Population-Based in-hospital stroke morbidity and mortality

To investigate the risk and prognostic factors for in-hospital incident stroke and stroke mortality, surveillance was conducted on Rochester, MN, residents free of stroke when admitted to the hospital in the years 1950 through 1984. The in-hospital first stroke attack rates remained stable for the calendar years 1970,1976, and 1980. Of the first strokes that occurred in the hospital, 42% were adventitious. In approximately 60% of patients with adventitious strokes, the strokes developed within 2 days after a hospital procedure. Patients with cardiovascular problems were the majority among the patients who had nonadventitious strokes. Four independently significant prognostic factors associated with stroke 30-day mortality included age (relative risk, 1.6 per 10 years), calendar year (relative risk, 0.7 per 10 years), atrial fibrillation or flutter (relative risk, 2.3), and intracranial hemorrhage compared with cerebral infarction (relative risk, 12.4). These findings emphasize the heterogeneity of the pathophysiologic mechanisms for in-hospital stroke.