METHODS: Low-risk women at 40 weeks' gestation undergoing membrane sweeping to expedite labor were recruited. Participants were scheduled for labor induction at 41 weeks' gestation. Transvaginal ultrasonography was performed immediately before and after membrane sweeping to measure the cervical length. Three presweep and postsweep cervical lengths were measured. The shortest lengths before and after the sweep were taken as the representative lengths. The effect of membrane sweeping on cervical length was analyzed. Multivariable logistic regression analysis was performed to evaluate the effect of cervical shortening on labor induction and the mode of delivery.

RESULTS: For the 160 participants, the mean presweep cervical length ± SD was 21.0 ± 10.0 mm; the postsweep length was 23.8 ± 10.9 mm, an average increase of 2.8 ± 0.6 mm (P < .001). Cervical shortening after membrane sweeping was noted in 53 of 160 cases (33%). Cervical shortening was associated with a reduction in all-cause cesarean delivery but not labor induction on bivariate analysis. After adjustment for maternal age, parity, presweep Bishop score, postsweep cervical length, oxytocin augmentation, epidural analgesia, and meconium-stained fluid, cervical shortening after membrane sweeping was independently predictive of a reduction in cesarean deliveries (adjusted odds ratio, 0.24; 95% confidence interval, 0.06-0.90; P = .034).