Abstract

Background Participatory learning and action (PLA) programmes are promoted as effective community-based intervention to improve maternal and child health. Evidence on their impact on knowledge and behaviour related to pregnancy and childcare is limited and mixed. Between 2015 and 2016, state-supported agencies implemented a large-scale PLA intervention in women's self-help groups in rural Bihar, India, with meetings facilitated by trained agency staff. We assessed the impact of this intervention using a cluster-randomized controlled trial. Methods In one implementation district, 68 village clusters were randomly assigned to receive the intervention or not. With survey data from 1612 women who were pregnant before the intervention, we evaluate programme impact on (a) beliefs about practices of feeding newborns, (b) knowledge about malaria, (c) childcare practices, (d) quantity and quality of antenatal care, and (e) attitude towards antenatal care and pregnancy. We estimate intention-to-treat effects in linear regression models using endline data and in a difference-in-differences model. Findings With no robust statistical significance across 50 outcomes, we find no evidence that the PLA intervention in Bihar had an impact on beliefs, attitudes, knowledge, behaviour related to pregnancy or childcare, or quality of antenatal care received. Less than one percent of the sample reported to have attended at least one PLA meeting. Conclusion Our results do not discredit the effectiveness of PLA interventions in general. They highlight that PLA interventions delivered through existing women's self-help groups need to consider the particular challenges to participation faced by pregnant women.