Institutionalization after stroke

Background and purpose: There are widespread regional variations in the institutionalization rate after stroke. This study sought to identify the factors that predict institutionalization after a primary diagnosis of stroke and determine whether institutionalization rate could be used to assess the quality of hospital care. Methods: A retrospective case note audit of 2778 consecutive admissions with stroke, between 1 September and 31 October 2000 from a randomly selected sample of 79 hospitals in England, Wales and Northern Ireland. Data were collected on demographics, case-mix, clinical outcome, organization of discharge, and place of residence at discharge and at three and six months. The figures from this audit were validated using data from the 1998 National Sentinel Audit of Stroke. Results: Overall 14% of patients previously living at home were discharged to nursing or residential homes. Using logistic regression, Barthel Index score at discharge accounted for 40% of variation. Two-thirds of patients with a discharge Barthel score of <5 were institutionalized. Age alone explained 14% and when taken together with discharge Barthel and length of stay it accounted for 54% of variation. On admission 22% of the variation in institutionalization rate could be accounted for by total Glasgow Coma Score (15, <15), age and ability to walk unaided. Regional variations in institutionalization rates are evident and may in part be explained by differences in ease of access to the institutions. Conclusions: Discharge disability and older age were the dominant factors determining admission to nursing and residential homes. It is not possible to predict sufficiently reliably for an individual patient the likelihood of institutionalization at admission or at 72 h after admission, to justify early resource allocation decisions.