Association Between Physical Therapy and Mortality in Older Patients with Heart Failure and Chronic Kidney Disease.

Background/Objectives: Evidence regarding the prognostic impact of outpatient physical therapy (PT) initiated shortly after discharge in older patients with chronic kidney disease (CKD) hospitalized for worsening heart failure (HF) is limited. We evaluated the association between outpatient PT initiation within 30 days of discharge and the primary endpoint of all-cause mortality within 1 year after discharge. Methods: In this retrospective multicenter cohort study, we applied a 30-day post-discharge landmark design. Patients who died or were rehospitalized within 30 days after discharge were excluded, and 6359 eligible 30-day event-free survivors were included in the analysis. Of these, 244 (3.8%) underwent outpatient PT within 30 days of discharge, and the remaining patients had no documented outpatient PT within that period. Propensity scores were estimated from prespecified covariates and used for 1:1 matching. We assessed associations with 1-year all-cause mortality, cardiovascular (CV)-related rehospitalization, and their composite using Cox proportional hazards models in the propensity score-matched cohorts and pooled estimates across the 50 imputed and matched datasets using Rubin's rules. Results: Within 1 year of discharge, 1099 participants died, 1586 were rehospitalized for CV-related reasons, and 2325 experienced either event. In pooled Cox regression analyses, outpatient PT initiation within 30 days was associated with lower all-cause mortality (hazard ratio [HR], 0.46; 95% confidence interval [CI], 0.22-0.97; p = 0.042) and the composite outcome (HR, 0.67; 95% CI, 0.48-0.94; p = 0.021), but not CV-related rehospitalization alone (HR, 0.78; 95% CI, 0.55-1.10; p = 0.178). Conclusions: Among patients with HF and study-defined CKD who were alive and free from CV rehospitalization at the 30-day landmark, early outpatient PT participation was associated with a lower risk of subsequent all-cause mortality. Given the observational design, treatment selection, and residual confounding, this finding does not establish a causal effect and requires confirmation in prospective controlled studies.