HFpEF-ABA Score Can Predict Long-Term Mortality Across All Heart Failure Categories in Acute Heart Failure

HFpEF-ABA score, based on age, body mass index, and history of atrial fibrillation, was initially developed as a diagnostic screening tool for heart failure with preserved ejection fraction (HFpEF). This study evaluated the prognostic significance of HFpEF-ABA score for all-cause mortality across HF phenotypes in patients hospitalized for acute HF (AHF). We analyzed 3,974 admitted patients in the STRATS-AHF registry (Korea, 2009-2016), classified into HFrEF (LVEF ≤40%; n=2,094), HFmrEF (LVEF 41-49%; n=609), and HFpEF (LVEF ≥50%; n=1,271). The primary endpoint was all-cause mortality. Over a median follow-up of 51.4 months, 1,642 patients (41.3%) died. HFpEF-ABA score was significantly higher in nonsurvivors than survivors (61.0±23.2 vs. 54.6±25.9; P<0.001), with a univariate hazard ratio (HR) of 1.009 per point (95% CI: 1.007-1.011; P<0.001). Restricted cubic spline analysis revealed a monotonically increasing, non-linear association between the score and mortality risk, and patients with a score <45 had much better 6-year survival compared to those with a score ≥45 in the overall group and in each HF phenotype (all P≤0.005). Adding HFpEF-ABA score to clinical, pharmacological, and echocardiographic variables produced a modest but significant improvement in individual risk reclassification (net reclassification improvement 0.198; integrated discrimination improvement 0.013), despite only a marginal increase in discrimination (C-index from 0.696 to 0.700, P<0.001), with an independent HR of 1.008 per point (95% CI: 1.005-1.012; P<0.001) and preserved calibration. HFpEF-ABA score independently predicted long-term all-cause mortality across HF phenotypes in AHF patients, offering modest incremental prognostic value beyond established clinical, echocardiographic, and therapeutic predictors.