Black-White Disparities in Motor Function Outcomes Taking into Account Patient Characteristics, Non-therapy Ancillaries, Therapy Activities, and Therapy Interventions.

Objective: Assess black-white differences in functional outcomes, controlling for patient characteristics, use of non-therapy ancillaries, and use of physical and occupational therapy activities and interventions.
Design: Multi-center prospective observational cohort study of post-stroke rehabilitation.
Setting: 6 US inpatient rehabilitation facilities.
Participants: 732 patients, subdivided in case-mix subgroups (CMGs): CMGs 104-107 for moderate strokes (N=397), and CMGs 108-114 for severe strokes (N= 335).
Intervention: Not applicable.
Main Outcome Measure: Discharge Motor Functional Independence Measure (FIM).
Results: Taking into account patient characteristics, non-therapy ancillaries, and therapy activities, multivariate regressions explained (R2) 54% and 69% of variation in outcomes among moderate and severe stroke patients, respectively. Black race was associated with lower outcomes than white race in the severe group. However, race was no longer associated with outcomes after including interventions used within physical and occupational therapy activities. Including interventions within therapy activities increased R2 to 64% and 74% for moderate and severe strokes, respectively.
Some PT and OT activities were provided more to blacks than whites and vice versa. Greater intensity was sometimes associated with better functional outcomes and sometimes associated with poorer functional outcomes.
Conclusions: After controlling for interventions within activities, no racial differences were found in functional outcome at discharge, despite racial differences in rehabilitation care, possibly because each racial group received a mixture of interventions that were negatively and positively associated with outcome. Clinicians should provide therapies associated with better outcomes, with high and similar intensities for black and white post-stroke patients.