Abstract
Background. The aim of this study was to compare the performance of the Bishop score and transvaginal ultrasonography to predict successful labor induction, and to estimate the most useful cut-off points for the two methods. Methods. The five components of the Bishop score were assessed by digital examination and cervical length was measured by transvaginal ultrasonography in 177 women with a single pregnancy, 36–42 weeks of gestation, and a live fetus in cephalic presentation before induction of labor. Results. Multiple regression analysis showed that the Bishop score, cervical length, and parity provided independent contribution in the prediction of the likelihood of delivering vaginally within 60 h. The only component of the Bishop score that was independently predictive of the probability of vaginal delivery within 60 h was station. The best cut-off points for predicting successful induction using receiver operating characteristic curves were 24 mm for cervical length and 4 for the Bishop score. Cervical length was a better predictor than the Bishop score (sensitivity and specificity of 66 and 77% versus 77 and 56%, respectively). Kaplan–Meier survival curves showed that cervical length was the best discriminator of successful induction. Conclusion. Measurement of cervical length by transvaginal ultrasonography is a better predictor of success in labor induction with both topical prostaglandin and oxytocin.