[Factors related to in-hospital mortality after stroke in Lazio region, Italy].

: The use of hospital discharge abstracts in estimating the outcome of hospital care represents an ongoing interest in public health. However standardized methodologies are still not available. We carried out a retrospective study to estimate the association between demographic and clinical characteristics and in-hospital mortality after stroke by using administrative data from the Hospital Information System in Lazio Region. We also assessed the relationship between the presence of neurology services and the outcome. We found 12,781 incident episodes of stroke (main diagnosis ICD-9: 430-431-434-436) (49.3% male, mean age = 74 years) admitted in 126 hospitals in the Lazio region for the period 1999-2000. From the hospital discharge abstracts we collected patient demographic and clinical data. The hospitals were classified in centres with and without neurology services. Admissions to hospitals with neurology services were evaluated as predictors of in-hospital mortality after adjustment for gender, age, residence, education, source of admission, type of stroke, heart disease, kidney disease and history of atrial fibrillation. In-hospital mortality (within 30 day) was 25.1%. Female gender, advanced age, residence in Rome, urgent transport, kidney disease and history of atrial fibrillation were associated with an increased risk. Hemorrhagic stroke (ICD-9 = 430-431) had a worse outcome than ischemic stroke (ICD-9 = 434) and acute undefined cerebrovascular disease (ICD-9 = 436). Patients admitted to hospitals with neurology services showed a significantly decreased risk (OR = 0.88, IC95% = 0.79-0.98), particularly in occlusion of cerebral artery (ICD-9 = 434) and in undefined cerebrovascular disease (ICD-9 = 436). Demographic and clinical variables are associated with the outcome of hospitalised stroke patients. Admissions of acute stroke patients in specialized hospitals seem to play a role in reducing the risk of in-hospital mortality.