Performance of the PREVENT-HF Equations Across Heart Failure Subtypes in Community-Based and Ambulatory Healthcare Samples.

PREVENT-HF predicts 10-year incident HF with preserved ejection fraction better than HFrEF, with AUROC 0.824 vs 0.789 in pooled community-based cohorts.

• Why it matters: Subtype-specific HF risk assessment is needed because HFpEF and HFrEF differ in risk factors and care, yet PREVENT-HF was built for overall HF risk.
• What they did: Adults aged 30–79 without prevalent cardiovascular disease or HF were followed for 10 years in five pooled longitudinal cohorts (n=23,652) and two EHR samples (BIDMC, MGB), evaluating discrimination (AUROC, AP/event rate) and calibration (ICI).
• The result: Discrimination favored HFpEF across pooled cohorts and both EHR datasets, while calibration was similar between subtypes (pooled ICI 0.041 for HFpEF vs 0.038 for HFrEF), supporting subtype-agnostic use and motivating subtype-specific thresholds.

AI summary of the abstract · may contain errors · not medical advice · read the paper.