Predictors of Early Preterm Birth Despite Vaginal Progesterone Therapy in Singletons with Short Cervix.

Abstract

This study aimed to identify the incidence of and risk factors for early preterm birth (PTB) (delivery <34 weeks) in women without prior PTB and current short cervix (≤20 mm) prescribed vaginal progesterone. Retrospective cohort study of singletons without prior PTB diagnosed with short cervix (≤20 mm) between 180/7 and 236/7 weeks. Women who accepted vaginal progesterone and had delivery outcomes available were included. Demographic/obstetric history, cervical length, and pregnancy characteristics compared between women with early PTB versus delivery ≥34 weeks. Multiple logistic regression analysis used to identify predictors; odds ratio for significant factors used to generate a risk score. Risk score and risk of early PTB assessed with receiver operating characteristic curve (ROCC). Perinatal outcomes compared by risk score. Among 109 patients included, 29 (27%) had a spontaneous PTB <34 weeks. In univariate analysis, only gestational age at ultrasound, presence funneling, and mean cervical length were significantly different between those with and without early sPTB. With multiple logistic regression analysis, only gestational age at diagnosis (odds ratio [OR]: 0.66; 95% confidence interval [CI]: 0.46-0.96; p = 0.028) and index cervical length (OR: 0.84; 95% CI: 0.76-0.93; p = 0.001) remained significantly associated with early PTB. ROCC for the risk score incorporating cervical length and gestational age was predictive of early PTB with an AUC of 0.76 (95% CI: 0.67-0.86; p < 0.001). A high-risk score was predictive of early PTB with a sensitivity of 79%, specificity of 75%, positive predictive value of 54%, and negative predictive value of 91%. Women with a high-risk score had worse perinatal outcomes compared with those with low-risk score. A total of 27% of patients with short cervix prescribed vaginal progesterone will have a sPTB < 34 weeks. Patients at high risk for early PTB despite vaginal progesterone therapy may be identified using gestational age and cervical length at diagnosis of short cervix. Given the narrow window for intervention after diagnosis of short cervix, this has important implications for clinical care.