Development and validation of a predictive model for delayed emergence from anesthesia in elderly patients undergoing video-assisted thoracoscopic surgery: a retrospective study

Delayed emergence from anesthesia is a common perioperative complication in elderly patients undergoing video-assisted thoracoscopic surgery (VATS), which may lead to prolonged recovery and increased healthcare burden. However, predictive tools specifically developed for this population remain limited. This study aimed to develop and validate a practical model to identify patients at increased risk of delayed emergence from anesthesia. This retrospective study included 1,400 elderly patients who underwent video-assisted thoracoscopic surgery (VATS) at a tertiary hospital in Beijing between January 2020 and April 2026. An additional temporally independent cohort of 300 patients was used for temporal validation. Perioperative variables, including baseline characteristics, comorbidities, intraoperative parameters, and laboratory indicators, were collected. Predictors were selected using least absolute shrinkage and selection operator (LASSO) regression, followed by multivariable logistic regression to construct the model. A nomogram was developed and internally validated using 1,000 bootstrap resamples. Model performance was assessed using discrimination, calibration, decision curve analysis (DCA), and temporal validation with subgroup analyses. Delayed emergence occurred in 210 patients (15.00%). Six independent predictors were identified: myocardial infarction, colloid solution use, bradycardia, end-tidal CO2 < 35 mmHg, hypothermia, and APTT. The model showed excellent discrimination in the derivation cohort (AUC = 0.909, 95% CI 0.888–0.929) with good calibration and clinical utility. In the temporal validation cohort, it maintained good discrimination (AUC = 0.863) and clinical usefulness, although calibration showed some deviation from ideal agreement. Subgroup analysis suggested better performance in patients aged ≥ 80 years. The nomogram demonstrated good discrimination and clinical utility for predicting delayed emergence from anesthesia in elderly patients undergoing VATS. It may facilitate perioperative risk assessment and individualized perioperative management, although further recalibration may be required before external application.