Establishment and validation of a prediction model for nonrecovery of left ventricular ejection fraction in acute myocardial infarction patients combined with decreased left ventricular ejection fraction.

ABSTRACT

BACKGROUND: This study aimed to investigate the risk factors for nonrecovery of left ventricular ejection fraction (LVEF) during follow-up in patients with acute myocardial infarction (AMI) who underwent percutaneous coronary intervention (PCI) combined with reduced LVEF, and establish and verify a risk prediction model based on these factors. METHODS: In this study, patients with AMI who underwent PCI in a high-volume PCI center between December 2018 and December 2021 were consecutively enrolled, screened, and randomly assigned to the model establishment and validation cohorts. A predictive model method based on least absolute shrinkage and selection operator regression was used for establishment and validation. RESULTS: Cardiac troponin I, myoglobin, left ventricular end-diastolic dimension, multivessel disease, and no-reflow were identified as potential predictors of LVEF recovery failure. The areas under the curve were 0.703 and 0.665 in the model establishment and validation cohorts, respectively, proving that the prediction model had some predictive ability. The calibration curves of the two cohorts showed good agreement with those of the nomogram model. In addition, the decision curve analysis showed that the model had a net clinical benefit. CONCLUSION: This prediction model can assess the risk of nonrecovery of LVEF in patients with AMI undergoing PCI combined with LVEF reduction during follow-up, and conveniently screen high-risk patients with nonrecoverable LVEF early.