RESUMEN
BACKGROUND: Stroke care across Australian hospitals is variable. The impact on health outcomes, in particular levels of disability for patients in rural areas, is unclear. The aim of this study was to determine whether geographic location and access to stroke units are associated with differences in health outcomes in patients with acute stroke. METHODS: Retrospective cohort study of consecutive eligible admissions from 32 hospitals (12 rural) in New South Wales between 2003 and 2007. Health status measured at discharge included level of independence (modified Rankin score: mRS) and frequency of severe complications during hospitalization. Multivariable analyses included adjustment for patient casemix and clustering. RESULTS: Among 2254 eligible patients, 55% were treated in metropolitan hospitals. Stroke unit treatment varied significantly (rural 3%; metropolitan 77%). Age, gender and stroke type did not differ by location (mean age 74, 50% female). After adjusting for age, gender, ethnicity, important risk factors and validated stroke prognostic variables, patients treated in rural hospitals had a greater odds of dying during hospitalization compared with those treated in metropolitan hospitals (adjusted odds ratio (aOR) 1.46, 95% confidence interval (CI) 1.03-2.05). There were no differences in mortality or frequency of severe complications between patients treated in rural and metropolitan hospitals when we adjusted for access to stroke units (aOR 1.00, 95% CI 0.62-1.61). Nevertheless, patients treated in rural hospitals were more dependent (mRS 3-5) at discharge (aOR 1.82, 95% CI 1.23-2.70) despite adjusting for stroke unit status. CONCLUSION: Patients with stroke treated in rural hospitals have poorer health outcomes, especially if not managed in stroke units.