Objective To determine the risk of preterm birth associated with uterine anomalies, and the clinical utility of serial measurement of cervical length.

Study design and results We carried out a retrospective cohort study of women with uterine anomaly referred to UK tertiary referral centres between 2001-2011. 91 women with uterine anomaliy were identified (61 bicornuate, 15 septate, 10 didelphis, 3 unicornuate and 2 arcuate). Preterm delivery rates are shown in the table. We found no association between type of anomaly and risk of preterm birth although the majority had a bicornuate uterus.

9 women received cervical cerclage. 6 were placed electively; only three of these had had a previous preterm delivery. 3 women had ultrasound indicated cerclage, at 19, 23 and 23 weeks, all 3 delivered before 34 weeks.

Conclusions The risk of preterm birth associated with uterine anomaly is high, especially with a history of preterm delivery or second trimester loss. The risk remains high even in women with a previous term delivery. However, serial measurement of cervical length fails to predict preterm birth in the majority of cases, and ultrasound indicated cerclage appears not to reduce the risk. The mechanism of preterm birth in women with uterine anomaly is unlikely to be primarily a cervical problem.