Development and Validation of a Nomogram for Predicting Epidural-Related Maternal Fever Following Labor Analgesia: A Retrospective Cohort Study

Abstract Objectives: Epidural-related maternal fever (ERMF) is a frequent complication following labor analgesia, posing diagnostic challenges due to its overlap with infectious causes. Reliable tools for predicting individualized risk are currently lacking. This study aimed to develop and validate a clinical nomogram to predict the risk of ERMF in parturients undergoing epidural labor analgesia. Methods: A retrospective cohort study was conducted involving parturients receiving epidural analgesia from January 2022 to December 2023 at a single university-affiliated hospital. Eligible participants (n=373) were randomly allocated using a 7:3 ratio into training and validation datasets. Potential predictor variables were screened using the LASSO regression, and significant predictors were identified via multivariable logistic regression analysis within the training set. The nomogram's performance was rigorously evaluated using the Area Under the Receiver Operating Characteristic curve (AUC) for discrimination, calibration plots with associated statistical tests for agreement, and Decision Curve Analysis (DCA) for clinical utility. Results: ERMF was observed in 321 (86.1%) parturients included in the study. Seven independent predictors were incorporated into the final nomogram: body weight, delivery history, number of antenatal vaginal examinations, uterine contraction intensity, white blood cell count, epidural fentanyl dosage, and baseline temperature. The nomogram demonstrated good discriminative performance with an AUC of 0.847 (95% CI: 0.778–0.917) in the training set and 0.772 (95% CI: 0.649–0.896) in the validation set. Calibration analyses indicated excellent agreement between predicted probabilities and observed ERMF frequencies in both datasets (P > 0.05). DCA confirmed positive net clinical benefit across wide and clinically relevant ranges of threshold probabilities (1.5%-71% in training; 4.5%-54.5% in validation). Discussion: This study successfully developed and internally validated the first nomogram designed to predict individualized risk for ERMF. By integrating readily available clinical factors, the nomogram demonstrates robust predictive accuracy and calibration, alongside potential clinical utility. It represents a valuable tool to support enhanced risk stratification and inform clinical management decisions for parturients receiving epidural labor analgesia, pending further external validation.