Effect of community health worker home visits on antenatal care and institutional delivery: an analysis of secondary outcomes from a cluster randomised trial in Mali

Abstract

Introduction: Though community health workers (CHWs) have improved access to antenatal care (ANC) and institutional delivery in different settings, it is unclear what package and delivery strategy maximises impact.

Methods: This study reports a secondary aim of the Proactive Community Case Management cluster randomised trial, conducted between December 2016 and April 2020 in Mali. It evaluated whether proactive home visits can improve ANC access at a population level compared with passive site-based care. 137 unique village clusters, covering the entire study area, were stratified by health catchment area and distance to the nearest primary health centre. Within each stratum, clusters were randomly assigned to intervention or control arm. CHWs in intervention clusters proactively visited all homes to provide care. In the control clusters, CHWs provided the same services at their fixed community health post to care-seeking patients. Pregnant women 15-49 years old were enrolled in a series of community-based and facility-based visits. We analysed individual-level annual survey data from baseline and 24-month and 36-month follow-up for the secondary outcomes of ANC and institutional delivery, complemented with CHW monitoring data during the trial period. We compared outcomes between: (1) the intervention and control arms, and (2) the intervention period and baseline.

Results: With 2576 and 2536 pregnancies from 66 and 65 clusters in the intervention and control arms, respectively, the estimated risk ratios for receiving any ANC was 1.05 (95% CI 1.02 to 1.07), four or more ANC visits was 1.25 (95% CI 1.08 to 1.43) and ANC initiated in the first trimester was 1.11 (95% CI 1.02 to 1.19), relative to the controls; no differences in institutional delivery were found. However, both arms achieved large improvements in institutional delivery, compared with baseline. Monitoring data show that 19% and 2% of registered pregnancies received at least eight ANC contacts in the intervention and control arms, respectively. Six clusters, three from each arm had to be dropped in the last 2 years of the trial.

Conclusions: Proactive home visits increased ANC and the number of antenatal contacts at the clinic and community levels. ANC and institutional delivery can be increased when provided without fees from professional CHWs in upgraded primary care clinics.

Trial registration number: NCT02694055.

Keywords: maternal health; public health.

Figure 1

Trial profile. Women were eligible if they were between 15 and 49 years and permanent residents with no plans to relocate out of the study area during the trial. !Med refers to the median cluster size in that arm in that year; range is the minimum and the maximum number of participants in the smallest cluster and the largest cluster, respectively. *Analytical sample for each arm is the sum of the eligible women who delivered since the 12-month and the 24-month survey, respectively. PHC, primary health centre.

Figure 2

Distribution of the maternal health indicators by trial arms. (A) Any antenatal care (ANC), institutional delivery, preventive treatment for malaria; (B) four or more antenatal visits and ANC initiated in the first trimester.