Tocolysis compared with no tocolysis in women with threatened preterm birth and ruptured membranes: A propensity score analysis

•The value of tocolytics for improving neonatal outcomes is globally disputed.•There is variety in local preference for treatment with/without tocolytics.•Therefore, it can be difficult to set up an international placebo controlled RCT.•A propensity score analysis creates a setting resembling an RCT.•No significant differences were found between tocolysis and no tocolysis. In women with preterm ruptured membranes and contractions, the administration of tocolysis is controversial. This study compares tocolysis with no tocolysis in women with threatened preterm birth and ruptured membranes. To compare tocolysis with no tocolysis in women with threatened preterm birth and ruptured membranes. Data from the APOSTEL III RCT was combined with data from the National Maternity Hospital, Dublin. In the APOSTEL III trial, women with threatened preterm birth were randomized to atosiban or nifedipine. Patient data from Ireland were obtained from a cohort of women with threatened preterm birth with ruptured membranes. The Irish women received no tocolytic treatment. Only women with ruptured membranes and contractions were selected. We studied women with singleton or twin pregnancies and a gestational age between 25+0 and 33+6 weeks. Propensity score matching was performed to create comparable groups. Primary outcome was a composite adverse neonatal outcome. Secondary outcomes were individual components of the primary outcome, as well as neonatal intensive care unit (NICU) admission, gestational age at delivery, prolongation of pregnancy and mode of delivery. 153 women from the Apostel III trial were compared with 51 eligible women of the Irish cohort. We could match 46 women who received tocolysis and 46 women who received no tocolysis. All women had ruptured membranes. Maternal age, BMI, parity and gestational age at study entry were comparable between the groups after matching. There were no statistically significant differences in neonatal composite outcome (9.6 % in the tocolysis group versus 18 % in the control group, OR 0.46, 95 % CI 0.13−1.63). We found a lower incidence of NICU admission in the tocolysis group (63 %) than in the control group (94 %; OR 0.11, 95 % CI 0.03−0.41), which could be explained by differences in national admission policies. There were no statistically significant differences between tocolysis and no tocolysis in any of the other outcomes including sepsis, gestational age at delivery and time to delivery. In this propensity score analysis