Induction of labor at term with vaginal misoprostol or a prostaglandin E2 pessary: a noninferiority randomized controlled trial.

Background: Induction of labor is among the most common procedures for pregnant women. Only a few randomized clinical trials with relatively small samples have compared misoprostol with dinoprostone. Although their efficacy seems similar, their safety profiles have not been adequately evaluated, and economic data are sparse.

Objective: This study aimed to test the noninferiority of vaginal misoprostol (prostaglandin E1) (25 μg) to a slow-release dinoprostone (prostaglandin E2) pessary (10 μg) for induction of labor with an unfavorable cervix at term.

Study Design: This was an open-label multicenter randomized noninferiority trial at 4 university hospitals of the Research Group in Obstetrics and Gynecology between 2012 and 2015. We recruited women who underwent induction of labor for medical reasons, those with a Bishop score of ≤5 at ≥36 weeks' gestation, and those with a cephalic-presenting singleton pregnancy with no previous cesarean delivery. Women were randomly allocated to receive either vaginal misoprostol at 4-hour intervals (25 μg) or a 10-mg slow-release dinoprostone pessary. The primary outcome was the total cesarean delivery rate. Noninferiority was defined as a difference in the cesarean delivery rates between the groups of no more than 5%. Secondary outcomes included neonatal and maternal morbidity, vaginal delivery at <24 hours after starting the induction of labor process, and maternal satisfaction.

Results: The study included 1674 randomized women. The per-protocol analysis included 790 women in each group. The total cesarean delivery rates were 22.1% (n=175) in the misoprostol group and 19.9% (n=157) in the dinoprostone group, a difference of 2.2% (with an upper-bound 95% confidence limit of 5.6%) (P=.092). Results in the intention-to-treat analysis were similar. Neonatal and maternal morbidity rates were similar between groups. Vaginal delivery within 24 hours was significantly higher in the misoprostol group (59.3% vs 45.7%; P<.001) as was maternal satisfaction, assessed in the postpartum period by a visual analog scale (mean score, 7.1±2.4 vs 5.8±3.1; P<.001).

Conclusion: The noninferiority of a 25-μg dose of vaginal misoprostol every 4 hours to the dinoprostone pessary for cesarean delivery rates after induction of labor at term could not be demonstrated, although the confidence limit of the difference barely exceeded the noninferiority margin. Nonetheless, given the small difference between these cesarean delivery rates and the similarity of neonatal and maternal morbidity rates in this large study, the clinical risk-to-benefit ratio justifies the use of both drugs.