What trans-vaginal ultrasound parameters are better correlated with a shorter labor induction to vaginal delivery interval? A prospective observational cohort study.

Abstract

Objective: During the induction of labor (IOL) planning, it is important to provide patients with information regarding how long the induction process might take. This study aimed to determine which ultrasonographic cervical parameters are independently associated with a shorter IOL-to-vaginal delivery (VD) interval.

Methods: This was a prospective observational cohort study. For enrollment purposes, women with single pregnancy, fetus in cephalic presentation, age between 18 and 45 years and good Italian proficiency were included. Women with a history of uterine surgery, in active labor, and cases of fetal growth abnormalities were excluded. The enrolled women underwent a transvaginal ultrasound within 7 days from the scheduled labor induction in order to measure the following parameters: the cervical length (CL), the utero-cervical angle (UCA), the cervical sliding sign (CSS) and the cervical consistency index (CCI). Before starting the labor induction process, patients were also digitally evaluated, acquiring the Bishop score (BS). The method of IOL was determined based on the BS. Ultrasound assessments and Bishop score evaluations were performed independently and in a blinded manner to reduce bias. Statistical analyses were performed using STATA 18.0.

Results: Between June 2023 and November 2024, 400 women were nonconsecutively enrolled in the study. Of these, 83 experienced spontaneous labor before the scheduled labor induction, resulting in 317 women who underwent IOL. The median IOL-to-VD interval was 1264 min (IQR 694-1940). Univariable regression analysis demonstrated significant associations between the IOL-to-VD interval and CL (β = 29.15; 95% CI 16.16, 42.23; p < 0.001), CCI (β = 12.60; 95% CI 3.93, 21.24; p = 0.004), and BS (β = -211.15; 95% CI -271.59, -150.71; p < 0.001). Multivariable analysis confirmed independent associations with CL (β = 13.89; 95% CI 0.35,27.44; p = 0.044) and BS (β = -183.96; -249.66, -118.27; p < 0.001). When stratified by parity, univariable regression in parous women showed significant associations between the IOL-to-VD interval and CL (β = 37.44; 95% CI 20.17, 54.72; p < 0.001), CSS (β = -582; 95% CI -1014.05, 151.20; p = 0.009), CCI (β = 15.43; 95% CI 1.75, 29.11; p = 0.027), and BS (β = -227.96; -315.57, -140.35; p < 0.001).

Conclusions: In summary, among the evaluated parameters, CL consistently showed the strongest and most independent association with a shorter IOL-to-VD interval across analyses, supporting its role as the most reliable predictor. Future research should explore multivariable prediction models incorporating various ultrasonographic cervical parameters to enhance the predictive accuracy of transvaginal ultrasound.