Comparing and combining evidence of treatment effects in randomized and nonrandomized studies on the use of misoprostol to prevent postpartum hemorrhage

OBJECTIVE: Postpartum hemorrhage (PPH) is a preventable condition and the main cause of maternal death worldwide. Evidence on the effectiveness of misoprostol in the prevention of PPH has been generated from both randomized controlled trials (RCTs) and nonrandomized studies (NRS). This study aimed to compare the results of RCTs and NRS, and to compare Classical and Bayesian approaches of combining the results of RCTs and NRS on the use of misoprostol versus placebo in the prevention of PPH. METHODS: We searched MEDLINE, EMBASE and the Cochrane Central Register of Controlled Trials for appropriate studies. We pooled estimates of effects from RCTs and NRS seperately, using random-effects models, then merged them using classical and Bayesian random effects meta-analysis. RESULTS: A total of 34 studies (20 RCTs and 14 NRS) involving 74 204 participants were identified. The summary odds ratio (OR) from RCTs for the use of misoprostol in the prevention of PPH was 0.69 (95% confidence interval [CI]: 0.59 to 0.80). The summary OR from NRS was 0.46 (95% CI: 0.36 to 0.63). Classical and Bayesian approaches of combining the two study designs both showed benefit of misoprostol in preventing PPH, with similar effects. CONCLUSIONS: Both RCTs and NRS show comparable significant benefit for the use of misoprostol in the prevention of PPH.