Accuracy of cervical transvaginal sonography in predicting preterm birth: a systematic review

This review investigates the accuracy with which transvaginal cervical sonography predicts spontaneous preterm birth. Published studies were identified without language restrictions through nine different databases and manual searching of bibliographies of known primary and review articles. Studies were selected if they undertook antenatal transvaginal sonographic cervical assessment among a population of pregnant women with known gestational age of delivery. There were 46 primary articles, which included a total of 31,577 women, consisting of 33 studies in asymptomatic and 13 studies in symptomatic women. Data were extracted for the studies' characteristics and quality. Accuracy data were used to form 2 x 2 contingency tables for various cervical length measurements with birth before 32, 34 and 37 weeks' gestation as the reference standards. Data were stratified according to singleton or twin pregnancy, gestational age at testing, cervical length threshold, and the various reference standards, and were pooled to produce summary estimates of likelihood ratios (LRs). Our review showed that transvaginal cervical sonography identifies women who are at higher risk of spontaneous preterm birth, although there was a wide variation amongst studies with respect to gestational age at testing, definition of threshold of abnormality and definition of reference standard. The most commonly reported sub-group was testing of asymptomatic women at < 20 weeks' gestation using a threshold cervical length of 25 mm with spontaneous preterm birth before 34 weeks' gestation as the reference standard. The summary LR+ for this group was 6.29 (95% CI, 3.29-12.02), with corresponding LR- of 0.79 (95% CI, 0.65-0.95). Both cervical length measurement and funneling, whether alone or in combination, appear to be useful (depending on the threshold chosen to define the abnormality) in predicting spontaneous preterm birth in asymptomatic women. For symptomatic women there was a paucity of data, although the degree of funneling appeared to be predictive of spontaneous preterm birth.