Title : Exploring the role of the red cell distribution width in predicting clinical outcomes among adult patients admitted for stroke in a resource-limited setting: A 3-year retrospective single-center study
Abstract:
Background: Stroke remains a leading cause of mortality and disability worldwide. Early identification of patients at risk for poor clinical outcomes is essential for timely intervention and improved management. Red cell distribution width (RDW), a routinely reported parameter complete blood count parameter, has emerged as a potential biomarker of systemic inflammation and adverse cardiovascular outcomes. This study aimed to determine the role of RDW in predicting clinical outcomes among adult patients admitted for stroke.
Methods: This retrospecEve study included adults paEent admi\ed with stroke from January 2023 through December 2025. Medical records were reviewed for age, sex, smoking history, alcohol intake, comorbidiEes, stroke type and chronicity, admission Glasgow Coma Scale score admission RDW, and clinical outcomes. LogisEc regression was used to assess RDW as a predictor of 7-day mortality, while Cox proporEonal hazards regression evaluated in-hospital mortality and nosocomial infecEon. Kaplan–Meier survival analysis and linear regression were used for survival and length of stay, respecEvely. SensiEvity, specificity, and area under the curve were calculated to assess the diagnosEc performance of RDW.
Results: A total of 115 participants were included in the study. The mean RDW was 13.36 ± 1.03%, and this was higher among those who expired(p-value 0.018). Among the participants, 6.09% died within 7 days of admission, and the overall mortality rate was 7.83%. Nosocomial infections occurred in 57.39%, most commonly pneumonia. The mean infection-free duration was 4.52 ± 3.70 days, and the mean hospital stay was 11.50 ± 9.05 days. The RDW was significantly associated with 7-day mortality (adjust odds ratio 2.44, 95% confidence interval 1.13 – 5.28, p=0.024) and in-hospital mortality (adjusted hazards ratio 3.08, 95% confidence interval 1.20 – 7.78, p=0.019). It demonstrated high specificity, high negative predictive value, and acceptable discrimination for both outcomes, with optimal cutoffs of 14.24 for 7-day mortality and 14.15 for in-hospital mortality.
Conclusion: Elevated RDW is an inexpensive and readily available biomarker that independently predicts early and in hospital mortality among patients with stroke. Incorporating RDW into the initial assessment may improve risk stratification, enabling closer monitoring, timely interventions, and more efficient allocation of healthcare resources for high-risk patients, ultimately contributing to improved clinical outcomes.
