Development and validation of a predictive model for intrapartum fever using indicators measured before the start of labor analgesia

Abstract Combined spinal-epidural anesthesia-related maternal fever has adverse effects on parturients and newborns. Early identification of high-risk febrile parturients can reduce the impact of fever. Herein, we aimed to establish an effective prediction model for combined spinal-epidural anesthesia-related maternal fever using indicators measured before the start of labor analgesia. In retrospective case–control study, 4538 parturients who received labor analgesia at Chengdu Jinjiang District’s Women & Children Health Hospital from January 2021 to March 2022 composed the development cohort, and 1325 parturients who received labor analgesia at Sichuan Jinxin Xinan Women & Children’s Hospital from January 2023 to June 2023 composed the validation cohort. Data were collected before the start of labor analgesia. The optimal predictive factors were identified using stepwise logistic regression and multivariate logistic regression analyses, and a nomogram was developed. The discriminability of the model was verified area under the receiver operating characteristic curve (AUROC). The consistency of the model was verified by constructing a calibration curve. The clinical validity of the model was determined through decision curve analysis (DCA). The model was further evaluated via external validation. A total of 5863 parturients were included, including 4538 parturients in the development cohort, and 1325 parturients in the validation cohort. The optimal predictors included height, neutrophil count, estimated fetal weight, primiparity, anemia, and degree of cervical dilatation measured before the start of analgesia. The model passed the Hosmer‒Lemeshow test. The AUROCs of the model were 0.699 and 0.701 in the development and validation cohorts, respectively. This predictive model can effectively predict the risk of intrapartum fever using indicators measured before the start of labor analgesia. This model will help clinicians implement early interventions.