Impact of the contraceptive implant on maternal and neonatal morbidity and mortality in rural Papua New Guinea: a retrospective observational cohort study

Abstract

ObjectivesUsing routinely collected birth data, this study sought to measure changes in maternal and neonatal morbidity and mortality after introduction of the levonorgestrel contraceptive implant into a large, rural island population in Papua New Guinea (PNG). Study designWe conducted a retrospective observational study of birth records from 4251 births that occurred between January 2010 and December 2016 on Karkar Island, PNG. The primary outcome was the change in crude birth rate (CBR) before (2010–2012) and after (2014–2016) introduction of the implant. Secondary outcomes were the change in rates (per year/1000 births) of severe postpartum hemorrhage, postpartum infection, hospital readmission, prematurity (<37 weeks), low birth weight (<2500 g) and maternal and neonatal mortality. We also studied changes in the number of pregnancies affected by grand multiparity (≥4) and short interpregnancy interval (<12 months) for the same time periods. Data were analyzed using interrupted time series and Poisson regression. ResultsCBR was stable until 2012 and then declined from 2014 (p<.0001). Following introduction of the implant, the annual rate/1000 births of selected adverse birth outcomes decreased between 56% and 74% (p<.0001). The number of women with parity ≥4 who gave birth decreased by 59% (p<.0001), and the number with interpregnancy interval <12 months decreased by 64% (p<.0001). ConclusionsIntroduction of the contraceptive implant was associated with reductions in CBR, maternal and neonatal morbidity, and the number of women with high-risk pregnancies giving birth. ImplicationsThese results encourage efforts to increase knowledge and availability of the contraceptive implant in low- and middle-income countries such as PNG. In cases where it reduces the CBR and the number of women with high-risk pregnancies birthing, the implant may have a beneficial impact on maternal and neonatal morbidity.