Identification of Disability Subgroups for Patients After Ischemic Stroke

Abstract

Abstract Objective The aims of this study were to identify disability subgroups among patients hospitalized for ischemic stroke and to determine the predictors and distal outcomes for the disability subgroups. Methods A retrospective, cross-sectional study design was used. Data on patients after ischemic stroke were collected from the electronic health records at 5 hospitals within a single health system. Covariates included social and demographic factors. Disability was characterized according to the Activity Measure for Post-Acute Care assessment. Distal outcomes were physical therapist treatment use, occupational therapist treatment use, and discharge disposition. Latent class analysis was used to identify disability subgroups of patients hospitalized for ischemic stroke. Results A total of 1549 patients were included in the analyses. Five disability subgroups were identified and labeled as follows: globally impaired (subgroup 1), impaired dynamic balance (subgroup 2), impaired self-care (subgroup 3), impaired mobility (subgroup 4), and independent (subgroup 5). Physical therapist treatment use (χ24 = 113.21 [P < .001]) and occupational therapist treatment use (χ24 = 122.97 [P < .001]) varied significantly across the disability subgroups. The globally impaired group had the highest probability of physical and occupational therapist treatment use. Similarly, discharge disposition varied across the subgroups (for home without services, χ24 = 246.61 [P < .001]; for home with home health care, χ24 = 35.49 [P < .001]; for institutional discharge, χ24 = 237.18 [P < .001]). The independent subgroup had the highest probability of discharge to home without services. Conclusion Five disability subgroups were identified for patients after ischemic stroke. The disability subgroups provide a common language for clinicians to organize the heterogeneity of disability after stroke. Impact Using the disability subgroups, the multidisciplinary team might be able to improve the accuracy and efficiency of care decisions. The number of current rehabilitation interventions is indeterminable; these subgroups may help to guide clinicians in selecting the most beneficial interventions for patients based on subgroup membership.