Introduction: End Stage Renal Disease (ESRD) is associated with accelerated vascular disease of the cerebral circulation. Several studies have established that patients with ESRD on hemodialysis (HD) are at much higher risk of ischemic stroke than general population.
Objective: To compare the outcomes of in-hospital stroke vs out of hospital stroke in maintenance hemodialysis and identify predictors of mortality.
Methods: Patients >18 years admitted between 2005-2014 were screened using ICD-CM-9 codes 39.9 for Hemodialysis and 430-438 for stroke. Of the 620 patients with stroke and HD, 135 patients who had “in-hospital” acute ischemic stroke were identified. Patient demographics, comorbidities, type of stroke, and outcomes including discharge disposition and mortality were compared between the two groups. Logistics regression analysis was performed to identify independent predictors of mortality in these patients.
Results: Rate of in hospital stroke was 28.8 % (135/468) and overall prevalence rate of infection was 47.3% (284/600) in overall cohort. Compared to out of hospital stroke, patients with in-hospital stroke had higher rate of atrial fibrillation (30.6 % vs 19.9%, p<0.05), anemia (74.0% vs 60.3%, p<0.05) and infection (69.7% vs 40.0%, p<0.001). They had higher mortality (35.5% vs 15.9 %, p<0.001) and longer length of stay (21.28±18.85 days vs 11.80±29.9 days, p<0.05). Using logistic regression, infection (OR 2.75, 95% CI 1.61- 4.09) anemia (OR 2.00, 95% CI 1.105-3.635) and carotid disease (OR 2.71 CI 1.11-6.63) were found to be independent predictor of in-hospital stroke. Similarly infection was found to be an independent predictor of mortality (OR 6.66, 95% CI 3.44-12.877).
Conclusion: There is a high incidence of “in-hospital” stroke in patients with ESRD on HD. Certain clinical parameters such as infections and anemia significantly impact the outcomes in these patients. Further studies should be directed looking at these parameters to establish guidelines for better in-patient management of these patients.