Development and validation of a postoperative hypothermia risk model for minimally invasive transurethral surgery under general anesthesia.

OBJECTIVE

To develop a risk prediction model for postoperative hypothermia in patients undergoing minimally invasive transurethral surgery under general anesthesia.

METHODS

This retrospective study collected data from hospital electronic medical records. The construction cohort included 1039 cases, and the validation cohort included 200 patients. Baseline characteristics and possibly significant preoperative and intraoperative factors were collected. Variables were selected using LASSO regression, followed by univariate and multivariate logistic regression to identify independent risk factors for postoperative hypothermia. A forest plot was created, and a predictive model was developed. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC) and the Hosmer-Lemeshow test.

RESULTS

Seven predictors were identified: age, BMI, glucocorticoid use, anesthesia medications (non-depolarizing muscarinics), bleeding > 50 ml, ASA classification, and use of intraoperative thermal blankets. In the construction cohort, AUC was 0.829 (95% CI 0.793-0.866; P < 0.001), with a sensitivity of 70.1%, specificity of 83.3%, and Youden's index of 0.541. The internal validation C-index was 0.85. In the external validation, AUC was 0.799 (95% CI 0.735-0.863; P < 0.001), with sensitivity of 72.4%, specificity of 73.9%, and Youden's index of 0.558. All Hosmer-Lemeshow tests showed P > 0.05.

CONCLUSION

The postoperative hypothermia risk prediction model for minimally invasive transurethral surgery demonstrated excellent discrimination and calibration by both internal and external validations, providing clinical value. It may aid clinicians in early identification of high-risk patients for personalized temperature management.