Transvaginal ultrasonography for cervical assessment before induction of labor

Objective. To evaluate the value of ultrasonographic cervical assessment in predicting the outcome of labor induction and to compare its performance against the Bishop score. Methods. The Bishop score was determined by digital examination, and transvaginal ultrasonography was performed in 105 women at 37 to 42 weeks' gestation scheduled for labor induction. Ultrasonographic parameters evaluated were cervical length, the presence of funneling, funnel width, and funnel length and were blinded to managing physicians. The primary outcome was the occurrence of active labor within 2 days (successful labor induction). The interval from the onset of induction to active labor (duration of induction) was the secondary outcome. Statistical analysis was performed by the chi(2) test, Wilcoxon rank sum test, Pearson correlation, receiver operating characteristic curves, logistic regression, Cox proportional hazards model, and generalized Wilcoxon test for survival data. Results. Induction of labor was successful in 93 women (89%). The area under the receiver operating characteristic curve for cervical length was greater than that of the Bishop score in predicting a successful labor induction (z = 2.18; P < .05). A cervical length of 3.0 cm or less had sensitivity of 75% (70 of 93) and specificity of 83% (10 of 12). Multiple logistic regression analysis showed a significant relationship between successful labor induction and cervical length but not the Bishop score (odds ratio = 0.24; 95% confidence interval, 0.096-0.59; P = .002). Only parity and cervical length had a significantly independent relationship with the duration of induction. Conclusions. Cervical length measured by transvaginal ultrasonography is a useful and independent predictor of successful labor induction and the duration of induction and provides better predictability of successful labor induction than the Bishop score does.