Methods of induction of labour: a network meta-analysis

RATIONALE: Induction of labour (IoL) aims to initiate labour when the risks of continuing pregnancy outweigh the benefits. Over 10 methods are currently available, yet the most effective and safest method remains unclear. OBJECTIVES: To compare the benefits and harms of various cervical ripening and IoL methods at or beyond term labour and to rank the methods. SEARCH METHODS: We searched Cochrane Pregnancy and Childbirth's Trials Register, ClinicalTrials.gov, and the WHO ICTRP until 1 February 2023. ELIGIBILITY CRITERIA: We included randomised controlled trials (RCTs) evaluating IoL methods among women with a live fetus at or beyond term labour (gestational age ≥ 37 weeks). We focused on the methods currently recommended by international guidelines and those proposed by previous Cochrane reviews. OUTCOMES: Our critical outcomes included failure to achieve vaginal delivery within 24 hours, caesarean section due to any causes, caesarean section due to non-reassuring fetal status, uterine hyperstimulation with changes in the heartbeat of the baby before birth, perinatal death, and severe neonatal morbidity. RISK OF BIAS: We assessed bias and trustworthiness using the Cochrane RoB 1 tool and the Cochrane Pregnancy and Childbirth Trustworthiness Tool (CPS-TST), respectively. The final analyses only included trials that met the CPS-TST. SYNTHESIS METHODS: We performed pair-wise meta-analyses using a random-effects model and network meta-analyses (NMA) with a frequentist approach. We reported analyses and results separately for two populations: 1) women without a previous caesarean section and a mix of women with or without a previous caesarean section (where more than 50% of participants had no previous caesarean section), and 2) women with a previous caesarean section. We used risk ratio (RR) and mean difference (MD) to present treatment effects with 95% confidence intervals (CIs). We assessed the certainty of evidence for critical outcomes using the GRADE approach for NMA. We used the surface under the cumulative ranking curve (SUCRA) to estimate treatment ranking. INCLUDED STUDIES: The analysis included 106 RCTs assessing 13 IoL methods among 30,348 women. Most trials (84.9%) were conducted in inpatient settings. Only two RCTs recruited women with previous caesarean section, and seven RCTs recruited a mix of women with or without a previous caesarean section. SYNTHESIS OF RESULTS: We evaluated the effects of the following IoL methods: vaginal misoprostol (≤ 50 μg), oral misoprostol (≤ 50 μg), sublingual or buccal misoprostol (≤ 50 μg), controlled-release misoprostol vaginal pessary, vaginal dinoprostone (tablet or gel), controlled-release dinoprostone vaginal pessary, oxytocin (alone), nitric oxide donors, balloon catheters, osmotic cervical dilators, oxytocin plus amniotomy (Oxytocin+Amniotomy), balloon catheters plus oxytocin (Balloon+Oxytocin), balloon catheters plus vaginal/oral misoprostol (≤ 50 μg) (Balloon+Misoprostol ≤ 50 μg), and inactive methods (placebo, no intervention, and expectant management). The following results show the relative effects of NMA for four of the six critical outcomes among women without previous caesarean section and a mix of women with or without previous caesarean section. The findings of women with previous caesarean section are not presented here due to limited evidence. We used vaginal misoprostol (≤ 50 μg) as the reference method for presenting results, where data for this method were available in the network. 1) Failure to achieve vaginal delivery within 24 hours There was no clear evidence that any of the methods were more effective than vaginal misoprostol (≤ 50 μg) in reducing the risk of this outcome. Oxytocin+Amniotomy (RR 0.41, 95% CI 0.14 to 1.24, 22.4% fewer, moderate-certainty evidence), Balloon+Misoprostol ≤ 50 μg (RR 0.85, 95% CI 0.60 to 1.19, 5.7% fewer, high-certainty evidence), and Balloon+Oxytocin (RR 0.94, 95% CI 0.71 to 1.26, 2.3% fewer, low-certainty evidence) likely resulted in little to no difference in this outcome compared with vaginal misoprostol (≤ 50 μg). Despite the comparable relative treatment effects between some methods and vaginal misoprostol (≤ 50 μg), these three methods ranked highest. 2) Caesarean section due to non-reassuring fetal status There was no clear evidence that any of the methods were more effective than vaginal misoprostol (≤ 50 μg) in reducing the risk of this outcome. Balloon+Oxytocin (RR 0.77, 95% CI 0.47 to 1.25, 2.0% fewer, moderate-certainty evidence), controlled-release dinoprostone vaginal pessary (RR 0.86, 0.62 to 1.18, 1.2% fewer, low-certainty evidence), and balloon catheters (RR 0.88, 95% CI 0.70 to 1.11, 1.0% fewer, low-certainty evidence) may result in little to no difference in this outcome when compared with vaginal misoprostol (≤ 50 μg). Despite the comparable relative treatment effects between all methods and vaginal misoprostol (≤ 50 μg), these three methods ranked highest. 3) Uterine hyperstimulation with changes in the heartbeat of the baby before birth Nitric oxide donors (RR 0.05, 95% CI 0.01 to 0.47, 3.9% fewer, moderate-certainty evidence), osmotic cervical dilators (RR 0.07, 95% CI 0.01 to 0.42, 3.8% fewer, moderate-certainty evidence), balloon catheters (RR 0.38, 95% CI 0.21 to 0.69, 2.5% fewer, moderate-certainty evidence), and oral misoprostol (≤ 50 μg) (RR 0.62, 95% CI 0.39 to 0.99, 1.6% fewer, moderate-certainty evidence) probably reduce this outcome compared with vaginal misoprostol (≤ 50 μg). When compared with inactive interventions (placebo, no intervention, or expectant management), vaginal misoprostol (≤ 50 μg) (RR 3.47, 95% CI 1.16 to 10.35) probably increases the risk of this outcome. The three highest-ranked methods were nitric oxide donors, osmotic cervical dilators, and inactive methods (placebo, no intervention, and expectant management). 4) Perinatal death There was no clear evidence for this outcome, as it was rare (10 cases reported across trials). AUTHORS' CONCLUSIONS: For women without previous caesarean section and a mix of women with or without previous caesarean section, there was no clear evidence that any of the IoL methods were more effective than vaginal misoprostol (≤ 50 μg) for the outcomes of failure to achieve vaginal delivery within 24 hours, caesarean section due to non-reassuring fetal status, and perinatal death. Nitric oxide donors, osmotic cervical dilators, balloon catheters, and oral misoprostol (≤ 50 μg) probably reduce the risk of uterine hyperstimulation with changes in the heartbeat of the baby before birth. FUNDING: This review had no dedicated funding. REGISTRATION: Protocol (2023): https://doi.org/10.1002/14651858.CD015234.