813: Sonographic cervical length for the short-term prediction of preterm birth in symptomatic patients: a meta-analysis

Abstract

To assess the predictive accuracy of transvaginal sonographic cervical length measurement as a short-term predictor of preterm birth in patients with signs and symptoms of preterm labor. We searched electronic databases and syllabi from scientific meetings. We also performed manual searches of references from textbooks and of known primary and systematic review articles and contacted authors and experts in the field. We targeted cohort studies, in any language, reporting data on the diagnostic accuracy of sonographic cervical length measurement for the prediction of preterm birth within 48 hours following testing in symptomatic women. Methodological quality was assessed using the QUADAS tool and 2x2 contingency tables were constructed for data abstraction and assessment of diagnostic measures. Diagnostic meta-analysis performed using bivariate methods. Pooled estimates of sensitivity, specificity, positive and negative likelihood ratios, area under the curve, and diagnostic odds ratios (DOR) were calculated. Tests for heterogeneity, publication bias and meta-regression were also performed. Ten studies were included for analysis (2,211 subjects). QUADAS scores ranged from 8-13. For delivery within 48 hours employing 15 mm as the selected cut off, the overall pooled estimates for sensitivity and specificity were 0.77 (95% confidence interval [CI] 0.54-0.90) and 0.88 (95% CI 0.84-0.91), respectively and a DOR of 24 (95% CI 9-65). Positive and negative likelihood ratios based on the selected 15 mm cut off were 6.4 (95% CI 4.7-8.7) and 0.26 (95% CI 0.12-0.58) respectively. The area under the curve was 0.90 (0.88-0.93). Based on a pre-test probability of 10% for delivery within 48 hours of testing, post-test probabilities (positive and negative) were 42% and 3% respectively. In women with signs and symptoms of preterm labor, TVS cervical length measurement has moderate accuracy for predicting delivery within 48 hours of testing.