Scaling Care, Saving Lives: Evidence from an Adaptive Trial in Uganda
Abstract:
Proof-of-concept studies suggest that community health worker (CHW) programs can significantly reduce child mortality in resource-poor settings. However, their effectiveness at scale remains largely untested. We evaluate the scaling up of a successful CHW program, implemented in Uganda by two NGOs and assessed using an adaptive trial design. After 15 months, interim results showed only modest improvements in service delivery. In response, NGO A substantially revised its program, while NGO B made only minor adjustments. The trial period was also extended to assess the impact of these changes. Over the full trial period, no significant reduction in child mortality was observed. Yet, following the release of the interim results, mortality declined significantly under Program A. Endline data show that improvements in households’ access to health services, health knowledge, and health behavior were substantially larger under Program A. These findings suggest that adaptive designs, when paired with implementers responsive to evidence, can help sustain the impact of successful pilot programs.
Organized by: Siwan Anderson
