Effects of family conversation on health care practices in Ethiopia: a propensity score matched analysis

Background Maternal and newborn mortality rates in Ethiopia are among the highest in sub-Saharan Africa. The majority of deaths take place during childbirth or within the following 48 h. Therefore, ensuring facility deliveries with emergency obstetric and newborn care services available and immediate postnatal follow-up are key strategies to increase survival. In early 2014, the Family Conversation was implemented in 115 rural districts in Ethiopia, covering about 17 million people. It aimed to reduce maternal and newborn mortality by promoting institutional delivery, early postnatal care and immediate newborn care practices. More than 6000 Health Extension Workers were trained to initiate home-based Family Conversations with pregnant women and key household decision-makers. These conversations included discussions on birth preparedness, postpartum and newborn care needs to engage key household stakeholders in supporting women during their pregnancy, labor and postpartum periods. This paper examines the effects of the Family Conversation strategy on maternal and neonatal care practices. Methods We used cross-sectional data from a representative sample of 4684 women with children aged 0–11 months from 115 districts collected between December 2014 and January 2015. We compared intrapartum and newborn care practices related to the most recent childbirth, between those who reported having participated in a Family Conversation during pregnancy, and those who had not. Propensity score matched analysis was used to estimate average treatment effects of the Family Conversation strategy on intrapartum and newborn care practices, including institutional delivery, early postnatal and immediate breastfeeding. Results About 17% of the respondents reported having had a Family Conversation during their last pregnancy. Average treatment effects of 7, 12, 9 and 16 percentage-points respectively were found for institutional deliveries, early postnatal care, clean cord care and thermal care of the newborn ( p < 0.05). Conclusion We found evidence that Family Conversation, and specifically the involvement of household members who were major decision-makers, was associated with better intrapartum and newborn care practices. This study adds to the evidence base that involving husbands and mothers-in-law, as well as pregnant women, in behavior change communication interventions could be critical for improving maternal and newborn care and therewith lowering mortality rates.