Objective:To construct a predictive model for postoperative cognitive dysfunction(POCD)in video-assisted thoracoscopic surgery(VATS)and improve the relevant assessment system.Methods:A cross-sectional survey was conducted to build a specialized optimized predictive model.The predic-tive model was transformed into a visualized optimized scoring scale to quantify the probability of VATS-related POCD.The efficacy of the optimized scoring in assessing VATS-related POCD was explored in comparison with the current predictive model.Results:① After sample size calculation and comprehensive inclusion-exclusion criteria,202 consecutive VATS patients were selected as the model group for this cross-sectional study,with 22 cases of postoperative POCD and 180 cases of nor-mal cognitive function.② Inter-group comparison of demographic features revealed a significant in-crease in BMI values,with statistically significant differences(P<0.05).Clinical data comparisons between groups showed significant differences in anesthesia duration,postoperative intubation time,postoperative to ambulation time,length of hospital stay,NSE,TNF-α,and TNM staging composi-tion(P<0.05).③ The above indicators were sequentially included in univariate and multivariate Pois-son regression analysis,and anesthesia duration,postoperative intubation time,length of hospital stay,NSE,and TNF-α were identified as independent influencing factors,all of which were risk fac-tors.The predictive model was transformed into a visualized optimized scoring scale in the form of a column chart,revealing the corresponding predicted probabilities for each score.④ The ROC curve was used to test the efficacy of the optimized scoring compared to the reference scoring in assessing VATS-related POCD.Patient data for the validation group were extracted based on both scoring sys-tems,and the final scores for each patient in the validation group were obtained.ROC curves were ap-plied to evaluate the predictive ability of different models.The cutoff values for optimized scoring and reference scoring were 172 points and 219 points,respectively.The AUC for optimized scoring was 0.811,and the Youden index was 0.52,both higher than the reference scoring.The efficacy of opti-mized scoring in assessing POCD was significantly better than that of reference scoring(Z=2.871,P<0.05).Conclusion:This study constructs a specialized predictive model for postoperative cogni-tive dysfunction in VATS patients and transforms it into an easily applicable optimized scoring system with reliable assessment efficacy.
NETL
NSTL
万方数据
