Abstract
Objective: To assess the independent effect of pre-morbid frailty on stroke rehabilitation outcomes, including mortality, function, and discharge destination.
Design: Observational cohort study.
Setting: Inpatient stroke rehabilitation.
Participants: Two hundred consecutive stroke patients, 65 years or older, admitted during 2021.
Main measures: Logistic regression models analysed binary outcomes (mortality, discharge destination, and significant functional gain), and multiple linear regression models analysed normally distributed continuous outcomes (function, measured by Functional Independence Measure (FIM)). The confounders of age, stroke severity, and pre-morbid dementia/cognitive impairment, were accounted for. Clinical Frailty Scale was the predictor (considered as categorical and dichotomous) for these models.
Results: After accounting for the confounders, increased frailty was independently associated with lower discharge FIM (B = -11.32; 95% confidence interval (CI) -14.46-8.18, p < 0.001), smaller FIM change (B = -4.21; 95% CI -6.58-1.84, p < 0.001) and smaller FIM improvements over time (B = -2.36; 95% CI -3.40-1.32, p < 0.001). Frailer patients were less likely to return home (Odds Ratio 0.26; 95% CI 0.17-0.41, p < 0.001). Most patients improved function, at all levels of frailty. Stroke severity (Odds Ratio 1.13; 95% CI 1.06-1.20, p < 0.001), but not frailty, was independently associated with 6-month mortality.
Conclusions: Pre-morbid frailty was independently associated with lower functional improvement and lower likelihood of discharge back home after stroke rehabilitation, but not with mortality up to 6 months. Measuring both pre-morbid frailty and stroke severity may be useful to provide prognostic information and set realistic expectations.