Cardiac Rehabilitation Participation in Heart Failure Over Time: An Analysis of the Colorado All Payer Claims Database

Background Despite CMS coverage expansion supporting cardiac rehabilitation (CR) in patients with heart failure (HF) in 2014, data suggest that utilization among patients with HF is low. We describe CR participation and adherence among HF patients in Colorado. Methods Data from the Colorado All-Payer Claims Database from 2010-2018 were used. Patients with HF were identified by ≥2 claims with a HF diagnosis code, then grouped by type of HF (HFrEF, HFpEF, unspecified). CR participation and adherence were identified using CR CPT codes. Participation rates were calculated by quarter of each year. Cochran-Armitage tests determined whether temporal trends were significant. Association between CR participation and payer source was examined in adjusted logistic regression models. Results 263,476 patients with HF were identified. 4.77% of all HF patients attended CR at least once; this result was similar for HFpEF (4.35%), unspecified HF (4.15%), and higher in the HFrEF group (8.25%). Overall adherence was poor (median 8 visits, IQR 3-18; full adherence=36 visits). CR participation over time increased (P<0.01) for all HF patients. Compared to patients with commercial insurance, patients with Medicare, Medicaid, or Medicare Advantage were less likely to have participated in CR at least once (P<0.01). Race, sex and presence of another indication for CR were also associated with at least one CR visit (P<0.01). Conclusions In the state of Colorado, CR participation improved from 2010-2018 among all patients with HF. Our data suggest that payer source, race, sex and presence of another indication for CR drive CR participation in patients with HF.