Key result
Machine learning models predict HF hospitalization in UA with HFpEF with ~0.84 AUC.
Why the study?
Does a multi-horizon machine learning prediction framework improve risk stratification for heart failure hospitalization at 12, 24, and 36 months compared to the MAGGIC score in patients with unstable angina and HFpEF?
Cohort (n=4,367)
Yes
Does a multi-horizon machine learning prediction framework improve risk stratification for heart failure hospitalization at 12, 24, and 36 months compared to the MAGGIC score in patients with unstable angina and HFpEF?
A machine learning framework integrating nine routine clinical and cardiometabolic variables accurately predicts 12-, 24-, and 36-month heart failure hospitalization risk in patients with unstable angina and HFpEF, substantially outperforming the conventional MAGGIC risk score.
Answers are grounded in this paper and cite their sources.
May support multi-horizon ML models for HF hospitalization risk stratification in UA with HFpEF; leaves open prospective validation versus MAGGIC before clinical adoption.
Heart failure with preserved ejection fraction (HFpEF) complicated by unstable angina (UA) represents a major challenge in chronic disease management with high risk of recurrent heart failure hospitalization (HFH), and despite the availability of multicenter real-world data, AI-driven tools for multi-horizon risk assessment remain limited. This multicenter retrospective cohort enrolled 4367 patients with UA and HFpEF, and an ensemble feature selection pipeline integrating LASSO, SVM-RFE and Boruta identified robust predictors, followed by training of twelve machine learning algorithms to estimate HFH risk at 12-, 24- and 36-month horizons with external validation. The optimal time-specific models (ET, GBDT and LightGBM) achieved strong discrimination in the training cohort with AUCs of 0.818, 0.834 and 0.863, and maintained performance in external validation with AUCs of 0.839, 0.836 and 0.851, respectively, while SHAP analysis highlighted diabetes mellitus, TyG-BMI and NT-proBNP as the most influential predictors. We developed and externally validated a multi-time-point machine learning framework for HFH risk stratification in UA and HFpEF, which showed acceptable calibration and potential clinical utility on decision-curve analysis. Further prospective and real-world validation across diverse populations and healthcare systems is needed to confirm clinical effectiveness and support translation into routine practice.
No takes yet. Share an insight, caveat, or question.
Huang et al. (2026) conducted a cohort in Heart failure with preserved ejection fraction (HFpEF) complicated by unstable angina (UA) (n=4,367). Multi-horizon machine learning risk stratification framework (ET, GBDT, and LightGBM models) vs. MAGGIC score was evaluated on Heart failure hospitalization (HFH) at 12, 24, and 36 months (AUC 0.818 (12m), 0.834 (24m), 0.863 (36m) in training; AUC 0.839 (12m), 0.836 (24m), 0.851 (36m) in external validation, 95% CI 0.845-0.880). Time-specific machine learning models (ET, GBDT, LightGBM) accurately predicted 12-, 24-, and 36-month heart failure hospitalization in UA with HFpEF (external validation AUCs 0.839, 0.836, and 0.851).
Synapse has enriched 10 related papers. Consider them for comparative context: