Development and Validation of a Risk Prediction Score for Patients with ST-Elevation Myocardial Infarction and Multivessel Disease

Abstract

Background: Patients with ST-elevation myocardial infarction (STEMI) and multivessel disease (MVD) are at a significantly higher risk of subsequent cardiovascular death (CV death) and recurrent myocardial infarction (MI). Although complete revascularization improves outcomes, baseline risk may differ significantly according to individual differences in clinical and anatomical characteristics. Currently available risk prediction tools were not optimized for this population. The goal of this study was to synthesize and internally validate a basic, point-based risk score that would predict 3-year cardiovascular events and guide treatment decisions in this high-risk population. Methods: Based on data from the COMPLETE trial (n=3,738 participants), we first screened potential clinical and angiographic variables for inclusion in the score using univariate. Key predictors of cardiovascular death or myocardial infarction over 3 years of follow-up were identified using a multivariate Cox proportional hazards model. Regression coefficients (β-coefficients) were converted into integer values. Model performance was evaluated by Harrell’s C-statistic for discrimination and by calibration plots to assess model fit. Results We identified the following six predictors for CV death or MI: age (≥65), sex, diabetes, chronic kidney disease (CKD), Killip class ≥2, and percentage diameter stenosis ≥80% of the non-culprit lesion (NCL) by visual estimation. The total score ranged from 0 to 30. The model demonstrated average discriminant performance, with a C-statistic of 0.60 when applied to the full cohort. Conclusions: We derived and internally validated a point-based risk score to risk stratify STEMI patients with MVD. However, further refinements to the model, including the use of a larger development cohort and external validation, are necessary to improve its performance and confirm its generalizability across a broader population of patients with STEMI and MVD.