Oxytocin via Uniject (a prefilled single-use injection) versus oral misoprostol for prevention of postpartum haemorrhage at the community level: a cluster-randomised controlled trial

Background: Access to injectable uterotonics for management of postpartum haemorrhage remains limited in Senegal outside health facilities, and misoprostol and oxytocin delivered via Uniject have been deemed viable alternatives in community settings. We aimed to compare the efficacy of these drugs when delivered by auxiliary midwives at maternity huts. Methods: We did an unmasked cluster-randomised controlled trial at maternity huts in three districts in Senegal. Maternity huts with auxiliary midwives located 3–21 km from the closest referral centre were randomly assigned (1:1; via a computer-generated random allocation overseen by Gynuity Health Projects) to either 600 μg oral misoprostol or 10 IU oxytocin in Uniject (intramuscular), stratified by reported previous year clinic volume (deliveries) and geographical location (inland or coastal). Maternity huts that had been included in a previous study of misoprostol for prevention of postpartum haemorrhage were excluded to prevent contamination. Pregnant women in their third trimester were screened for eligibility either during community outreach or at home-based prenatal visits. Only women delivered by the auxiliary midwives in the maternity huts were eligible for the study. Women with known allergies to prostaglandins or pregnancy complications were excluded. The primary outcome was mean change in haemoglobin concentration measured during the third trimester and after delivery. This study was registered with ClinicalTrials.gov, number NCT01713153. Findings: 28 maternity hut clusters were randomly assigned—14 to the misoprostol group and 14 to the oxytocin group. Between June 6, 2012, and Sept 21, 2013, 1820 women were recruited. 647 women in the misoprostol group and 402 in the oxytocin group received study drug and had recorded pre-delivery and post-delivery haemoglobin concentrations, and overall 1412 women delivered in the study maternity huts. The mean change in haemoglobin concentrations was 3·5 g/L (SD 16·1) in the misoprostol group and 2·7 g/L (SD 17·8) in the oxytocin group. When adjusted for cluster design, the mean difference in haemoglobin decreases between groups was not significant (0·3 g/L, 95% CI −8·26 to 8·92, p=0·71). Both drugs were well tolerated. Shivering was common in the misoprostol group, and nausea in the oxytocin group. Postpartum haemorrhage was diagnosed in one woman allocated to oxytocin, who was referred and transferred to a higher-level facility for additional care, and fully recovered. No other women were transferred. Interpretation: In terms of effects on haemoglobin concentrations, neither oxytocin nor misoprostol was significantly better than the other, and both drugs were safe and efficacious when delivered by auxiliary midwives. The programmatic limitations of oxytocin, including short shelf life outside the cold chain, mean that misoprostol could be more appropriate for community-level prophylaxis of postpartum haemorrhage. Funding: Bill & Melinda Gates Foundation