Community health workers

What the evidence says

A 2010 Cochrane review of 82 studies of lay health workers, 55 of them in high-income countries, found moderate-quality evidence that they increase childhood immunization (risk ratio 1.22) and breastfeeding. It found low-quality evidence that they may reduce child deaths: the estimate was a 25 percent reduction, but its confidence interval included no effect.

A randomized trial in Uganda found a large effect on child deaths. Locally recruited health promoters received incentives to visit homes, teach health behaviors, give medical advice and referrals, and sell preventive and curative products. After three years, under-five mortality was 27 percent lower in program villages, at an estimated $68 per life-year saved.

A later trial in Uganda, presented at a 2023 research seminar, tested two programs at scale. The trial design let researchers review results midway and change course. One program was redesigned and the trial extended by two years. Among children fully exposed to the revised version, under-five mortality fell by 29 percent. The other showed no significant difference in child mortality. The idea alone doesn’t guarantee results: design and management vary, and so do outcomes.

WHO’s 2018 guideline on community health worker programs covers how to educate, deploy and manage them and how to fit them into the wider health system.

The grade is moderate. Several trials show gains in vaccination and breastfeeding, and the evidence on deaths is promising but less consistent. The $68 figure comes from one trial of one model, and these programs vary too much for one number to stand for the whole category, so this entry has no Impact Calculator figure.

Where it sits on the depth scale

Most of this work is prevention (depth 2): vaccination, breastfeeding and products that keep children from getting sick. Some of it is relief, such as caring for or referring a child who is already ill. When a government builds community health workers into its health system and budget, the work moves toward structure.

Open questions

• Who should pay community health workers, and what happens to a program when its grant ends? See the government question and the Short Horizon.
• Most studies in the Cochrane review came from high-income countries. How well do their results carry over to poorer ones?
• These workers come from the communities they serve. How much say do those communities have in what the programs do? See the Missing Voice.