Abstract 16524: Inpatient Cardiac Rehabiliation is Not Associated With Improvement in Hospital Outcomes Among Patients With Recent Cardiac Surgery
Introduction: Inpatient cardiac rehabilitation (ICR) provides risk factor education, daily ambulation, and facilitation of referral to outpatient cardiac rehabilitation, but little is known about the impact of ICR on hospital outcomes such as length of stay (LOS), discharge disposition, or hospital readmission.
Hypothesis: We hypothesized that ICR would be associated with improved hospital outcomes.
Methods: We examined the association of receipt of ICR with hospital LOS, discharge disposition, and 1 and 3-month hospital readmission between January 2007 and June 2011, in a geographically and structurally diverse sample of US hospitals (Premier, Inc.) that perform cardiac surgery. We developed multivariable hierarchical models (HGLM) and adjusted for patient demographics, comorbidities, hospital characteristics, and patient clustering within hospitals. To address the threat of residual confounding due to unmeasured factors, we performed an instrumental variable (IV) analysis using the hospital preference for ICR as the instrument.
Results: At 170 hospitals, there were 186,875 admissions for cardiac surgery (age 67 years, 68% bypass surgery, 69% male, 73% White) in which a total of 82,755 (44.3%) received ICR. Median LOS was 8 days, 20% were discharged to a nursing home, and 7 and 14% were readmitted within 1 and 3 months, respectively. In adjusted analyses, receipt of ICR was associated with a minimally shorter hospital LOS (-0.16 days, p = 0.001), a lower odds of discharge to a nursing home (OR 0.70, p <0.001) but no difference in 1 or 3 month hospital readmission. In the IV analysis, receipt of ICR was not significantly associated with any outcome (+ 0.26 days LOS, +0.1% readmission rate, and +3.0% discharge rate to nursing home, p > 0.05 for all).
Conclusions: In a large sample of cardiac surgery patients in hospitals across the U.S. using robust analytic techniques, we found that the receipt of ICR was not associated with clinically significant changes in LOS, discharge disposition, or readmission.
Author Disclosures: Q.R. Pack: None. A. Priya: None. T.C. Lagu: None. P.S. Pekow: None. P.K. Lindenauer: None.
- © 2016 by American Heart Association, Inc.