Abstract
<title>Abstract</title> <p>The prognostic nutritional index (PNI), derived from serum albumin and absolute lymphocyte count, is widely used for risk stratification in heart failure, but whether its lymphocyte component adds information beyond albumin remains unclear. We decomposed the prognostic contribution of PNI in a single-centre cohort of hospitalised chronic heart failure patients outside critical care (n = 3,643; 1,591 deaths; median follow-up 22.4 months) and an external MIMIC-IV critical-care cohort (n = 1,972; 722 deaths within 365 days). Cox models compared albumin, lymphocyte count and PNI per 1-standard-deviation decrease, with model fit and discrimination assessed by AIC/BIC, Harrell's C-index and paired bootstrap delta C-index. In the primary cohort, albumin was independently associated with mortality (HR 1.12, 95% CI 1.05–1.20), whereas lymphocyte count was not after NT-proBNP adjustment; PNI remained modestly associated but did not improve discrimination over albumin. In MIMIC-IV, both albumin and lymphocyte count were associated with 365-day mortality, and the lymphocyte association persisted in the NT-proBNP-measured subset. Across all settings, PNI did not improve C-index over albumin. These findings suggest that PNI has context-dependent composition and should not be framed as superior to serum albumin for mortality discrimination.</p>