Cervical Stiffness Measured by the Cervisense Intravaginal Probe for Identification of Imminent Delivery in Symptomatic Threatened Preterm Labor

Abstract

Background and Objectives: Risk stratification in symptomatic threatened preterm labor (TPTL) remains challenging, particularly for clinically actionable time horizons. We evaluated whether quantitative cervical stiffness measured with the Cervisense Intravaginal Probe discriminates between women who will and will not deliver within 7, 10, or 14 days after presentation and assessed its discriminative performance alone and in combination with transvaginal cervical length. Materials and Methods: We conducted a multicenter, non-interventional study across 10 obstetric emergency units in Spain. Eligible participants were women aged ? 18 years with a live singleton pregnancy at 28(+0) to 36(+6) weeks, regular uterine contractions, intact membranes, and cervical dilatation < 2 cm on speculum examination. Cervical stiffness was measured at presentation using torsional wave elastography with the Cervisense Intravaginal Probe. Cervical length was measured by transvaginal ultrasonography following international standards. Outcomes were delivery within 7, 10, and 14 days after presentation. Discrimination was quantified using the area under the receiver operating characteristic curve (AUC). An explanatory multivariable logistic regression model for delivery within 14 days included cervical stiffness, cervical length, and covariates selected as potential confounders. Results: Among 305 participants, 17 (5.6%) delivered within 7 days, 19 (6.2%) within 10 days, and 24 (7.9%) within 14 days. Mean cervical stiffness was lower among women who delivered within 14 days than among those who did not (8.63 vs. 14.82 kPa; p = 0.048). Cervical length was shorter among women who delivered within 7, 10, and 14 days (all p < 0.001). Discrimination for cervical stiffness was moderate (AUC, 0.66, 0.64, and 0.73 for ?7, ?10, and ?14 days) and higher for cervical length (AUC, 0.80, 0.77, and 0.78). The combined model achieved the highest discrimination (AUC, 0.85, 0.85, and 0.86). In the ?14-day explanatory model, higher log-transformed cervical stiffness was associated with lower odds of delivery (OR, 0.907; 95% CI, 0.828 to 0.995; p = 0.039), as was longer cervical length (OR per 1 mm, 0.910; 95% CI, 0.876 to 0.946; p < 0.001). Conclusions: In symptomatic TPTL, Cervisense-derived cervical stiffness showed moderate discrimination for short-term delivery and provided complementary information to cervical length, with improved discrimination when both measures were combined. These findings support future studies developing multivariable prediction models that incorporate quantitative cervical consistency.