Background — Better outcomes have been found among hospitals treating higher volumes of patients for specific surgical and medical conditions. We examined hospital ischemic stroke (IS) volume and 30-day mortality to inform regionalization planning.
Methods — Using a population-based hospital discharge administrative database (2005/06 to 2011/12) average annual IS patient volumes were calculated for 162 Ontario acute hospitals. Hospitals were ranked and classified as small (<126), medium (126 – 202) and large (>202). Hierarchical multivariable logistic regression was used to estimate the odds of death within seven and 30 days to account for the homogeneity in outcomes for patients treated at the same hospital.
Results — Overall, 73,368 patients were hospitalized for IS and 30-day mortality was 15.3%. The mean (+/- SD) of annual hospitalizations for IS was 29 (31) for small hospitals, 156(20) for medium hospitals and 300(78) for high volume hospitals. High volume hospitals admitted younger patients, mean (+/- SD) age 73.0 (13.9) years compared to medium and small hospitals, 74.0 (13.2) and 75.5 (12.5) years respectively (p <0.0001). Patients at low volume hospitals were more likely than patients at high volume hospitals to die at 30-days following an acute IS (adjusted OR 1.37; 95% CI 1.14 to 1.65).
Conclusions — Hospital IS volume is associated with 30-day mortality in Ontario. Patients admitted to hospitals with annual IS volumes less than 126 annually are more likely to die within 30-days than patients admitted to hospitals that see on average 300 patients annually. This finding supports centralizing care in stroke specialized hospitals.
View full text
Health care evaluation