Combination Foley catheter and prostaglandins or Foley and oxytocin for cervical ripening: a network meta-analysis

Author Department


Document Type

Article, Peer-reviewed

Publication Date




Trial and meta-analysis data show a reduction in time to delivery for Foley and prostaglandins or Foley and oxytocin versus Foley alone. However, there is scant data for comparison of the two combination methods against each other.


To determine whether Foley and prostaglandins or Foley and oxytocin decrease the time to vaginal delivery via network meta-analysis.


A network meta-analysis (PROSPERO CRD42018081948) was performed comparing Foley and prostaglandins (PGE1 or PGE2) and Foley and oxytocin for cervical ripening. Foley alone and prostaglandins (PGE) alone were used as nodes for indirect comparison. Database searches were performed from inception to March 2020 with data abstracted from published manuscripts. Eligibility criteria included randomized trials comparing Foley and oxytocin to Foley and prostaglandins (PGE1(misoprostol) or PGE2 (dinoprostone)). Trials were also included if they compared the Foley catheter or PGE to a combination of Foley and PGE or Foley and concurrent oxytocin. Nulliparous and multiparous women were analyzed together. Foley catheters of any catheter material or size, gestational age greater than 24 weeks with a live fetus were included. Quasi-randomized, cohorts and other combination methods for cervical ripening were not included. PGE1 and PGE2 combined methods were analyzed separately in a planned subanalysis. The primary outcome was the mean time from induction to vaginal delivery in hours. Secondary outcomes included time from induction to delivery, delivery within 24 hours, cesarean delivery, chorioamnionitis, endometritis, epidural use, tachysystole, postpartum hemorrhage, meconium, NICU admission, and 5 minute APGAR <7. Data was analyzed as a network meta-analysis (NMA) using multivariate meta-regression.


30 randomized controlled trials with a total of 6,465 women were considered eligible for inclusion in this network meta-analysis. When compared to Foley alone, the use of Foley/oxytocin reduced the time to vaginal delivery by 4.2 hours (MD -4.2 hours [95% CI -6.5 to -1.9]). Foley/PGE reduced time to vaginal delivery compared to Foley but did not meet statistical significance (MD -2.9 hours [95% CI -5.7 to 0.0], p=0.05). When compared head to head, there was no difference in the time to vaginal delivery between Foley/PGE and Foley/oxytocin (MD 1.3 hours [95% CI -2.0 to 4.7]). There was no difference in the rate of cesarean delivery, chorioamnionitis , epidural, tachysystole, postpartum hemorrhage, meconium, NICU admissions, or 5min APGAR<7 for Foley/PGE vs. Foley/oxytocin, although the rate of endometritis was high for Foley/PGE. In the sub-analysis by PGE type, there was no difference in time to vaginal delivery for Foley/PGE1 vs Foley/PGE2 vs Foley/oxytocin. However, Foley/PGE2 had a definite trend towards longer time to all deliveries compared to both Foley/PGE1 and Foley/oxytocin (p=0.05).


Time to vaginal delivery was similar when comparing Foley with combined misoprostol, combined dinoprostone, and combined oxytocin. Dinoprostone comparisons are limited by small sample size, but suggest longer time to delivery compared with Foley and misoprostol or oxytocin. No significant differences were seen in maternal or neonatal adverse events with the exception of endometritis, but this is limited by the sample size, varied reporting of studies used in the indirect comparisons, and definitions of infectious morbidity use in the studies.