Abstract 13536: Cardiac Rehabilitation Enrollment Among Medicare Beneficiaries: Surgical Hospitals Play a Critical Role!

Introduction: Hospitalization for myocardial infarction (MI), percutaneous coronary intervention (PCI), or cardiac surgery [coronary artery bypass surgery (CABG) or valve surgery] represents a key opportunity to refer and recruit eligible patients to cardiac rehabilitation (CR.) However, little is known about how hospitals influence CR enrollment at the national level. Methods: Using Medicare fee-for-service claims from 2017, we identified all patients with a principal discharge diagnosis of MI, PCI, CABG, or valve surgery. We defined CR enrollment as attending ≥ 1 session within 90 days and calculated hospital-specific CR enrollment rates. To assure stability of our statistical results, we limited our sample to hospitals where cardiac surgery was performed, and then examined the association between patient demographics, procedures, comorbidities, and hospital characteristics on CR enrollment using mixed effects hierarchical logistic regression. We calculated the median odds ratio (MOR) to estimate the extent to which the hospital where a patient was cared for influenced their chance of enrolling in CR. Results: At 3,420 hospitals, we identified a total of 264,970 eligible patients (age 77 ± 8 yrs., 58% male, 85% white, 39% surgical), of which 25% attended CR. While a minority (38%) of hospitals performed cardiac surgery - these hospitals cared for the vast majority (92%) of all eligible patients. At these hospitals, the median CR enrollment rate was low (22%) and varied 10-fold across hospitals at the 10 th and 90 th percentiles (3%, 42%). After multivariable adjustment, the factors most associated with CR enrollment were having cardiac surgery or undergoing PCI (OR 5.8, 2.6 vs. MI respectively), advancing age (OR 0.77 per 5-year increase), low socioeconomic status (OR 0.33), smoking (OR 0.59), or living in the Midwest (OR 2.3 vs. South), p <0.001 for all. The hospital MOR was 2.1, similar in effect size to PCI or Midwest location. Conclusions: In this large national study of older adults, participation rates in CR were low, and the hospital where patients receive their care was strongly associated with CR enrollment. As surgical hospitals care for the majority of all eligible patients, they appear to be the most logical place to focus improvement efforts.