Ultrasonographic trans-abdominal measurement of uterine cervical length for the prediction of a midtrimester short cervix

Abstract

<h3>Objectives</h3> There is a controversy regarding the use of routine midtrimester trans-vaginal (TV) ultrasound for the screening of short cervix and the risk of preterm birth. We estimated the value of trans-abdominal (TA) ultrasound measurement of cervical length (CL), as an alternative of TV ultrasound, for universal screening of short cervix. <h3>Methods</h3> We conducted a prospective cohort study of nulliparous women with singleton pregnancy at 20 to 24 weeks gestation. All participants underwent TA ultrasound followed by TV ultrasound with acquisitions of images and videos of the uterine cervix. A second sonographer, blinded to the participants' data and pregnancy outcomes, measured the CL using TA and TV images and videos. Pearson's correlation test and ROC curves analyses were performed. <h3>Results</h3> 805 participants were recruited, including 780 (97%) where TA-CL measurement was feasible. We observed a strong correlation between TA-CL and TV-CL (correlation coefficient: 0.57; p<0.0001) with a TA-CL being 4 mm (95% CI -6 to 14 mm) below the TV-CL (mean of differences: 5 mm ± 4 mm). We observed that a TA-CL <30 mm was highly predictive of a short cervix defined as a TV-CL≤25 mm (area under the ROC curve: 0.97; 95% CI 0.95–0.99; p<0.0001) with a sensitivity of 100% and a false-positive rate of 22%. <h3>Conclusions</h3> Universal short cervix screening in nulliparous women could be performed using TA ultrasound, which could allow the avoidance of TV ultrasound in more than three quarter of women. In low-risk population, TV ultrasound could be reserved to women with TA-CL <30 mm.