Home visits versus fixed-site care by community health workers and child survival: a cluster-randomized trial, Mali.

Objective: To test the effect of proactive home visits by trained community health workers (CHWs) on child survival. Methods: We conducted a two arm, parallel, unmasked cluster-randomized trial in 137 village-clusters in rural Mali. From February 2017 to January 2020, 31 761 children enrolled at the trial start or at birth. Village-clusters received either primary care services by CHWs providing regular home visits (intervention) or by CHWs providing care at a fixed site (control). In both arms, user fees were removed and primary health centres received staffing and infrastructure improvements before trial start. Using lifetime birth histories from women aged 15-49 years surveyed annually, we estimated incidence rate ratios (IRR) for intention-to-treat and per-protocol effects on under-five mortality using Poisson regression models. Findings: Over three years, we observed 52 970 person-years (27 332 in intervention arm; 25 638 in control arm). During the trial, 909 children in the intervention arm and 827 children in the control arm died. The under-five mortality rate declined from 142.8 (95% CI: 133.3-152.9) to 56.7 (95% CI: 48.5-66.4) deaths per 1000 live births in the intervention arm; and from 154.3 (95% CI: 144.3-164.9) to 54.9 (95% CI: 45.2-64.5) deaths per 1000 live births in the control arm. Intention-to-treat (IRR: 1.02; 95% CI: 0.88-1.19) and per-protocol estimates (IRR: 1.01; 95% CI: 0.87-1.18) showed no difference between study arms. Conclusion: Though proactive home visits did not reduce under-five mortality, system-strengthening measures may have contributed to the decline in under-five mortality in both arms.