Abstract
Objective: To evaluate whether the admission hemoglobin-albumin-lymphocyte-platelet (HALP) score was independently associated with all-cause in-hospital mortality and whether it provided incremental prognostic information beyond a clinical model including the National Institutes of Health Stroke Scale (NIHSS) in adults with acute ischemic stroke (AIS).Methods: This retrospective single-center cohort included consecutive eligible adults with AIS treated in the emergency department of a tertiary-care referral hospital in Türkiye between January 2020 and December 2025. The primary outcome was all-cause in-hospital mortality. HALP was modeled continuously. Multivariable logistic regression adjusted for age, sex, NIHSS, atrial fibrillation, and intravenous thrombolysis. Incremental performance was assessed using area under the receiver operating characteristic curve (AUC), DeLong testing, calibration, bootstrap optimism correction, 10-fold cross-validation, and sensitivity analyses.Results: Among 511 patients, 81 (15.9%) died in hospital. Median HALP was lower in non-survivors than survivors (15.78 vs 32.27; p<0.001). Each 5-point decrease in HALP was independently associated with higher mortality odds (adjusted odds ratio 1.69, 95% confidence interval 1.41-2.03; p<0.001). Adding HALP increased AUC from 0.867 to 0.908 (ΔAUC 0.040, 95% confidence interval 0.010-0.071; p=0.010). The extended model retained an AUC of 0.898 after bootstrap correction and 10-fold cross-validation, and HALP remained significant across sensitivity analyses.Conclusion: Lower admission HALP was independently associated with in-hospital mortality and provided modest incremental prognostic information beyond NIHSS-based clinical assessment. External validation in prospective multicenter cohorts with standardized 90-day mortality and functional outcomes is required before clinical implementation.