Predictive models for intrapartum maternal fever: Development and validation of pre-analgesia and labor process indicators

Combined spinal-epidural anesthesia is effective for labor pain relief but is associated with increased rates of intrapartum maternal fever, which can negatively impact maternal and neonatal outcomes. This study aimed to develop and validate 2 predictive models: one to assess the risk of fever before labor analgesia (model B) and another to evaluate the risk of fever throughout the labor process (model W). This retrospective case-control study was conducted at Chengdu Jinjiang District Maternal & Child Health Hospital, including 2783 parturients who received labor analgesia between January 2021 and March 2022. Stepwise logistic regression was used to identify clinical predictive indicators, followed by multivariate logistic regression to determine intrapartum fever predictors. Model performance was assessed using the Hosmer-Lemeshow test and areas under the receiver operating characteristic curves (AUROCs). A total of 2276 patients were included in the development cohort and 507 in the validation cohort. Optimal predictors for model B included primiparity, neutrophil count, anemia, estimated fetal weight, body surface area, and cervical dilation before analgesia. For model W, predictors included height, primiparity, anemia, neutrophil count, estimated fetal weight, total duration of labor, and time from rupture of membranes to delivery. AUROCs for models B and W were 0.698 and 0.740, respectively; external validation showed AUROCs of 0.703 and 0.797. In conclusion, model B effectively predicts fever risk before labor analgesia, though its predictive efficiency is lower than model W, which better predicts fever risk after analgesia. The combination of these 2 models will aid in the early identification and management of high-risk parturients, thereby reducing the incidence of intrapartum fever and improving maternal and neonatal outcomes.