AI Summary & Key Takeaways
AI InsightIts low cost and routine availability support further evaluation as a complementary prognostic marker; however, external prospective validation and evidence of clinical utility are required before its incorporation into risk-guided management strategies.
📌 3 Key Takeaways
Abstract
Aims: This study evaluated the prognostic value of several novel laboratory-derived indices: red blood cell distribution width (RDW), blood urea nitrogen-to-albumin ratio (BAR), Systemic Immune-Inflammation Index (SII), Model for End-Stage Liver Disease excluding the international normalized ratio (MELD-XI), uric acid-to-albumin ratio (UAR), and estimated plasma volume status (ePVS) in patients with heart failure with reduced ejection fraction (HFrEF). Their incremental contributions beyond conventional risk factors were compared. Methods: This retrospective single-center study included 1.314 patients with HFrEF. RDW, BAR, SII, MELD-XI, UAR, and ePVS were each added separately to a fixed reference model including age, left ventricular ejection fraction (LVEF), log10 transformed N-terminal pro-B-type natriuretic peptide (NT-proBNP), estimated glomerular filtration rate (eGFR), New York Heart Association (NYHA) functional class, angiotensin-converting enzyme inhibitor (ACEI)/angiotensin receptor blocker (ARB) or angiotensin receptor-neprilysin inhibitor (ARNI) use, beta-blocker use, mineralocorticoid receptor antagonist (MRA) use, and loop diuretic use. Age and LVEF were modeled with restricted cubic splines. Prognostic performance was assessed using Harrell’s concordance index (C-index), continuous net reclassification improvement (NRI), and integrated discrimination improvement (IDI), with bootstrap internal validation and 3-year calibration. Results: During follow-up, 428 patients (32.6%) died from all causes. Among the evaluated biomarkers, addition of log10transformed BAR produced the largest improvement in discrimination, increasing the C-index from 0.7489 (95% CI: 0.7256-0.7725) to 0.7585 (95% CI: 0.7355-0.7816), with a paired ΔC-index of 0.0096 (95% CI: 0.0021-0.0168). Bootstrap internal validation showed that BAR was the most frequently selected biomarker, although the optimism-corrected incremental improvement remained modest. Kaplan-Meier analysis demonstrated progressively worse survival across increasing L_BAR quartiles (log-rank p<0.001). Conclusion: In this cohort, BAR showed the largest incremental improvement in discrimination among the evaluated laboratory derived biomarkers, although the absolute gain was modest. Its low cost and routine availability support further evaluation as a complementary prognostic marker; however, external prospective validation and evidence of clinical utility are required before its incorporation into risk-guided management strategies.