An individual participant data network meta-analysis of the APOSTEL trials on the effect of tocolysis in threatened preterm birth between 30–33+6 weeks of gestation in twin pregnancies

OBJECTIVE: Twin pregnancies have a substantial risk of preterm birth, but evidence on tocolytics is limited. We evaluated whether nifedipine or atosiban improves neonatal outcomes compared with placebo. DESIGN: Individual participant data network meta-analysis (IPD-NMA) of two international multicentre randomised controlled trials (APOSTEL 3 and 8). SETTING: The APOSTEL-3 compared nifedipine with atosiban, while the APOSTEL-8 compared atosiban with placebo. POPULATION: weeks. METHODS: Analyses were performed at the infant level using generalised estimating equations to account for clustering within twin pairs. MAIN OUTCOME MEASURES: Adverse neonatal outcome: composite of neonatal morbidity or mortality. RESULTS: A total of 179 women were included: 94 allocated to atosiban, 27 to nifedipine, and 58 to placebo. Prolongation beyond 48 h was not significantly increased with nifedipine (RR 0.96, 95% CI 0.68-1.36) or atosiban (RR 1.10, 95% CI 0.88-1.38). The composite adverse neonatal outcome occurred in 3.7% of neonates in the nifedipine group (2/54), 7.5% in the atosiban group (14/188), and 5.2% in the placebo group (6/116). Relative risks were 0.72 (95% CI 0.15-3.43) for nifedipine vs placebo, 1.44 (95% CI 0.57-3.64) for atosiban vs placebo, and 2.01 (95% CI 0.47-8.58) for atosiban vs nifedipine. CONCLUSION: weeks was found.