Transvaginal Ultrasonography for Cervical Assessment Before Induction of Labor

Abstract

Although transvaginal ultrasonography is widely used to predict preterm delivery, its usefulness in predicting successful induction of labor at or after term has not been demonstrated. This study was planned to show whether cervical sonography could independently predict successful induction of labor in 105 consecutive parturients, all of whom had completed 37 weeks of a singleton pregnancy. All those studied had vertex presentation, intact membranes, and no more than 3 cm of cervical dilation. The most common reason for induction was postterm pregnancy of 41 completed weeks or more followed by a large-for-gestational-age fetus. The entire endocervical canal was imaged in the sagittal plane to estimate cervical length, funneling, and funnel width and length. Induction used oral and intravaginal prostaglandin E 2 and, if needed, oxytocin infusion. Only oxytocin was used if the Bishop score exceeded 4. Univariate logistic regression analysis indicated that successful induction correlated significantly with parity, the Bishop score, funnel width, and cervical length. Maternal age, gestational age, and the presence of funneling were not significant predictive factors. On multiple logistic regression analysis, only cervical length independently predicted successful induction. Induction required less time when the cervical length was 3 cm or less (Fig. 1). In women with Bishop scores less than 4, induction succeeded 97% of the time when cervical length was 3 cm or less and 78% of the time when it exceeded 3 cm. A cervical length of 3 cm or less was 75% sensitive and 83% specific in predicting successful induction. Transvaginal sonographic estimates of cervical length help to predict successful induction of labor. When active labor is used as the final outcome of induction, sonography performs better than the Bishop score.