Predictive Value of Midtrimester Universal Cervical Length Screening Based on Parity.

Abstract

To evaluate the effect of parity on performance characteristics of midtrimester cervical length (CL) in predicting spontaneous preterm birth (sPTB) before 37 weeks. This was a retrospective cohort study of 13,508 women with no history of sPTB undergoing universal transvaginal CL screening at 17 to 23 weeks' gestation from 2011 to 2016. Patients who declined screening or with unknown delivery outcomes were excluded. Areas under the receiver operator characteristic curves were used to assess and compare the predictive ability of CL screening for sPTB. The sensitivity, specificity, and positive and negative predictive values were estimated for specific CL cutoffs for prediction of sPTB. There were 20,100 patients, of whom 2087 (10%) declined screening and 4505 (22%) did not meet inclusion criteria. Of the remaining 13,508 patients, 43% were nulliparous. The incidence rates of sPTB were 6.5% in nulliparas and 4.9% in multiparas (P < .001). The mean CLs were 39.9 mm in nulliparas and 41.8 mm in multiparas (P < .001), and those of the first percentiles were 19.0 mm in nulliparas and 24.0 mm in multiparas. Cervical length was significantly more predictive of sPTB in nulliparas (area under the curve, 0.67; 95% confidence interval, 0.63-0.70; versus 0.61, 95% confidence interval, 0.57-0.63; P = .008). At CL cutoffs of 10, 15, 20, and 25 mm or less, the sensitivity was lower in multiparas, and the specificity was comparable between the groups. Midtrimester CL is less predictive of sPTB in multiparas compared to nulliparas. The poor predictive ability, especially in multiparas, calls into question the value of universal CL screening in this population.