Aim: The prediction of stroke outcome in the elderly can be rather difficult, due to the potential interference into disability and handicap development of several clinical modifiers, such as comorbidity, medical complications, neuropsychological impairment proper to the aging brain and social issues. These factors can strongly affect old patient response to rehabilitation and need to be taken into account, along with ageing per se, to optimize health resource efficiency for the care of disability due to stroke. In this study, we tried to identify outcome determinants of stroke rehabilitation specific for the elderly.
Methods: A total of 359 first-stroke patients aged 75 years, admitted for active rehabilitation treatment to hospital rehabilitation wards, were enrolled into a multicenter cohort (prospective) study. They all underwent a comprehensive medical rehabilitation program. We considered as primary outcomes the frequency of home discharge and the extent of functional recovery, assessed by Functional Independence Scale (FIM) and expressed as the Montebello Rehabilitation Factor Score (MRFS) efficacy. Each measure of outcome was related to age, as well as other potential clinical and functional confounders, according to a multivariate model. For each dependent variable, two models were developed, using either FIM total score or FIM domains scores at admission among predictors.
Results: FIM total score increased from 55.8+/-24 to 75.3+/-30 (P<0.0001), with a mean MRFS efficacy of 0.33+/-0.25. Most patients (79.9%) were discharged home. Age turned out as independently and inversely related to MRFS, explaining at the most 3.6% of its variance, although FIM at admission was its most powerful predictor. Home discharge was not related to age, but to social issues, such as living in family before admission, and cognition.
Conclusion: The present study suggests that rehabilitation can be effective in elderly stroke patients, in improving function as well as in favorably affecting discharge destination. In fact, age per se predicts the outcome at a lesser extent than other clinical covariates, such as functional and cognitive status at admission and social situation.