A prognostic model including age, stroke history, NYHA class, hemoglobin, eGFR, and PPCI predicted all-cause mortality with C-index 0.795 and 0.741 in training and validation cohorts, respectively, over a 33-month follow-up.
Observational (n=873)
None
Randomized allocation to training and validation cohorts
No
A novel 6-variable prognostic model incorporating age, stroke history, NYHA class, hemoglobin, eGFR, and PPCI accurately predicts post-discharge all-cause mortality in patients with HFmrEF following acute myocardial infarction.
Effect estimate: C-index 0.795 training cohort; 0.741 validation cohort; AUC 0.861 (6 months), 0.805 (2 years), 0.815 (3 years) training; AUC 0.722 (6 months), 0.742 (2 years), 0.736 (3 years) validation (95% CI C-index training cohort 0.758-0.832; validation cohort 0.672-0.81)
Absolute Event Rate: 20.9% vs 21%
p-value: p=Not applicable
Background: Accurately assessing mortality risk in patients with heart failure with mildly reduced ejection fraction (HFmrEF) after acute myocardial infarction (AMI) remains challenging. This study developed and validated a mortality risk predictive model for such patients. Methods: In this single-center retrospective study of 873 hospitalized patients with HFmrEF after AMI, 611 patients were included in the training cohort and 262 in the validation cohort. The primary outcome was all-cause mortality over an average 33-month follow-up. Least absolute shrinkage and selection operator (LASSO) regression identified predictive variables for post-discharge mortality, with model performance assessed via receiver operating characteristic (ROC) analysis and decision curve analysis (DCA). Results: Six mortality risk predictors were identified: age, stroke history, New York Heart Association (NYHA) classification, hemoglobin (Hb) levels, estimated glomerular filtration rate (eGFR), and primary percutaneous coronary intervention (PPCI) implementation. The C-index for training and validation cohorts was 0.795 (95% confidence interval (CI), 0.758–0.832) and 0.741 (95% CI, 0.672–0.81), respectively. Training cohort area under the curve (AUC) metrics for 6-month, 2-year, and 3-year survival were 0.861, 0.805, and 0.815; for the validation cohort, they were 0.722, 0.742, and 0.736. Conclusions: A validated predictive model assessing mortality risk in HFmrEF patients post-AMI was established. External validation in future studies is recommended.
Liu et al. (Sun,) conducted a observational in Hospitalized patients with heart failure with mildly reduced ejection fraction (LVEF 41-49%) after acute myocardial infarction (n=873). Prognostic model based on clinical variables vs. None was evaluated on All-cause mortality post-discharge (C-index 0.795 training cohort; 0.741 validation cohort; AUC 0.861 (6 months), 0.805 (2 years), 0.815 (3 years) training; AUC 0.722 (6 months), 0.742 (2 years), 0.736 (3 years) validation, 95% CI C-index training cohort 0.758-0.832; validation cohort 0.672-0.81, p=Not applicable). A prognostic model including age, stroke history, NYHA class, hemoglobin, eGFR, and PPCI predicted all-cause mortality with C-index 0.795 and 0.741 in training and validation cohorts, respectively, over a 33-month follow-up.