Cervical Length Measurement for the Prediction of Preterm Birth in Multiple Pregnancies: A Systematic Review and Bivariate Meta-Analysis

Transvaginal sonography of the cervix can identify pregnancies at increased risk for preterm birth in both symptomatic and asymptomatic pregnant women. A previous randomized controlled trial has shown that vaginal progesterone is effective in preventing preterm birth in singleton women with asymptomatic cervical shortening during the second trimester. Randomized trials in women with multiple pregnancies have failed to show a significant reduction in preterm birth after treatment with progesterone. Overdistention of the cervix in previous pregnancies may limit the predictive value of cervical length measurement during the second trimester for preterm birth in multiple pregnancies. The aim of this meta-analysis was to review the predictive value of cervical length assessment for preterm birth in asymptomatic women with multiple pregnancies. The investigators conducted electronic searches of MEDLINE and Embase for published articles to 2009 as well as searches of reference lists of relevant articles not found with the electronic search. Two-by-two tables cross-classifying cervical length and gestational age were constructed from articles to score extracted descriptive data and study characteristics. The meta-analysis used a bivariate regression model analysis to construct summary receiver-operating characteristic (ROC) curves estimating sensitivity and specificity for changes in cervical length. A total of 21 studies involving 2757 women were identified Although there was a large variation in the gestational age, cutoff points of cervical length, and the exact definition of preterm birth, summary ROC curve estimates indicated that progressive shortening of cervical length was a good predictor of preterm birth before 34 weeks of gestation. ROC summary estimates showed sensitivity and specificity of 78% and 66%, respectively, at a cervical length of 35 mm, 41% and 87% at 30 mm, 36% and 94% at 25 mm, and 30% and 94% at 20 mm. These findings show that assessment of second-trimester cervical length in women with multiple pregnancies is a strong predictor of preterm birth