Outcomes at 12 months in a population of elderly patients discharged from a Rehabilitation Unit

Objectives: This study investigates the cognitive, functional, and clinical variables associated with the risk of institutionalization, rehospitalization, and death at 12 months among a population of elderly discharged from a Rehabilitation and Aged Care Unit (RACU) within a 1-year period (May 1, 2004 to April 30, 2005). The RACU is a relatively new setting of care providing intensive rehabilitation and clinical support to elderly with highly heterogeneous reasons for admission. Methods: There were 1303 patients (>= 65 years old) contacted 12 months after discharge from the RACU. We obtained information about institutionalization, rehospitalization, and death. Predictors were all the demographic and clinical variables potentially related to these outcomes. The relationship among predictors and outcomes was tested with multiple stepwise logistic regression models. Results: Among the 1072 patients alive at the 12-month follow-up, 90 (8.4%) were institutionalized (3.4% early at discharge and 4.9% within the next period). The logistic regression analysis showed that 2 ranges of age (78 to 83 years and 84 years or more), living alone, occurrence of delirium, cognitive impairment (Mini Mental State Examination lower or equal to 20/30), and poor functional status at discharge (Barthel Index scores ranging from 69 to 85 and Barthel Index scores lower than 68/100) were independently and significantly associated with the risk of institutionalization during the 12 months following discharge from the RACU. Three hundred and twenty-three (30.1%) patients had been rehospitalized onceand 86 (8.0%) patients twice at the 12-month follow-up. In the multivariate analysis, comorbidity (Charlson Index scores ranging from 2 to 3 and Charlson index scores higher than 4) and delirium were significantly and independently associated with this outcome. One hundred and thirty-six (11.3%) patients had died by the 12-month follow-up. The stepwise logistic regression analysis showed that age greater than 83 years, poor functional status (Barthel Index lower than 60/100 at discharge), high comorbidity (Charlson Index scores ranging from 3 to 4 and Charlson Index scores higher than 4, respectively), and albumin serum levels ranging from 3.2 to 2.9 mg/dL and lower than 2.9 mg/dL independently and significantly predicted the 12-month risk of death. Absence of depressive symptoms (Geriatric Depression Scale