Development and validation of a predictive model for forceps delivery risk in term singleton primiparas for early...

This study aimed to establish and validate a predictive model for forceps-assisted delivery in term singleton primiparas during the early second stage of labor, enabling early identification of high-risk cases and targeted intrapartum care to reduce maternal and neonatal risks associated with prolonged labor.

The study retrospectively analyzed clinical data from pregnant women who delivered at The Second People's Hospital of Yibin City between January 2017 and December 2024. Term singleton primiparas who underwent forceps delivery were designated as the forceps delivery group (n = 346), while each corresponding case of spontaneous vaginal delivery (SVD) was included in the vaginal delivery group (n = 346) using adjacent case matching. Maternal, fetal, and labor characteristics were collected and compared between groups. Univariate and multivariate logistic regression analyses were conducted to identify independent risk factors for forceps delivery and to construct a predictive model. Model performance was comprehensively evaluated in terms of discrimination, calibration, and clinical utility using the area under the receiver operating characteristic curve (AUC), the Hosmer-Lemeshow test, the Brier score, and clinical decision curve analysis (DCA).

Independent risk factors for forceps delivery in term singleton primiparas identified at the onset of the second stage of labor included: pre-delivery body mass index (BMI) ≥ 28 kg/m², fetal abdominal circumference ≥ 340 mm, grade II or III meconium-stained fluid (MSF), non-occipital anterior (OA) fetal position, and active phase duration ≥ 120 min. This predictive model for early risk identification during the second stage of labor achieved an AUC of 0.775 (95% CI: 0.741-0.810), with an optimal cutoff value of 0.413, corresponding to a sensitivity of 0.688 and a specificity of 0.725. The Hosmer-Lemeshow test indicated a good model fit. After 500 internal bootstrap validations, the corrected C-index was 0.771. The model demonstrated acceptable calibration and provided a net benefit over the treat-all strategy for clinically plausible threshold probabilities up to 0.35.

The proposed model, based on routine intrapartum predictors at the onset of the second stage, enables early risk stratification for forceps delivery in term singleton primiparas and supports anticipatory intrapartum planning. Further prospective, multicenter external validation is warranted to determine its real-world clinical impact and generalizability.