It has been reported that the risk of spontaneous preterm delivery is increased in women who have shorter cervices on vaginal ultrasonography during pregnancy. However, the appropriate timing of the measurement of cervical length and the best cutoff value are controversial in Japan. To evaluate the relation between the cervical length and the risk of preterm delivery, we conducted this study by retrospectively reviewing the cases treated in our hospital. Subjects were 1184 patients who had

more »

... gone the examination of the cervical length at both 16-19 and 20-24 weeks of gestation and were delivered in Showa University Hospital from January 2008 to December 2011. Multiple pregnancies were excluded. The patients were hospitalized when they had cervices shorter than 25 mm before 26 weeks of gestation. Hospital Ethics Committee approved this study. The mean gestational weeks of delivery was 38.6 ± 2.4, percentage of preterm delivery < 37 weeks and < 35 weeks were 5.9% and 2.4%, respectively. The cervical length at 16-19 and 20-24 weeks was shorter in the preterm delivery at < 35 weeks group compared with that in the delivery at ≧ 35 weeks group (mean [± SD] , 37.1 ± 8.9 mm v.s. 40 .2 ± 7.3 mm * p = 0.0271 and 32.7 ± 12.6 mm v.s. 38 .3 ± 7.6 mm * p = 0.0001, respectively) . The relative risks of preterm delivery at < 35 weeks among the women with cervix shorter than 25 mm at both 16-19 and 20-24 weeks were 5.8 and 7.4, respectively. The risk of preterm delivery was increased in patients with short cervix at both 16-19 and 20-24 weeks of gestation. However, we cannot treat patients merely based upon cervical length, because sensitivity and positive predictive value for preterm delivery are both insufficient even when we use the clinically best cutoff value of cervical length.