Effect of integrating postpartum family planning into the health extension program in Ethiopia on postpartum adoption of modern contraception.

Background: Ethiopia has large unmet need for contraception among postpartum women. Community-level services may improve postpartum contraceptive use in Ethiopia and other contexts where home childbirth is common. This study estimated the additional effect of systematically integrating messages on postpartum family planning (PPFP) into community contacts with pregnant and postpartum women, on top of integrated facility contacts. Methods: The quasi-experimental study was conducted in two districts in Oromia Regional State and used a controlled trial design with random assignment of primary health care units-a health center and surrounding health posts-to intervention and comparison arms. We used the log-rank test and fitted a multivariate Cox proportional hazards regression model to estimate the adjusted hazard ratio (adjHR) and 95% confidence interval (95% CI) for differences in contraceptive uptake by arm. Time from delivery to uptake of modern contraception was the outcome variable. We separately analyzed subsets of women by place of delivery, as this variable interacts with the study arm. Results: A total of 772 pregnant women were enrolled February-March 2017 and re-interviewed 15 months later (May 2018). Loss to follow-up was 10% in intervention, 7% in comparison areas. Among women who delivered at home, there was higher PPFP adoption by one year postpartum in the intervention arm (35.2%, 95% CI: 28.8-42.4%) versus comparison arm (27.8%, 95% CI: 22.2-34.4%). In the adjusted Cox regression model, women who delivered at home in the intervention arm were 45% more likely to adopt contraception (adjHR1.45, CI: 1.01-2.07). There was no difference by arm for women who delivered in a facility. Conclusions: Integrating PPFP into community-level services for pregnant and postpartum women and infants may have additional benefit on top of PPFP services at facilities. The intervention benefited women who delivered at home, an important target population in countries like Ethiopia, where many women do not deliver in a facility. This study, implemented under real world conditions, informs the PPFP body of evidence and fills a gap in research on the contribution of community-based PPFP in contexts where services are integrated within maternal, newborn, and child health care in facilities.