Community-based rehabilitation services, including physical therapy (PT) and occupational therapy (OT), play an important role in recovery after a stroke, yet these services remain underused in routine practice. A secondary analysis of the Comprehensive Post-Acute Stroke Services (COMPASS) pragmatic trial published in Circulation evaluated whether a transitional care model could improve utilization of post-acute rehabilitation.

Forty hospitals were randomized to implement the COMPASS transitional care program (COMPASS-TC) or maintain usual care for patients hospitalized with stroke or transient ischemic attack. During a crossover phase, hospitals initially assigned to usual care subsequently implemented COMPASS-TC. Participants were linked to administrative claims to evaluate PT/OT use after hospital discharge. Outcomes included therapy utilization at 30 and 90 days, time to first therapy visit, number of therapy visits, and receipt of combined PT and OT versus a single therapy. Adjusted generalized estimating equations were used to compare outcomes between COMPASS-TC and usual care groups.

A total of 8,377 patients were enrolled between July 2016 and March 2019, of whom 5,261 were successfully linked to administrative claims. Thirty-day PT/OT utilization ranged from 22.6% in usual care to 37.5% in the COMPASS-TC group. Therapy utilization did not differ significantly between groups in the randomized trial cohort. However, in the crossover analysis, COMPASS-TC was associated with higher therapy use at 30 days (increase of 9.4%; 95% CI 5.6-13.3%). COMPASS-TC was also associated with a shorter time to therapy initiation (mean difference −0.16; 95% CI −0.03 to −0.29).

COMPASS-TC was associated with earlier initiation of rehabilitation therapy and increased therapy utilization in crossover analyses. Differences between trial and crossover findings may reflect confounding or hospital adoption patterns.