Research
Community health workers cut child mortality by 31% across six districts
A four-year randomised evaluation across six districts found that door-to-door screening by trained community members reduced under-five deaths by nearly a third.
6 min read
The result that surprised us was not the size of the effect. It was where it came from. We had assumed the gains would concentrate in the villages furthest from a clinic, where the marginal value of a health worker is obviously highest. They did not. The largest reductions were in villages roughly forty minutes from a facility — close enough that families believed care was available, far enough that they consistently arrived too late.
That finding has changed how we allocate. Distance to a clinic is a poor proxy for need, because it does not capture the cost of a journey that a family will attempt only once they are certain it is necessary. Certainty arrives late. A health worker in the village removes the decision entirely.
The full evaluation, including the pre-registration, the analysis code and the two secondary outcomes where we found no effect, is published in our research library. We would encourage anyone planning a similar programme to read the null results first.
Work like this is funded almost entirely by monthly gifts from individuals.
Support our work