Challenges

Shortages of Health Care Personnel

The government implemented several initiatives to alleviate shortages of health care personnel during the study period. Serious challenges remain however. Among the most serious of these is the highly uneven distribution of the health workforce.

Despite the government’s use of a lottery system to deploy physicians and nurses to rural areas, providers are still more likely to work in urban areas. This uneven distribution leaves other regions, particularly pastoralist communities, with far too few health workers scattered across vast geographic areas.

For example, in Addis Ababa the ratio was one physician per 3,056 people in 2009 – whereas in Afar Region, in the heavily pastoralist northeast, the ratio was one physician per 98,258, and in Oromia it was one per 76,075.¹

Even the health workers who do work in rural areas often transfer to urban locations within three years, again leaving vacancies in the rural facilities.¹

Ethiopia’s efforts to strengthen and sustain sufficient numbers of medical professionals throughout the country is compromised by high levels of dissatisfaction among health workers in the public sector. In 2012, 70 percent of doctors and nurses in the public sector reported being “completely unsatisfied” with their salaries.¹

Key informants expressed concern about the potential for widespread burnout and turnover among health extension workers (HEWs), whose vital contributions to public health have expanded over the years as additional interventions have been added to their responsibilities.

Inequity of Coverage

The successful expansion of various interventions – some of which had very low coverage levels at the start of the study period – likely contributed to impressive reductions in U5M. However, numerous gaps remain in both the level and equity of coverage. Many interventions are now implemented at a national scale, but still have not achieved ideal reach.

For example, the measles vaccine has been implemented since 1980, but Demographic and Health Surveys estimated national coverage to be only 54 percent in 2016, with a sizeable equity gap, ranging from 43 percent in the lowest wealth quintile to 74 percent in the highest.⁶

One factor in uneven coverage is geography. In particular, uptake of several interventions remains low in northeastern regions such as Afar and Somali, especially compared with other regions that have achieved high coverage.

In 2005, Ethiopia’s Federal Ministry of Health (FMOH) introduced health care financing reforms, with the specific aim to improve equity in the availability of health services. Maternity and family care services, including antenatal care, were made free of charge at health posts, health centers, and primary hospitals.³

A 2008 study revealed that many facilities did not adhere to these reforms. Findings from this study showed that 68 percent of facilities offering delivery services either charged a fee for normal delivery or required women to buy their own supplies. This findings indicated that although the FMOH had enacted policies to reduce financial barriers to facility-based delivery, enforcement of those policies had to improve if Ethiopia was to achieve its goal of reducing inequity.⁴ As late as 2013, participants were still noting high costs as a reason for not delivering at a facility.⁵

Other geographic inequalities remained. The 2016 Demographic and Health Survey (DHS) reported that only 33 percent of Ethiopian women who had given birth in the preceding five years had done so at a health facility. Only 26 percent of women in rural areas delivered at a facility, compared with 86 percent in urban areas.⁶

Low Use of Health Care Services Despite Improved Access

Use of health care services in Ethiopia is traditionally low. The government used many strategies during the study period to increase use, such as establishing new health facilities, creating the Health Extension Program, deploying ambulances for deliveries and emergencies, and removing user fees for many health services.

Despite this enhanced availability, use of services remained modest. From 2005 to 2011, outpatient department attendance per capita hovered around only 0.3 visits per person each year.⁷

Because many interventions to decrease child mortality are delivered through health facilities, low use of health care services likely affected the coverage of some U5M interventions.

A 2012 cluster-randomized study assessed the performance of an Integrated Community Case Management (iCCM) program in Oromia Region. Findings from the study showed that although iCCM implementation was generally strong in intervention areas, the program did not significantly increase care-seeking for – and appropriate treatment of – pneumonia, diarrhea, and fever in children under five. In addition, although U5M had decreased by 12.7 percent in the study area, the difference in mortality decreases between the study area and comparison areas were not statistically significant.

Four years later, the 2016 DHS found that advice or treatment was sought for fewer than half (46 percent) of children under five with diarrhea.⁶ The same study reported that only 19 percent of mothers who gave birth within the preceding two years had received postnatal care. For women who had given birth outside of facilities, the rate was worse still – just 1.9 percent, compared with 42 percent for those who had delivered at a facility.

Furthermore, only 27 percent of newborns born within the preceding two years had at least two of seven key services (“signal functions”) of postnatal care performed within two days after birth.⁶ This likely represented low overall use of health care services, as well as potential gaps in quality of care.

In addition, the low rates of facility-based delivery may have contributed to low coverage of key neonatal interventions such as neonatal resuscitation and neonatal intensive care units, thereby affecting progress in reducing neonatal mortality.

In a 2016 literature review conducted by UNICEF and PATH on barriers to use of Health Extension Program services in Ethiopia, it identified both demand-side and supply-side barriers. The review found that care-seeking was likely limited by poor knowledge of caregivers about the signs of childhood illnesses. Lack of awareness of services offered by HEWs, including iCCM, probably also contributed to low use. Multiple studies found that the distance of some communities from health posts – as well as a lack of transportation – made care-seeking more difficult.⁸

In addition to these demand-side barriers, the review identified supply-side barriers to use of services. Availability of services at health posts was identified as a frequent barrier due to their limited hours of operation and the absence of HEWs from health posts during working hours. Supply chain issues resulting in lack of medicines at health posts was another key supply-side barrier.⁸

Overall, while improving access to key interventions to reduce U5M is essential, increasing use of these services remains an area for improvement in Ethiopia.

Gaps in Quality of Care and Fidelity to Evidence-Based Interventions

Throughout the study period, Ethiopia encountered substantial challenges in maintaining quality of care and ensuring systemwide fidelity to intervention guidelines.

A 2011 study of children admitted to outpatient therapeutic programs for severe acute malnutrition in 94 health districts in Ethiopia’s Southern Nations, Nationalities, and People's (SNNP) Region noted that upon discharge from the program, only 33 percent of children met the recovery criteria of 15 percent weight gain or resolution of edema. Fourteen weeks after admission to the outpatient therapeutic program, 40 percent of children were still classified as having severe acute malnutrition and 37 percent were classified with moderate acute malnutrition; 23 percent had attained normal nutritional status.⁹

The 2014 Ethiopian Service Provision Assessment Plus (ESPA+) survey found ongoing gaps in the coverage and quality of Integrated Management of Childhood Illness (IMCI) services in facilities across Ethiopia, more than a decade after the program’s introduction.

For example, only 5 percent of observed consultations for ill children under the age of five included the requisite assessment of all danger signs indicated by IMCI standards, including such basic conditions as convulsions or an inability to eat. Only 40 percent of children diagnosed with malaria received antimalarial drugs of any kind.¹⁰

Fewer than half (43 percent) of providers in health posts were receiving in-service training on diarrhea diagnosis and treatment – among all other facility types, the rate was 7 percent, although this figure may underestimate actual readiness because it does not include preservice training.¹⁰

A 2016 FMOH assessment found that only 9 percent of health facilities had provided antenatal corticosteroids (ACS) in the preceding three months. Of the 91 percent of facilities that had not provided ACS during that period, 36 percent cited a lack of corticosteroids as a reason for not administering ACS, and 23 percent cited insufficient training.¹¹

Several quality-of-care issues have arisen in the area of newborn health. A 2006 study found a low level (24 percent) of correct usage of the Active Management of the Third Stage of Labor (AMSTL) protocol. Partograph use – a central component of AMSTL – remained low, as shown by a study conducted in early 2012. All providers knew what a partograph was, but just over half (57 percent) were actually using one, and only 20 percent knew the correct function of a partograph alert line. As late as 2014, only 10 percent of providers interviewed for the Ethiopia Service Provision Assessment Plus (ESPA+) survey had received in-service training on AMTSL.¹⁰

That same assessment found that although 67 percent of facilities (excluding health posts) providing normal delivery services also provided kangaroo mother care (skin-to-skin care for low-weight infants), only 15 percent of providers reported receiving in-service training in the practice.¹⁰

High Level of Dependence on Donor Financing

The FMOH leveraged donor funding to introduce and expand several interventions to reduce U5M. Donor resources have increased the feasibility of rapidly expanding interventions across a large country; however, continued dependence on donor funding may threaten sustainability of this work, in spite of the government’s strong leadership in the health sector.

Uneven and Low Quality of Data

Ethiopia’s use of data facilitated reductions in U5M, but key informants still expressed concerns about the quality of data, particularly in the Health Management Information System. During the study period (2000–2015), the problem of low-quality data was most acute with regard to vaccinations. For example, the FMOH reported coverage of the third dose of pneumococcal conjugate vaccine to be 96 percent in 2015. However, WHO and UNICEF estimated coverage to be only 68 percent, and the 2016 DHS reported an even lower level – 49 percent.^2,6   These differences in coverage based on different data sources were seen in a number of vaccination metrics.

As data use has been an important element in Ethiopia’s strategies for reducing U5M, inaccurate data may affect implementation of these interventions, and the ability to accurately assess their effectiveness.

Persistent Supply Chain Weaknesses

Gaps in Ethiopia’s medical supply chain have diminished the availability of medicines and other important health supplies, potentially reducing the effective coverage of some interventions.

The 2014 ESPA+ survey identified low availability of many essential medicines (e.g., amoxicillin, paracetamol, zinc, and artemisinin-based combination therapy) for iCCM at health posts. Availability of these medicines for facility-based IMCI was higher, though zinc and vitamin A capsules were available at fewer than half of facilities (excluding health posts).

As of 2014, only 22 percent of facilities carried magnesium sulfate, a treatment for eclampsia. This was double the rate seen in 2008, but still very low.¹² In 2016, a national assessment found stockouts of magnesium sulfate across all facility levels, primarily due to inadequate supplies at regional stores and inadequate transport of supplies.¹¹

Supply chain challenges also hampered the distribution of vaccines. Most facilities providing child immunization services had individual vaccines in stock, but availability of the entire set of basic child vaccines (pentavalent, oral polio, measles, Bacillus Calmette–Guérin, and pneumococcal conjugate) was much lower.

Just under a quarter (24 percent) of health posts had all basic child vaccines at the time of the assessment, though availability was higher at health centers (58 percent).¹⁰

Climate Change

Climate change appears likely to have been a factor in temporarily increasing threats of severe acute malnutrition during the study period, and it threatens to pose future U5M-related challenges as well.

In May 2008, the Oromia and SNNP regions reported substantial increases in severe acute malnutrition due to inadequate spring rains (known as Belg). Precipitation during this time of year is crucial to the generation of adequate crop yields for Ethiopia’s subsistence farmers. The diminished rainfall and resulting high food prices affected nutrition in 193 districts (woredas), with an estimated 110,000 children in these areas affected by severe acute malnutrition.¹³

In 2010, Oxfam International produced a report on climate variability and food security in Ethiopia. It noted that while the national effects of climate change on precipitation were not yet clear, one trend stood out – rains have decreased during the Belg season (February to May). If Belg rains continue to decline, food insecurity will increase.”¹⁴

This impact was seen at the end of the study period, in 2015, when the Ethiopian government once again declared the failure of the Belg rains. After a government-led assessment the following October found that the expected harvest was below expectations due to El Niño weather patterns, the government increased the number of top-priority woredas for food assistance from 142 to 186.

At the time, 400,000 children were estimated to have severe acute malnutrition. In response to these events, the government worked with UNICEF to increase (1) malnutrition screening efforts, (2) the number of health facilities capable of treating severe acute malnutrition, and (3) the number of health workers and HEWs trained to manage the condition.

Admissions for severe acute malnutrition increased accordingly, with nearly 300,000 cases treated in 2015. UNICEF reported that the same year, 88 percent of cases were considered cured.¹⁵