ABSTRACT
OBJECTIVE: Our objective was to determine if transvaginal cervical length at 16-20 weeks is predictive of preterm birth <34 weeks in a large cohort of twin pregnancies. STUDY DESIGN: This is a secondary analysis from a randomized trial of 17 alpha-hydroxyprogesterone caproate in twins to prevent preterm birth. Transvaginal cervical length was performed at 16-20-week gestation. The inclusion criteria were non-anomalous twins with transvaginal cervical length at 16-20 weeks. Receiver-operating characteristic (ROC) curves were generated to determine the transvaginal cervical length associated with preterm birth. RESULTS: Of 655 pregnancies, 27% (N = 178) women met our inclusion criteria. The rate of spontaneous preterm birth before 34 weeks was 16% (N = 29). A receiver operator characteristic curve was generated for all preterm birth <34 weeks (spontaneous and indicated) which demonstrated an area under the curve of 0.51, 95% CI (0.41-0.61). When indicated preterm birth (n = 15) were excluded, the area under the curve was 0.59 (95% CI 0.47-0.70), indicating that transvaginal cervical length values were not a clinically useful test for the prediction of spontaneous preterm birth. A transvaginal cervical length of 30 mm from this model would produce a sensitivity of detecting spontaneous preterm birth of 95% and a specificity of 14%. CONCLUSION: In an asymptomatic twin population, a single transvaginal cervical length between 16 and 20 weeks was not predictive of spontaneous preterm birth before 34 weeks. Thus, our findings suggest that routine transvaginal cervical length screening of twins should not be performed between 16-20 weeks.