Introduction Aiming for universal health coverage (UHC) as a country-level goal requires that progress is measured and tracked over time. However, few national and subnational studies monitor UHC in low-income countries and there is none for Ethiopia. This study aimed to estimate the 2015 national and subnational UHC service coverage status for Ethiopia. Methods The UHC service coverage index was constructed from the geometric means of component indicators: first, within each of four major categories and then across all components to obtain the final summary index. Also, we estimated the subnational level UHC service coverage. We used a variety of surveys data and routinely collected administrative data. Results Nationally, the overall Ethiopian UHC service coverage for the year 2015 was 34.3%, ranging from 52.2% in the Addis Ababa city administration to 10% in the Afar region. The coverage for non-communicable diseases, reproductive, maternal, neonatal and child health and infectious diseases were 35%, 37.5% and 52.8%, respectively. The national UHC service capacity and access coverage was only 20% with large variations across regions, ranging from 3.7% in the Somali region to 41.1% in the Harari region. Conclusion The 2015 overall UHC service coverage for Ethiopia was low compared with most of the other countries in the region. Also, there was a substantial variation among regions. Therefore, Ethiopia should rapidly scale up promotive, preventive and curative health services through increasing investment in primary healthcare if Ethiopia aims to reach the UHC service coverage goals. Also, policymakers at the regional and federal levels should take corrective measures to narrow the gap across regions, such as redistribution of the health workforce, increase resources allocated to health and provide focused technical and financial support to low-performing regions.
Ethiopia is the second most populous country in Africa, with a total population of about 105 million in 2016.16 Administratively, Ethiopia is divided into nine regional states (Tigray, Afar, Amhara, Oromia, BenshangulGumuze, Southern Nations, and National Region (SNNPR), Somali and Harari) and two chartered cities (Addis Ababa and Dire Dawa). Healthcare delivery in Ethiopia is organised in a three-tier system.8 The first, at the district level, is the primary healthcare unit (PHCU). The PHCU comprises one primary hospital, which can serve a population of about 60 000–100 000; four health centres (each serving a population of 15 000–25 000) and five health posts are attached to each health centre (each health post serving 3000–5000 people). The second level comprises general hospitals, each serving a population of 1–1.5 million, while the third level comprises specialised hospitals for a population of 3.5–5 million. While the Federal Ministry of Health is responsible for the formulation and harmonisation of health programmes and strategies, the RHBs are mostly responsible for actual implementations. The budget flows to RHBs in two ways. From one side, the RHBs receive about 5%–10% of the total annual regional budget. This part of the budget is mainly spent on salaries for health professionals, procurement of medical supplies and procurement of drugs. The regions also use this part of the budget for construction and expansion of health centres and primary hospitals. The RHBs have a full mandate on this part of the budget. On the other side, RHBs receive an additional earmarked budget for specific programmes from external sources via the Federal Ministry of Health. In addition, the Ministry of Health also distributes un-earmarked funds from the SDG pool fund. In this study, we applied the approaches described by Hogan et al and the WHO/WB report on tracking progress towards UHC, with some modifications.3 11 We used local data sources and checked that the indicators were also relevant for Ethiopia and that the data were available for all nine regions and the two city administrations. The selected indicators were well aligned with Ethiopia’s priorities, set by the health sector transformation plan.8 Sixteen indicators are from four major categories: RMNCH, infectious diseases, NCDs and service capacity and access. Tracer indicators in the area of RMNCH were as follows: family planning (FP) demand satisfied with a modern method among married women or in a union; pregnancy care (PC); immunisation for infants with three doses of pentavalent vaccine and care-seeking for children with suspected pneumonia. For infectious diseases, tracer indicators were tuberculosis treatment coverage (TB cases detected and cured); HIV treatment coverage; use of insecticide-treated bed nets among populations at risk of malaria and household access to at least basic sanitation services. For NCDs, we used the following: prevalence of non-raised blood pressure (BP), mean fasting plasma glucose, cervical cancer screening and prevalence of tobacco non-smoking. To assess service capacity and access, we used hospital bed density, health worker density, access to essential medicines and the International Health Regulations core capacity index. We used a variety of data sources from Ethiopia (table 1), namely Ethiopia’s Health Management Information System (HMIS),17 Ethiopia’s 2016 DHS (EDHS),18 the 2015 Malaria Indicator Survey (MIS),19 the NCD STEPwise approach to noncommunicable disease risk factor surveillance (STEPS) survey,20 the 2016 Service Readiness and Availability (SARA) survey21 and a Human Resource Information System (HRIS). Also, health security (HS) information was collected from administrative records at the Federal Ministry of Health and regional health offices. Since this indicator is only available at national level only, it is excluded from subnational analysis. Sources of data and indicator description for the UHC service coverage tracer indicators CVD, cardiovascular disease; DM, diabetes mellitus; EDHS, Ethiopia’s Demographic and Health Survey; HMIS, Health Management Information System; HRIS, Human Resource Information System; MIS, Malaria Indicator Survey; NCDs, non-communicable diseases; RMNCH, reproductive, maternal, neonatal and child health; SARA, Service Readiness and Availability; TB, tuberculosis; UHC, universal health coverage. The Ethiopian 2016 DHS data were used to estimate UHC service coverage for RMNCH indicators. For measurement of the malaria prevention indicator, the 2015 Ethiopian MIS survey was used. The 2016 SARA survey was used to estimate the coverage of essential medicine. The SARA survey generates a set of core indicators on key inputs and outputs of the health scheme, which can be applied to assess progress in the health system, strengthening over time.22 To estimate service coverage for the prevention of cardiovascular disease (CVD), management of diabetes mellitus (DM), cervical cancer screening and tobacco control, the NCD STEPS survey was applied. The Ethiopia STEPS are a nationally representative survey to gather comprehensive data on risk factors for NCDs, injuries and violence in Ethiopia. To estimate HIV treatment coverage, HMIS data, which is routinely collected from service provision at each facility, was used. The data source for health workforce (HWF) density are the HRIS of the Ministry of Health. The UHC coverage index was constructed from geometric means of the four major component indicators.3 For the RMNCH category, the geometric mean of FP, PC, immunisation and child healthcare (CHC) were taken; for FP, contraceptive prevalence rate; for PC, a combination of prevalence of births attended by a skill birth attendant and prevalence of antenatal care coverage (ANC4+); for immunisation, DPT3 coverage and for CHC, treatment for childhood pneumonia in the last 2 weeks were used as follows: RMNCH =(FP*PC*DPT3*CHC)1/4 For measurement of UHC service coverage in the infectious disease category, tuberculosis treatment (TB) was measured using the TB case detection rate and cure rate; antiretroviral treatment (ART) coverage was measured using people living with HIV who are currently on ART; water and sanitation (WASH) was measured using the average coverage of households with access to improved water and sanitation and Long-lasting insecticidal nets (LLIN) coverage was used. Infectious = (TB∗ART∗WASH∗ITNright)1/4 LLIN coverage was not accounted for in Addis Ababa since the area is malaria-free. NCD service coverage was calculated by a geometric mean of non-raised BP, fasting blood plasma glucose level (FPG), cervical cancer screening coverage and prevalence of non-tobacco users. We used the 2015 STEPS survey to compute the four tracer indicators in this category. The non-raised BP rate was measured by a prevalence of systolic BP<140 mm Hg or diastolic BP<90 mm Hg among adults aged 18 years and older. The FPG rate was measured by a prevalence of fasting plasma glucose of ≥7.0 mmol/L or those on medication for raised blood glucose among adults aged 18 years or older. The cervical cancer screening rate was measured by a proportion of women aged 30–49 years who reported ever having had a screening test for cervical cancer using any of the methods (visual inspection with aceticaccede, pap smear and human papillomavirus test). For measurement of no tobacco use, the proportion of adults 15 years and older who have not smoked tobacco in the last 30 days of survey time was applied: NCD=(BP∗FPG∗CancerScreening∗Tobaccoright)1/4 We used the prevalence of non-raised BP to estimate the service coverage for ‘prevention of CVD’ and the prevalence of non-raised blood glucose to estimate the service coverage for ‘management of DM’. Since these two indicators are not measured in a proportion, we used the rescaling formulas recommended by WHO/WB:3 For measurement of health service capacity and access (HSCA) coverage, hospital access (HP), HWF density, HS and data on the availability of essential medicine were used. For hospital access, we used annual in-patient admission or discharge per capita. For HWF density, we used the availability of health professionals: physicians, psychiatrists and surgeons per capita. For HS, we used the International Health Regulation core capacity index. Since this indicator is only available at the national level, it was excluded from the subnational analysis. For the measurement of essential medicines, we calculated the availability of the 14 WHO-recommended core list of essential medicines (ie, glibenclamide, beta-blocker, ACE inhibitor, simvastatin, amitriptyline, ciprofloxacin, co-trimoxazole, amoxicillin, ceftriaxone injection, diazepam tablet, diclofenac/ibuprofen, paracetamol and omeprazole) at health facilities: HSCA=(Hospitalaccess∗HWF∗Essentialmedicines∗HSright)1/4 Therefore, the overall UHC service coverage was computed within each of the four categories and then across those category-specific means to obtain the final summary index.3 UEHSC=(RMNCH∗Infectious∗NCD∗HealthServiceCapacityright)1/4 We computed the regional-level UHC service coverage status in the same way as the national index and compared their distributions to the national coverage and among regions. In this analysis, the geometric mean was applied instead of the arithmetic mean because the geometric mean is less sensitive to extreme values.19 We tested the sensitivity of the index to see how the indicators were combined into a summary measure by recomputing the index, using the arithmetic means in addition to the geometric means that was performed in the base case. We also assessed the sensitivity of the index by dropping one indicator at a time: first, deleting the HS variable and then deleting the entire ‘health service and capacity’ component. The study was approved by the Institutional Review Board (IRB) of the Ethiopian Public Health Institute (Ref: EPHI/6.13/607). No patient level data were used in this study.
N/A