Association of Baseline Leukocyte Count with 90-Day Mortality in Incident Hemodialysis Patients

Abstract

Background: Early mortality remains high among incident hemodialysis (HD) patients. Routine laboratory parameters may provide prognostic information. Objective: To investigate associations between baseline laboratory parameters and 90-day mortality and assess the discriminatory ability of leukocyte count. Methods: This retrospective cohort study included adults initiating HD at Grandmed Hospital, Lubuk Pakam, during October–December 2025. Baseline measurements included leukocyte count, neutrophil-to-lymphocyte ratio (NLR), hemoglobin, sodium, potassium, urea, and creatinine. Mortality was assessed over 90 days. Analyses included bivariate tests, logistic regression, receiver operating characteristic (ROC) analysis, and crude relative risk (RR) estimation. Results: Among 71 patients, 26 (36.6%) died within 90 days. Median leukocyte count was higher in nonsurvivors than survivors (14.17 vs. 9.24 × 10³/µL; p = 0.024), while other parameters showed no significant differences. Leukocyte count remained associated with mortality after adjustment for potassium (OR 1.088; 95% CI 1.013–1.169; p = 0.020). ROC analysis yielded an AUC of 0.662, with a cutoff of 10.6 × 10³/µL (sensitivity 73.1%; specificity 57.8%). Patients above this cutoff had higher mortality (50.0% vs. 21.2%; crude RR 2.36; 95% CI 1.14–4.90; p = 0.012). Conclusion: Higher baseline leukocyte counts were associated with increased 90-day mortality, but limited discrimination precludes their use as a standalone prognostic biomarker.