Detailed findings

ACCEPTABILITY

Lower care seeking (9%) from HEWs compared to higher-level public facilities (16%) in 2013² 

FEASIBILITY

By 2014,

-- Nearly 30,000 HEWs in 8 regions participated in iCCM

-- iCCM implemented in 86% of districts nationally³

High availability of curative child services at health posts (96%) per 2014 ESPA+

FIDELITY

 A 2012 study found that HEWs in areas implementing iCCM correctly classified 53% of children for all major iCCM illnesses. Similar gaps in quality were reported by the 2014 SPA.

EFFECTIVENESS/REACH

Care-seeking behavior impacts utilization and coverage of iCCM services

ACCEPTABILITY

Ongoing community perception of diarrhea as disease not requiring care at a health facility

FEASIBILITY

Low availability of ORS and zinc at health facilities, although some increase by the end of the study period and high rates of ORS

High availability of FB-IMCI and iCCM services at facility levels

Variable rates of recent provider training (possibly underestimated due to pre-service training)

Challenges with maintaining stocks

FIDELITY

Low rates of providers receiving in-service training on diarrhea diagnosis and treatment in 2014 – 43% in health posts and 7% in all other facility types

Low provider adherence to national guidelines and preference to prescribe antibiotics in facilities, varying use of zinc

EFFECTIVENESS/REACH

2014 – ORS given to only 29% of children with diarrhea without dehydration and 66% of those with dehydration treated at health facilities excluding health posts and zinc given to only 16% and 8%, respectively

Higher coverage of ORS at health posts (100%), but no administration of zinc

Per key informant, growing proportion of diarrhea cases treated in private facilities but tendency to prescribe antibiotics rather than ORS and zinc at these facilities, requiring future efforts to address treatment in the private sector

Remaining inequities in coverage – per KI, possibly due to higher health literacy and access in urban areas, as well as cultural practices in rural areas:

“We come to the cultural or traditional practices, the more rural I would say people may stick to the tradition and avoiding that meal or this meal and giving that tradition-based intervention…”

ACCEPTABILITY

No evidence of rejection by the community

FEASIBILITY

Available in all regions as planned and stocked

Outreach activities conducted

Sustainment or increase in rates with UNICEF/WHO estimated coverage increased to 80% in 2016 and 2017⁴

Integrated into national EPI program

FIDELITY

EFFECTIVENESS/REACH

UNICEF/WHO coverage estimates of infants receiving the 2nd dose of rotavirus vaccine:

• 2014 – 56%
• 2015 – 79%⁴

The 2016 DHS reported the following. This shows some remaining inequities in coverage

• Overall coverage of 64% for 1 dose and 56% for 2 doses of rotavirus vaccine
• Regional variation in coverage – from 23% in Afar to 92% in Addis Ababa for 2 doses
• Higher coverage in urban areas compared to rural ones (79% and 53%, respectively, for 2 doses)

18% reduction in proportion of diarrhea hospitalizations due to rotavirus in children under 1 year, decline in seasonal peaks of disease⁵

IHME DECOMPOSITION RESULTS

21.4% of the reduction in under-five mortality attributed to vaccines including rotavirus, PCV, Hib, Measles, DTP3, and Meningitis.¹

ACCEPTABILITY

High levels (nearly 100%) of general acceptability of vaccines in the community, according to interviewees in a 2015 study

FEASIBILITY

PCV successfully introduced and no stockouts noted

Integrated into national EPI program and strengthening of surveillance planned

FIDELITY

EFFECTIVENESS/REACH

Increases in coverage for vaccination in all geographic districts and both urban and rural areas but remained below targets

Challenges with achieving coverage goals

WHO/UNICEF estimates of coverage of PCV3 increasing from 12% in 2011 to 76% in 2015⁶

The 2016 DHS reported:

• PCV3 coverage varied by region, ranging from 18% in Afar to 91% in Addis Ababa
• Urban areas (73%) had higher PCV3 coverage than rural ones (46%)
• Variable coverage by wealth quintile, ranging from 36% in the lowest quintile to 71% in the highest

ACCEPTABILITY

Delayed introduction following global shortage

Introduction in all regions achieving coverage shown in Effectiveness and Coverage (Reach)

 No evidence of stockouts seen and ongoing commitment to continued vaccination

FIDELITY

2012 cluster survey indicated 26% dropout between Penta 1 and Penta 3⁷

EFFECTIVENESS/REACH

High Penta 3 coverage in 2016 per administrative estimates (96%), though lower coverage estimated by WHO/UNICEF (73%) and DHS (53%) with significant gaps

Variability of coverage at regional level – from 20% in Afar to 96% in Addis Ababa per 2016 DHS

Lower coverage in rural and lower wealth populations in 2016

ACCEPTABILITY

Remaining belief in communities that measles is not severe per key informant

Required separate visit at 9 months of age

FEASIBILITY

Measles vaccines given both as routine and through catch-up as well as targeted to meet needs of refugees

Large numbers vaccinated

Measles vaccination is an integral part of the national U5M reduction strategy and in the HMIS and EPI schedule

FIDELITY

EFFECTIVENESS/REACH

2015 coverage: 65% per WHO/UNICEF estimates, 92% per FMOH

See equity

Increased in number of measles cases over study period, believed to be due partly to increased sensitivity of surveillance system

Decrease in U5 proportion of measles cases from 56% in 2008 to 30% in 2014⁸

Acceptability issues limit coverage

Per 2016 DHS, gaps in coverage between wealth quintiles (43% in lowest quintile to 74% in highest)

Large regional variation in coverage per 2016 DHS – from 30% in Afar to 93% in Addis Ababa

ACCEPTABILITY

High vaccine acceptance per FMOH

FEASIBILITY

All 3 phases completed as planned as a single dose campaign

FIDELITY

EFFECTIVENESS/REACH

 High coverage in target areas per FMOH post-campaign coverage surveys:⁹

• Phase 1: 98.4%
• Phase 2: 97.6%
• Phase 3: 93%

 High coverage across the country, with limited geographic variation (from 88% in parts of Afar to 97% in parts of Tigray)

ACCEPTABILITY

Per National Malaria Indicator Surveys (2007, 2015) only 2/3 of nets were used the night before the survey

FEASIBILITY

Widespread distribution was feasible

Delivery achieved through using multiple approaches including outreach to ensure coverage

 Reliance on external funding and procurement of nets limits sustainability

FIDELITY

Targeting for higher risk areas

EFFECTIVENESS/REACH

 Per MIS: Household ownership of ITNs reached 66% in areas <2000m in 2007 and remained steady (64%) through 2015

MIS: In 2015, almost half (49%) of the population had access to an LLIN. LLIN access varied by region – from 24% in Harari to 59% in Tigray.

MIS: decline in all-age parasitemia (by RDT) from 4.5% in 2011 to 1.2% in 2015, under-5 parasitemia of 1.4% in 2015

Access to an LLIN in the household in 2015 varied by region, from 24% in Harari to 59% in Tigray

 Access to an LLIN in 2015 varied moderately by wealth quintile:

Lowest – 35%

Highest – 50%

ACCEPTABILITY

No refusal data found

FEASIBILITY

IRS was able to be implemented (although see reach)

HEWs able to support expansion of community-based spraying to targeted communities

FIDELITY

In 2013, compliance with standard spraying procedures was 81% in districts using community-based IRS and 92% in those using district-based IRS¹⁰ 

2013 spray quality assessment in Oromia found mortality of mosquitoes exposed to sprayed walls to be 99.5% in district-based IRS districts and 99.9% in community-based IRS districts¹¹

EFFECTIVENESS/REACH

 2001-2005 target for IRS (60%) not met (30%) due to lack of funds

Coverage in targeted regions was higher – 92% of targeted structures were sprayed in 2015/16

significant regional variability

ACCEPTABILITY

Increased uptake of ART and adherence reported by 2009 MSG program evaluation suggested high acceptability

High acceptance of provider-initiated testing – 92% of counseled mothers receiving ANC at facilities providing PMTCT services accepted testing in 2010

FEASIBILITY

PMTCT implemented although challenges with reach

ART coverage remained <90% and even saw drop in ART coverage for pregnant women to 62% in 2015-2016¹²

Significant reliance on donor funding

FIDELITY

 52% of facilities assessed in 2014 ESPA+ reported having at least one staff member recently trained on PMTCT

The HIV/AIDS Strategic Plan for 2015-2020 identified several existing gaps in the PMTCT program, including inadequate integration of option B+ into the MNCH platform, weak monitoring and evaluation system for PMTCT, poor access to and utilization of facilities for early infant diagnosis, and low skilled birth attendance coverage

EFFECTIVENESS/REACH

2016 DHS reported 34% of pregnant women were tested for HIV during ANC or labor and received results

70% ARV coverage for PMTCT in 2014

New pediatric cases averted due to PMTCT

WHO estimated that the final MTCT rate (including the breastfeeding period) in Ethiopia was 16% in 2016¹³

 Regional variation in testing for HIV during ANC or labor in 2016 DHS – from 6% in Somali to 78% in Addis Ababa

Regional differences in MTCT, from 4.2% in SNNPR to 15.7% in Dire Dawa¹⁴

IHME DECOMPOSITION RESULTS¹

<1% pf the reduction in under-five mortality attributed to PMTCT and ART

ACCEPTABILITY

Per key informant, coverage may be impacted due to stigma around HIV and household prioritization of care for adults rather than children.

FEASIBILITY

Expansion of HIV testing and ART facilities including infants and children

FIDELITY

Below-target coverage of testing (timing and testing) of HIV-exposed infants

EFFECTIVENESS/REACH

2012 survey of five facilities in Amhara found that only 41% of HIV-exposed infants had blood taken for testing at or prior to 6 weeks of age.

31% estimated coverage of early infant diagnosis in 2015

Improved coverage of children under 15 receiving ART – 1% in 2005/6 to 25% in 2014/15, but still major gaps

Substantial decrease in related mortality in children between 2002 and 2016

ACCEPTABILITY

Data not found

FEASIBILITY

National scale-up of SAM management through national scale of FB-IMCI and iCCM

FIDELITY

 2011 study found that a large portion of children admitted to health post treatment programs with SAM were discharged while still classified as having SA M¹⁵

EFFECTIVENESS/REACH

High availability of SAM diagnosis and treatment at public health facilities (84%) in 2016

SAM diagnosis and treatment widely available at the community level – offered in 82% of health posts

Variable recovery rates reported

IHME DECOMPOSITION RESULTS

32% of the reduction in under-five mortality attributed to reductions in child growth failure and reductions in vitamin A and zinc deficiency.¹

ACCEPTABILITY

Data not found

FEASIBILITY

Scale-up of Enhanced Outreach Strategy

High coverage across regions as seen below

FIDELITY

EFFECTIVENESS/REACH

94% national coverage in 2008, ranging from 82% in Dire Dawa to 99% in Tigray

DHS: coverage of 53% in 2011 and 45% in 2016

Coverage varying by region in 2016 DHS (from 35% in Afar to 74% in Tigray)

ACCEPTABILITY

Although data on acceptability was not found, the uptake of HEW-delivered ANC was limited despite increased access to these services at the community level, through inclusion of ANC in the scope of HEWs

FEASIBILITY

87% of facilities assessed in 2014 ESPA+ reported providing ANC services

FIDELITY

In 2014, 54% of facilities providing ANC reported having at least one staff member who received in-service training for ANC care

Low availability of guidelines at facilities (2014 SPA)

EFFECTIVENESS/REACH

Per DHS, utilization of any ANC visits increased from 27% in 2000 to 62% in 2016.

Low remaining utilization rates

Persistent low ANC4+ rates with only 32% reach by 2016

Improvement in ANC1+ rate with 62% reach by 2016

According the 2000 DHS and 2016 DHS, women in rural areas were less likely to receive ANC from skilled provider (22% and 58%) in comparison to those in urban areas (68% and 90%) respectively

Regional variation from 51% in Afar to 97% in Addis Ababa (DHS 2016)

ANC by a skilled provider remained at 76% in urban areas versus 26% in rural areas.

IHME DECOMPOSITION RESULTS

2% of the reduction in U5M attributed to increased use of skilled birth attendants. An additional 2% of the reduction in U5M is attributed to reductions in low birth weight and short gestational age

ACCEPTABILITY

Data not found

FEASIBILITY

Intervention implemented although challenges with availability

FIDELITY

Low availability of magnesium at facilities with frequent stock outs

• Only 22% of facilities carried magnesium sulfate in 2014
• In 2016, 53% of facilities reported experiencing a stock out

High availability (92%) of antihypertensives at health facilities in 2016

2008 the national EmONC assessment found that anti-hypertensives were available at 75% of health centers and 99% and

in 2016 EmONC assessment showed better availability of anti-hypertensives, with drugs available at 92% of hospitals visited.

in 2016 the national EmONC assessment found stock-outs across all facility levels, primarily due to stock out at central or regional stores or inadequate transport of supplies.¹⁶

EFFECTIVENESS/REACH

High facility readiness (72%) but low provision (26% of facilities) of parenteral anticonvulsants with fairly low adherence to national guidelines - magnesium sulfate only used at 55% of hospitals that administered parenteral anticonvulsants in the 3 months preceding national EmONC survey in 2016

Provision of anti-hypertensives documented in less than half (48%) of pre-eclampsia and eclampsia cases assessed in 2016

ACCEPTABILITY

High coverage suggested high acceptability- see effectiveness and coverage

FEASIBILITY

Immunization of 15 million women of childbearing age between 1999 and 2009

FIDELITY

Data not found

EFFECTIVENESS/REACH

Maternal and neonatal tetanus elimination achieved in all regions

High TT2 coverage achieved during supplemental immunization activities conducted in high-risk zones (76-94%)¹⁷

Elimination achieved and validated in 2017

The National Strategy for Newborn and Child Survival in Ethiopia (2015/16 – 2019/20) identified tetanus toxoid immunization during pregnancy as a high impact intervention for improving child survival. It set a coverage target of 90% of newborns protected against tetanus by 2020.

In 2017, all regions in the country were validated for maternal and neonatal tetanus elimination.^18,19

ACCEPTABILITY

Limited uptake of delivery services at facilities

Perceptions of institutional delivery remained negative

FEASIBILITY

By 2010, 33,819 HEWs had been trained and deployed and were estimated to reach 89% of communities

increase in number of trained midwives from 1,275 in 2008 to 9,244 in 2014

 health centers (99%) and all hospital types (98%, 93%, 88% for primary, general, and referral hospitals, respectively

 Increase in facilities providing delivery

Ambulances deployed across country

women believe that delivering at a health facility was:

- not necessary 42%,

- not customary 36%

- very costly 22%.²⁰

FIDELITY

Policies enacted to reduce financial barriers to institutional delivery were not enforced. User fees for delivery services remained in place in many facilities²¹

EFFECTIVENESS/REACH

 Some improvement in institutional delivery and assistance by a skilled provider per DHS

Per 2014 ESPA+, 99% of public facilities offered normal delivery services

26% of facilities charged a fee for normal delivery.²¹

In 2000:

1. 32% in urban areas gave birth in a health facility, compared to only 2% in rural areas
2. Those who live in Addis Ababa were most likely to deliver in a facility (67%), in Dire Dawa (31%), in Amhara (3%).

In 2016, proportions of facility-based over the previous five years vary widely by location, wealth, and urban/rural location delivery ranged from 14.7% in Afar to 97% in Addis Ababa²²

Equity gap widened notably from 2000 to 2016 (IHME)

ACCEPTABILITY

Data not found

FEASIBILITY

increased number of trained health workers

Improvement in proportion of UN-recommended standard number of EmONC facilities increased from 11% in 2008 to 40% in 2016

FIDELITY

Data not found

EFFECTIVENESS/REACH

Very low coverage of births taking place in BEmONC facilities

Estimated only 16% met need for BEmONC services in 2015

Geographic disparities in met BEmONC need

ACCEPTABILITY

In 2016, low C-section national rates achieved at 3%, compared to the FMOH target of 7% by 2015.²³

FEASIBILITY

High availability of C-section delivery in 2014 (77% in primary hospitals, 86% in general hospitals, 84% in referral hospitals)

Variable- rates remained low in many settings but higher in urban

FIDELITY

C-section national rates at 3% in 2016, (FMOH) or 1.9% (DHS) is very low due to low rates of facility-based deliveries compare to the FMOH target of 7% by 2015.²³

 Rate of 21.4% of C-sections in Addis Ababa suggest unnecessary C-section

EFFECTIVENESS/REACH

 Deployment of 136 IESOs throughout Ethiopia and 462 in training in 2015²⁴

 Increase in C-section rates in a few regions, per 2000 and 2016 DHS surveys but larger equity gap and decrease in some areas

Inequities in C-section rates between regions and urban/rural

In 2016 the equity gap has increased 10.6% of in urban versus 0.9% in rural

• 0.4% in the Somali

21.4% in Addis Ababa.

ACCEPTABILITY

Data not found

FEASIBILITY

Expansion of CBNC to 70% of health posts nationally

FIDELITY

Uptake for deliveries remained < 100%

Limited improvement was seen in all clean cord care practices, with the exception of cutting the cord with a sterile instrument

Overall, 46% took clean cord care, compared to 36% at baseline.²⁵

EFFECTIVENESS/REACH

Increase in clean delivery practices demonstrated by L10K program (46% vs 36% at baseline

Expansion to CBNC to 70% of health posts nationally

Limited progress seen in estimated mortality attributed to neonatal sepsis

ACCEPTABILITY

Data not found

FEASIBILITY

Administered by few facilities at only at higher levels of care

FIDELITY

Stock outs

EFFECTIVENESS/REACH

National EmONC assessment in 2016 found only 9% of facilities had provided ACS over the previous three months

 Availability of ACS at health facilities in 2016 varied by region – from 5% in Benishangul-Gumuz to 27% in Harari

ACCEPTABILITY

Data not found

FEASIBILITY

Low levels of training in health providers in 2014

 High availability of oxytocin

FIDELITY

Low correct usage (29%) of AMTSL observed in 2006 study

Lack of training – only 10% of providers interviewed in 2014 SPA+ received in-service training on AMTSL

EFFECTIVENESS/REACH

2006 study: low coverage of correct usage of AMSTL

ACCEPTABILITY

Data not found

FEASIBILITY

Increased knowledge of providers

70% of facilities nationally staffed and equipped for neonatal resuscitation in 2016

FIDELITY

98% of facilities in 2016 had at least one staff member capable of providing neonatal resuscitation

Though data on quality of care was not found, a key informant noted “whenever we have a supportive supervision we collected some data and that data showed us good number of babies are salvaged and more than 90% are resuscitated and saved”

EFFECTIVENESS/REACH

Higher rates of neonatal resuscitation reported by key informant, but still low across regions per 2016 assessment¹⁶

Reach of training was limited (SPA 2014)

Regional differences in facility readiness to provide resuscitation

2016 EmONC assessment: resuscitation provided in highest percentage of facilities in Benishangul-Gumuz (88%) but lowest in Afar (42%)

ACCEPTABILITY

Per KI, high willingness of facilities to prioritize establishment of NICUs and staff training due to prioritization by FMOH

FEASIBILITY

49% of NICUs functional by mid-2015

Challenge in lack of physical space at health facilities

FIDELITY

100% of target hospitals received NICU training by 2015

EFFECTIVENESS/REACH

Mid-2015, 49% of planned NICUs were operational in Ethiopia

ACCEPTABILITY

High acceptability by mothers following education per KI:

“Mothers are very willing but they have fear. The fear is they think that they can kill their babies because babies are there when they sleep and they could suffocate their babies… In our case, we have counseling methods like videos, audios, and posters in Amharic and other local languages.”

FEASIBILITY

See coverage below

FIDELITY

Only 15% of providers surveyed in ESPA+ reported receiving in-service training on KMC

2014 assessment: only 14% of eligible neonates born in a facility offering KMC were actually enrolled

EFFECTIVENESS/REACH

2014 ESPA+: 67% of facilities providing normal delivery services offered KMC

Significant regional variation in availability of KMC at health facilities in 2014 - from only 2% of facilities in Afar to 84% in Benishangul-Gumuz²⁶

ACCEPTABILITY

Data not found

FEASIBILITY

Expansion of CBNC to 70% of health posts in the country by 2015

FIDELITY

Fairly low adherence to national treatment protocols in program midline assessment

EFFECTIVENESS/REACH

In program midline assessment, 94% of infants with very severe disease were given a pre-referral dose of antibiotics

ACCEPTABILITY

Low uptake of PNC remains despite increased access through HEWs

 Cultural beliefs and traditional practices such as the traditional 40-day confinement period intended to protect the mother and newborn from malevolent spirits limits care-seeking

FEASIBILITY

Expansion of PNC service delivery through HEP

FIDELITY

 2016 DHS: only 27% of newborns born within the previous 2 years had at least 2 signal functions performed within 2 days after birth

EFFECTIVENESS/REACH

 Coverage remains low – per 2012 DHS, 86% of newborns did not receive a post-natal check at any point after delivery

Lower coverage of postnatal checks for newborns who delivered outside of a health facility (1% vs 34% of those delivered in a facility), despite service delivery by HEWs to provide home-based care

Regional differences in coverage of postnatal checks within 2 days after delivery per 2016 DHS – ranging from 9% in Oromia to 55% in Addis Ababa