Background
Acute myocardial infarction (AMI) with consequent heart failure is one of the leading causes of death in humans. The aim of this study was to develop a prediction model to identify heart failure risk in patients with AMI during hospitalization.
Methods
The data on hospitalized patients with AMI were retrospectively collected and divided randomly into modeling and validation groups at a ratio of 7:3. In the modeling group, the independent risk factors for heart failure during hospitalization were obtained to establish a logistic prediction model, and a nomogram was constructed. The receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis (DCA) were used to evaluate the predictive performance and clinical value. Machine learning models with stacking method were also constructed and compared to logistic model.
Results
A total of 1875 patients with AMI were enrolled in this study, with a heart failure rate of 5.1% during hospitalization. The independent risk factors for heart failure were age, heart rate, systolic blood pressure, troponin T, left ventricular ejection fraction and pro-brain natriuretic peptide levels. The area under the curve (AUC) of the model in modeling group and validation group were 0.829 and 0.846, respectively. The calibration curve showed high prediction accuracy and the DCA curve showed good clinical value. The AUC value of the ensemble model by the stacking method in the validation group were 0.821, comparable to logistic prediction model.
Conclusions
This model, combining laboratory and clinical factors, has good efficacy in predicting heart failure during hospitalization in AMI patients.