The Effectiveness of Oxytocin for Preventing Postpartum Haemorrhage: An Individual Participant Data Meta-Analysis.

Abstract

Post-partum haemorrhage (PPH) is a common complication of labour. To assess the effectiveness of oxytocin in comparison to no treatment for preventing PPH. Published and unpublished randomised controlled trials (RCTs) comparing systemic oxytocin to placebo or no intervention for preventing PPH were included. We did not apply language restrictions. We identified RCTs from the Cochrane network meta-analysis on uterotonics for preventing PPH and updated the search via: Ovid MEDLINE, Embase via Ovid, Web of Science, CENTRAL, CINAHL Plus and clinicaltrials.gov. An Individual participant data (IPD) meta-analysis. Of 14 eligible RCTs, four provided IPD (n = 4304; 51.7% received oxytocin and 48.4% received placebo or no intervention). Meta-analysis of IPD showed that oxytocin decreased the risk of PPH ≥ 500 mL (aOR 0.59; 95% CI 0.46 to 0.74) and PPH ≥ 1000 mL (aOR 0.51; 95% CI 0.32 to 0.80). Of 10 RCTs that did not share data, seven met trustworthiness criteria while three did not. Trustworthy IPD and aggregate data (AD) from RCTs meeting trustworthiness criteria (n = 6003) showed that oxytocin significantly reduced the rate of PPH ≥ 500 mL (aOR 0.53; 95% CI 0.45 to 0.62) and PPH ≥ 1000 mL (aOR 0.59; 95% CI 0.48 to 0.71). RCTs not meeting trustworthiness criteria reported a larger risk reduction of oxytocin for PPH ≥ 500 mL (n = 1027; aOR 0.37; 95% CI 0.03 to 4.03) and PPH ≥ 1000 mL (n = 1157; aOR 0.13; 95% CI 0.01 to 1.45). Prophylactic oxytocin reduces the risk of PPH ≥ 500 mL and PPH ≥ 1000 mL compared to no treatment. Twenty-one percent of RCTs did not meet our pre-defined trustworthiness criteria, underlining the importance of integrity assessment in evidence synthesis.