Ethiopia's first full DHS in nine years: what moved and what stalled

Ethiopia went nine years between full Demographic and Health Surveys. The 2024–25 EDHS Key Indicators Report, published in January 2026 from fieldwork run between August 2024 and April 2025 across 22,540 households 1, is therefore the first hard national picture since 2016 — and the first one that reflects everything that happened in between.

The headline is a split verdict. The indicators tied to reaching a facility improved, some of them dramatically. The indicators tied to sustained, routine contact did not.

What moved

Delivery care. 64% of live births in the two years before the survey were assisted by a skilled provider, and 61% took place in a health facility 1. The trend line is the striking part: skilled attendance was 12% in 2011 1. Few countries have moved that indicator that far, that fast.

Maternal survival. The pregnancy-related mortality ratio for the seven years preceding the survey is 169 deaths per 100,000 live births, down from 412 in the 2016 EDHS and 871 in the 2000 EDHS 1.

Child survival.

Under-5 mortality was 166 per 1,000 in the five years preceding the 2000 survey 1. Neonatal mortality has moved least — 49 to 25 over the same span — which is the usual pattern: the deaths hardest to prevent are the ones closest to birth.

Fertility. Total fertility is 4.0 children per woman, down from 5.5 in 2000, and splits sharply by residence: 3.2 urban, 4.5 rural 1.

What stalled

Nutrition. 40% of children under 5 are stunted, and 15% severely stunted 1. That is essentially unchanged from 2016. Nine years of falling mortality did not move the chronic-malnutrition line.

Routine immunisation went backwards. Among children aged 12–23 months 1:

Full basic coverage was 39% in 2016 — the historical high — so this is a decline, not a plateau 1. Coverage also splits hard by residence: 50% urban versus 22% rural 1.

Why the two halves point in different directions

Skilled delivery is, from the household's perspective, a single high-stakes event. Immunisation is five or six separate visits spread over a year, none of which feels urgent on the day.

The EDHS data shows exactly that shape: 72% of children reach a facility at least once for BCG, but only 40% complete the third DTwP-Hib-HepB dose and 30% finish the basic schedule 1. The system is demonstrably able to make contact. It is losing children between contacts.

That distinction matters for anyone running a clinic, because the two problems have different fixes. Getting a first contact is an access and outreach problem. Keeping a schedule is a recall problem — knowing who is due, who did not come, and who to call. See the companion note on the immunisation dropout curve.

What this note does not claim

The Key Indicators Report is a preliminary release. Final EDHS reports add regional disaggregation, confidence intervals for most indicators, and methodological detail that can shift point estimates slightly; figures here should be read as the survey's preliminary national estimates 1.

The survey also measures the population, not facilities. It says nothing about how many of these births, visits or vaccinations happened in private clinics versus public ones, and nothing about the record-keeping behind them. Nor does it establish causes: the immunisation decline overlaps a period of conflict, displacement and pandemic disruption, and this report does not separate those effects.