Related Experiment Video
Updated: Jan 10, 2026
Randomized, Triple-Blind, and Parallel-Controlled Trial of Transcranial Direct Current Stimulation for Cognitive Rehabilitation after Stroke
Published on: June 6, 2025
Ethnic Differences in Postacute Stroke Rehabilitation
Imadeddin Hijazi1, Amanda Malingagio1, Emily Anderson2
1Stroke Program, Department of Neurology, University of Michigan Medical School, Ann Arbor, MI.
Objective:
To investigate ethnic differences in poststroke rehabilitation between Mexican American (MA) and non-Hispanic White (NHW) stroke patients.
Design:
Prospective cohort study.
Setting:
Community-based (December 2019 to August 2024).
Participants:
Of 1453 stroke patients (N=1453), the median age was 66 years, 46% were women, and 66% were MA.
Interventions:
Not applicable.
Main Outcome Measures:
Stroke patients or their proxies were contacted every 2 weeks for 90 days after stroke to identify their rehabilitation provider: inpatient rehabilitation facility (IRF), skilled nursing facility, home health agency (HHA), or outpatient rehabilitation. Binomial regression was used to model the association of ethnicity (MA vs NHW) with IRF as the first rehabilitation provider. Sequential modeling and propensity-score adjustment were used to investigate demographic and clinical variables impacting ethnic disparities.
Results:
Rehabilitation services were received by 974 patients in the 90 days after stroke; 34% of patients received no rehabilitation for the first 2 calls. The most common transitions between the first 2 providers were HHA to home without rehabilitation (14%), IRF to HHA (10%), and IRF to home without rehabilitation (7%). NHW patients had higher IRF usage as the first provider compared with MA patients (NHW vs MA risk difference=7.4%, 95% CI, 0.8-14.0). This difference remained after adjustment for patient and clinical factors (risk difference=6.5%, 95% CI, -0.1 to 13.0) and was attenuated by 74.6% after further adjustment for socioeconomic status (risk difference=1.9%, 95% CI, -5.2 to 9.0). Propensity-score methods implemented in a separate model confirmed a smaller, nonsignificant effect (risk difference=2%, 95% CI, -6 to 9) given observed covariates.
Conclusions:
MA stroke patients are less likely to receive rehabilitation in an IRF as the first provider after stroke compared with NHW stroke patients. This difference is largely explained by socioeconomic status.
