(131) Mid-pregnancy cervical length and pregnancy outcomes in low-risk singletons: a prospective cohort (QP screening study)

Objective: Short midtrimester cervical length (CL) increases the risk for spontaneous preterm birth (sPTB), but sensitivity of this screening tool is low. This study aims to improve sPTB prediction by identifying which factors, combined with CL most accurately predict sPTB.

Study Design: This multicenter prospective cohort study included singletons with no prior sPTB < 34 weeks of whom CL was measured at 18-22 weeks gestation. Primary outcome was sPTB < 37 weeks. Secondary outcomes were (s)PTB < 28, 32, 34 and 37 weeks. CL predictive capacity was assessed with likelihood ratios (LR). Multivariable cox regression (MCR) combined CL with other risk factors (e.g. smoking, cervical surgery, interpregnancy interval (IPI) and was expressed as Hazard Ratios (HR) and 95% confidence intervals (CI)). For PTB risk factors, we calculated population-attributive risk (PAR) for sPTB < 37 and < 32 weeks. Nulliparous and multiparous women were assessed separately.

Results: We included 14,718 singletons between 2015 – 2021. Nulliparous women (N=7532) with a CL ≤ 25, 26-30, 31-35 and > 35mm had sPTB < 37 rates of 28.6%, 9.0%, 9.4% and 3.4%, respectively. In multiparous (N=7186) women, rates were 17.5%, 7.9%, 4.9% and 2.2%. Rates of sPTB < 32 weeks for CL ≤ 25 mm were 20.9% in nulliparous and 12.7% in multiparous women. MCR in nulliparous women showed that CL ≤ 25, 26-30 and 31-35 mm remained significantly associated with increased sPTB hazard (all p < 0.001). In multiparous women, CL ≤ 25mm (p < 0.001), smoking (p = 0.049) and IPI < 6 months (p = 0.002) increased sPTB hazard. In total, 1.0% of women had a CL ≤25mm, yet the PAR of short cervix for sPTB < 37 and < 32 weeks was 7.0% and 29.3% in nulliparous vs 5.1% and 25.1% in multiparous women.

Conclusion: Asymptomatic short midtrimester CL is a significant predictor of sPTB, highlighting the need for implementation of measurement in routine care. A multifactorial approach may enhance sPTB prediction, yet the impact of these additional factors seems limited. CL measurement remains the most effective tool available, especially for identifying those at high risk for very sPTB.