Construction and validation of a predictive model for the risk of postoperative adverse cardiovascular events in arthroplasty based on CCTA examination

SHAO Linquan; LYU Meng; WANG Yuelan

doi:10.3969/j.issn.1002-266X.2025.05.015

Abstract：Objective To construct and verify a prediction model for the risk of postoperative adverse cardiovascular events in patients undergoing arthroplasty based on coronary computed tomography angiography(CCTA).Methods To-tally 799 patients who underwent knee/hip arthroplasty surgery were retrospectively collected.The composite endpoint events were used to determine postoperative adverse cardiovascular events.The Coronary Artery Calcification Score(CACS)and the number of coronary artery occlusions were determined according to the results of CCTA.Clinical indica-tors related to adverse cardiovascular events after arthroplasty were collected,including demographic information,anamne-sis,preoperative laboratory test parameters,preoperative cardio-pulmonary function parameters,intraoperative parame-ters,and preoperative CCTA examination-related parameters.Least Absolute Shrinkage and Selection Operator(LASSO)regression analysis was used to screen the influencing factors of adverse cardiovascular events after arthroplasty.Patients were randomly divided into the training set and validation set according to the ratio of 7∶3.In the training set,the influenc-ing factors screened by LASSO regression analysis were further subjected to multivariate Logistic regression analysis to draw a nomogram risk prediction model for adverse cardiovascular events after arthroplasty.In the validation set,Hosmer-Lemeshow test was used to evaluate the goodness-of-fit of the model,the calibration curve was used to evaluate the calibra-tion,and the decision curve was used to evaluate the clinical benefit.The discrimination of the model was evaluated by drawing the receiver operating characteristic(ROC)curve and calculating the area under the curve(AUC).Results A total of 80 of 799 patients undergoing arthroplasty experienced postoperative adverse cardiovascular events,with an inci-dence of 10.0%.LASSO regression analysis showed that age,American Society of Anesthesiologists(ASA)classifica-tion,surgical site,blood glucose,albumin,D-dimer,left ventricular ejection fraction(LVEF)and the number of coro-nary artery obstructions were the influencing factors for the postoperative adverse cardiovascular events.The patients un-dergoing arthroplasty were randomly divided into the training set(n=559)and validation set(n=240)according to the ratio of 7∶3.The influencing factors of the training set and the validation set were comparable.Multivariate Logistic regres-sion analysis identified surgical site,ASA classification,blood glucose,and coronary artery obstructions as independent risk factors for postoperative cardiovascular events.According to the independent influencing factors screened by multivari-ate Logistic regression analysis,a nomogram prediction model for adverse cardiovascular events after arthroplasty was con-structed.The results of calibration curve analysis showed that the calibration curve basically fitted the ideal curve.The Hosmer-Lemeshow test revealed no significant difference between the nomogram model and the ideal model(P=0.831).The decision curve analysis revealed that the threshold of the nomogram prediction model ranged from 0.01 to 0.75.ROC curve analysis showed that the AUC of the nomogram prediction model in predicting adverse cardiovascular events after ar-throplasty was 0.76(95%CI:0.68-0.83)for the training set and 0.76(95%CI:0.66-0.86)for the validation set.Conclusions The number of coronary artery obstructions,surgical site,ASA classification and blood glucose are inde-pendent influencing factors for postoperative adverse cardiovascular events in patients undergoing arthroplasty.The predic-tion model for postoperative adverse cardiovascular events in patients undergoing arthroplasty based on the above influenc-ing factors has good discrimination,calibration,goodness of fit and clinical practicality.