Cervical length surveillance for predicting spontaneous preterm birth in women with uterine anomalies: A cohort study

Introduction: Uterine anomalies occur in an estimated 5% of women and have been shown to confer a higher risk of spontaneous preterm birth (SPTB). A sonographically short cervix (<25 mm) is a risk indicator for SPTB, although its predictive utility has been little studied in this specific high-risk population. We aimed to assess the pregnancy outcomes and predictive ability of short cervix in a cohort of women with uterine anomalies attending a high-risk antenatal clinic.

Material and methods: This historical cohort study assessed all pregnancies in women with congenital uterine anomalies referred to the Preterm labor Clinic at the Royal Women's Hospital, Melbourne, Australia, between 2004 and 2013. Logistic and linear regressions and receiver-operator curves were used to examine associations between cervical length and preterm birth.

Results: SPTB (<37 weeks' gestation) occurred in 23% of the 86 pregnancies (n = 20); rates by subgroup were: unicornuate uterus 60% (n = 3/5), uterus didelphys 40% (n = 6/15), bicornuate uterus 18% (n = 9/51), septate uterus 13% (n = 2/15). Preterm prelabor rupture of membranes occurred in 55% of spontaneous preterm births and was not independently associated with the presence of cervical cerclage or ureaplasma urealyticum. Short cervical length was associated with SPTB in women with septate uterus. Short cervix at 24 weeks (not at 16 or 20 weeks) was moderately predictive of SPTB < 34 weeks.

Conclusions: Women with uterine anomalies are at increased risk of spontaneous preterm birth, particularly those with unicornuate uterus or uterus didelphys, but cervical surveillance did not identify these cases. Short cervix may be associated with SPTB in women with septate uterus. Preterm prelabor rupture of membranes occurred in 55% of SPTB. More research is required into etiology to help determine appropriate monitoring and treatment.