Estimation of fine-scale spatial heterogeneity in treatment effects in a large cluster-randomized trial

In Dosso, Niger (2020–2023), mass azithromycin distribution shows spatially varying child mortality benefits, with treatment effects concentrated in the southeast along a northwest–southeast gradient.

• Why it matters: Geographic variation in randomized treatment effects is hard to quantify in large cluster trials, limiting localized understanding of where benefits occur.
• What they did: In a secondary analysis of AVENIR (1390 villages; 276,452 person-years; 3635 deaths), Poisson models estimated spatially continuous mortality and average treatment effects.
• The result: Overall mortality was 13.1 and the ATE was -2.6 deaths per 1000 person-years; mortality patterns were preserved at 160 health catchment areas, and underlying mortality explained much but not all treatment-effect heterogeneity.

AI summary of the abstract · may contain errors · not medical advice · read the paper.