Outcome and pathogenesis of stroke is conditioned by place of stroke occurrence

To investigate whether stroke pathogenesis and sequelae are conditional on place of initial stroke occurrence, 2,238 incident stroke cases among Rochester, Minnesota, residents in the period from 1950 to 1984 were evaluated. Virtually compete case ascertainment of first strokes in Rochester nursing homes (153 strokes), hospitals (194), and homes (1,891) was possible through a medical record linkage system. Stroke cases were reviewed by a neurologist applying uniform criteria. Although most patients had the first stroke at home (84%), there was an increasing trend over the study period of first strokes occurring while patients were hospitalized. For nursing home incident strokes, there was a linear increase over calendar time of patients being hospitalized. This paralleled an overall increase in patients being hospitalized for the first stroke. Persons with initial stroke in the home who were hospitalized had lower 5-year observed survival than those who were not hospitalized (p 0.0067), an indication that those with more severe stroke were more likely to be hospitalized. The 5-year relative survival was highest for initial stroke occurrence in the home (57%), intermediate for in-hospital strokes (35%), and lowest for nursing home strokes (11%). Comorbid conditions undoubtedly contributed to the difference in survival patterns. A cardiac source of embolus was present in more than 40% of strokes occurring in the nursing home and in over half the strokes that occurred in the hospital.