The Barthel index in clinical practice: use on a rehabilitation ward for elderly people.

Primary activities of daily living (ADL) were monitored weekly in 102 patients admitted to a rehabilitation ward for elderly people using the Barthel index. The three commonest diagnoses were 'stroke', 'fractured neck of femur' and 'dementia recovering from acute illness'. Multiple disabling diagnoses were common: 60% of patients had dementia and 23% had a live-in carer; mean (median) length of stay in the rehabilitation ward was 98 (62 days). Over 18 months, the weekly assessment of patients in the ward was omitted once. No extra resources were needed. There was a significant rise in Barthel scores between admission to the rehabilitation ward (median Barthel 6) and discharge (median 13) for the group as a whole (median change 6, 95% CI 5-7; p < 0.001) and for each of the three main diagnostic groups. Barthel scores on discharge were significantly lower than in patients discharged from an acute ward for elderly people. Barthel scores and mental test scores (MTS) at discharge were significantly related to destination on discharge, with a characteristic pattern for patients unable to return home and having to be placed in nursing homes (Barthel < 10, MTS < 7). Our experience confirms that routine clinical use of the Barthel in this setting is feasible and responds to clinically important change, at least in group evaluation. It suggests that the Barthel may be useful in outcome measurement, case-mix adjustment and audit of discharge practices.