Risk prediction model and intervention strategies for intraoperative hypothermia in robotic-assisted urological surgery

To develop a risk prediction model for intraoperative hypothermia in patients undergoing robotic-assisted urological surgery and formulate hierarchical warming strategies. This single-centre retrospective study calculated the sample size using G*Power 3.1. A total of 487 patients admitted to our hospital from May 2023 to May 2025 were included after correcting for 10% invalid samples (minimum required: 438). Hypothermia was defined as an intraoperative nasopharyngeal temperature < 36 ℃. Patients were divided into hypothermia (n = 61) and non-hypothermia (n = 426) groups. Lasso regression was used to screen variables, and a binary logistic regression model with a nomogram was constructed and internally validated by 1,000 bootstrap resamples. Operating room temperature, the absence of active warming measures, age, body mass index and preoperative baseline core temperature were independent risk factors (P < 0.05). The model had an area under the curve (AUC) of 0.928 (sensitivity 91.8%, specificity 80.5%, Youden index 0.723). The corrected AUC was 0.915 after validation. The Hosmer-Lemeshow test (χ²=7.514, P = 0.0897) showed good calibration, and the decision curve analysis demonstrated significant clinical net benefit. The prediction model has excellent performance, which can accurately identify high-risk patients and guide personalised perioperative warming to reduce intraoperative hypothermia incidence.