Abstract:Objective To construct a predictive model for postoperative hypoxaemia in general anesthesia patients in the post - anesthesia care unit（PACU）using LASSO - logistic regression and to validate the efficacy. Methods The clinical data of 100 general anesthesia patients（modeling group）who were transferred to the PACU for observation after surgery in Zhongshan Hospital Affiliated to Xiamen University from December 2020 to December 2022 were retrospectively analyzed. According to whether hypoxaemia occurred in PACU，patients were divided into occurrence group and non-occurrence group. The clinical data of the two groups were collected，and LASSO regression were carried out for preliminary screening. Then the predictive variables were determined by multivariate logistic regression analysis to establish a prediction model of hypoxaemia. In addition，30 patients transferred to PACU undergoing general anesthesia after surgery in Zhongshan Hospital Affiliated to Xiamen University from January 2024 to June 2024 were prospectively selected as the verification group，and the predictive efficacy of the model was validated by using receiver operating characteristic（ROC）curves，calibration curves and decision curves. Results Among 100 patients，21patients developed hypoxemia. Multivariate logistic regression analysis showed that age ≥70 years，preoperative peripheral, capillary oxygen saturation（SpO2）< 95%，thoracic surgery，and operation time ≥120 min were independent risk factors for hypoxaemia in PACU（P<0.05）. The C-index of the nomogram prediction model was 0.811，and the area under the ROC curve was 0.833（95%CI：0.758-0.892）. The calibration curve was close to the ideal curve，and the decision curve showed that the model had high predictive net benefit. Conclusion Hypoxaemia in PACU is closely related to age，and preoperative SpO2，surgical site，operation time，and the nomogram model based on this has good prediction efficiency.