Introduction Cervical length (CL) is a potent predictor of preterm birth even in asymptomatic women. Very high risk women often have cervical cerclage, owing to a short cervix or previous adverse events. The prediction of CL with cerclage in situ has not been evaluated. Cerclage may influence outcome/alter the ability of cervical change to predict delivery. Our aim was to compare the value of CL in predicting preterm birth in high risk women with a history-indicated cerclage and those without cerclage.

Methods We performed a retrospective analysis of asymptomatic women high risk women in a preterm surveillance clinic. All women had at least one transvaginal ultrasound measurement of CL between 20+⁰ and 23+⁶ weeks. We compared women with a history-indicated cerclage to a control group with no cerclage. ROC curves and statistical tests were performed on Stata.

Discussion The incidence of preterm delivery was higher in the cerclage group versus non-cerclage group at <30 weeks (11.4% vs 2.8%) and <37 (26.7% vs 20.4%).There was strong evidence that CL performed well at all endpoints in both subgroups (p<0.001 for all ROC areas). There was no statistical evidence that CL performed differently between the two groups (p>0.1 in all cases).

Conclusions: We have shown CL is a valid predictor of preterm birth before 30, 34 and 37 weeks in women with an elective cerclage in situ.

Funding Tommy's (1060508).