Construction and validation of a predictive model for major adverse cardiovascular events in the long term after percutaneous coronary intervention in patients with acute ST-segment elevation myocardial infarction

Abstract

Purpose Construction of a prediction model to predict the risk of major adverse cardiovascular events (MACE) in the long term after percutaneous coronary intervention (PCI) in patients with acute ST-segment elevation myocardial infarction (STEMI). Method Retrospective analysis of STEMI patients treated with PCI from April 2018 to April 2021 in Fuyang People’s Hospital. Lasso regression was used to screen the risk factors for the first occurrence of MACE in patients, and multifactorial logistic regression analysis was used to construct a prediction model. The efficacy was evaluated by area under the ROC curve (AUC), Hosmer–Lemeshow deviance test, calibration curve, clinical decision curve (DCA) and clinical impact curve (CIC). Results Logistic regression results showed that hypertension, diabetes mellitus, left main plus three branches lesion, estimated glomerular filtration rate and medication adherence were influential factors in the occurrence of distant MACE after PCI in STEMI patients (P < 0.05). The AUC was 0.849 in the modeling group and 0.724 in the validation group; the calibration curve had a good fit to the standard curve, and the result of the Hosmer–Lemeshow test of deviance was x2 = 7.742 (P = 0. 459); the DCA and the CIC indicated that the predictive model could provide a better net clinical benefit for STEMI patients. Conclusion A prediction model constructed from a total of five predictor variables, namely hypertension, diabetes, left main + three branches lesions, eGFR and medication adherence, can be used to assess the long-term prognosis after PCI in STEMI patients and help in early risk stratification of patients.