Referral to Cardiac Rehabilitation after Percutaneous Coronary Intervention, Coronary Artery Bypass Surgery, and Valve Surgery

Background - Despite guideline recommendations that patients undergoing percutaneous coronary intervention (PCI), coronary artery bypass surgery, or valve surgery be referred to cardiac rehabilitation, cardiac rehabilitation is underused. The objective of this study was to examine hospital-level variation in cardiac rehabilitation referral after PCI, coronary artery bypass surgery, and valve surgery. Methods and Results - We analyzed data from the Clinical Outcomes Assessment Program, a registry of all nonfederal hospitals performing PCI and cardiac surgery in Washington State. We included eligible PCI, coronary artery bypass surgery, and valve surgery patients from 2010 to 2015. We analyzed PCI and cardiac surgery separately by performing multivariable hierarchical logistic regression for the outcome of cardiac rehabilitation referral at discharge, clustered by hospital. Patient-level covariates included age, sex, race/ethnicity, comorbidities, and procedure indication/status. Cardiac rehabilitation referral was reported in 48% (34 047/71 556) of PCI patients and 91% (21 831/23 972) of cardiac surgery patients. The hospital performing the procedure was a stronger predictor of referral than any individual patient characteristic for PCI (hospital referral range 3%-97%; median odds ratio, 5.94; 95% confidence interval, 4.10-9.49) and cardiac surgery (range 54%-100%; median odds ratio, 7.09; 95% confidence interval, 3.79-17.80). Hospitals having an outpatient cardiac rehabilitation program explained only 10% of PCI variation and 0% of cardiac surgery variation. Conclusions - Cardiac rehabilitation referral at discharge was less prevalent after PCI than cardiac surgery. The strongest predictor of cardiac rehabilitation referral was the hospital performing the procedure. Efforts to improve cardiac rehabilitation referral should focus on increasing referral after PCI, especially in low referral hospitals.