In three towns in Ethiopia's Somali region, 30% of children had had no vaccine at all

The mental map of who misses out on childhood vaccines is usually rural: a remote village, a long walk, no cold chain. That map is increasingly wrong, and a new mixed-methods study from Ethiopia's Somali region is a sharp illustration of why. In three urban towns, it found that 30% of the children surveyed had received no vaccines at all — the group global health now calls "zero-dose" [s1].

What the study measured

The researchers collected data from 842 households across three urban towns, combining community surveys with six key-informant interviews [s1]. Of the 842 households targeted, 827 participated, a 98.2% response rate [s1] — high enough that the finding is unlikely to be an artefact of who agreed to be counted.

The primary outcome was defined narrowly and meaningfully: the proportion of zero-dose children, meaning those who had not received the first dose of the DTP-containing vaccine (Penta-1) during their first year of life, based on vaccination-card records or caregiver recall [s1]. That definition matters, because Penta-1 is the standard tripwire for zero-dose status worldwide — a child who misses it has usually touched the immunisation system not at all, rather than partially.

Why urban is the surprising part

Ethiopia has a high zero-dose burden, and the Somali region is a recognised hotspot, but urban areas within it remain understudied [s1]. The study's framing is that urbanisation has widened disparities, particularly among children in informal settlements [s1] — and its own numbers bear that out, with zero-dose children concentrated in urban peripheries [s1].

This is the counter-intuitive core of the result. Cities are where the clinics, cold chains and health workers are densest, and urban children are routinely assumed to be the reached ones. A 30% zero-dose rate in three towns [s1] says that physical proximity to services is not the binding constraint. Living near a clinic in an informal urban settlement is not the same as being connected to it.

What predicted a missed child

The study's regression analysis points away from geography and toward the encounters a mother does or does not have before and after birth. Children were significantly more likely to be zero-dose if their mothers had received no antenatal care (adjusted odds ratio around 5.0), if vaccination required someone's permission (around 3.4), if the birth happened at home (around 3.1), or if the mother had not received the second dose of tetanus toxoid (around 2.7) [s1].

Read as a set, these are not four separate risk factors so much as one story told four ways. Antenatal care, a facility birth and maternal tetanus vaccination are all points where a mother enters the health system during pregnancy — and each missed contact predicts a child who then never enters it either. The system tends to reach mothers and their infants together, or miss them together.

The permission finding is the one that names a social barrier rather than a service gap. That a child was more likely to be unvaccinated when vaccination required permission [s1] locates part of the problem in household autonomy — who decides whether a child is taken to be immunised — which no amount of clinic density addresses on its own.

The protective factors mirror the risks. Higher perceived household wealth and at least one postnatal care visit reduced the likelihood of a child being zero-dose by about 45% and 55% respectively [s1]. Postnatal care shows up on both sides of the ledger: its presence roughly halves the odds, its absence compounds them.

The global frame

The Somali-region finding lands inside a global picture that has stopped improving. Globally in 2025, there were 13.5 million children missing out on any vaccination — the zero-dose children — while coverage of a third dose of the diphtheria, tetanus and pertussis vaccine held at 85% and first-dose measles coverage stood at 84% [s2]. WHO's own summary is that coverage has held steady since 2024 while progress toward the Immunization Agenda 2030 targets stalls, and that global figures hide significant disparity among countries of different income strata [s2].

A stalled global average and a 30% zero-dose pocket inside a single country's cities are two views of the same problem. The children left behind are not distributed evenly; they concentrate in specific places and specific households, and the averages that look merely flat at the top are hiding gaps like this one underneath.

What the study implies for action

The authors analysed their qualitative findings using the Gavi IRMMA framework — identify, reach, monitor, measure, advocate — for zero-dose children [s1], which signals what kind of response the data call for. The predictors it found are not about building more clinics in the three towns; the clinics are already there. They are about connecting mothers to antenatal and postnatal care, addressing the permission barrier, and treating a facility birth as an immunisation opportunity rather than a separate event.

What to watch

The operational test is whether zero-dose programmes in urban hotspots start being designed around the antenatal and postnatal contact points this study flags, rather than around distance to a clinic. The measurable signal would be a fall in the urban-periphery concentration the study documented [s1] — the clearest sign that the reached and the unreached are no longer being decided by which informal settlement a child is born into.

Sources

• Prevalence of zero-dose vaccination and its associated factors in urban areas of Somali region, Ethiopia: a mixed-methods approach, BMJ Public Health, 15 July 2026
• Immunization coverage (fact sheet), World Health Organization