Zhongguo linchuang yanjiu (Aug 2024)
Construction of a predictive model for fever in elderly patients after endoscopic mucosal dissection surgery
Abstract
Objective To analyze the risk factors for postoperative fever in elderly patients after endoscopic mucosal dissection (ESD), and construct a predictive model. Methods The research subjects were 500 elderly patients who received upper gastrointestinal ESD treatment in The First Affiliated Hospital with Nanjing Medical University from January 2023 to January 2024. They were divided into a modeling group (n=350) and a validation group (n=150) in a 7∶3 ratio. The modeling group of patients was divided into fever group and non-fever group based on whether they experience fever after ESD surgery. Data were collected from inpatient electronic medical record systems. Univariate and multivariate logistic regression analysis were used to screen factors and a predictive model was established in elderly patients after ESD. ROC curve was used to evaluate the discriminability of the predictive model. Results A total of 90 elderly patients who underwent ESD treatment experienced postoperative fever (T≥37.3℃), with a postoperative fever incidence rate of 25.71%. The proportion of women, history of diabetes, postoperative bleeding, maximum tumor diameter>3cm, and catheter placement in the fever 〖JP+1〗group were higher than those in the non-fever group (P<0.05). Multivariate binary logistic regression analysis showed that women, history of diabetes, postoperative bleeding, maximum tumor diameter>3cm, and placement of urinary catheter were independent risk factors for postoperative fever in elderly patients with ESD (P<0.05). The predictive performance and calibration of the ESD postoperative fever prediction model were good, and with clinical practicality. Conclusion Women, history of diabetes, postoperative hemorrhage, maximum tumor diameter>3cm, and placement of urinary catheter are risk factors for postoperative fever in elderly patients with ESD. Medical personnel need to focus on these factors to predict the risk of postoperative fever. By building a prediction model, it can provide practical tools for early identification and prevention of postoperative fever in elderly patients with ESD.
Keywords