|| Checking for direct PDF access through Ovid
Background: Cardiac rehabilitation (CR) is recommended for patients with cardiovascular disease, however the impact of a region-wide, centrally managed CR model with local non-medical community-based delivery of CR is unknown in a universal health care system setting.Objectives: To determine the impact of a regional 6-month, once per week supervised exercise session CR model on mortality and re-hospitalization compared to a matched control within a regional health care system.Methods: The regional CR service comprised: a standardized regional referral strategy for patients with established cardiac disease including stable heart failure, local (within a 30-minute drive time from place of residence) CR service delivery by trained CR professionals. Patient data that was prospectively entered into a regional CR database for patients referred to CR from 2012 through 2014, was linked to government mandated administrative databases and all-cause mortality and hospitalization rates were compared between patients who were referred to CR vs. a matched cohort who were not. Matching included age, sex, year of referral, socioeconomic status, and cardiac condition at hospital discharge. Mortality and re-hospitalization were assessed at 1-yr post index cardiac event excluding the initial 6-month period to account for immortal time bias.Results: Fewer patients who completed CR compared to matched controls died (14/1,318 or 1.1% vs. 31/1,318 or 2.4%, p=0.011), or were readmitted to hospital (128 or 9.7% vs. 190 or 14.4%, p<0.001). Fewer patients who started but did not complete CR (CR non-completers) died compared to matched controls (10/639 or 1.6% vs. 21/639 or 3.3%, p=0.045), however there was no difference in hospital readmission rates between CR non-completers compared to matched controls (18.0% vs. 14.9%, p=0.13). More patients who were referred to CR but did not enroll in CR, died compared to matched controls (241/3,933 or 6.1% vs. 196/3,933 or 5.0%, p=0.027), or were readmitted to hospital (932 or 23.7% vs. 643 or 16.3%, p<0.001).Conclusions: In a universal health-care system, patients who completed a regionally coordinated, locally delivered, standardized CR program close to home, experienced lower mortality or re-hospitalization rates compared to matched patients not referred to CR. This association was not likely due to a referral bias as patients who were referred to CR but did not enroll in CR faired worse, not better, compared to matched non-CR referred patients. Health care system decision makers (e.g. provincial governments in Canada) should strongly consider funding regional CR strategies.