Abstract DP309: Higher Physical Therapy Utilization is Associated with Fewer Secondary Strokes in the All of Us Research Cohort

Background: Physical therapy (PT) is a key component of stroke rehabilitation, yet the association between PT and secondary stroke prevention is underexamined, particularly among groups historically observed to have elevated stroke rates. This study examines whether PT utilization after stroke is associated with reduced secondary stroke risk after adjusting for participant self-reported demographic variables and clinical risk factors for stroke. Methods: This retrospective observational cohort study analyzed electronic health records and survey data from the All of Us research cohort (data through October 2023). Participants were included if they had an inpatient ICD stroke diagnosis code and ≥1 healthcare visit within two years following the index stroke. The exposure was poststroke PT utilization, ascertained through CPT procedure codes and categorized by rehabilitation visits expected during an acute inpatient stay (low PT, 1-3 visits), acute inpatient stay plus post-acute rehabilitation (high PT, ≥ 4 visits) or no PT (0 visits). The primary outcome was secondary stroke, defined as a stroke diagnosis ≥30 days after the index event. CatBoost gradient boosting was used to generate propensity scores for inverse probability of treatment weighting, adjusting for age, sex, race, ethnicity, insurance type, income, hypertension, diabetes, atrial fibrillation, and smoking. Missing values were encoded as separate categories. Weighted Kaplan-Meier curves and Cox proportional hazards models were used to estimate the association between PT utilization and secondary stroke risk. Results: The analytic cohort consisted of 13,607 individuals who had an index stroke and encompassed 13 patient-years of follow-up. 4,657 (34%) utilized PT and 2,012 (14.8%) had a secondary stroke. High PT recipients had a longer median time to secondary stroke (3.0 years) than those who received low PT (1.8 years, p=0.0015) or no PT (1.1 years, p<0.001). High PT was also associated with lower adjusted secondary stroke risk compared to no PT (HR 0.57, 95% CI 0.50-0.64, p<0.001) and low PT (HR 0.61, 95% CI 0.52-0.71, p<0.001). Conclusions: Poststroke PT was associated with reduced secondary stroke risk. This association persisted after controlling for demographic and clinical factors, suggesting an important but previously unrecognized benefit of rehabilitation across diverse patient populations. Future study is needed to determine whether this association is causal.