruitment and deployment of health workers to rural areas is a challenge and grossly affects the provision of quality health care. While the annual number of health cadres graduating from health training institutions in Uganda is still low, most of them are not absorbed into the system. This could be attributed to: (i) the inefficient decentralized system of recruiting health workers with several districts failing to fully absorb their wage budgets each year; (ii) continued training of health workers that the health sector does not need; (iii) weak system for redistributing existing health workers across the districts to optimize the numbers; and (iv) ineffective Human Resources for Health (HRH) retention and management strategies for staff working in rural areas. Sub-regions with higher levels of per capita public spending on health generally have higher levels of skilled health providers. This means that the distribution of health workers is a key factor in how financial resources are distributed in the public health sector in Uganda. It also implies that geographical allocation of public funds in the health sector could be improved by redistributing health workers. However, as provided above and in more detail in the main report, a high number of skilled health providers in a sub-region is not necessarily associated with more institutional deliveries and low institutional maternal deaths. Health workers in some sub-regions are more efficient in translating available resources into outputs; and more effective at translating the services into better outcomes. This means that there are implementation gaps across the sub-regions in Uganda and this leads to the provision of poor- quality maternal healthcare services. The low productivity of health workers and poor maternal health outcomes in Uganda can also be attributed to the high level of absenteeism. In Uganda, absenteeism is estimated at 46 percent, which implies that almost half of the health workers are absent from duty most of the time. Having half of the health workforce absent on any given day creates a significant challenge for providing timely, efficient, high-quality health services. The absentee staff place a heavy workload on the staff on duty which could lead to the provision of poor-quality health care and ‘burn-out’. Ultimately, there is wastage of resources, and this contributes to poor health outcomes. Given the high proportion of the public expenditure on HRH in Uganda, this study estimates direct annual losses from absenteeism at about UGX292 billion (US$78.5 million) through “wasted” payments for salaries to absentee health care workers. A study by the Uganda Inspectorate of Government estimated the total annual direct losses due to corruption in the health sector at US$180.2 million, which is equivalent to 25 percent of the annual government spending on health. UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. X A multiple case study in three districts (Ntungamo, Mukono, and Nwoya) revealed the following: a) Gaps in expenditure tracking at the health facilities; and a disconnect between the level of financing, expenditure on drugs and medical supplies, patient data and results; b) Significant deficiencies in the management of the payroll for health workers. A number of health workers were being paid through non-health facility cost centers and non-existent health centers. The underlying problem is the inconsistency across key government documents and databases, namely: the master facility list, staff list, payroll register, and the NMS database; c) ‘Non-existent’ health centers in Ntungamo district are being used to pay health workers’ salaries. Though the health workers exist, and they are serving at other health centers, it wasn’t clear why their pay points are under non-existent health centers; d) Health facility managers have limited control of their budgets for drugs and medical supplies. This leads to limited transparency on procurements and deliveries, and it contributes to the expiry of drugs and medical commodities. It was revealed that NMS often delivered some commodities that are not needed and/or ordered by the health facilities; e) The online system for making orders for drugs and medical supplies at NMS was commendable but the system didn’t allow the health facilities to retain a copy of their requisitions. Furthernore, NMS was reported to be undermining the oversight role of health facility managers, district authorities, and community leaders because they rendered very little time for a thorough inspection of the deliveries. Equity in Financing and Access to Health Services Do the Poor Benefit from Public Healthcare Services in Uganda? Health services provided at public health centers in both rural and urban settings are accessed more by the poor than the non-poor. This implies that the GoU’s free healthcare policy and other UHC reforms have been positive. However, some parts of the country still don’t have access to quality health services because: (i) the distribution of health facilities is still skewed towards the urban areas, (ii) public health centers are underrepresented in urban areas, (iii) there are accessibility challenges, especially for people living in remote and marginalized communities (Batwa, Ik, islands, etc), and (iv) the quality of health services offered at public health centers is still below par. Poor and non-poor households accessed private health centers almost equally. This means that both the poor and the non-poor are exposed to OOP payments for health services. For the private health centers in urban settings, results show that they were mostly accessed by the poor. This raises concern on the exposure of the urban poor to OOP payments on health in urban areas. The trend in the health benefits received compared to the health need has progressed in favor of households in the poorest, poorer, and middle-class income groups. Over the period 2012/13–2019/20, the health benefits and needs of the households in the poorest, poorer, and middle-income groups increased. In addition, households in these three income groups benefited more than their health needs in 2012/13 and 2019/20. Catastrophic Health Spending and Impoverishment due to Health Spending The incidence of catastrophic health spending (CatHE) at the 10 percent threshold has declined from 22.4 percent in 2005/6 to 11.9 percent in 2019/20. Furthermore, the incidence of CatHE is higher among the non-poor than the poor households. However, 11.9 percent of the 8.93 million households in Uganda in 2019/20 is equivalent to 1.1 million households. This is equivalent to 4.9 million people, and it means that millions of Ugandans are still experiencing CatHE. Furthermore, the relatively low incidence of CatHE among the poor could also mean that some poor households have forgone health care and not necessarily that there is increased financial protection for the poor. Thus, it is possible that their health needs are not met, even within the context of a “free healthcare” policy. Therefore, lower incidence of CatHE among the poor may not be because there is a protection mechanism for them, but most likely because they are too poor to seek care when they need it. The incidence of CatHE varies across the sub-regions of Uganda, with the highest incidence in Teso and West Nile sub-regions and the lowest incidence in Kigezi and Acholi sub-regions. CatHE was also higher in rural than urban areas. The main factors associated with CatHE were gender and age (having children below the age of five and XI UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. households headed by old people (>60yrs)). The number of households falling into poverty due to healthcare expenditures reduced during the period under review. The percentage of households that fell below the national poverty line due to healthcare expenditures decreased from 4.6 percent in 2005/6 to 2.3 percent in 2019/20. Despite this reduction, about 206,405 households (949,464 people) were pushed into poverty. Using the international poverty line, the percentage of households that have been impoverished due to healthcare expenditures reduced from 5.2 percent in 2005/6 to 2.6 percent in 2019/20, which translates into about 233,328 households (1.1 million people). Although there has been a decline in impoverishment due to healthcare expenditures, households in rural areas are more likely to be impoverished. The incidence of impoverishment among households in rural areas declined from 5.6 percent in 2005/6 to 2.9 percent in 2019/20, which is 181,374 households (870,597 people). On the other hand, the incidence of impoverishment among households in urban areas declined from 2.9 percent in 2005/6 to 1.9 percent in Key Recommendations Based on the findings, Table 1.1 provides the key recommendations that the government and other stakeholders operating in the health sector could use to address the identified challenges. Table 1.1 Summary of Key Issues and Recommendations from this Public Expenditure Review Issue Theme Key Findings Short Term (0-2 yrs) Medium Term (3-5 yrs) Long Term (6 yrs plus) 1. Health Uganda has made significant Strengthen primary Strengthen primary Strengthen primary outcomes and achievements in health and healthcare including at the healthcare including at the healthcare including at the readiness of the nutrition outcomes. Despite community levels, health community levels, health community levels, health health system to these improvements, service promotion and disease promotion and disease promotion and disease deliver services availability readiness at prevention. prevention. prevention. health facilities is low with an average of 59 percent. Cautiously scale up priority Cautiously scale up priority Cautiously scale up priority Key readiness challenges specialized healthcare specialized healthcare specialized healthcare include low staffing levels, programs while taming the programs while taming the programs while taming the diagnostics capacity, cost of such scale up. cost of such scale up. cost of such scale up. perennial medicine stock Establish and/or update Implement, monitor, and Implement, monitor, and outs, and inadequate service delivery performance regularly adapt service regularly adapt service amenities (water and agreements with private delivery performance delivery performance electricity). health providers. agreements with private agreements with private While there have been health providers. health providers. Establish and strengthen significant public patient-focused feedback Implement, monitor, Implement, monitor, investments in the health and grievance redress and regularly adapt and regularly adapt sector, these investments mechanisms across patient-focused feedback patient-focused feedback have been skewed towards the health system to and grievance redress and grievance redress curative, hospital-based monitor patient-reported mechanisms across the mechanisms across the care. experiences and outcomes. health system. health system. Private sector involvement in Uganda is high and MoH has long-term service agreements with the private-not-for-profit sector 2. Adequacy of The overall level of Increase and ring-fence Increase and ring-fence Increase and ring-fence total spending spending on health in government spending government spending government spending on health in Uganda is inadequate, with on health to (i) reduce on health to (i) reduce on health to (i) reduce Uganda per capita expenditure at dependence on dwindling dependence on dwindling dependence on dwindling US$50.5. This is less than and uncertain external and uncertain external and uncertain external the average spending in funding amid global crises funding amid global crises funding; (ii) reduce the risk aspirational peer countries threatening further cuts in threatening further cuts in of catastrophic spending on (US$75.6 per capita) and development assistance; development assistance for health and impoverishment. the recommended US$86 (ii) reduce the risk of health; (ii) reduce the risk per capita for low-income catastrophic spending on of catastrophic spending on countries to provide basic health and impoverishment health and impoverishment. services. due to high OOP. UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. XII Issue Theme Key Findings Short Term (0-2 yrs) Medium Term (3-5 yrs) Long Term (6 yrs plus) GoU is not adequately Develop a health financing Implement the health Implement the health prioritizing funding to the sustainability plan aimed at financing sustainability plan financing sustainability plan health sector; GGHE-D as sustaining the financing of with an increasing share of with an increasing share of a share of TGE has declined health services in the face domestic resource allocation domestic resource allocation from 6.5 percent in 2014/15 of dwindling and uncertain to health. to health. to 3.9 percent in 2020/21; external funding. and development partners Implement, monitor, and Implement, monitor, and and households provide Establish the national social adapt the SHI scheme. adapt the SHI scheme. 85 percent of the total health insurance (SHI) scheme with the primary Engage development Engage development CHE (55 and 30 percent, focus on enhancing strategic partners and ensure partners and ensure respectively). purchasing, protecting alignment of development alignment of development While development partners people from catastrophic assistance to national health assistance to national health are still the main source spending on health, and policy priorities policy priorities of funding to the health to some extent, raising sector, 84 percent of the revenues as experienced total external spending in has shown elsewhere . the health sector over the period 2014/15–2020/21 Engage development was off-budget. Off- partners and ensure budget funding reduces alignment of development government’s flexibility in assistance to national health resource allocation and is policy priorities through unsustainable. signing compacts and strengthening the existing accountability systems. 3. Composition Public spending on health Prioritize spending on PHC, Prioritize spending on PHC, Prioritize spending on PHC, of health is skewed towards higher disease prevention and disease prevention and disease prevention and spending in the level health facilities (i.e., health promotion to reduce health promotion. health promotion. public sector regional, specialized, the higher risks and costs and national referral associated with curative hospitals). Comparatively, care. these facilities make up one percent of all Adequately resource PHC Adequately resource PHC the government health facilities and improve facilities and improve Adequately resource PHC facilities whereas district- availability of inputs, availability of inputs, facilities and improve level facilities make up leadership and governance, leadership and governance, availability of inputs, 99 percent of all the functionality of facilities, functionality of facilities, leadership and governance, government health facilities. and quality of care at PHC and quality of care at PHC functionality of facilities, However, only 17 percent of facilities. facilities. and quality of care. This will the total public spending at help to decongest referral government health facilities facilities and allow them to goes to district-level execute their primary role at facilities. lesser costs Spending on salaries and wages increased from 24 to 37 percent of total public expenditure between 2016/17 to 2020/21 mainly due to increases in staff recruitments. The level of spending is similar to that of Kenya and Tanzania, but below Southern African countries and the expected range for African countries (45-60 percent). In nominal terms, there has been a 65 percent increase in spending on medicines and medical commodities between 2016/17 and 2020/21. XIII UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. Issue Theme Key Findings Short Term (0-2 yrs) Medium Term (3-5 yrs) Long Term (6 yrs plus) In per capita terms, the increase was from US$2.2 in 2016/17 to US$3.2 in 2020/21. This is still lower than the recommended spending of US$13-US$25 per capita on medicines and medical commodities for low-income countries and the expected range for African countries (45-60 percent). In nominal terms, there has been a 65 percent increase in spending on medicines and medical commodities between 2016/17 and 2020/21. In per capita terms, the increase was from US$2.2 in 2016/17 to US$3.2 in 2020/21. This is still lower than the recommended spending of US$13-US$25 per capita on medicines and medical commodities for low- income countries. 4. Efficiency Government budget Increase the availability and Increase the availability and Increase the availability and and performance is good with 97 productivity of the health productivity of the health productivity of the health effectiveness of percent of allocated funds workers by: (i) improving the workers by: (i) improving the workers by: (i) improving the spending in the released and 97 percent of recruitment and retention recruitment and retention recruitment and retention health sector released funds used. in local governments; (ii) in local governments; (ii) in local governments; (ii) re-distributing the available re-distributing the available re-distributing the available Aggregated results at health workforce across the health workforce across health workforce across national level shows that sub-regions and districts; the sub-regions and the sub-regions and Uganda is more effective (iii) scaling-up RBF with the districts; (iii) scaling-up districts; (iii) scaling-up than aspirational peer potential to address staff RBF with the potential; (iv) RBF with the potential; (iv) countries at using the absenteeism; (iv) enhancing enhancing mentorship and enhancing mentorship and existing health services to mentorship and supervision, supervision, accreditation, supervision, accreditation, produce better maternal accreditation, and regular and regular competence- and regular competence- health outcomes. However, competence-based training based training of health based training of health within the country at sub- of health workers to match workers; (v) developing workers; (v) developing regional level, availability the evolving disease and institutionalizing a and institutionalizing a of skilled health providers burden; (v) developing performance management performance management doesn’t necessarily translate and institutionalizing a system that promotes system that promotes into better maternal health performance management accountability; and (vi) accountability; and (vi) outcomes. system that promotes leveraging local community leveraging local community accountability; and (vi) groups and users of health groups and users of health The number of skilled leveraging local community services to monitor and services to monitor and health workers in service groups and users of health report on health facility report on health facility is critically low (5.9 skilled services to monitor and performance. performance. health workers per 10,000 report on health facility population) and their Implement the Implement the performance supported by distribution is skewed recommendations from recommendations from effective enforcement of towards a few sub-regions. the human resource audit the human resource audit standards. However, presence of a report by the Office of report by the Office of the high number of skilled Implement the the Auditor General to Auditor General to address health providers in a sub- recommendations from the address discrepancies in discrepancies in staffing, region is not necessarily human resource audit by the staffing, employment status, employment status, associated with more Office of the Auditor General renumeration of health renumeration of health institutional deliveries and to address discrepancies in workers, cost centers, and workers, cost centers, and low institutional maternal staffing, employment status, inconsistencies in HRH data inconsistencies in HRH data deaths.This means that renumeration of health across databases. across databases. there are implementation workers, cost centers, and gaps across the sub- inconsistencies in HRH data regions in Uganda and across databases (master this leads to the provision facility list, payroll register, of poor-quality maternal cost centers, staffing list). healthcare services UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. XIV Issue Theme Key Findings Short Term (0-2 yrs) Medium Term (3-5 yrs) Long Term (6 yrs plus) Provision of quality Ensure that every health Strengthen forecasting, Strengthen forecasting, healthcare is hindered by worker is allocated to a cost quantification, procurement, quantification, procurement, ineffective management center of their workstation storage and distribution storage and distribution of health inputs (financial and “non-existent” health of medicines and medical of medicines and medical and human resources, centers are removed from commodities. Key actions commodities. Key actions medicines, and medical the payroll. to achieve this include include this include (i) supplies); fraud/corruption; this include (i) linking the linking the electronic and absenteeism of health Strengthen forecasting, electronic procurement procurement system at workers. quantification, procurement, system at the NMS to the NMS to the IFMIS at storage and distribution the IFMIS at MoFPED; MoFPED; (ii) adopting The annual direct losses of medicines and medical (ii) adopting electronic electronic systems that link due to corruption in the commodities. systems that link medicines medicines consumption and/ health sector is estimated consumption and/or or issuance to patient data; at US$180.2 million, which Key actions to achieve issuance to patient data; and (iii) procuring and delivering is equivalent to 25 percent this include (i) linking the (iii) procuring and delivering medicines according to of the annual government electronic procurement medicines according to facility needs. spending on health. system at the NMS to facility needs. Meanwhile, the annual cost the IFMIS at MoFPED; Enhance transparency and of absenteeism is estimated (ii) adopting electronic Enhance transparency and accountability through (i) at US$78.5 million. systems that link medicines accountability through (i) strengthening financial consumption and/or strengthening financial reporting and tracking of There are significant issuance to patient data; reporting and tracking of health inputs; (ii) displaying deficiencies in payroll (iii) procuring and delivering health inputs; (ii) displaying financing and performance management in some medicines according to financing and performance data at notice boards; and districts with discrepancies facility needs; and (iv) data at notice boards; and (iii) ensuring interoperability in data between budgets, providing full access to (iii) ensuring interoperability between the various payroll registers, health health facilities for their between the various resources management and facilities and cost centers. online NMS requisitions; and resources management and service delivery information (iv) providing sufficient time service delivery information systems. Access to medicines and for inspecting and signing- systems. medical products in the off the deliveries. Implement and update country is low with only Implement the virtual the virtual pooling 46 percent of facilities Enhance transparency pooling mechanisms for mechanisms for alignment having 95 percent of and accountability alignment of development of development partner tracer medicines in through (i) strengthening partner support. support. 2020/21. Further, there financial reporting and are discrepancies between tracking of health inputs; spending on medicines and (ii) displaying financing and performance data at their availability at facilities partly because the IFMIS notice boards; and (iii) and the procurement system upgrading connectedness at the NMS are not linked. and interoperability between the electronic information While development partners systems for better alignment contribute significantly e.g., alignment of the MoH to the health sector, the master facility list to the overall impact of their IFMIS, payroll register, and investments is not optimized the NMS database. because of fragmentation in financing, implementation, Develop virtual pooling and reporting. Use of mechanism for development parallel systems also partner financing in line with compromises government Compact (s) between GoU stewardship, ownership, and and development partners. accountability. XV UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. Issue Theme Key Findings Short Term (0-2 yrs) Medium Term (3-5 yrs) Long Term (6 yrs plus) 5. Equity of There are variations in the MoH and MoFPED to fully MoH and MoFPED to MoH and MoFPED to fully health spending level and growth in per apply and monitor the update the RAF ensuring implement the updated and use capita public expenditure resource allocation formula that the other health RAF covering all key on health across the sub- (RAF) for non-wage PHC systems inputs (such inputs and monitor its regions in the country. recurrent grants to assure as human resources, performance in terms of Over the period 2012/13- equity. drugs, infrastructure, and promoting equity. 2019/20, the health benefits medical equipment) too are and needs of the households allocated equitably. Expand social protection in the poorest, poorer, and for the poor and vulnerable middle-income groups Expand social protection linked to the SHI scheme. increased. In addition, for the poor and vulnerable households in these three through cash transfers, Adopt service level income groups benefited vouchers, cash for work, agreement with the private more than their health needs community-level pre- providers in areas focusing in 2012/13 and 2019/20. payment and/or micro- on hospital level care financing linked to the SHI which are less pro-poor. CatHE has declined from 22.4 percent in 2005/06 to scheme. 11.9 percent in 2019/20; and is higher among the Adopt service level non-poor than the poor agreements with the households. Despite this private providers with a significant reduction, about focus on both PHC service 1.1 million households (4.9 and hospital level care, million people) across the which is less pro-poor. country still experience The social health protection CatHE. Furthermore, the system could be linked to relatively low incidence the SHI scheme. of CatHE among the poor could also mean that some poor households have forgone health care and not necessarily that there is increased financial protection for the poor. Similar to the results on CatHE, there has been a declining trend in the incidence of impoverishment due to OOP expenditure on health. The percentage of households that fell below the national poverty line due to OOP expenditures decreased from 4.6 percent in 2005/6 to 2.3 percent in 2019/20. Despite this reduction, about 206,405 households (949,464 people) were pushed into poverty (fell below the national poverty line) due to OOP expenditures. UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. XVI XVII UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 2019/20 which is equivalent to 51,677 households (206,709 people). 1. Introduction 1.1 Public Expenditure Reviews in the Health Sector in Uganda 1. The GoU with support from the World Bank Group has conducted a Public Expenditure Review (PER) of the health sector as part of the broader PER which also covers macro-fiscal adjustment, education, public investment management, and intergovernmental fiscal transfers. The overall PER seeks to provide evidence on the financing and spending in the country in order to inform the GoU on areas for fiscal savings and expenditure rationalization, raising the equity and efficiency of spending, rebalancing expenditures between hard infrastructure, investments in quality service delivery and human capital development, and strengthening institutional aspects of public financial management. The health component of the PER builds on four pieces of work that have previously been undertaken by the GoU with support from the World Bank that look at public spending on health, efficiency, resource mobilization, and service delivery. These are: (i) the 2008 PER that focused on affordability of pay reform and the health sector; (ii) the 2010 Fiscal Space for Health analysis; (iii) the 2013 service delivery indicator survey in the health and education sectors; and (iv) the 2021 study on Raising Taxes for Improving Health in Uganda. The findings from these studies have been used to inform policy and planning as provided in Table 1.2. However, some challenges still persist which require more evidence for the GoU to design and implement the most appropriate interventions. Table 1.2 Implementation of Recommendations from Previous PERs in the Health Sector Issues Identified Recommendation Status of Implementation  Overall spending on health is inadequate  Increase public spending on health for  Nominal government expenditure on for Uganda to meet sectoral and national non-salary costs at health facilities health has been increasing but health targets expenditure as a share of GDP has de-  Develop an appropriate health financing creased from 6.4 percent in 2014/15 to  Government spending on health unlikely strategy 4.1 percent in 2018/19 to increase but increases expected from external donor assistance  Health Financing Strategy was developed in 2016. The strategy includes various  High population growth mitigates the mechanisms for enhancing domestic effect of budgetary increases in per resource mobilization and strategic pur- capita terms chasing through RBF  RBF in the health sector designed and implemented within the government system to make it easier to be adopted and sustained by government. i.e., pro- posal for funding through the Uganda Intergovernmental Fiscal Transfer Reform Program.  Low efficiency in the allocation and use  Strengthen programming of donor fi-  RBF implementation has shown the of both government and donor resources nancing. ability to address service delivery ef- ficiencies by enhancing financial and  Enhance planning and monitoring of managerial autonomy on-budget donor support  Aid Management Platform has been  Establish systems to capture and monitor established at the MoFPED level but not off-budget donor support well implemented  Low efficiency in the allocation and use  Establish systems to capture and monitor  MoH has a Department of Health Sector of both government and donor resources off-budget donor support Partners and Multi-sectional Coordina- tion, and partner reporting platform, but  Pursue avenues for improving donor they are not very visible coordination and harmonization e.g., clear procedures for receiving earmarked funds and for districts and facilities to receive direct external support UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 1 Issues Identified Recommendation Status of Implementation  Weak procurement and logistics man-  Improve drug procurement and logistics  Public financing for medicines is 50 agement for drugs and medical supplies management percent of requirement Harmonize procurement of third-party  Only 46 percent of facilities have avail- commodities ability of >95 percent of essential med- icines  Clarify roles of main stakeholders in- volved in drug procurement and logistics  Funds for medicines is centralized and a management Review government pro- push system is still being used at lower cedures on drug procurement to remove levels of the healthcare system rigidities at the NMS.  RBF implementation has shown ability  Review drug financing through PHC and to increase efficiency in the purchase of credit lines and PHC conditional grants medicines, and availability and access to medicines  Low efficiency in the allocation and use  Strengthen governance and anti-cor-  Several avenues for addressing corrup- of both government and donor resources ruption mechanisms: tion in the sector (Health Monitoring Unit, supervision), but limited improvements Develop a governance and anti-corrup- tion strategy  Guidelines on requirements for each level of facility and catchment population are  Develop a policy to guide construction, available but there is a political drive to expansion, and maintenance of public have health centers in all administrative health facilities units, and this undermines allocative efficiency  Poor management and low performance  Address delays in recruitment, payroll  Increase in health worker recruitment of health workers entry, confirmation, promotion etc. over the years, but increases are not matched to needs. This is due to inade-  Health sector performance not aligned to  Improve health workforce management quate funds to recruit needed workforce; defined results/outputs (reduce absenteeism, drug leakages etc.) late releases from central government; inadequacies at district level  Attract and retain staff in rural and re- mote areas  New staffing norms have been developed  Align sector performance to outputs and  Allowances are provided to medical results (explore output-based financing) officers deployed in rural areas but there is need to review impact  Address differences in salaries for health workers at public vs. private-not-for-  RBF has been implemented in the health profit facilities sector countrywide but it needs to be sustained  Health worker salaries have been revised upwards 1.2 Specific Objectives of the Analyses in the Health Sector 2. This PER generally covers the period 2014/15–2020/21 and builds on previous PERs and other analyses that have been undertaken in the health sector in Uganda. Information from the benefit incidence analysis covers the period 2012/13 to 2019/20; and the catastrophic health expenditure analysis 2005/6 to 2019/20. The overall study examines trends in the level and composition of financing in the health sector and efficiency, equity, and effectiveness of spending in the health sector. The specific objectives of the PER are to: i. Assess the trends in the overall level and adequacy of spending on health in the public sector; ii. Assess the allocative and technical efficiency of spending benchmarked against national targets, peer and regional countries, and global and regional norms; iii. Assess the allocative and technical efficiency of spending by economic and administrative functions, and by level of healthcare delivery system; iv. Establish the efficiency of external financing in the health sector, particularly what external resources are financing and their alignment with government priorities; 2 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. v. Examine the functionality of the health system at local government level through an in-depth multiple case study review in selected districts and health facilities; vi. Assess equity in health financing by geographical region, rural/urban mix, type of facility and provider, outpatient and inpatient care, and socio-economic status; and vii. Determine whether the poor are benefiting from the free PHC policy. 1.3 Methodology and Data Sources 3. The framework for preparing PERs for human development sectors (World Bank, 2009) and revised analytical tools for preparing PERs in the health sector were used for the analyses.1 This was complemented by the ‘benchmarking plus’ approach for measuring health system efficiency (Hafez, 2020); and case study research design and methods by Yin (2013). Information on health financing and expenditure was obtained from the IFMIS database at the MoFPED, Medium Term Expenditure Framework (MTEF) reports, health financing reports from development partners, National Health Accounts survey database, and the Uganda National Household Survey database. To fully understand the data, the study team engaged with the health desk officer at the MoFPED and officials from the planning and budgeting units at the Uganda MoH. Additional data on financing, service delivery and performance was extracted from annual health sector reports, Health Management Information System, Annual HRH Audit Reports, NMS database, service availability readiness assessments, the Harmonized Health Facility Assessment (HHFA) report, statistical abstracts from the Uganda Bureau of Statistics, National Integrity Surveys by the Uganda Inspectorate of Government, Uganda Demographic and Health Survey reports, Institute for Health Metrics and Evaluation database, and the World Development Indicator database. 4. To assess the functionality of the health system at local government level, an in-depth multiple case study was conducted in three districts: Ntungamo, Mukono, and Nwoya. These districts were purposively selected based on population densities, number of health facilities, and level of economic development. In each district, the study team visited six public health facilities, namely: one district hospital, two level 4 Health Centers, and three level 3 Health Centers. The health facilities were randomly selected using the official MoH master facility list. Data was collected through a comprehensive review of documents, extraction of data from official government databases, and interviews with key stakeholders at district and facility levels. 1.4 Key Features of the Health System in Uganda 5. The Ministry of Health (MoH) is responsible for policy formulation, strategic planning, and specialized health services while the delivery of PHC services is largely undertaken by the local governments. The Constitution of Uganda 1995, the Public Health Act 1935, and the Local Government Act 1997 mandates the MoH to provide the leadership and stewardship role in policy and planning, and monitoring and evaluation in the health sector. At policy and planning level, Uganda’s strategy for the development of the health sector is part of the Human Capital Development Program (HCDP) under the third National Development Plan (NDP III) which covers the period 2020/21–2024/25 (NPA 2020). The HCDP aims at addressing human development challenges including health, nutrition, education, gender, and women and youth empowerment. On health, the HCDP seeks to increase the proportion of the population in Uganda accessing UHC. The NDP III is aligned to the long-term national development agenda for Uganda – the Vision 2040. 6. Uganda runs a three-tier health system with a total of 8,386 public and privately managed health facilities at primary, secondary, and tertiary levels of healthcare. At PHC level, the Health Center IIs are the first points of entry to the health system, and they are designed to provide outpatient and outreach services. The next types of primary health facilities in the referral system are Health Center IIIs, Health Center IVs, and general hospitals, respectively. At the secondary level of healthcare there are regional referral hospitals, and at tertiary level there are national and specialized hospitals. By 2022/23, private-for-profit providers were operating 46 percent (3,856) of the overall health facilities in Uganda, followed by the GoU at 41 percent (3,448) and private-not-for-profit providers at 13 percent (1,082) (Table 1.3). Nevertheless, the majority of the country’s largest and busiest health facilities are owned and operated by the GoU. As a result, the GoU is the leading health provider and employer in the health sector (Table 1.3). 1 The World Bank has refined the analytical tools for PERs in the health sector. UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 3 Table 1.3 Health Facilities in Uganda – 2022/23 Ownership Share of Facility Type Type of Facility Govt PFP PNFP Total Govt PFP PNFP Health Center II 1,815 3,418 558 5,791 31% 59% 10% Health Center III 1,357 321 389 2,067 66% 16% 19% Health Center IV 197 41 30 268 74% 15% 11% Special Clinic 0 2 28 30 0% 7% 93% General Hospital 55 72 73 200 28% 36% 37% Regional Referral Hospital 16 0 0 16 100% 0% 0% Specialized Hospital 3 2 4 9 33% 22% 44% National Referral Hospital 5 0 0 5 100% 0% 0% Total 3,448 3,856 1,082 8,386 Share of Total 41% 46% 13% Source: World Bank staff construction from MoH 2022/23 Master Facility List. Govt=government, PFP=private-for-profit, and PNFP=private-not-for-profit. Health Center II includes clinics. 7. Over the past 10 years, the number of health facilities in Uganda increased significantly. Overall, the number of health facilities increased by 68 percent between 2011/12 and 2022/23 (Figure 1.1). The number of facilities managed by private-for-profit providers increased the most by 170 percent, followed by GoU at 29 percent, and private-not-for-profit providers at 24 percent. Even though the GoU has constructed a large number of health facilities in the last ten years, the additional number of new government facilities is lower than that of private-for- profit providers. Nevertheless, the GoU’s dedication to expanding physical access to healthcare is demonstrated by the growing number of new government health facilities. The GoU has also modernized and restored a number of existing government health facilities. The increasing quantity of privately run healthcare institutions is another indication of GoU’s leadership in creating an atmosphere that encourages private sector investment in the health sector. Conversely, the growing proportion of private healthcare institutions could have been necessitated by GoU’s incapacity to build health facilities in underserved regions and/or from its inability to deliver high-quality healthcare at existing health facilities. Whatever the case, given the large proportion of private health facilities in Uganda, there is need for strong public- private partnerships in health service delivery. Figure 1.1 Trends in Ownership of Health Facilities in Uganda 68% 4,981 5,229 5,007 6,404 6,937 6,904 6,232 8,386 170% 1,430 1,488 1,202 1,809 2,373 2,795 2,793 3,856 874 873 871 1,229 947 1,009 982 1,082 2,680 2,867 2,932 3,194 24% 3,084 3,133 3,129 3,448 29% 2011/12 2012/13 2015/16 2016/17 2017/18 2018/19 2019/20 2022/23 GROWTH Govt PNFP PFP TOTAL Source: World Bank staff construction from MoH Reports and 2022/23 Master Facility List. Govt=government, PNFP= private-not-for-profit, and PFP=private-for-profit 4 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 1.5 Health Status in Uganda 8. Uganda has made significant progress towards the attainment of UHC over the past two decades. As shown in Table 1.4, the UHC service coverage index2 score for Uganda increased by 123 percent between 2000 and 2021. This is the third highest increase in comparison to peer countries with the largest magnitude of increase being in Ethiopia (169 percent) and Rwanda (158 percent). The UHC effective coverage index measures essential service coverage across population health needs and how much the services contribute to improved health. This means that investments in the health sector in Uganda are bearing fruit. Nonetheless, the UHC service coverage index score for 2021 for Uganda is still less than that of Kenya. This means that additional investments are needed to improve access to health care, and to attain better health and nutrition outcomes. Table 1.4 UHC Index Scores: Uganda vs Benchmark Countries Peer Countries 2000 2015 2021 Increase (2021 vs 2000) Sudan 25 43 44 76% Kenya 28 50 53 89% Chad 15 27 29 93% Congo, Rep. 21 38 41 95% Cameroon 22 42 44 100% Uganda 22 43 49 123% Rwanda 19 44 49 158% Ethiopia 13 34 35 169% Source: World Bank staff construction from World Development Indicators data 9. Increased coverage and access to health services in Uganda over the period 2011 to 2022 have contributed to better health and nutrition outcomes. As shown in Table 1.5, Uganda has achieved better health and nutrition outcomes in several areas than the averages in low-income countries. However, the percentage of children aged 12-23 months with all basic vaccinations estimated at 54 percent in 2022 is very low. Secondly, the total and adolescent fertility levels are higher in Uganda than the average levels in low-income countries. Furthermore, service availability and readiness at health facilities are very low, and these contribute to the provision of poor-quality health care. According to the 2019 Service Availability and Readiness Assessment (MoH 2019b), the readiness or capacity of health facilities in Uganda to provide general services is 52 percent. The mean availability of functional medical equipment was 37 percent and none of the facilities reported availability of all the diagnostic tests. Results from the 2022 Harmonized Health Facility Assessment (HHFA) (MoH and WHO 2023) estimate the general service readiness index at 0.59, meaning that only 60 percent of the facilities were ready to provide services. These gaps are symptomatic of a poorly financed or managed health system. This suggests that Uganda has to do much more to achieve the health- related SDGs. Thus, Uganda requires additional investments in the health sector to improve the quality of health service delivery, which will in turn lead to improved health and nutrition outcomes. 2 The UHC service coverage index combines 16 tracer indicators of service coverage organized by four components of service coverage (reproductive, maternal, newborn and child health; infectious diseases; non-communicable diseases) into a single summary measure. UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 5 Table 1.5 Key demographic and health indicators Uganda LIC Indicator 2011 2016 2022 Population, total (millions)** 32.3 40.1 47.2 - Immunization, all basic vaccinations (% of children aged 12–23 52 55 54 months)* Fertility rate, total (births per woman)* 6.2 5.4 5.2 4.6 Adolescent fertility rate (births per 1,000 women ages 15-19) 128.5 117.8 107.9 95.6 Teenage mothers (% of women ages 15–19 who have had 23.8 24.8 24.0 children or are currently pregnant) Prevalence of stunting, height for age (% of children under 5 33 29 26 33.5 years)* Mortality rate, under-5 (per 1,000 live births)* 90 64 52 67.4 Maternal Mortality Ratio (deaths per 100,000 live births)* 438 336 189 409 Life expectancy at birth, total (years)^ 58.0 61.6 62.7 62.5 Data sources: Uganda data from *Uganda Demographic and Health Survey (UBOS 2022), **World Population Review, ^World Development Indicators. LIC=Low- Income Country. LIC data from World Development Indicators. LIC data is latest year available. 6 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 2. Level and Composition of Total Current Health Expenditure 2.1 Overview of Total Current Health Expenditure 10. The overall level of spending on health in Uganda is inadequate. As shown in Table 2.1, Uganda’s total CHE per capita at US$50.5 in 2020/21 is below the US$86 per capita spending per year which is required by low-income countries to provide basic health services (McIntyre et al 2017). Furthermore, at US$50.5, total CHE per capita in Uganda is also lower than the average spending of US$75.6 per capita in aspirational peer countries (Ethiopia, Rwanda, Sudan, Chad, Kenya, Cameroon, Congo Rep) (Table 2.2). However, the total CHE as a share of the GDP at 5.2 percent in 2020/21 in Uganda is in line with the recommendation of 5 percent of GDP by WHO, and average spending of 5.2 percent and 4.7 percent of GDP in structural and aspirational peer countries, respectively (Table 2.2). This means that population growth is outstripping the available funding for the health sector. Table 2.1 Key Health Finance Indicators for Uganda – 2020/21 CHE per CHE as % GGHE-D as % GGHE-D as % GGHE-D as % Health wage bill Country capita of GDP of GDP of CHE of GGE as % of PEH Uganda 50.5 5.2 0.7 13.4 3.9 37 Benchmarks 86a 5.0b 15.0c 45-60d Data source: MoH 2023; Government financial data; World Development Indicators; Vujicic, Ohiri, and Sparkes 2009. GGHE-D = General government health expenditure (domestic sources); PEH=Public Expenditure on Health. a= McIntyre et al (2017), b=WHO, c=Abuja target (WHO), and d=Chisholm and Evans (2010). Table 2.2 Current Health Expenditure and Per Capita Expenditure CHE per capita Country CHE (% of GDP) (current US$) Ethiopia 3.3 25.3 Chad 4.4 31.4 Structural Peers 5.2 47.1 Uganda 2020/21* 5.2 50.5 Rwanda 6.5 51.2 Cameroon 3.5 52.0 Congo, Rep. 2.2 52.4 Kenya 4.4 73.6 Aspirational Peers 4.7 75.6 Lower Middle Income 3.8 92.0 Sudan 5.0 116.2 Upper Middle Income 5.8 508.7 Data source: World Bank staff construction from World Development Indicators and *MoH 2023 11. Households and external development partners contribute 85 percent of the total CHE in Uganda. As shown in Figure 2.1, the major contribution to the total CHE over the period 2014/15–2020/21 was from external development partners, and by 2020/21, external development funding was 55 percent of total CHE. This was followed by households spending through OOP spending on health at 30 percent, government at 13 percent, and corporations (employers) at 1.8 percent (Figure 2.1). What is also worth noting is the declining level of health spending in Uganda between 2015/16 and 2018/19 after which there was an increase mostly from external funding from development partners. Over the period, total CHE as a share of the GDP declined from 6.2 percent in 2014/15 to 5.7 percent in 2020/21 (Figure 2.1). UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 7 Figure 2.1 Trends in Total CHE in Uganda 100% 7.0% 6.2% 90% 5.7% 6.0% 80% 5.1% 45% 55% 70% 5.0% 4.1% 47% Share of CHE (%) CHE as % GDP 60% 4.0% 50% 3 .0% 40% 39% 37% 30% 3 0% 2.0% 20% 1.0% 10% 15% 15% 13% 0% 0.0% 2014/15 2015/16 2016/17 2017/18 2018/19 2019/20 2020/21 Average Government Corporations Households Development Partners CHE % GDP Source: World Bank staff construction from MoH 2020a; 2023. 2.2 Government Health Expenditure 12. There is low prioritization of health by the GoU. As shown in Figure 2.2, in nominal terms, the general government health expenditure from domestic sources (GGHE-D) as a share of the total government expenditure (TGE), declined from 6.5 percent in 2014/15 to 3.9 percent in 2020/21. Declining GGHE-D as a share of TGE implies that the GoU is not adequately prioritizing funding to the health sector. Consequently, funding of health services in Uganda is dependent on households and external development partners who provide 85 percent of the total CHE. By 2020/21, the results show that in nominal terms, GGHE-D per capita was US$6.8 and for the period 2014/15–2020/21, it was US$6.2 on average per year. Uganda’s GGHE-D per capita is also below the average level of spending in peer countries in Africa and other continents (Figure 2.3). To show the actual value in spending over the period under review, the GGHE-D was adjusted for inflation by using the consumer price index with 2010 being the base year. The results show that the value of GGHE-D has been almost half the nominal amounts throughout the period 2014/15–2020/21 (Figure 2.4). 13. The negative balance of payments position in Uganda over the years has been affecting the level of government funding to the health sector (Figure 2.4). Low and declining government domestic spending on health makes it very difficult to procure sufficient goods and services in the health sector because Ugandan shillings have to be converted to foreign currencies when importing drugs, vaccines, and medical equipment. Additionally, in the event of a crisis, the already low level of financing for health could be diverted, and this could lead to further gaps in service delivery and increased financial burden on poor households to access quality health care. This could lead to forgone health care and reverse the gains in health service delivery and health outcomes which have been achieved over the years. It is important to note that the slowdown in economic growth between 2019/20 and 2020/21 amidst a high population growth had a negative effect on Uganda’s capacity to increase financing for health from domestic resources. Studies by Burt et al (2021) and Ahmed et al (2022) observe that the COVID-19 pandemic and its response in Uganda negatively affected the provision of maternal, neonatal, and child healthcare services depicting health system resilience challenges. 8 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. Figure 2.2 Trends in General Government Health Expenditure – Nominal Terms 1,100,000 12.0% 1,000,000 9.7% 10.0% 900,000 8.0% 800,000 8.0% 700,000 6.8% 6.2% UGX Millions Percent 600,000 6.0% 6.5% 5.9% 4.5% 4.4% 500,000 3.9% 3.8% 4.0% 3.3% 4.0% 400,000 4.0% 3.2% 1,027,885 300,000 739,041 813,087 728,556 856,316 803,339 812,962 2.0% 200,000 100,000 0.0% 2014/15 2015/16 2016/17 2017/18 2018/19 2019/20 2020/21 Nominal GGHE-D (UGX Million ) GGHE as share of TG E GGHE-D as share of TGE Source: World Bank staff construction from government financial reports. Note: GGHE=General Government Health Expenditure; GGHE-D=General Government Health Expenditure – Domestic; TGE=Total Government Expenditure Figure 2.3 Regional Comparison of GGHE-D – 2020/21 45 40 Kenya Congo, Rep. 35 GGHE-D per capita (current US$) 30 Aspirational Peers Rwanda 25 20 15 Sudan Structural Peers Cameroon 10 LICs Ethiopia 5 Chad Uganda 0 0 500 1,000 1,500 2,000 2,500 GDP per capita (current US$) Source: World Bank staff construction from Uganda government financial reports and World Development Indicators data. Note: GGHE-D=General Government Health Expenditure – Domestic; LICs=Low Income Countries Figure 2.4 Nominal vs Real General Government Health Expenditure – Domestic 1,200,000 1,027,885 1,000,000 856,316 813,087 803,339 812,962 800,000 739,041 728,556 UGX Million 600,000 400,000 527,152 558,975 506,215 506,815 462,765 454,255 200,000 447,930 0 2014/15 2015/16 2016/17 2017/18 2018/19 2019/20 2020/21 Nominal GGHE-D (UGX Million) Real GGHE-D (UGX Million) Source: World Bank staff construction from government financial reports UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 9 2.3 External Development Partners (Donor) Expenditure on Health 14. External development partners are still the largest source of funding to the health sector in Uganda (Figure 2.5). In nominal terms, external funding to the health sector was relatively the same over the period 2014/15–2018/19 but increased substantially by 56 and 26 percent in 2019/20 and 2020/21, respectively. However, the large share of external development funding poses a potential risk of making health financing unsustainable if the GoU has no capacity to take up the responsibility. Given the unpredictability of external financing (i.e., pandemics, wars in Ukraine and Israel, donor fatigue), the Ugandan government must urgently create ways of increasing its share of spending to the health sector. If this is not done, the financial gap which will be left by the development partners is most likely going to be funded by households and/or the households will forgo health care. 15. Secondly, a significant amount of the funds from the development partners that are allocated and spent in the health sector in Uganda are off-budget. In other words, most of the donor funding for health programs in Uganda are not reflected in the IFMIS and MoFPED financial reports. On- budget donor expenditure on health as a share of the total donor spending in the health sector in Uganda was 16 percent on average over the period 2014/15–2020/21 (Figure 2.6). In other words, 84 percent of the total donor spending in the health sector in Uganda over the period 2014/15–2020/21 was off- budget. The consistently high off-budget donor funding to the health sector in Uganda could be attributed to inadequate confidence in the public financial management system in the country. However, off-budget funding reduces government’s flexibility in resource allocation, its ability to re-prioritize funding to emergent needs, and is a missed opportunity to improve the public financial management system in the country. Aid agencies and non-governmental organisations (NGOs) administer the majority of the off-budget donor funds through their own systems of planning, financing, procurement, and monitoring. Donors will be required to utilise government systems more frequently in the future, as doing so will promote national ownership, increase the efficiency and effectiveness of the resources, facilitate financial sustainability, and strengthen the government’s stewardship role. Figure 2.5 Trends in Total External (Donor) Expenditure on Health – Nominal Terms 4,500 1,400 $1,151 4,000 1,200 $917 3,500 1,000 3,000 $723 $648 800 $608 2,500 $588 $571 4,204 USD Millions UGX Billions 2,000 600 3,402 1,500 400 2,232 2,183 2,157 2,113 2,009 1,000 200 500 0 0 2014/15 2015/16 2016/17 2017/18 2018/19 2019/20 2020/21 Total (UGX Bn) Total (US$ M) Source: World Bank staff construction from previous National Health Account survey reports and MoH 2023. 10 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. Figure 2.6 Trends in On- and Off-Budget Donor Funding 55% 74% 84% 96% 96% 93% 98% 99% 45% 26% 2% 1% 4% 4% 16% 7% 2014/15 2015/16 2016/17 2017/18 2018/19 2019/20 2020/21 PERIOD AVERAGE On-Bu dget Off-Budget Source: World Bank staff construction from previous National Health Account survey reports and MoH 2023. 2.4 Households Expenditure on Health 16. Households in Uganda spend a significant share of their income on health care. In 2020/21, the households contributed about 30 percent of the total CHE, and over the period 2014/15–2020/21, they contributed 37 percent of the total CHE on average per year (Figure 2.1). Households mostly spend on health through OOP payments at the time of accessing health care. Such high levels of spending can lead to catastrophic consequences, including forgone health care and impoverishment. These are analyzed in detail in chapter 5. A breakdown of the household expenditure on health shows that most of the money was spent on medicines (49 percent) (Figure 2.7). This can be attributed to the low availability of medicines and medical commodities at public health facilities in Uganda. Only 46 percent of the health facilities had over 95 percent of the tracer drugs and commodities in 2019/20 and 2020/21 (MoH, 2021). Results from the HHFA (MoH and WHO, 2023) also show that only 49 percent of the health facilities had all the essential medicines available at the time of the survey. This amplifies the need to provide more funding for medical supplies coupled with enhanced medicines management practices to meet the needs of the people. Figure 2.7 Composition of Households Spending on Health Facility charges 10% Other expenses (food, lodging, etc) 31% Transport Medicines 8% 49% Traditional doctors 2% Source: World Bank staff construction from Uganda National Household Survey 2019/20. UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 11 2.5 Alignment of the overall spending on health to the disease burden 17. In general, health expenditures from the primary sources of financing in Uganda are aligned to the disease burden. This was established after reviewing the trend in the distribution of the total CHE by diseases and conditions, and the main causes of death and disabilities (Figures 2.8 and 2.9). As shown in Figure 2.8, communicable, maternal, neonatal, and nutritional diseases have been the main causes of death and disabilities in Uganda over the period 2009–2019. Specifically, the top 10 causes of death and disabilities in Uganda over the years have been: neonatal disorders, malaria, HIV/AIDS, lower respiratory infect, diarrheal diseases, congenital defects, tuberculosis, sexually transmitted infections, measles, and road injuries (Figure 2.8). Comparing this to the overall spending patterns on diseases and conditions (Figure 2.9) shows that financing in Uganda is aligned to the disease burden. Focusing health expenditures on the leading causes of mortality and disability has enabled Uganda to substantially reduce the prevalence of its most common diseases and conditions (Figure 2.8). For example, even though neonatal disorders, malaria, and HIV/AIDS are still the top three causes of death and disabilities in Uganda, their prevalence reduced by 7.1, 43.4 and 69.2 percent, respectively, over the period 2009–2019. Meanwhile, between 2009 and 2019, there was an increase in the prevalence of sexually transmitted infections, measles, and road injuries (Figure 2.8). This could be attributed to inadequate funding and health systems related challenges. Figure 2.8 Top 10 Causes of Death and Disabilities in 2019 and percent change 2009–2019, all ages Neon atal disorders -7.09 1 Malaria -43.4 2 HIV/AIDS -69.2 3 Lower respiratory in fect -21.4 4 Diarrheal diseases -16.1 5 Con gen ital defects -3.45 6 Tubercu losis -8.07 7 STIs 92.6 8 Measles 4.37 9 Road inju ries 18.4 10 -8 0 -6 0 -4 0 -2 0 0 20 40 60 80 100 Source: World Bank staff construction from IHME (2023). Figure 2.9 Total CHE by Diseases and Conditions, 2016/17–2020/21 100% 18% 14% Other diseases 14% 90% 20% 19% 80% 6% RH 70% 6% 10% 11% 11% 60% HIV/AIDS 27% 20% 16% 50% 28% 23% 40% NCDs 20% 22% 30% 25% 25% 27% 20% Malaria 22% 30% 10% 17% 14% 13% 0% 2016/17 2017/18 2018/19 2019/20 2020/21 Malaria NCDs HIV/AIDS RH RI TB ND VPD Other diseases Source: World Bank staff construction from previous National Health Account survey reports and MoH 2023.RH=Reproductive Health, RI=Respiratory Infections, TB=Tuberculosis, Nutritional Deficiencies=ND, VPD=Vaccine Preventable Diseases, NCDs=Noncommunicable Diseases (includes injuries) 12 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 3. Composition of Public Expenditure on Health 3.1 Overview 18. Public expenditure on health includes spending by the government through domestically generated funds and external funds channeled through the government system. This chapter provides the distribution of the health expenditures across the sub-regions, different levels of the health system, by facility type, and by key health system inputs. The analysis excludes all off-budget expenditure on health from all sources. 3.2 Public Expenditure on Health at Sub-Regions 19. Uganda is administratively divided into regions, sub-regions, districts, counties, sub-counties, parishes, and villages to facilitate service delivery nearer to the people. By January 1, 2021, there were 4 regions, 15 sub-regions, 135 districts, 11 cities, 322 counties, 1,488 sub-counties, 7,553 parishes, and 58,197 villages. In addition to Kampala (capital city), the following ten secondary cities were established over the past five years: Arua, Fort Portal, Gulu, Hoima, Jinja, Lira, Masaka, Mbale, Mbarara, and Soroti. Given the differences in the population, the level and growth in public expenditure on health at sub-regional level was analyzed in per capita terms. The results show variations in the level and growth in per capita public expenditure on health across the sub-regions (Figure 3.1). The largest increase in per capita public expenditure on health was observed in North Central which received 6.6 times more in 2018/19 as compared to 2016/17, followed by Busoga (5.3), and Tooro (4.5). The lowest gaining sub-regions were Bukedi (1.4 times more), Bunyoro (1.39 times more), and South Central (1.16 times more). 20. Per capita expenditure on health varies among sub-regions suggesting that the resources allocation formula is ineffective in addressing variations in the distribution of financial resources across the sub-regions. Looking closely at the level of per capita expenditures on health across the sub-regions, North Central still ranked third from the lowest at UGX10,221 in 2018/19 despite the huge increase in per capita spending between 2016/17 and 2018/19. On the other hand, the increases in sub-regions which already receive more per capita spending (Karamoja, Bugisu, and Kampala) ranged from 2.9 to 4.4 and these sub-regions had the largest level of per capita expenditures on health in 2018/19 at UGX26,745, UGX 23,273, and UGX19,752, respectively. These results suggest that the existing formula for allocating financial resources from the center to the local government level was ineffective at adjusting the variations across the sub-regions. However, variations in funding by sub-regions can also be affected by differences in the number and type/size of the health facilities, and the number and skills-mix of health workers in each sub-region. These are examined in subsequent sections below. Figure 3.1 Growth in Per Capita Expenditure on Health at Sub-Regions 30,000 6.61 7 x-fold increase: 2016/17 - 2018/19 25,000 6 5.33 4.49 5 20,000 Per Capita (UGX) 4 15,000 1.39 3 10,000 1.41 1.16 2 5,000 1 0 0 a li di ga ile le al zi Ka ro o so Ka u h C ro la al ho oj gis ng ge tr ke pa ko o o so Te tr tN m en So uny To La Ac en Ki Bu Bu m An Bu ra es C B W rth ut No 2016/17 2018/19 Increase Source: World Bank staff construction from government financial reports and IFMIS data UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 13 21. From the above, it is clear that the variations in health spending at sub-national level are associated with the differences in the availability and/or distribution of health infrastructure. Through the Uganda Intergovernmental Fiscal Transfer (UgIFT) program and the Uganda Reproductive, Maternal and Child Health Services Improvement Project (URMCHIP), the government made efforts to improve physical access by upgrading existing health facilities and constructing new ones. Furthermore, formulae for allocating development and non-wage recurrent grants to health facilities at local government level were introduced in 2018/19 and 2019/20, respectively. After the introduction of the new formulae, it was reported that there has been an improvement in the distribution of financial resources at local government level (World Bank 2022b). For example, in the health sector, the Gini coefficient of the distribution of per capita local government allocations decreased from 35 percent to 29 percent and the 90/10 ratio dropped from 5.3 to 4. The 90/10 ratio of the non-wage recurrent grant, dropped from 6.5 to 2.6 and the Gini coefficient dropped from 38 percent to 22 percent (World Bank 2022b). 22. In order to ensure equity and efficiency in the allocation of resources across the sub-regions and districts, the MoH needs to ensure that the formulae for allocating development and non-wage recurrent grants are fully applied. Secondly, there is need for continuous monitoring of the performance of the formulae in enhancing equity and efficiency. Thirdly, the MoH needs to ensure that the other health systems inputs outside the formulae (such as human resources, drugs, medical equipment) are allocated equitably. 3.3 Public Expenditure on Health by Level of Healthcare 23. While there has been an increase in the level of spending at central and local government levels, the increases have been more at headquarters and district/local government offices. As shown in Table 3.1, at central level, there was a 216 percent increase in spending at headquarters as compared to the hospitals and other government health institutions between 2018/18 and 2020/21. The increase at the central level is partly due to projects that are centrally coordinated but actually implemented by local governments and/or implementation of some of the decentralized services by the center on behalf of the local governments. At national and specialized hospitals, there was a 4 percent decline in spending over the same period. At local governments level, there was a 55 percent increase in spending at district/local government offices while at health center IIIs the increase was 37 percent. On the other hand, spending at district/general hospitals, and health center IVs and IIs declined by 8 percent. The high expenditure at headquarters and district/local government offices could be due to bulk procurements on behalf of the health facilities. Such expenditures were evident during the COVID-19 outbreak. The declining and increasing expenditure at health center IIs and health center IIIs, respectively, is due to the upgrading of health center IIs to health center IIIs. 14 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. Table 3.1 Growth in Healthcare: 2020/2021 vs 2018/19 Growth 2018/19 2020/21 (2) vs (1) (UGX, millions) (1) (UGX, millions) (2) Central Government 635,852.88 1,123,426.90 77% Headquarters 118,711.67 374,561.29 216% Government Health Institutions 206,891.73 417,869.32 102% National and Specialized Hospitals 164,485.61 158,584.16 -4% Regional Referral Hospital 145,763.87 172,412.13 18% Local Government 393,593.99 580,618.91 48% District/Local Government Health Office 329,831.75 512,757.92 55% District/General Hospitals 19,457.00 17,900.30 -8% Health Center IV 12,782.00 11,759.44 -8% Health Center III 20,360.00 27,931.20 37% Health Center II 11,163.24 10,270.04 -8% TOTAL 1,029,446.87 1,704,045.81 66% Source: World Bank staff construction from government financial reports and IFMIS data 3.4 Distribution of Public Expenditure on Health at Public Health Facilities by Facility Type 24. Public spending on health in Uganda is concentrated at higher levels of healthcare. Firstly, a review of the growth in health expenditures across all the different types of health facilities, shows that only regional referral hospitals and health center IIIs had increased expenditures in 2020/21 as compared to 2018/19 (Table 3.2). The increased spending on health center IIIs coincided with the surge in the upgrade of health center IIs to health center IIIs under the UgIFT program. Secondly, looking at the shares of expenditures in 2018/19 and 2020/21, the top two highest expenditures were at national and specialized hospitals, and regional referral hospitals. Thirdly, comparing the public expenditure on health at public health facilities at district level with the facilities at regional and national level shows that Uganda focuses more on funding higher level healthcare as compared to PHC. As shown in Figure 3.2, while there are only 24 government-owned regional, specialized, and national referral hospitals (which is one percent of the 3,448 government health facilities), their share of the total public expenditure at government health facilities is 83 percent. On the other hand, government facilities at district level (general hospitals, and health centers II, III, and IV) are 3,424 (99 percent of the 3,448 government health facilities) yet their share of the total public expenditure at government health facilities is only 17 percent. While PHC services can be provided at all levels of the health system, the problem then is that higher-level facilities are not being used as intended. Therefore, the government must examine the functionality of different levels of health facilities in the country. Table 3.2 Trends in Public Expenditure at Public Health Facilities by Facility Type   2018/19 2020/21 2018/19 2020/21 Growth (UGX, millions) (UGX, millions) Share Share National and Specialized Hospitals 164,486 158,584 -4% 44% 40% Regional Referral Hospitals 145,764 172,412 18% 39% 43% District/General Hospitals 19,457 17,900 -8% 5% 4% Health Center IV 12,782 11,759 -8% 3% 3% Health Center III 20,360 27,931 37% 5% 7% Health Center II 11,163 10,270 -8% 3% 3% TOTAL 374,012 398,857 7%     Source: World Bank staff construction from government financial reports and IFMIS data UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 15 Figure 3.2 Public Expenditure on Health at Public Health Facilities Source: World Bank staff construction from government financial reports and MoH Master Facility List. 3.5 Health Expenditure by Key Health Systems Inputs 25. The share of public expenditure on health on salaries and wages in Uganda has been growing over the years and is within the historical spending patterns in Eastern Africa. Expenditure on the health wage bill as a share of the total public expenditure on health increased from 24 percent in 2016/17 to 37 percent in 2020/21 (Figure 3.3). The level of spending on health workers in Uganda is similar to spending patterns in other countries in Eastern Africa (Kenya and Tanzania) but lower than other countries in Southern Africa (Malawi, Seychelles, Zimbabwe and Zambia) (Figure 3.4). Uganda’s health wage bill as a share of the total public expenditure on health is also below the recommended range of 45 to 60 percent for low-income countries (Chisholm and Evans 2010). Though still low, the growth in the health workforce wage bill in Uganda is partly due to the recruitment of additional health workers. Therefore, while recruitment and retention of health workers in rural areas could increase the wage bill it would help to improve the staff-to-population ratio for skilled health providers, which is currently below expected norms. See subsequent sections on staffing levels in Uganda. Figure 3.3 Breakdown of Expenditure by Key Health Systems Inputs 44% 41% 41% 42% 37% 29% 29% 27% 26% 24% 19% 18% 16% 15% 16% 14% 15% 15% 11% 13% 2% 3% 2% 1% 1% 0.3% 0.4% 0.3% 0.2% 0.2% 2016/17 2017/18 2018/19 2019/20 2020/21 Personal Emoluments Goods and Services Training Drug s and Medical Commodities Capital Expenditure Operational Grants Source: World Bank staff construction from government financial reports and IFMIS data 16 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. Figure 3.4 Health wage bill as a share of the total public expenditure on health 70 62 60 60 54 51 50 43 41 Percentage 40 37 30 20 10 0 Uganda Kenya Tanzania Malawi Seychelles Zimbabwe Zambia Source: World Bank staff construction from World Bank Health PERs. 26. Despite the increase in the volume of public expenditure on drugs and medical commodities in nominal terms, the level of spending in Uganda is still significantly below expected norms. Between 2016/17 and 2020/21, the GoU increased its spending on drugs and medical commodities in nominal terms by about 65 percent from US$77.5 million (UGX265.1 billion) in 2016/17 to US$105.9 million (UGX393.6 billion) in 2020/21 (see Figure 4.8). However, expenditure on drugs and medical commodities as a share of the total public expenditure on health decreased from 44 percent in 2016/17 to 29 percent in 2020/21 (Figure 3.5). The proportion of expenditure in 2020/21 is lower than the African regional average of 33 percent (Bennett et al 1997). Secondly, the per capita spending on drugs and medical commodities – estimated at US$3.2 in 2020/21 in Uganda (Figure 3.5) – is significantly below the recommendation by the Lancet’s Commission on Essential Medicines for low-income countries to finance a basic package of 201 essential medicines at US$13 to US$25 per capita (Wirtz et al 2016). Inadequate spending on drugs and medical commodities in Uganda is one of the factors contributing to the low availability of tracer drugs and commodities at health facilities in the country. Figure 3.5 Adequacy of Public Spending on Drugs and Medical Commodities 50% 3.5 3.2 Share of total public spending on health 3.0 Per capita spending (US$) 40% 44% 2.5 41% 2.5 2.1 30% 2.0 2.2 2.2 29% 29% 26% 1.5 20% 1.0 10% 0.5 0% 0.0 2016/17 2017/18 2018/19 2019/20 2020/21 Drugs as a Share of Total Publi c Spending on Health Per Capita Spending on Drugs (US$) Source: World Bank staff construction from government financial reports and IFMIS data. UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 17 Nurse in charge at Buyobo Millenium Health Center, (Amanya Brian, 2022) 18 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 4. Efficiency and Effectiveness of Spending in the Health Sector 4.1 Benchmarking Uganda with Peer Countries 27. A benchmarking approach (Hafez 2020) was used to analyze potential inefficiencies by looking at: (a) the ability of Uganda to produce health services at its level of per capita spending; and (b) the ability of Uganda to translate the health services that it produces into better health outcomes. For the first part of the analysis, a comparison between the health access and quality (HAQ) index and the total CHE per capita were used (Figure 4.1). The HAQ index (Fullman et al 2018) incorporates 32 causes of disease and injury considered amenable to health care. In other words, death is not supposed to occur from the 32 causes if there is effective care. The results show that though Uganda’s total CHE per capita of US$50.5 is lower than that of Rwanda, Cameroon, Kenya, and aspirational peers, its HAQ score of 32.4 is comparatively the same as these countries. This suggests that Uganda is more efficient than its aspirational peers and some countries in East Africa (Rwanda and Kenya) at using its limited resources to produce quality health services (Figure 4.1). However, this analysis doesn’t take into consideration differences in the cost-of-service provision. Considering that health service provision in Uganda is generally less expensive, this could be the reason why Uganda is relatively more efficient than peer countries. For example, salaries and wages for health workers in Uganda are much lower than in Kenya and Rwanda. However, Congo Republic, Sudan, and Ethiopia are more efficient than Uganda at using their limited resources to produce quality health services. Figure 4.1 Producing quality services with available funding – 2020/21 50 45 Sudan Health Access and Quality Index 40 Aspirational Peers 35 Congo, Rep. Cameroon Uganda Kenya Rwanda 30 Ethiopia Structural Peers 25 Chad 20 15 25 35 45 55 65 75 85 CHE per capita (current US$) Source: World Bank staff construction from Uganda government financial reports and World Development Indicators data. 28. Uganda is more effective than its peers at transforming the available health services into better maternal health outcomes. As shown in Figure 4.2, at a HAQ score of 32.4, Uganda’s Maternal Mortality Ratio (MMR) is 189 deaths per 100,000 live births. Despite Uganda having a lower HAQ score than Sudan, aspirational peer countries, Congo Republic, Cameroon, and Kenya, it has a lower (better) MMR. Overall, the results suggest that Uganda is more effective at using the existing health services in the country to produce better maternal health outcomes than most countries in sub-Saharan Africa. Nonetheless, there are variations in maternal mortality within the country particularly across the sub-regions. See the results in the sub-section 4.3. UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 19 Figure 4.2 Translating available services into better health outcomes – 2020/21 1,200 Maternal Mortality Ratio (Deaths per 100,000 Live Births) 1,000 Chad 800 600 Kenya Structural Peers Cameroon 400 Ethiopia Congo, Rep. Sudan 200 Rwanda Aspirational Peers Uganda 0 20 25 30 35 40 45 50 Health Access and Quality index Source: World Bank staff construction from UBOS (2023), Uganda government financial reports, and World Development Indicators data. 4.2 Budget Performance 29. The other way of assessing efficiency in resource allocation and utilization is by looking at the performance of the budget. Budget performance was computed as the proportion of the actual amounts spent out of the amounts allocated, budget execution as the proportion of amounts disbursed out of the amounts allocated, and budget absorption as the proportion of money actually spent from the amounts disbursed. The results show that budget performance for the health sector over the period 2016/17–2020/21 was fairly good. On average, about 97 percent of the budgeted funds were released during the period under review, 97 percent of which were used (Figure 4.3a). Given the need for additional funding in the health sector, it is expected that 100 percent of the budgeted funds are released and used each year. For drugs and medical commodities, the budget execution rate was about 100 percent on average during the same period (Figure 4.3b). The absorption rate for drugs and medical commodities was also around 100 percent during the period under review. However, as explained in subsequent sections, the availability of drugs and medical commodities in Uganda is still low despite the high budget performance. Figure 4.3 Health Sector Budget Performance A. Budget Execution and Absorption Rates - Total 2,000 98.8% 100% 97.7% 97.8% 1,800 96.8% 96.9% 98% 97.7% 1,600 97.1% 96.8% 96% 96.3% 1,400 94% 1,200 92% Billions (UGX) 92.2% 1,000 90% 800 88% 600 86% 400 84% 200 82% 0 80% 2016/17 2017/18 2018/19 2019/20 2020/21 Budget Disbursed Expenditure Absorption Rate Budget Execution Source: World Bank staff construction from government financial reports and IFMIS data 20 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. B. Budget Execution and Absorption Rates − Drugs and Medical Supplies 600 1 00% 90% 500 80% Execution and Absorption Rates 70% 400 Billions (UGX) 60% 300 50% 40% 200 30% 20% 1 00 1 0% 0 0% 2016/17 2017/18 2018/19 201 9/20 2020/21 Budget Disbursed Expenditure Budget Executi on Absorption Rate 4.3 Availability and Productivity of the Health Workforce 30. The increasing share of public expenditure on health devoted to salaries and wages demonstrates the GoU’s commitment to improving health outcomes. The goal of Uganda’s HCDP (which includes strategies on health) is to enhance the productivity and social wellbeing of the population (NPA 2020). The GoU has been making strategic plans specifically for the improvement of human resources in the health sector, and the latest one is the 2020-2030 HRH strategic plan (MoH 2021a). The goal for the HRH strategic plan is “to develop and maintain a well-performing appropriately skilled health workforce, equitably deployed and accessible at all levels of the health care system, providing quality health services.” The various interventions on HRH in Uganda have led to some improvements in the health workforce in-post from 59,105 in 2016/17 to 64,808 in 2019/20 (Figure 4.4). However, this increase is below the targeted 85 percent fill-rate of the health sector staffing norm (staff establishment). As shown in Figure 4.4, by 2019/20, only 74 percent of the required staff positions had been filled. Furthermore, though the total number of health training institutions (public, mission, and private) in Uganda have increased over the years, the annual number of health workers graduating from these institutions has only increased slightly from 5,911 in 2016/17 to 6,481 in 2019/20 (Figure 4.4). Figure 4.4 Trends in training outputs and staff in-post 70,000 90% 75% 78% 74% 80% 60,000 Staff in-post and training output 70% Percentage of filled positions 50,000 60% 4 0,000 50% 64,808 62,445 59,105 30,000 4 0% 30% 20,000 20% 6,481 6,245 5,911 10,000 10% 0 0% 2016/17 2017/18 2019/20 Clinical heal th workers in-post Training output Percentage of filled positions Source: World Bank staff construction from HRH system data, and MoH 2020b; 2021a. UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 21 31. The number of skilled health providers in Uganda is critically low and the distribution is skewed towards some sub-regions. As shown in Figure 4.5, some sub-regions such as Bugisu, Kigezi, and Kampala have more skilled health providers in the public sector (doctors, nurses, midwives, clinical officers) per 10,000 population than the other sub-regions. The differences in the distribution of public sector skilled health providers per 10,000 population by sub- region could be explained by the population density, socio-economic conditions, and the number and size of health facilities in each sub-region. In 2022, the highest number of skilled health providers per 10,000 population was in Kampala at 13.6 which was more than twice the national level (Uganda) estimate of 5.9. This means that none of the sub-regions, nor Uganda as a whole, had met the required standard of 23 skilled health providers per 10,000 population. However, the data only included skilled health providers in the public sector. Thus, the density of skilled health providers in Uganda can increase if data from the private sector is included. Nonetheless, the national target of 23 skilled health providers per 10,000 population will still not be attained because the government is the largest employer in the health sector, and it has most of the largest health facilities in the country. Furthermore, considering that Uganda’s ambition is to meet the health-related SDGs, the SDG target of 44.5 doctors, nurses, and midwives per 10,000 population is far-off. Therefore, there is an urgent need for Uganda to invest in HRH – including training, recruitment, and retention. 32. A major factor contributing to the low progress in the HRH staffing level in Uganda is the high population growth rate in the country. In addition, recruitment and deployment of health workers to rural areas is a challenge and grossly affects the provision of quality health care. Thus, while the annual number of health cadres graduating from health training institutions in Uganda is still low (Figure 4.4), most of them are not recruited. This could be attributed to: (i) inefficiencies in the decentralized system of recruitment whereby some districts fail to fully absorb their wage budgets each year; (ii) weak coordination between the health and education sectors, leading the continued training of health workers that the health sector does not need; (iii) a weak system for redistributing existing health workers across the districts to optimize the numbers and skills-mix; and (iv) ineffective HRH retention and management strategies for staff working in rural areas. The recent increase in the salaries and wages of the existing health workers in Uganda rather than an increase in recruitment is also likely to affect the number and distribution of health workers in-post. In other words, the health wage bill will increase but the number of health workers in-post will remain the same or even shrink due to attrition. Figure 4.5 Distribution of Skilled Health Providers by Sub-Region in Uganda - 2022 14.0 13.6 12.0 Number per 10,000 Population 10.0 9.3 7.5 8.0 6.8 6.8 6.4 6.1 6.0 5.9 5.7 5.7 6.0 5.3 5.1 4.9 4.3 3.6 4.0 2.0 0.0 ile li la le a zi l da so l o o ga di o su ra ra ho oj ng or or ge pa ko ke Te tN gi so an nt nt m To ny Ac La Ki Bu m An Bu Bu Ce Ce Ug ra es Bu Ka Ka W h h rt ut No So Source: World Bank staff construction from HRH system data. 33. There is a direct relationship between the number of skilled health providers per 10,000 population and the level of public spending on health (Figure 4.6a). Sub-regions with higher levels of per capita public spending on health generally have higher levels of skilled health providers. This means that the distribution of health workers is a key factor in how financial resources are distributed in the public health sector in Uganda. It also implies that the geographical allocation of public funds in the health sector could be improved by redistributing health workers. However, 22 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. there are some sub-regions like Bukedi which have low levels of spending but a higher level of skilled health providers. Furthermore, sub-regions like Lango have higher spending than North Central, Teso, Busoga, and Bunyoro but a lower concentration of skilled health providers. This means that the type of skilled health provider (doctor vs clinical officer, nurse, and midwife) matters due to differences in salaries. Figure 4.6b examines the productivity of the skilled health providers and shows that sub-regions with higher numbers of skilled health providers have more institutional deliveries. However, there are some outliers like Karamoja and Kigezi, where institutional deliveries are much lower than other sub-regions besides having higher numbers of skilled health providers. This suggests that some sub-regions are more efficient or productive in translating resources into outputs than others. 34. Having a high number of skilled health providers in a sub-region doesn’t lead to reduced institutional maternal deaths. As shown in Figure 4.7, institutional maternal mortality is more in sub-regions with high numbers of skilled health providers in the public sector. The exceptions are Bukedi, Tooro, and Kigezi sub-regions where the number of skilled health providers is relatively high and maternal deaths relatively lower than other sub-regions. However, Bukedi is an even better performer than Tooro and Kigezi. Overall, these results suggest that there are implementation gaps, including poor quality of maternal healthcare services in almost all the sub-regions in Uganda. Figure 4.6 Productivity of Skilled Health Providers by Sub-Region, 2022 A. Staffing Level vs Per Capita Spending B. Institutional Deliveries vs Staffing Level 15.0 0.055 Kampala Kampala Skilled Health Provider per 10,000 Population 13.0 0.050 Per capita Institutional Deliveries Bugisu 11.0 0.045 Kigezi 9.0 Ankole 0.040 Bukedi Tooro Kigezi Acholi Bugisu North West Nile North Central 7.0 Central 0.035 Ankole Acholi Teso Tooro Karamoja Bunyoro West Nile Bukedi Busoga 5.0 0.030 Lango Teso Bunyoro Karamoja Lango Busoga 3.0 0.025 South Central South Central 1.0 0.020 0 5,000 10,000 15,000 20,000 25,000 30,000 2 4 6 8 10 12 14 Per capita Public Health Expenditure (UGX) Skilled Health Providers per 10,000 Population Source: World Bank staff construction from HRH, government financial reports, and HMIS data Figure 4.7 Effectiveness of Maternal Healthcare by Sub-Region, 2022 Institutional MMR vs Staffing Levels 250 Kampala 200 Deaths per 100,000 population 150 North Central Bunyoro Acholi 100 Ankole South Central Busoga Karamoja Lango Bugisu Teso 50 Tooro Kigezi Bukedi West Nile 0 2 4 6 8 10 12 14 Skilled Health Providers per 10,000 Population Source: World Bank staff construction from HRH and Maternal and Perinatal Death Surveillance and Response (MPDSR) data. MMR=Maternal Mortality Ratio UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 23 The low productivity of health workers and poor maternal health outcomes in Uganda can also be attributed to the high level of absenteeism. According to Hafez (2020), absenteeism refers to health professionals who are absent and not on-duty from the facility on an unannounced visit. A high absenteeism rate can reduce the number of health workers available to provide services on a particular day leading to increased workload and poor quality of service delivery. In Uganda, absenteeism was estimated at 46 percent in 2013 (Wane and Martin 2013) which implies that almost half of the health workers were not at the health facility at the time of the survey. Having half of the health workforce absent on any given day creates a significant challenge for providing timely, efficient, high-quality health services. It also leads to wastage of resources and poor health outcomes. The absenteeism rate in Uganda (46 percent) is comparable to the rate in Kenya (53 percent) but it is much higher than other countries in eastern and southern Africa such as Madagascar (27 percent), Mozambique (24 percent), Malawi (46 percent), Zambia (16 percent), and Tanzania (14 percent) (Di Giorgio et al 2020; Hafez 2020; World Bank 2018; World Bank 2019b; World Bank 2019c;). Given the high proportion of the public expenditure on HRH in Uganda, the direct annual loss from absenteeism is estimated at about UGX292 billion (US$78.5 million). This means that the government loses about UGX24.3 billion (US$6.5 million) each month due to absenteeism i.e., “wasted” payments for salaries to absentee health care workers. A study by the Uganda Inspectorate of Government puts the annual loss due to absenteeism at UGX495 billion (US$133.1 million) (Inspectorate of Government 2021). 4.4 Supply and Availability of Drugs and Medical Products 35. Access to quality and affordable essential medicines and medical products is key in promoting health and achieving the SGDs. In nominal terms, the GoU increased its spending on drugs and medical commodities by about 65 percent from US$77.5 million (UGX265.1 billion) in 2016/17 to US$105.9 million (UGX393.6 billion) in 2020/21 (Figure 4.8). This is equivalent to an annual average increase of about 14 percent. Despite the high and increasing level of spending on drugs and medical products in Uganda, expenditure on drugs and medical commodities as a share of the total public expenditure on health decreased from 44 percent in 2016/17 to 29 percent in 2020/21 (see Figure 3.5). This is lower than the African regional average of 33 percent (Bennett et al 1997). Uganda’s per capita spending on drugs and medical commodities in 2020/21 was only US$3.2 and this too, is lower than the recommended per capita spending of US$13 to US$25 on drugs and medical commodities for low-income countries by the Lancet’s Commission on Essential Medicines (Wirtz et al 2016). 36. The supply and access to drugs and medical products in Uganda are poor. According to the annual health sector performance report for 2020/21, only 46 percent of the health facilities had over 95 percent of the tracer drugs and commodities in 2019/20 and 2020/21 (MoH 2021b). Results from the HHFA (MoH and WHO, 2023) also shows that only 49 percent of the health facilities had all the essential medicines available at the time of the survey. Low availability of drugs and medical supplies could be due to poor accountability. For example, the NMS procurement system is not linked to the IFMIS at MoFPED. As a result, contracts for drugs and medical supplies are signed outside the IFMIS. Consequently, contract management is conducted outside the IFMIS, and each contract must be checked manually to ensure compliance with available budgetary allotments. Figure 4.8 Expenditure on Drug and Medical Commodities: 2016/17–2020/21 140,000,000 35.0% 27.3% 28.7% 120,000,000 30.0% 25.0% 100,000,000 20.0% 80,000,000 US$ 15.0% 60,000,000 10.0% 40,000,000 2.1% 5.0% 20,000,000 -1.5% 0.0% 0 -5 .0% 2016/17 2017/18 2018/19 2019/20 2020/21 Expen diture Annual increase Source: World Bank staff construction from government financial reports 24 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 4.5 Findings From Existing Studies on Efficiency in the Health Sector in Uganda 37. In sub-section 4.1, results from the benchmarking analysis showed that Uganda is more efficient than its peers at: (i) using its limited resources to produce quality health services, and (ii) using the existing health services to produce better maternal health outcomes. However, this analysis doesn’t take into consideration differences in the cost-of-service provision. Considering that health service provision in Uganda is generally less expensive, this could be the reason why Uganda is relatively more efficient than peer countries. The analysis also masks variations at regional and sub-regional level within a country. Furthermore, the benchmarking analysis focuses on the total amount spent and doesn’t look at the actual amount of the resources used to achieve the results. What if half of the resources are lost and/or misapplied? We have already established that absenteeism in Uganda is high, and this leads to a loss in financial resources and lives due to forgone health care and/or unmet health needs. As a matter of fact, several studies have highlighted significant inefficiencies and gaps in the delivery of services in the health sector in Uganda. The studies include: previous PER reports by the World Bank; national health account surveys by the Uganda MoH; service availability readiness assessments, cross-programmatic efficiency analysis, and HHFA by the World Health Organization; and National Integrity Surveys and a study on the cost of corruption in the health sector by the Uganda Inspectorate of Government. 38. The study on the extent and cost of corruption in the health sector in Uganda (Inspectorate of Government 2021) revealed four issues that contribute to the loss of resources in the health sector, and hinder access to essential health services. These are: (a) bribes or gifts to healthcare providers; (b) absenteeism of health workers; (c) theft of medicines and equipment; and (d) corruption in the procurement of medicines, infrastructure, and services. The study estimates the annual direct costs for these four vices at UGX670 billion (US$180.2 million) which is equivalent to 25 percent of the annual government spending on health. As part of the total annual direct costs, the annual cost of absenteeism was estimated at UGX495 billion (US$133.1 million). The study further reveals that patients were made to pay bribes in order to receive good quality care and/or to speed up a medical procedure. The unfortunate thing is that poor people had a greater probability of paying for the bribe, particularly for households with patients with long-term health conditions and pregnant mothers. Therefore, bribery and corruption in the health sector in Uganda increases the costs that patients pay to access health care and also hinder access to essential health services. This ultimately affects people’s wellbeing and health outcomes—especially for the poor. 39. The cross programmatic efficiency assessment of the health system in Uganda (WHO 2022) identified five cross-programmatic inefficiencies. The study focused on maternal and child health, tuberculosis, HIV/AIDS, and malaria programs; and the health system as a whole. The study revealed that there were: (i) Multiple, fragmented funding sources with limited pooling arrangements, (ii) Limited coordination mechanisms and separate governance arrangements; (iii) Fragmentation in service delivery models across health programs; (iv) Lack of coordination across program-specific procurement and supply chain systems; and (v) Fragmentation in the recruitment, training and renumeration of health workers which are not linked to the government system. Among others, the study called for the government, development partners and other stakeholders to: (a) Strengthen coordination and governance arrangements; (b) Deliver integrated person-centered health services at each level of the health system; (c) Align national procurement and supply chain systems for all programs; and (d) Strengthen HRH planning and implementation across the health sector. 4.6 Detailed Review of the Health System at Local Government Level 40. As part of this PER, a detailed review of the functionality of the health system at local government level was undertaken. The main aim of this analysis was to examine potential inefficiencies in the health sector, focusing on planning, budgeting, and financial management; effectiveness of key health system delivery inputs such as human resources, and drugs and medical supplies; and functionality of health centers. Using a mixed-methods multiple case study design, three districts (Ntungamo, Mukono, and Nwoya) were purposively selected based on population densities, number of health facilities, and level of economic development. In each district, the study team visited six public health facilities, namely: one district hospital, two level 4 Health Centers, and three level 3 Health Centers. The health facilities were randomly selected using the official MoH master facility list. Data was collected through a comprehensive review of documents, extraction of data from official government databases, and interview of key stakeholders at district and facility levels. UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 25 4.6.1 Planning, Budgeting and Financial Management Annual Planning and Budgeting Process 41. The annual planning and budgeting process in the health sector is guided by the Public Financial Management Act No. 3 of 2015. Interviews with district authorities and health facility workers revealed that the process is well-defined and understood by all the stakeholders at district and facility levels. It includes top-down guidance from MoFPED (which provides the annual indicative planning figures) and annual budgeting guidelines by MoH; and a bottom-up planning process by districts and health facilities which is supported by local government authorities. The annual budgeting guidelines by MoH specify the formula for allocating the non-wage PHC grants to health facilities. These guidelines promote allocative and technical efficiency as resources are allocated based on a multi-criteria formula which has equity and efficiency enhancing measures. Authorities from the district also indicated that the Program Budgeting System (PBS) had greatly enhanced the budgeting process. However, some district officials felt that using the PBS was a challenge, and that there was a need for MoH and MoFPED to provide training on the PBS. Adequacy and Predictability of Funding 42. In nominal terms, there has been an increase in the level of the PHC non-wage grants. Nevertheless, district authorities and health facility workers articulated that the funds were still insufficient to meet all the needs of the health facilities. Secondly, it was reported that the PHC non-wage grant for drugs and medical supplies are disbursed directly to the National Medical Store (NMS), and that the districts and health facilities were not involved in the procurement process. According to the district authorities and managers at the health facilities, this was undermining financial and managerial autonomy at district and facility levels. See more details in sub-section 4.6.3 below. 43. The PHC non-wage grants are disbursed on time, and the amounts that are released are generally comparable to the budgets. The funds are released directly into health facility bank accounts, and online systems (IFMIS and PBS) are used to monitor usage of the funds. Direct transfer of the funds into health facility bank accounts has enhanced efficiency in disbursement and adsorption of the funds. Notwithstanding the above, for 2020/21, the amounts which were released in the third quarter were very low while in the fourth quarter there were huge disbursements. This could be attributed to the COVID-19 pandemic which led to the diversion of government resources to health and non- health interventions at the center. Uganda recorded the first case of COVID-19 on March 21, 2020. Low and untimely disbursements can affect health service delivery as unspent funds are returned to the treasury after the fourth quarter. Budget Monitoring and Financial Reporting 44. District quarterly budget performance reports are generally accessible at the district offices and online websites. Secondly, to enhance transparency, financial data is displayed at public noticeboards. However, this was mostly done at district level and not at the health facilities. Generally, accessing health expenditure data at the health facilities was challenging. Some of the reasons for this are: regular transfer of staff, poor record keeping, and unstable internet connection to the online PBS. Furthermore, the quarterly district expenditure and budget performance reports do not disaggregate expenditures at the health facilities. There is also a need to increase the effectiveness of the non-wage PHC grants. While it was observed that the PHC non-wage grants have increased in absolute terms over the years, there has been no commensurate increase in service outputs. Therefore, there is need to continue aligning financing to results through supply- and demand-side RBF schemes. More broadly, there is need to address encumbrances in the public financial management system. 4.6.2 Health Resources Management and Functionality of Health Facilities 45. There were variations in the number and mix of health workers at the same level of health facilities. According to the staffing norms for the health sector in Uganda, each type and/or level of health facility is supposed to have the same number and mix of health workers. Health facilities located in densely populated areas generally had more health workers which is reflective of the need to respond to the high workload. Some facilities had a progressive increase in staffing numbers over the past three fiscal years (2019/20 to 2021/22) partly due to the ongoing process 26 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. of upgrading facilities from lower to higher-level status. Despite the variations in staffing levels, all the health facilities in the three districts (Mukono, Ntungamo, and Nwoya) adhered to the staffing norms. However, there are no medical officers with specializations in community health, internal medicine, and surgery in the three districts. In line with the staffing norms, all level 4 health centers and hospitals are supposed to have these cadres. This raises questions on the quality of health care that is provided in the three districts. And while community health workers are often engaged through village health teams (VHTs), the recruitment and remuneration of these cadres is haphazard. This often leads to duplication of work and double payments. 46. There are significant deficiencies in the management of the payroll for health workers. This is due to errors that were made when the Integrated Payroll and Pension System was set up in 2014, and unsatisfactory efforts to clean up the system over the years. A number of health workers were being paid through non-health facility cost centers and non-existent health centers. This is further explained in Box 1. In particular, only 42 percent of the staff in Ntungamo district were being paid under their designated health facility cost centers. In Mukono and Nwoya districts, only 12 percent and 24 percent of the staff were being paid under their designated health facility cost centers, respectively. This means that the majority of the health workers in the three districts were paid outside the cost centers of their duty stations. These anomalies can lead to wastage of resources because health facility cost centers have higher pay grades. For example, in Ntungamo district, some staff from other departments (administration and education) were getting higher salaries through the health center payrolls than their job allows. 47. There are some ‘non-existent’ health centers in Ntungamo district. Review of the official MoH master facility list, IFMIS data, payroll register, and the NMS database highlighted some discrepancies across these key government documents and databases. While the list of health facilities on the MoH master facility list was similar to the list on the NMS delivery list, some health facilities on the MoH master facility list were not reflected on the payroll register and vice versa. In Ntungamo district, while there are 37 health facilities reflected on the master facility list, there are only 21 health facilities on the payroll register. Further, there are inconsistencies on the designated level and physical location of the health facilities between what is contained in the MoH master facility list and what is on the ground. In Ntungamo district, Itojo is labelled as a level two health center instead of Itojo Hospital; and Rwamabondo as a level four health center but it exists as a level two health center. Furthermore, several health centers are listed under Ntungamo district even though they belong to other districts. 48. ‘Non-existent’ health centers in Ntungamo district are being used to pay health worker salaries. It was confirmed that the health workers who were being paid through the non-existent health facilities existed, and their names were registered on the district staff lists. These health workers were serving at various health centers even though their salaries were being paid under non-existent health centers. The explanation from the district leadership was that it was not possible to put all the health workers on the payroll and that payrolls at non-existent health centers were created in order to put some health workers on the payroll. However, it is inconceivable for such things to happen because it implies that it is easier to put health workers on a payroll at a non-existing health facility than an existing facility. Does this mean that the staff establishments at existing health centers are already filled? There is need for the MoH to fully investigate this matter. Furthermore, there is a need to harmonize the staff list with the cost centers. All health workers must be paid from the cost centers at their duty stations. 4.6.3 Drugs and Medical Supplies Overview of supply chain management in Uganda 49. The NMS is mandated to procure and distribute drugs and medical commodities to public health facilities countrywide. The NMS conducts deliveries every two months which sums up to six deliveries per year. The budgets for drugs and medical commodities are part of the PHC non-wage budgets, and each year; health facilities are provided with information on the annual allocations for drugs and medical commodities. For level two and three health centers, NMS uses a push system while at higher level facilities a pull system is used. Health facilities track usage of their credit lines based on invoices provided by NMS. The interviewees appreciated the last-mile delivery arrangement by NMS and use of an online system for making orders. However, it was reported that the online system didn’t allow the health facilities to retain a copy of their requisitions. UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 27 Expenditure on Drugs and medical Supplies 50. Review of the invoices from NMS shows that expenditure on drugs and medical supplies has been increasing since 2019/20, especially for level four health centers where there was a 66 percent absolute increase. However, increased expenditure on drugs and medical supplies was not associated with increased patient caseloads or service delivery outputs at all the health facilities which were visited. For instance, at one of the level three health centers, there was a 13 percent decrease in the patient caseload despite a 250 percent increase in the expenditure on drugs and medical supplies. This suggests that there are some inefficiencies in the management of drugs and medical supplies. This could be attributed to theft of medicines as observed in a report on corruption in the health sector in Uganda (Inspectorate of Government 2021). 51. Health facility managers have limited control of the budgets for drugs and medical supplies, and this contributes to expiry of drugs. The MoFPED disburses the funds directly to the NMS which makes all the decisions on the procurement and distribution processes. Even if health facility managers track the usage of their credit lines for drugs and medical supplies by looking at the invoices provided by NMS, some items supplied by NMS are not needed. For lower-level facilities (Level 2 and 3 health centers), where a push system for drugs and medical supplies is used, it was indicated that NMS would deliver some medicines that are not required. In contrast, at higher-level facilities (from level 4 health centers to hospitals) where the pull system is used, facility managers reported that NMS would deliver some medicines that they did not order. This ultimately pushes a facility’s budget for drugs and medical supplies to 100 percent but in actual fact, some of the medicines which are delivered aren’t required and they remain unused. This further shows that a 100 percent utilization of the budget for drugs and medical supplies masks underlying inefficiency issues. The unwanted medicines would eventually expire, leading to a waste of resources. Several medicines were found to have expired with the common ones being antiretroviral drugs, anti-TB drugs, antimalarials, and antidepressants. Across the health facilities, there is an arrangement to supply or receive medicines from neighboring facilities to meet short-term shortages and/or to prevent medicines from expiring. However, additional funds are needed to redistribute the unwanted medicines to other health facilities. Performance of NMS 52. Managers at health facilities and districts are discontented with the performance of NMS. When NMS delivers medicines at health centers, the Health Unit Management Committees, the facility in-charges, stores managers and the district leadership inspect the consignments to ensure consistency between the invoices and deliveries. However, the inspection time was considered inadequate as the NMS teams were said to be always in a rush. Subsequently, there are some discrepancies between what is indicated on the invoices, what is delivered, and what is recorded in the NMS forms. Broader implications could be elicited from the dynamics within national and local governments regarding the (re)centralization versus decentralization of some management functions. The centralization of the function for the procurement and distribution of drugs and medical supplies was considered necessary to enhance scale efficiency and to provide quality medicines. However, these gains could be undermined by market failures arising from the NMS’ monopoly. This manifests in limited transparency over procurement rates for medical supplies. Thus, the pursuit of efficiency through the last-mile model by NMS tends to undermine the oversight role of facility and community leaders. The challenges could be addressed by: (i) procuring and delivering medicines that are needed, (ii) providing requisitions through the online system, (iii) having regular discussions on the procurement process and pricing of medicines and medical supplies, and (iv) having sufficient time for inspecting and signing-off the deliveries. Data management 53. There is poor management and use of data for decision making at the health facilities which were visited. For example, records for medicines required sorting before data extraction could be conducted. This implies that health facilities rarely review their past performance to inform future decisions. 28 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 4.6.4 Recommendations - Health System at Local Government Level 54. The district and other structures at the local government level play vital roles in ensuring the efficient allocation and utilization of scarce resources. These structures have to be adequately resourced for them to play their roles effectively. Based on our findings from Ntungamo, Mukono, and Nwoya districts, we make the following recommendations: a) Strengthen financial reporting and expenditure tracking at the health facilities; b) Improve data management and use in decision making. This includes alignment of financing and drug expenditure data to patient data and performance of health facilities; c) Link financing to results by implementing a robust RBF system at facility and district levels; d) Implement the recommendations from the human resource audit report by the Office of the Auditor General to address discrepancies in staffing, employment status, renumeration of health workers, cost centers, and inconsistencies in HRH data across databases (master facility list, payroll register, cost centers, staffing list); e) Every health worker should be allocated to a cost center that is specifically defined for their location of employment, and all “non-existent” health centers should be removed from the payroll; f) Improve transparency and accountability in the procurement and distribution of drugs and medical supplies by: (i) having regular discussions between NMS, district authorities, and health facility managers on the procurement process and pricing of medicines and medical supplies; (ii) procuring and delivering medicines that are needed; (iii) providing requisitions through the online system; and (iv) having sufficient time for inspecting and signing- off the deliveries. UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 29 Midwife assessing patient at Kashozi HC3, Sheema District (Amanya Brian, 2022) 30 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 5. Equity Analyses 5.1 Introduction 55. In Uganda, health outcomes are affected by various factors, including poverty, low levels of education, inadequate healthcare infrastructure, and high disease burden, among others. Therefore, equity in public spending on health is a critical issue for policymakers in Uganda, where health outcomes are poor and resources are limited. The GoU has made significant efforts to improve access to health services for its citizens, but questions remain on whether public spending on health is distributed equitably across different population groups. 56. There is no clear pattern in the relationship between health spending and health outcomes across sub-regions in Uganda. As shown in Figure 5.1, while there is more per capita expenditure in Elgon sub-region than in Karamoja, Bunyoro, and West Nile sub-regions, Elgon has a lower U5MR than Karamoja, Bunyoro, and West Nile. It is important to note that the variation in per capita expenditure and health outcomes has important allocative efficiency consequences in that as outcome measures improve in some sub-regions where historically there has been more financing, other sub-regions are left behind. As the marginal return to health investments are higher in lagging regions, there are allocative inefficiencies if this variation is not corrected overtime. Thus, there is need to improve efficiency in resource allocation by prioritizing needy areas. Figure 5.1 Per Capita Public Spending and Under-Five Mortality Rate 30,000 120 25,000 100 Deaths per 1000 live births 20,000 80 UGX Million 15,000 60 10,000 40 5,000 20 0 0 li le a Ka z i Bu o o o di h B ala ga a da ile so on ho ng or or nd oj ge ko ke an so Te tN Elg p m ny To La Ac Ki ga m Bu An Bu ug ra es Bu Ka W rth ut No So GGHE-D Per Capita (US$) Under-5 Mortality Rate Source: World Bank staff construction from government financial reports and Demographic Health Survey 2016. 5.2 Benefit Incidence Analysis 57. One of the aspirations listed in the Uganda Vision 2040 is for all Ugandans to have access to affordable quality health services. Recognizing that UHC is a critical component of the country’s development agenda, Uganda is committed to ensuring that all its citizens have access to quality and affordable health services. To achieve UHC, the Ugandan government has implemented various policies and programs aimed at improving access to health services for all the citizens. In particular, Uganda has developed a robust UHC roadmap covering the period 2019-2030 which, among other things, outlines the priorities and sequencing of implementation for the main interventions and milestones for attaining UHC. The UHC roadmap also defines the key investments within and outside the health sector which are required to positively influence the determinants of health (MoH 2019a). Furthermore, as part of the NDP III, Uganda seeks to increase the proportion of the population accessing UHC (NPA 2020). UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 31 58. The government has also prioritized PHC as a key strategy for achieving UHC. This includes strengthening the PHC system, improving access to essential medicines and health technologies, and promoting community participation in healthcare delivery. Over the years, the government has also invested in building new health facilities and upgrading existing ones in order to improve availability of health services in all parts of the country. As a result, about 91 percent of the population lives within a radius of 5 kilometres from a health facility. Furthermore, there have been several interventions aiming to enhance the capacity and number of health workers in the country. In addition, the government has implemented programs aimed at reducing the burden of infectious diseases, such as HIV/AIDS, malaria, and tuberculosis. The government has also prioritized maternal and child health by implementing demand- and supply- side RBF programs which focus on reproductive, maternal, newborn, child health, and nutrition. Uganda has been implementing RBF since 2003 mainly as pilots but scaled it up between 2019 and 2022 when it was implemented in 131 districts and 1,400 health facilities (HCIIIs, HCIVs, Hospitals) under the URMCHIP. To sustain achievements from the RBF under the URMCHIP, the government of Uganda mainstreamed the RBF mechanism into the PHC grant mechanisms in July 2023 through the UgIFT program (MoH 2022). But while the mainstreamed RBF model covers more technical areas, there is less managerial autonomy by health facilities; no payment of incentives to staff; less frequent performance assessments; and it excludes hospitals. 59. A benefit incidence analysis (BIA) was conducted to access whether the various policies, programs and interventions – which have been implemented in the health sector in Uganda – have been reaching the poor. Specifically, this study looked at the distribution of benefits in both the public and private sectors, at different levels of care, across the different socio-economic groups, by urban and rural setting, by sub-regions, by type of service (outpatient vs inpatient), and by comparing the benefits received to the needs of each socioeconomic group. The main data sources for the study are the Uganda National Household Survey (UNHS) for the years 2012/13, 2016/17 and 2019/20. 5.2.1 Trends in Benefit Incidence 60. At public health centers, there was a shift from nearly equally distributed benefits between all socioeconomic groups in 2012/13 and 2016/17 to a distinctly pro-poor distribution in 2019/20 (Figure 5.2). This is a positive finding that encourages GoU to strengthen the efforts of providing comprehensive and high-quality services at public health centers across the country. At public health centers, the poorest, poorer, and the middle class benefit more overtime. The richest and richer benefit less overtime. On the other hand, results in Figure 5.3 show that at public hospitals, the greatest proportion of benefits went to the non-poor in 2012/13, 2016/17 and 2019/20, and it is noted with concern that the proportion of benefits going to the poorest declined from 20 percent in 2016/17 to 14.1 percent in 2019/20. Also, at public hospitals, the poorest benefit less overtime but the poorer benefit more overtime. For the middle class and richer there is no change while the richest benefit more overtime. The trend for benefits at public hospitals is disconcerting because it shows an inequitable distribution of benefits, with the poor possibly having inadequate hospital services over a period of about 8 years. This calls for GoU to improve access for the poor and to remove barriers that prevent them from accessing hospital and secondary care services by increasing the functionality of general hospitals and health center IVs. Figure 5.2 Trends in Benefit Incidence: Public Health Centers 30 24.8 25.6 25 23.2 23.1 20.6 21.1 21.4 20.4 20.0 20.1 20 18.9 17.0 17.5 16.5 15 10.0 10 5 0 Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Pub lic Health Center 2 012/13 Pub lic Health Center 2 016/17 Public Health Center 2 019/20 Source: World Bank staff computations based on UNHS 2012/13, 2016/17 and 2019/20 datasets 32 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. Figure 5.3 Trends in Benefit Incidence: Public Hospitals 30 25.9 24.6 25 23.1 21.6 21.6 20.0 20.5 20.6 18.8 19.3 19.1 19.4 20 15.8 15.8 14.1 15 10 5 0 Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Public Hospital 2012/13 Pub lic Hospital 2016/17 Pub lic Hospital 2019/20 Source: World Bank staff computations based on UNHS 2012/13, 2016/17 and 2019/20 datasets. 61. The poor have had reasonable benefits at private health centers over the years but less benefits at the private hospitals. Results in Figure 5.4 show an almost equal distribution of benefits across the socioeconomic groups in 2012/13 and 2016/17 at private health centers, and a change in distribution with the middle groups benefitting the most in 2019/20, and a marked decline in the benefits of the richest group from 21.8 percent in 2016/17 to 15.9 percent in 2019/20. At private health centers, there was no change in benefit for the poorest overtime but for the poorer and middle class, they benefit more overtime. The richest and richer benefit less overtime. On the other hand, results in Figure 5.5 show that the trends in the distribution of benefits at private hospitals was almost equal for all socioeconomic groups in 2012/13 and 2016/17. However, there was a marked shift to having the largest share of benefits at private hospitals going to the non-poor, with a notable reduction in benefits for the poorest group from 19.4 percent in 2016/27 to 8.6 percent in 2019/20, and a pronounced increase in the benefits of the richest group from 21.1 percent in 2016/27 to 31.1 percent in 2019/20 (Figure 5.5). At private hospitals, the poorest and poorer benefit less overtime while the middle class, richer and richest benefit more overtime. The reason for this shift in distribution of benefits in 2019/20 cannot be deduced from the UNHS data. However, it is important to note that, in general, the use of the private health services (from both private health centers and private hospitals), especially by the poor is confirmation of a lack of financial risk protection and leaves them vulnerable to potentially experience both catastrophic health spending, impoverishment and foregone care. Figure 5.4 Trends in Benefit Incidence: Private Health Centers 25 23.0 22.1 22.1 22.2 21.8 21.5 21.1 20 18.8 19.1 19.3 18.8 18.3 17.8 18.3 15.9 15 10 5 0 Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Private Health Center 2012/13 Private Health Center 2016/17 Private Health Center 2019/20 Source: World Bank staff computations based on UNHS 2012/13, 2016/17 and 2019/20 datasets. UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 33 Figure 5.5 Trends in Benefit Incidence: Private Hospitals 35 31.1 30 25 23.6 21.6 22.0 21.8 20.2 21.1 18.9 19.1 19.4 19.7 19.7 20 18.5 14.9 15 10 8.6 5 0 Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Private Hospital 2012/ 13 Private Hospital 2016/ 17 Private Hospital 2019/ 20 Source: World Bank staff computations based on UNHS 2012/13, 2016/17 and 2019/20 datasets. 62. Over the years at public health centers, benefits to the poorest have increased in both rural and urban areas with the rural poor having more benefits than the urban poor. For the richest, overtime, benefits have decreased in both rural and urban areas with the rich households in rural areas consistently benefitting less than the rich in urban areas. These findings are shown in Figure 5.6. This means that the urban poor are not benefitting from public health services yet there are many poor people in the urban areas. Urban areas like the Kampala have a lower density of public PHC facilities hence people benefit significantly from private facilities. At public hospitals, overtime, benefits to the poorest have decreased in both rural and urban areas but the rural poor still have had more benefits than the urban poor. For the richest, overtime, benefits have increased in both rural and urban areas with the rich households in the urban areas consistently benefitting more than the rich in rural areas. These findings are shown in Figure 5.7. Figure 5.6 Trends in Benefit Incidence in Urban and Rural Areas: Public Health Centers 60 50 14.8 16.9 40 26.4 25.6 23.4 20.2 15.5 30 20.3 17.5 23.1 17.8 23.7 22.2 42.0 23.4 20 36.5 9.1 22.0 21.3 21.2 21.0 19.6 19.2 10 17.3 16.6 14.2 14.5 13.1 12.4 9.1 0 Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Pub lic Health Center 2 012/13 Public Health Center 2 016/17 Public Health Center 2 019/20 urban rural Source: World Bank staff computations based on UNHS 2012/13, 2016/17 and 2019/20 datasets. 34 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. Figure 5.7 Trends in Benefit Incidence in Urban and Rural Areas: Public Hospitals 70 60 22.0 50 17.2 18.1 40 19.1 21.2 19.3 20.8 30 21.2 19.5 20.3 23.0 24.1 20 38.6 38.3 37.0 18.4 18.8 5.0 16.8 24.2 22.8 22.4 20.7 10 18.7 18.8 14.8 11.7 9.6 9.2 8.1 0 Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Public Hospital 2012/13 Public Hospital 2016/17 Public Hospital 2019/20 urban rural Source: World Bank staff computations based on UNHS 2012/13, 2016/17 and 2019/20 datasets. 63. The distribution of benefits at private health centers in urban areas was distinctly favoring the non- poor in 2012/13 and 2016/17. However, between 2016/17 and 2019/20 there was a re-distribution of benefits with increasing benefits for the poorest group from 8.1 percent to 13 percent; and a decline in benefits for the richest group from 41.7 percent to 30.9 percent, respectively (Figure 5.8). Interestingly, while we found that benefits at public hospitals in rural areas were initially almost equal and later largely going to the non-poor in 2019/20, results in Figure 5.8 show that benefits at private health centers in rural areas are largely going to the poor, with a steady decline in the benefits going to the richest group from 16.9 percent (2012/13) to 14.7 percent (2016/17) and to 12.6 percent (2019/20). At private health centers, overtime, benefits to the poorest have consistently increased in urban areas while there has been a reduction in rural areas. Nevertheless, the rural poor have benefitted more than the urban poor. For the richest, overtime, benefits have decreased in both rural and urban areas with the rich households in the rural areas consistently benefitting less than the rich in urban areas. These findings point to the undesirable position of the poor seeking health services from the private sector then having to make OOP payments. This could expose them to catastrophic health spending and impoverishment. Figure 5.8 Trends in Benefit Incidence in Urban and Rural Areas: Private Health Centers 60 50 14.7 16.9 40 12.6 19.7 19.3 20.8 23.1 30 24.1 19.6 20.1 21.1 19.5 41.7 22.9 23.1 20 38.0 22.7 30.9 26.0 26.1 24.5 10 18.4 15.4 15.0 13.2 12.9 13.0 9.1 7.7 8.1 0 Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Private Health Center 2012/13 Private Health Center 2016/17 Private Health Center 2019/20 urban rural Source: World Bank staff computations based on UNHS 2012/13, 2016/17 and 2019/20 datasets. UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 35 64. The benefits at private hospitals in urban settings are disproportionately going to the non-poor, with a consistent increase in the share going to the richest group from 35.5 percent (2012/13) to 39.3 percent (2016/17) to 52.1 percent (2019/20), and a reduction of the share for the poorest group from 9.4 percent (2016/17) to 4.8 percent (2019/20) and a reduction for the poorer group too (Figure 5.9). The sharp increase for the richest group may be associated with the high costs of hospitalization during the COVID-19 period. Obviously, the decline in the share of benefits going to the poorest group at private hospitals is not associated with improvement in financial risk protection since we see a decline in their share of benefits at public hospitals over the same period. These results may simply imply that the poorest are foregoing care because of their low ability to pay in private hospitals. This assertion is confirmed by the results for the distribution of benefits at private hospitals in rural areas, where initially the largest share of benefits was accruing to the poor in 2012/13 and 2016/17 and then shifted significantly with the largest share (25.1 percent) going to the richest group in 2019/20 (Figure 5.9). Results show that between 2016/17 and 2019/20, the benefits for the ‘poorest’ declined from 23.5 percent to 9.7 percent and for the ‘poorer’ from 23/7 percent to 17.3 percent, while the share of benefits for the ‘richer’ increased from 18.5 percent to 24.7 percent and the ‘richest’ from 13.5 percent to 25.1 percent (Figure 5.9). In conclusion, at private hospitals, overtime, benefits to the poorest have decreased in both rural and urban areas, but the rural poor still have had more benefits than the urban poor. For the richest, overtime, benefits have increased in both rural and urban areas with the rich households in the urban areas consistently benefitting more than the rich in rural areas. Figure 5.9 Trends in Benefit Incidence in Urban and Rural Areas: Private Hospitals 80 70 25.1 60 50 13.5 15.8 40 18.8 18.5 24.7 19.5 30 23.3 20.7 52.1 22.8 23.7 23.5 23.1 39.3 20 35.5 17.3 24.4 24.1 20.0 20.1 10 9.7 17.2 16.5 11.2 10.0 9.4 9.0 6.6 4.8 0 Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Private Hospital 2012/13 Private Hospital 2016/17 Private Hospital 2019/20 Urban Rural Source: World Bank staff computations based on UNHS 2012/13, 2016/17 and 2019/20 datasets 65. Another way to present benefit incidence is to look at the distribution of benefits for the outpatient and inpatient services accessed by the different socioeconomic groups at the different levels of care, in both the public and private sectors, over the study period. Figures 5.10, 5.11, and 5.12 show the distribution of benefits for outpatient services for 2012/13 to 2019/20, at health centers and hospitals, in both public and private facilities. Results show that the distribution of benefits for outpatient services in public health centers were distinctly pro-poor in 2012/13, but this trend changed slightly in 2016/17 and 2019/20 when the benefits for the poorest group declined. 36 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. Figure 5.10 Trends in Total Benefits for Outpatient Services: 2012/13 Public Private 40 30 30 20 20 10 10 0 0 Health Center Hospital Health Center Hospital Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Source: World Bank staff computations based on UNHS 2012/13, 2016/17 and 2019/20 datasets Figure 5.11 Trends in Total Benefits for Outpatient Services: 2016/17 Public Private 30 30 25 20 20 15 10 10 5 0 0 Health Center Hospital Health Center Hospital Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Source: World Bank staff computations based on UNHS 2012/13, 2016/17 and 2019/20 datasets Figure 5.12 Trends in Total Benefits for Outpatient Services: 2019/20 Public Private 30 30 25 20 20 15 10 10 5 0 0 Health Center Hospital Health Center Hospital Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Source: World Bank staff computations based on UNHS 2012/13, 2016/17 and 2019/20 datasets UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 37 66. Results also show the same trends for outpatient services at private health centers, where the distribution of benefits for outpatient services in private health centers was distinctly pro-poor in 2012/13, but this trend changed slightly in 2016/17 and 2019/20 when the benefits for the poorest group declined. In 2016/17 and 2019/20, the groups that benefited the most from outpatient health centers (both public and private) were the poorer, the middle and the richer. On the other hand, the distribution of the benefits from outpatient public hospital services were almost equal in 2012/13, while they were distinctly in favor of the non-poor for outpatient public hospital services in 2016/17 and 2019/20. 67. The distribution of benefits for outpatient services in private hospitals is in favor of the non-poor over the study period, with a significant reduction in outpatient benefits at private hospitals, for the poorest group over the years. Similar trends are noted for the distribution of benefits for inpatient services (Figures 5.13, 5.14, and 5.15). Results show that the distribution of benefits for inpatient services in public health centers were distinctly pro-poor in 2012/13 and 2016/17, but this trend changed slightly in 2019/20 when the benefits for the poorest group declined from 23.5 percent in 2016/17 to 15.1 percent in 2019/20. Results also show that the same trends for inpatient services at private health centers, where the distribution of benefits for inpatient services was leaning towards being pro-poor in 2012/13 and 2016/17, but this trend changed tremendously in 2019/20 when the benefits for the poorest group declined sharply from 21 percent and 22 percent, respectively, for the poorest and poorer groups in 2016/17 to 4.3 percent and 5.3 percent for these groups in 2019/20. In 2019/20, the groups that benefited the most from inpatient services at health centers (both public and private) are the middle, the richer and the richest. 68. On the other hand, the distribution of the benefits from inpatient public hospital services were almost equal in 2012/13, while they were distinctly in favor of the non-poor for inpatient public hospital services in 2016/17 and 2019/20. There was a marked increase in the benefits of inpatient services in public hospitals for the richest group in 2019/20. The distribution of benefits for inpatient services in private hospitals is in favor of the non-poor over the study period, with a significant reduction in inpatient benefits at private hospitals for the poorest group over the years (e.g., a sharp reduction from 11.9 percent in 2016/17 to 4.9 percent in 2019/20). Figure 5.13 Trends in Total Benefits for Inpatient Services: 2012/13 Public Private 40 30 30 20 20 10 10 0 0 Health Center Hospital Health Center Hospital Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Source: World Bank staff computations based on UNHS 2012/13, 2016/17 and 2019/20 datasets 38 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. Figure 5.14 Trends in Total Benefits for Inpatient Services: 2016/17 Public Private 30 30 25 20 20 15 10 10 5 0 0 Health Center Hospital Health Center Hospital Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Source: World Bank staff computations based on UNHS 2012/13, 2016/17 and 2019/20 datasets Figure 5.15 Trends in Total Benefits for Inpatient Services: 2019/20 Public Private 40 40 30 30 20 20 10 10 0 0 Health Center Hospital Health Center Hospital Poorest Poorer Middle Richer Richest Poorest Poorer Middle Richer Richest Source: World Bank staff computations based on UNHS 2012/13, 2016/17 and 2019/20 datasets 5.2.2 Trends in Benefits Received Compared to Needs 69. Figure 5.16 compares total benefits with total needs. Results show that between 2012/13 and 2019/20, the health benefits and needs of the households in the poorest and poorer groups increased. In addition, households in these two groups benefited more than their needs in 2012/13 and 2019/20. Based on data in 2012/13 and 2019/20, this was also observed for households in the middle-income group. On the other hand, the health benefits and needs of the households in the richest group reduced between 2012/13 and 2019/20. Additionally, in both 2012/13 and 2019/20, households in the richest group received less benefits as compared to their needs. UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 39 Figure 5.16 Comparing Total Benefits with Total Needs: 2012/13–2019/20 16.9 21.3 19.6 21.9 22.8 26.3 19.4 20.8 20.1 20.7 20.9 20.7 21.8 19.3 19.7 21.2 19.3 18.0 19.2 21.9 19.2 20.2 18.7 17.2 19.0 18.2 20.1 17.8 19.1 19.0 2012/13 2016/17 2019/20 2012/13 2016/17 2019/20 BENEFIT NEED Poorest Poorer Middle Richer Richest Source: World Bank staff computations based on UNHS 2012/13, 2016/17 and 2019/20 datasets 5.3 Catastrophic Health Spending and Impoverishment due to Health Spending 70. Despite the removal of user fees in public facilities, Uganda households OOP spending was estimated at about 40 percent of total CHE in 2018/19.21 This means that households still incurred significant OOP spending on health due to the high cost of services in the private sector and payments incurred at public facilities against the policy on free health care for all.8 Previous studies9,10 such payments comprise a large portion of total health financing. This study assesses the catastrophic and impoverishing impact of paying for health care out-of-pocket in Uganda. Methods: Using data from the Uganda National Household Surveys 2009/10, the catastrophic impact of out-ofpocket health care payments is defined using thresholds that vary with household income. The impoverishing effect of out- of-pocket health care payments is assessed using the Ugandan national poverty line and the World Bank poverty line ($1.25/day have shown significant inequities in health service utilization in Uganda. Uganda’s heavy reliance on OOP payments to finance its health system limits risk pooling, increases fragmentation in health financing, and does not offer financial protection to its population. It defeats the country’s agenda to expand the share of population participating in the money economy. Households which incur high OOP expenditures on health are normally forced to sacrifice other basic necessities and can be pushed into or further into poverty.14 71. In several low- and middle-income countries, OOP payments on health have been found to push households into poverty. It is estimated that 6.5 percent of households in low- and middle-income countries face CatHE and this figure rises to 11.7 percent for the poorest households.3 High OOP costs are associated with reduced medication adherence and increased emergency department visits and hospitalizations.4 an employer, or a Medigap plan. Near-poor Medicare beneficiaries (with incomes more than 100 percent but less than 200 percent of the federal poverty level Households facing CatHEs are more likely to have lower consumption of food and other basic necessities, and more likely to take on debt or sell assets to pay for health care. As a result, CatHEs are associated with a higher risk of falling into poverty.5 the expenses of noncommunicable diseases (NCDs Generating evidence on financial risk protection is important in Uganda because households contribute about 38.7 percent of the total CHE. 72. This section examines the trends in financial burden of OOP spending on health in Uganda. Previous studies9,10such payments comprise a large portion of total health financing. This study assesses the catastrophic and impoverishing impact of paying for health care out-of-pocket in Uganda. Methods: Using data from the Uganda National Household Surveys 2009/10, the catastrophic impact of out-ofpocket health care payments is defined using thresholds that vary with household income. The impoverishing effect of out-of-pocket health care payments is assessed using the Ugandan national poverty line and the World Bank poverty line ($1.25/day have examined equity in health service delivery in Uganda using UNHS datasets for 2005/6, 2009/10, 2012/13, and 2016/17. This study builds on the findings of these previous studies and uses the same methodology to allow a comparison between 2005/6 and 2019/20. The analysis covers: (i) the incidence of catastrophic health expenditure at household level, region/sub-region, 40 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. and rural/urban setting. Secondly, an assessment of impoverishment due to OOP spending on health is conducted by using the Uganda national poverty line and the international poverty line. 5.3.1 Trends in Incidence of Catastrophic Health Spending 73. The results show that at both the 10 percent and 25 percent thresholds, the share of households incurring CatHE has declined over the years (Figure 5.17). Specifically, this means that 11.9 percent of households in Uganda incurred CatHE by spending 10 percent of their total household expenditure on health in 2019/20. This was almost half the 22.4 percent of the households that incurred CatHE in 2005/06. However, 11.9 percent of the 8.97 million households3 in Uganda in 2019/20 is equivalent to 1.07 million households. Thus, if expressed in numbers of people, the 1.07 million households equate to 4.9 million people. This means that many people in Uganda are still experiencing CatHE. Figure 5.17 Trends in the Incidence of CatHE 25 22.4 21.4 20 15 13.8 14.2 11.9 10 5.9 5.4 5 2.6 2.7 2.2 0 Headcount (%) at 10% Headcount (%) at 25% 2005/06 2009/10 2012/13 2016/17 2019/20 Source: World Bank staff computation based on the UNHS 2019/20. Findings for 2005/06-2016/17 (Kwesiga et al 2015; 2020). 74. The incidence of CatHE is higher among the non-poor compared to the poor (poor and poorest socioeconomic groups). See Figure 5.18 and Figure 5.19. This trend holds true for all the data points being compared between 2005/6 and 2019/20. The finding reflects the middle-income vulnerabilities in the country and calls for an urgent need to enhance financial risk protection. Figure 5.18 Incidence of CatHE by Socioeconomic Quintiles - 10 Percent Threshold 25 22.4 21.4 20 15 13.8 14.2 11.9 10 5.9 5.4 5 2.6 2.7 2.2 0 Headcount (%) at 10% Headcount (%) at 25% 2005/06 2009/10 2012/13 2016/17 2019/20 Source: World Bank staff computation based on the UNHS 2019/20. Findings for 2005/06-2016/17 (Kwesiga et al 2015; 2020) 3 In 2019/20, there were 8,974,142 households in Uganda with a national average household size of 4.6 (UBOS, 2020). UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 41 Figure 5.19 Incidence of CatHE by Poor and Non-Poor (% households): 10 Percent Threshold 23.7 22.6 19.5 17.2 14.8 14.7 13.3 12.8 9.8 6.4 Non-poor Poor 2005/06 2009/10 2012/13 2016/17 2019/20 Source: World Bank staff computation based on the UNHS 2019/20. Findings for 2005/06-2016/17 (Kwesiga et al 2015; 2020). 75. The incidence of catastrophic health spending has been declining in both rural and urban areas, but it is still more prevalent in rural than in urban areas (Figure 5.20). This finding is aligned to what UBOS reported for 2019/20, with CatHE of 14.1 percent and 12.1 percent in rural and urban settings, respectively, at the 10 percent threshold. Figure 5.20 Incidence of CatHE by Rural and Urban Setting - 10 Percent Threshold 25 23.5 21.7 20 19.5 15.3 15 14.9 16.2 12.5 13.5 13 10 10.1 5 0 2005/06 2009/10 2012/13 2016/17 2019/20(a) Rural Urban Source: World Bank staff computation based on the UNHS 2019/20. Findings for 2005/06-2016/17 (Kwesiga et al 2015; 2020). 76. The incidence of CatHE varies across regions and sub-regions. As shown in Figure 5.21, the likelihood of experiencing CatHE has been declining across the regions, but the levels differ. For example, in 2019/20, households in the Eastern, Northern, and Western regions of Uganda were more likely to experience CatHE than those in the Central region. At sub-regional level, the highest incidence of CatHE in 2019/20 was in Teso, followed by the West Nile and Toro sub-regions, respectively. The lowest incidence of CatHE in 2019/20 was in the Kigezi and Acholi sub-regions (Figure 5.22). The variations at sub-region are quite significant with Teso (the highest) having 2.6 times the proportion of households experiencing CatHE than Kigezi (the lowest). These results may be modulated by differences in the concentration of different types of services providers across the country; in some regions like in northern Uganda, a big proportion of services are provided by NGOs and low cost PNFP facilities. Results also show that out of the 15 sub- 42 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. regions, eight are below the average incidence of CatHE for 2019/20 (11.9 percent) while seven sub-regions are above the average. This implies that when implementing interventions to address CatHE, there is need to look at variations by the different socioeconomic groups, rural/urban setting, and sub-regions. Figure 5.21 Incidence of CatHE by Region - 10 Percent Threshold 30 27.8 25 22.4 20 21.1 20.3 20.2 Percentage 15 13.4 12.8 12.1 11.9 10 9.6 5 0 Central Eastern Northern Western National 2005/06 2009/10 2012/13 2016/17 2019/20 Source: World Bank staff computation based on the UNHS 2019/20. Findings for 2005/06-2016/17 (Kwesiga et al 2015; 2020). Figure 5.22 Incidence of CatHE by sub-region in 2019/20: 10 Percent Threshold 25 20.3 20 16.4 15 12.9 13.2 13.3 11.7 12.3 12.4 10.8 10.9 9.1 9.8 10 7.9 8.1 8.2 5 0 ile n ga i th ro so e o li di a th o la ez go ho l or oj ng pa ke ko ou To or Te tN so g m ny Ac El La Ki An Bu m aN aS Bu ra es Bu Ka Ka W nd nd ga ga Bu Bu Source: World Bank staff computation based on the UNHS 2019/20 data. 77. The low incidence of CatHE among the poor could be attributed to the poor having a lower ability to spend on health, which suggests that the higher the poverty levels, the lower the incidence of CatHE. This seems to be true for some sub-regions and not for others. Figure 5.22 seems to show mixed results where, on one hand, some of the sub-regions with the highest level of poverty have a relatively low incidence of CatHE (e.g., Acholi, Karamoja, Kigezi), and on the other hand some sub-regions with the high poverty levels also have a high incidence of CatHE (e.g., Busoga, Lango, Teso and West Nile). To investigate the factors influencing the incidence of CatHE, the data for 2019/20 were analyzed further. These are presented in Table 5.1. The results show that the factors that influence the incidence of CatHE among households in Uganda are poverty levels, residing in Teso and West Nile regions, having children below the age of five, and households headed by old people (>60yrs). UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 43 Table 5.1 Factors Influencing CatHE - 10 Percent Threshold Independent Variables Odds Ratio Std. Err. t P>t 95% CI Poverty (poor = 1 , non- poor = 0)* 0.409 0.047 -7.80 0.000 0.3262 - 0.5117 Residence (urban = 1, rural = 0) 0.857 0.083 -1.60 0.110 0.7094 - 1.0355 Region (R = Kampala) Buganda South 0.818 0.198 -0.83 0.407 0.5097 - 1.3144 Buganda North 1.217 0.265 0.90 0.367 0.7943 - 1.8647 Busoga 1.425 0.316 1.60 0.111 0.9223 - 2.2003 Bukedi 1.254 0.293 0.97 0.332 0.7934 - 1.9818 Elgon 1.092 0.252 0.38 0.704 0.6937 - 1.7185 Teso* 2.376 0.518 3.97 0.000 1.5495 - 3.6445 Karamoja 1.408 0.367 1.31 0.189 0.8446 - 2.346 Lango 1.498 0.341 1.78 0.076 0.9592 - 2.3398 Acholi 1.188 0.330 0.62 0.535 0.6889 - 2.0498 West Nile* 1.790 0.409 2.55 0.011 1.1436 - 2.8028 Bunyoro 1.285 0.309 1.04 0.297 0.8019 - 2.0585 Toro 1.437 0.330 1.58 0.115 0.9154 - 2.2543 Ankole 1.284 0.287 1.12 0.264 0.8280 - 1.9922 Kigezi 0.869 0.229 -0.53 0.595 0.5192 - 1.4560 - hsize 1.000 0.017 -0.01 0.989 0.9675 - 1.0331 Employment (R = Unemployed) Formal 0.951 0.078 -0.62 0.538 0.8091 - 1.1169 Casual/Subsistence 1.020 0.108 0.19 0.853 0.8288 - 1.2549 Children below 5 (yes = 1, no = 0)* 1.383 0.117 3.84 0.000 1.1716 - 1.6315 Adults above 60 (yes = 1, no = 0) 1.060 0.106 0.58 0.563 0.8707 - 1.2896 Marital status - (married = 1, not married = 0) 0.922 0.084 -0.89 0.372 0.7721 - 1.1017 Age group (Reference=<30yrs)* 30-39 0.816 0.083 -2.00 0.046 0.6684 - 0.996 40-49 0.675 0.078 -3.41 0.001 0.5378 - 0.8461 50-59 0.755 0.089 -2.38 0.018 0.5988 - 0.9523 60+ 0.910 0.129 -0.67 0.505 0.6884 - 1.2019 _cons 0.132 0.032 -8.39 0.000 0.082 - 0.2115 Source: World Bank staff computation based on the UNHS 2019/20 data. 44 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 5.3.2 Incidence of Household Impoverishment due to OOP Spending on Health 78. Though the incidence of household impoverishment due to OOP spending on health in Uganda has reduced over the years, many people are still being pushed into poverty. Poor households and those living in rural areas are particularly vulnerable to impoverishment. Based on Uganda’s national poverty line, the percentage of households that have been impoverished due to OOP spending on health has decreased from 4.6 percent in 2005/6 to 2.3 percent in 2019/20 (Figures 5.23 and Table 5.2). Considering the 2019/20 impoverishment result of 2.3 percent, this translates to about 206,405 households (949,464 people) pushed into poverty. When the international poverty line is used, the percentage of households that were impoverished due to OOP spending on health reduced from 5.2 percent in 2005/06 to 2.6 percent in 2019/20 (Figure 5.23 and Table 5.3). This translates to 233,328 households (1.07 million people) pushed into poverty. Figure 5.23 Incidence of Impoverishment due to OOP Spending on Health (% households) 7.0 6.0 5.6 4.9 5.0 4.1 4.0 Percentage 3.1 2.9 2.9 2.9 3.0 2.2 1.9 2.0 1.6 1.0 0.0 2005/06 2009/10 2012/13 2016/17 2019/20 Rural Urban Source: World Bank staff computation from UNHS 2019/20. Findings for 2005/06-2016/17 (Kwesiga et al 2015; 2020). UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 45 Table 5.2 Impoverishment at Uganda’s National Poverty Line (2005/6 - 2019/20) Pre-Payment Post-Payment Difference (%)     Poverty (%) (A) Poverty (%) (B) (B-A) 2005/06 Poverty headcount 31.1 35.6 4.6 Normalized mean positive poverty gap 35.2 37.0 2009/10 Poverty headcount 23.2 27.2 4.0 Normalized mean positive poverty gap 27.6 28.3 2012/13 Poverty headcount 19.7 21.7 2.0 Normalized mean positive poverty gap 26.4 26.7 2016/17 Poverty headcount 21.5 24.1 2.7 Normalized mean positive poverty gap 5.3 6.0 2019/20 Poverty headcount 29.5 31.8 2.3 Normalized mean positive poverty gap 28.5 28.7 Source: World Bank staff computation based on the UNHS 2019/20. Findings for 2005/06-2016/17 (Kwesiga et al 2015; 2020). Table 5.3 Impoverishment at International Poverty Lines (2005/6 - 2019/20) Pre-Payment Post-Payment Difference (%)   Poverty (%) (A) Poverty (%) (B) (B-A) 2005/06 (PPP = 513.95) Poverty headcount 51.8 57.0 5.2 Normalized mean positive poverty gap 35.2 37.0 2009/10 (PPP = 741.33) Poverty headcount 46.3 50.8 4.5 Normalized mean positive poverty gap 33.4 34.9 2012/13 (PPP = 1,043.08) Poverty headcount 64.0 67.2 3.2 Normalized mean positive poverty gap 39.4 40.2 2016/17 (PPP = 1,161.99) Poverty headcount 51.8 57.0 5.2 Normalized mean positive poverty gap 35.2 37.0 2019/20 (PPP = 1,331.07) Poverty headcount 55.9 58.5 2.6 Normalized mean positive poverty gap 36.9 38.5   Source: World Bank staff computation based on the UNHS 2019/20. Findings for 2005/06-2016/17 (Kwesiga et al 2015; 2020). 46 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 79. Based on the international poverty line, the OOP spending on health among households in rural areas in Uganda between 2005/06 and 2019/20, the level of impoverishment in rural areas has remained above the level in urban areas. Results show that in 2019/20, about 2.9 percent of the households in the rural areas were pushed into poverty due to health spending compared to 1.9 percent of the households in urban areas (Figure 5.24). In absolute numbers, about 181,374 households (870,597 people) in rural areas experienced impoverishment due to OOP spending on health which is significantly higher than 51,677 households (206,709 people) in urban areas who experienced the same fate.4 Figure 5.24 Incidence of Impoverishment due to OOP Spending on Health - Rural vs Urban (% households) 6 5 5.6 4.9 4 4.1 3 3.1 2.9 2.9 2.9 2 2.2 1.9 1 1.6 0 2005/06 2009/10 2012/13 2016/17 2019/20 Ru ral Urban Source: World Bank staff computation based on the UNHS 2019/20. Findings for 2005/06-2016/17 (Kwesiga et al 2015; 2020). 80. Using the international poverty line, results show that CatHE is more prevalent in some groups than others. As shown in Figure 5.25, while impoverishment due to CatHE has reduced over the period under review, the ‘middle’ and ‘second richest’ groups are the most affected – more than the poorest, second poorest and richest groups. The impoverishment results for 2019/20 require critical attention for the ‘middle’ and ‘second richest’ socioeconomic groups where 7 percent and 5.7 percent of the households, respectively, were impoverished because of CatHE. However, the consistently low incidence of impoverishment due to CatHE among the ‘poorest’ and ‘second poorest’ groups between 2005/06 and 2019/20 could also mean that they are too poor to be further impoverished because they don’t have the financial ability to spend on health services. Analysis of impoverishment due to CatHE by region shows declining trends in all the regions except for the Northern region (Figure 5.26). Between 2005/06 and 2019/20, the Central and Western regions had the largest reduction in the incidence of impoverishment due to CatHE while in the Northern region there was a slight increase. 4 In 2019/20, there were 6,254,292 households in rural areas and 2,719,850 households in urban areas in Uganda. The average household sizes in rural and urban areas were 4.8 and 4.0, respectively (UBOS, 2020). UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 47 Figure 5.25 Incidence of Impoverishment Due to OOP Spending on Health by Socioeconomic Groups (% households) 25 20 18.6 16.7 14.9 15 11.0 10 8.3 7.0 5.7 5 2.5 2.5 1.0 1.5 1.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.3 0 Poorest Second poorest Middle Second richest Richest 2005/06 2009/10 2012/13 2016/17 2019/20 Source: World Bank staff computation based on the UNHS 2019/20. Findings for 2005/06-2016/17 (Kwesiga et al 2015; 2020). Figure 5.26 Incidence of impoverishment due to OOP Spending on Health by Region (% households) 6.9 5.4 5.2 5.3 4.8 4.9 4.4 3.6 3.4 3.4 3.0 2.8 3.0 2.5 2.5 2.4 2.7 2.7 2.0 1.7 Central Eastern Northern Western 2005/06 2009/10 2012/13 2016/17 2019/20 Source: World Bank staff computation based on the UNHS 2019/20. Findings for 2005/06-2016/17 (Kwesiga et al 2015; 2020). 48 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 49 6. References Ahmed, T.; Roberton, T.; Vergeer, P.; Hansen, P.M.; Peters, M.A.; Ofosu, A.A.; Mwansambo, C.; Nzelu, C.; Wesseh, C.S.; Smart, F.; and Alfred, J.P. 2022. “Healthcare Utilization and Maternal and Child Mortality During the COVID-19 Pandemic in 18 Low-And Middle- Income Countries: An Interrupted Time-Series Analysis With Mathematical Modeling of Administrative Data.” PLoS Medicine, 19(8), p.e1004070. Bennett, S.; Quick, J.D.; and Velasquez G. 1997. “Public-Private Roles in the Pharmaceutical Sector: Implications for Equitable Access and Rational Drug Use.” Geneva: World Health Organization. Burt, J.F.; Ouma, J.; Lubyayi, L.; Amone, A.; Aol, L.; Sekikubo, M.; Nakimuli, A.; Nakabembe, E.; Mboizi, R.; Musoke, P.; and Kyohere, M. 2021. “Indirect Effects of COVID-19 on Maternal, Neonatal, Child, Sexual and Reproductive Health Services in Kampala, Uganda.” BMJ Global Health, 6(8), p.e006102. Chisholm D. and Evans D.B. 2010. “Improving Health System Efficiency as a Means Of Moving Towards Universal Coverage.” World Health Report 2010 Background Paper No. 28. Accessed from https://cdn.who.int/media/docs/default-source/health- financing/technical-briefs-background-papers/whr-2010-background-paper-28.pdf on June 16, 2023. Di Giorgio, L.; Evans, D.K.; Lindelow, M.; Nguyen, S.N.; Svensson, J.; Wane, W.; Tärneberg, A.W. 2020. Analysis of clinical knowledge, absenteeism and availability of resources for maternal and child health: a cross-sectional quality of care study in 10 African countries. BMJ Global Health, 5(12), p.e003377. Fullman, N.; Yearwood, J.; Abay, S. M.; Abbafati, C.; Abd-Allah, F.; Abdela, J.; ... and Chang, H. Y. 2018. Measuring performance on the Healthcare Access and Quality Index for 195 countries and territories and selected subnational locations: a systematic analysis from the Global Burden of Disease Study 2016. The Lancet, 391(10136), 2236-2271. Hafez, R., ed. 2020. “Measuring Health System Efficiency in Low- and Middle-Income Countries: A Resource Guide.” Joint Learning Network for Universal Health Coverage. Inspectorate of Government. 2021. “Extent and costs of corruption in the health sector in Uganda.” Kampala: Inspectorate of Government. IHME (Institute for Health Metrics and Evaluation). 2023. “Uganda. What causes the most death and disability combined?” https:// www.healthdata.org/research-analysis/health-by-location/profiles/uganda Kwesiga, B.; Zikusooka, C.M. and Ataguba, J.E., 2015. “Assessing catastrophic and impoverishing effects of health care payments in Uganda.” BMC health services research, 15(1). https://doi.org/10.1186/s12913-015-0682-x Kwesiga, B.; Aliti, T.; Nabukhonzo, P.; Najuko, S.; Byawaka, P.; Hsu, J.; Ataguba, J.E.; and Kabaniha, G., 2020. What has been the progress in addressing financial risk in Uganda? Analysis of catastrophe and impoverishment due to health payments. BMC health services research, 20, pp.1-8. https://doi.org/10.1186/s12913-020-05500-2 Mathonnat J. 2010. “Disponibilité des ressources financières pour la santé dans les pays d’Afrique subsaharienne.” Paris: Agence Française de Développement Département de la Recherche. McIntyre, D., Meheus, F., and Røttingen, J.A., 2017. “What Level of Domestic Government Health Expenditure Should we Aspire to for Universal Health Coverage?” Health Economics, Policy and Law, 12(2), pp.125-137. MoH (Ministry of Health). 2019a. “A Roadmap Towards Universal Health Coverage in Uganda: 2019-2030.” Kampala: Ministry of Health. MoH (Ministry of Health). 2019b. “Service Availability and Readiness Assessment and Data Quality Review.” Kampala: Ministry of Health. MoH (Ministry of Health). 2020a. “National Health Accounts 2016–2019.” Kampala: Ministry of Health. MoH (Ministry of Health). 2020b. “Human Resources for Health Audit Report 2019/2020.” Kampala: Ministry of Health. MoH (Ministry of Health). 2021a. “Health Resources for Health Strategic Plan 2020–2030.” Kampala: Ministry of Health. MoH (Ministry of Health). 2021b. “Annual Health Sector Performance Report 2020/21.” Kampala: Ministry of Health. MoH (Ministry of Health). 2022. “Strategy for Mainstreaming Results-Based Financing into the Uganda Intergovernmental Fiscal Transfers Program for Results in the Health Sub-Program.” Kampala: Ministry of Health. 50 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. MoH (Ministry of Health). 2023. “Draft National Health Accounts 2019/20 and 2020/21.” Kampala: Ministry of Health. MoH (Ministry of Health) and WHO (World Health Organization). 2023. “Harmonized Health Facility Assessment in Uganda, 2022.” Kampala: Ministry of Health. NPA (National Planning Authority). 2020. “Third National Development Plan (NDP III) 2020/21- 2024/25.” Kampala: National Planning Authority. UBOS (Uganda Bureau of Statistics). 2016. Uganda National Household Survey Report 2016. Kampala: Uganda Bureau of Statistics. UBOS (Uganda Bureau of Statistics). 2017. Demographic and Health Survey 2017. Kampala: Uganda Bureau of Statistics. UBOS (Uganda Bureau of Statistics). 2020. Uganda National Household Survey Report 2020. Kampala: Uganda Bureau of Statistics. UBOS (Uganda Bureau of Statistics). 2023. Uganda Demographic and Health Survey 2022: Key Findings. Kampala: Uganda Bureau of Statistics. Vujicic, M., Ohiri, K., and Sparkes, S. 2009. “Working in health: financing and managing the public sector health workforce.” Washington, DC: World Bank. Wane, W. and Martin, G. 2013. “Education and health services in Uganda: Data for Results and Accountability.” Washington, DC: World Bank. WHO (World Health Organization). 2004. “Making Pregnancy Safer: The Critical Role of the Skilled Attendant.” A Joint Statement by WHO, ICM, and FIGO. Geneva: WHO. WHO (World Health Organization). 2013. “A Universal Truth: No Health Without a Workforce.” Geneva: WHO. WHO (World Health Organization). 2016. “Global strategy on human resources for health: workforce 2030.” Geneva: WHO. WHO (World Health Organization). 2022. “Cross programmatic efficiency assessment for selected health programs in Uganda: 2015-2020.” Geneva: WHO. Wirtz, V.J., Hogerzeil, H.V., Gray, A.L., Bigdeli, M., de Joncheere, C.P., Ewen, M.A., Gyansa-Lutterodt, M., Jing, S., Luiza, V.L., Mbindyo, R.M., and Möller, H. 2016. Essential medicines for universal health coverage. The Lancet, 389(10067), pp.403-476. World Bank. 2009. “Core Guidance: Preparing Public Expenditure Reviews for Human Development.” Washington, DC: World Bank World Bank. 2016. “Fiscal Space for Health in Malawi and Revenue Potential of Innovative Financing.” Washington, DC: World Bank. World Bank. 2017. “Uganda Economic Update 10th edition: Accelerating Uganda’s Development: Ending Child Marriage, Educating Girls.” Washington, DC: World Bank. World Bank. 2018. “Zambia Health Sector Public Expenditure Review.” Washington, DC: World Bank. World Bank. 2019a. “Kenya Health Service Delivery Indicator Survey 2018 Report.” Washington DC: World Bank. World Bank. 2019b. “Uganda Poverty Monitoring and Analysis. Uganda Poverty Update Note.” Washington, DC: World Bank. World Bank. 2019c. “Malawi Harmonized Health Facility Assessment: 2018-2019 Main Report.” Washington, DC: World Bank World Bank. 2020a. “Malawi Public Expenditure Review 2020: Strengthening Expenditure for Human Capital.” Washington, DC: World Bank. World Bank. 2020b. “Strengthening Social Protection to reduce Vulnerability and Promote Inclusive Growth.” Uganda Economic Update 14th edition. Washington, DC: World Bank. World Bank. 2022a. “Ghana Health Public Expenditure Review 2022.” Washington, DC: World Bank. World Bank. 2022b. “Uganda Intergovernmental Fiscal Transfer Program. Impact and Lessons Learned Report.” Kampala: World Bank. World Bank. 2022c. World Development Indicators Database. Accessed April 2022. https://databank.worldbank.org/source/world- development-indicators Yin R, K. 2013. Case study research: Design and Methods. 5th edition. Thousand Oaks, CA: Sage Publications. UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending. 51 52 UGANDA PUBLIC EXPENDITURE REVIEW 2022–23 Enhancing, Efficiency, Effectiveness & Equity in Health Spending.