| Authors: |
Zhao, Xinrui1,2 (AUTHOR) 330619030@qq.com, Yang, Lijun1 (AUTHOR) 553717280@qq.com, Peng, Jing1 (AUTHOR) 15670513217@163.com, Zhao, Kai3 (AUTHOR) kai_zhao@hust.edu.cn, Xia, Weina1 (AUTHOR) 13554330315@163.com, Zhao, Yun1 (AUTHOR) zhao020060@163.com |
| Abstract: |
Background: With the implementation of China's three-child policy, an increasing number of Chinese women are not opting for cesarian delivery (CD), with particular concern about intrapartum CD. At the same time, reducing the rate of intrapartum CD is closely associated with a decrease in maternal and neonatal complications, as well as an improvement in maternal satisfaction. Therefore, it is essential to develop a predictive model specifically tailored to Chinese pregnant women to reduce the rate of intrapartum CD. Objective: This study aimed to develop a predicting model for intrapartum CD at admission using the coefficients of Lasso regression analysis, multi-factor logistic regression and a nomogram. Methods: This single-center retrospective cohort study involved singleton pregnancies of women willing to undergo vaginal delivery (VD) at admission between August 2021 to March 2022 at the Department of Obstetrics in our hospital. The study cohort comprised 3,025 pregnant women who underwent a trial of vaginal labor at admission, with 378 cases in the intrapartum CD group and 2,647 cases in the VD group. These cohorts were divided into a training and a test cohort (7:3 ratio). A predictive model was developed using the training cohort to estimate the risk of intrapartum CD, incorporating coefficients of Lasso regression analysis, a nomogram and multivariate logistic regression (MLR). A Decision Curve Analysis (DCA) was conducted to evaluate the clinical utility and net benefit of the nomogram in the training cohort. The receiver operating characteristic (ROC) curve, along with sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), accuracy and precision were used to evaluate clinical utility of the prediction models in both the training and test cohorts. Results: Nine factors in the training cohort were identified as independent predictors of intrapartum CD, including BMI before labor (OR = 1.07, p = 0.006), maternal height(OR = 0.94, p < 0.001), gestational age at delivery(OR = 1.13, p = 0.120), fetal weight by last ultrasound before labor (OR = 1.00, p = 0.003), previous VD history (OR = 0.09, p < 0.001), previous CD history (OR = 2.53, p = 0.020), spontaneous labor(OR = 2.01, p < 0.001), cervical Bishop scores(OR = 0.73, p < 0.001), and Hypertensive disorder complicating pregnancy(OR = 2.08, p = 0.001). Additionally, a DCA was conducted to evaluate the clinical utility and net benefit of the nomogram in the training cohort. The area under the curve (AUC), sensitivity, specificity, PPV, NPV, accuracy and precision were 0.793 (95%CI 0.768–0.819), 84.9%, 60.1%, 23.8%, 96.4%, 63.2% and 23.8%, respectively in the training cohort, and 0.753(95%CI 0.703–0.798), 81.3%, 59.3%, 21.0%, 96.0%, 61.9% and 21.1% respectively in the test cohort. These results indicate that the mode demonstrated good predictive performance in both datasets. Conclusion: Our model, which utilized admission indicators, performed well in predicting the risk of intrapartum CD using a nomogram. These findings have significant practical implications for physicians seeking to reduce the rate of intrapartum CD. [ABSTRACT FROM AUTHOR] |
