Comparison on predictive efficacy of two models for MACE in elderly patients with coronary artery calcification
Li Chuanbo
Li Xiding
Ji Miaomiao
Wang Yuekun
Abstract：Objective To compare the efficacy of multivariate logistic regression and XGBoost models in predicting major adverse cardiovascular events(MACE)after percutaneous coronary in-tervention(PCI)in elderly patients with coronary artery calcification(CAC).Methods A total of 120 elderly patients with CAC lesions undergoing PCI in our hospital from June 2020 to June 2023 were retrospectively enrolled in this study.The incidence of MACE was observed during 1 year of follow-up.Nine patients were lost during the period,and the left patients were divided into MACE group(28 patients)and non-MACE group(83 patients).Multivariate logistic regression analysis and XGBoost model were used to screen the influencing factors of MACE in elderly CAC patients after PCI.ROC curve and calibration curve were drawn to compare the predictive efficiency of the two models.Results The MACE group had significantly advanced age,larger proportions of smoking and diabetes,higher LDL-C and Gensini score,and increased ratios of diseased vessels ≥3,severe calcification,combined rotary grinding and number of stent implantation when compared with the non-MACE group(P＜0.05,P＜0.01).Multivariate logistic regression model showed that smoking,diabetes,LDL-C,Gensini score,and number of stents implanted were independent risk factors for MACE in CAC patients after PCI(P＜0.05,P＜0.01).XGBoost model indicated that the top five important feature scores were Gensini score of 35,number of stent implantation score of 25,combined diabetes score of 22,smoking score of 18,and LDL-C score of 15.ROC curve analysis revealed that the AUC value of multivariate logistic regression model in predicting MACE in elderly CAC patients after PCI was 0.925(95％CI:0.859-0.966),with a sensitivity of 82.14％and a specificity of 97.59％,and the value of the XGBoost model was 0.918(95％CI:0.850-0.961),with a sensitivity of 89.29％and a specificity of 78.31％.There was no significant difference in predictive efficacy between the two models(Z=0.148,P=0.8823).Conclusion Multiple logistic regression model and XGBoost model show equally efficacy in predicting MACE in elderly CAC patients after PCI.Smoking,diabetes,LDL-C,Gensini score and number of stents implanted are independent risk factors for MACE in the patients.
Keywords：risk factorsmajor adverse cardiovascular eventscoronary artery calcification
Publication Date：2025-01-14
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：5( 48-52 )