Predicting Category II and III foetal heart rate patterns during epidural analgesia: a retrospective cohort study

BACKGROUND: This study aimed to identify predictors and develop a multivariable prediction model for Category II and III foetal heart rate (FHR) patterns in parturients undergoing labour epidural analgesia (LEA). METHODS: A retrospective cohort study of 237 parturients receiving LEA was conducted. To address the multicollinearity of 18 intrapartum variables, least absolute shrinkage and selection operator (LASSO) regression and cross-validation were used for feature selection. A predictive nomogram was constructed and internally validated. Model performance was evaluated by the area under the receiver operating characteristic curve (AUC), calibration plots, and Decision Curve Analysis (DCA). RESULTS: Abnormal FHR patterns (Category II and III) occurred in 168 (70.9%) parturients. The LASSO algorithm identified 11 predictors. Multivariable logistic regression demonstrated that maternal intrapartum temperature (OR = 3.518), initial LEA bolus count (OR = 3.625), body mass index (BMI), and oxytocin dosage were independent predictors associated with abnormal FHR patterns. The constructed nomogram demonstrated good discrimination (AUC = 0.800) and good calibration. DCA demonstrated potential clinical net benefit across a wide range of threshold probabilities (0.02-0.99). CONCLUSIONS: We developed and internally validated an 11-variable nomogram (AUC = 0.800) to predict abnormal FHR patterns during LEA. By integrating routinely available clinical variables, the nomogram facilitate early risk stratification and support individualised intrapartum management. TRIAL REGISTRATION: Clinical trial registration: (https://www.chictr.org.cn ChiCTR2300073493; registered 12th July 2023).