Effect of second-trimester sonographic cervical length on the risk of spontaneous preterm delivery in different risk groups: A prospective observational multicenter study

Abstract

Introduction: The aim of the study is to compare the effect of cervical length measured with transvaginal ultrasound in the second trimester on the risk of spontaneous preterm delivery (PTD) between different risk groups of asymptomatic women with a singleton pregnancy. Material and methods: This is a pre-planned exploratory analysis of the CERVIX study, a prospective blinded multicenter diagnostic accuracy study. Asymptomatic women with a singleton pregnancy were consecutively recruited at their second-trimester routine ultrasound examination at seven Swedish ultrasound centers. Cervical length was measured with transvaginal ultrasound at 18–20 weeks (Cx1; n = 11 072) and 21–23 weeks (Cx2, optional; n = 6288). The effect of cervical length on the risk of spontaneous PTD and its discriminative ability was compared between women with: (i) previous spontaneous PTD, late miscarriage or cervical conization (high-risk group; n = 1045); (ii) nulliparae without risk factors (n = 5173); (iii) parae without risk factors (n = 4740). Women with previous indicated PTD were excluded (n = 114). Main outcome measures were: effect of cervical length on the risk of spontaneous PTD expressed as odds ratio per 5-mm decrease in cervical length with interaction analysis using logistic regression to test whether the effect differed between groups, area under the receiver operating characteristic curve (AUC), sensitivity, specificity, number needed to screen to detect one spontaneous PTD. Results: The effect of cervical length at Cx2 on the risk of spontaneous PTD <33 weeks was similar in all groups (odds ratios 2.26–2.58, interaction p value 0.91). The discriminative ability at Cx2 was superior to that at Cx1 and was similar in all groups (AUC 0.69–0.76). Cervical length ≤25 mm at Cx2 identified 57% of spontaneous preterm deliveries <33 weeks in the high-risk group with number needed to screen 161. The number needed to screen for groups (ii) and (iii) were 1018 and 843. Conclusions: The effect of cervical length at 21–23 weeks on the risk of spontaneous PTD <33 weeks is similar in high- and low-risk pregnancies. The differences in number needed to screen should be considered before implementing a screening program.