Methods of cervical ripening in induction of labour: an individual participant data network meta-analysis of...

Induction of labour (IOL) is a commonly performed obstetric intervention, particularly when delivery is deemed more beneficial than continuing the pregnancy due to maternal or fetal indications. When the cervix is unfavourable for delivery, cervical ripening is performed prior to IOL. A wide variety of mechanical, pharmacological and combination methods are used, but the optimal approach balancing efficacy, safety and patient experience remains uncertain. Conventional aggregate data (AD) meta-analyses lack individual-level data, limiting exploration of patient-level factors for personalised medicine and do not address concerns about the trustworthiness of data presented in peer-reviewed randomised controlled trials (RCTs). This protocol describes an individual participant data (IPD) network meta-analysis (NMA) designed to evaluate and rank cervical ripening methods for IOL using only high quality, trustworthy data.

We will identify eligible parallel-group RCTs enrolling pregnant women with a singleton, cephalic fetus at ≥34 weeks' gestation requiring cervical ripening, through comprehensive searches of Ovid MEDLINE, Embase, Emcare, Scopus, Cochrane Pregnancy and Childbirth Register, WHO International Clinical Trials Registry Platform, clinicaltrials.gov and reference lists of prior reviews. The interventions we consider will be selected via Delphi consensus with international clinical experts. Eligible trial investigators will be invited to contribute de-identified IPD; AD will be used if IPD is unavailable. Trials will be assessed for trustworthiness using the Trustworthiness in RAndomised Clinical Trials checklist and the IPD Integrity Tool, with only eligible studies included in the primary analysis. All statistical analyses will follow a pre-specified statistical analysis plan (SAP) finalised before any analyses are conducted. A two-stage, contrast-based, frequentist IPD-NMA will compare cervical ripening methods for three co-primary outcomes: vaginal birth, composite adverse perinatal outcomes and composite adverse maternal outcomes. Subgroup analyses will assess effect modifiers (eg, parity, age and previous caesarean), with treatment rankings presented using the surface under the cumulative ranking curve and rank-heat plots. Sensitivity analyses will examine the impact of bias, missing data and population criteria.

This study has been approved by the Monash University Human Research Ethics Committee (No. 48189). IPD will be de-identified and securely transferred for storage on a Monash University-hosted shared network drive. Findings will be disseminated via peer-reviewed publications, conference abstracts and the Cervical Ripening for Induction of Labour Collaborative Evidence Network Meta-Analysis (CIRCLE-NMA) website (https://circlenma.com). Patient and public involvement will guide the communication and interpretation of results.

CRD420251077464.