Cardiac rehabilitation following an acute coronary syndrome: Trends in referral, predictors and mortality outcome in a multicenter national registry between years 2006-2013: Report from the Working Group on Cardiac Rehabilitation, the Israeli Heart Society.

Background Utilization of cardiac rehabilitation is suboptimal. The aim of the study was to assess referral trends over the past decade, to identify predictors for referral to a cardiac rehabilitation program, and to evaluate the association with one-year mortality in a large national registry of acute coronary syndrome patients. Design and methods Data were extracted from the Acute Coronary Syndrome Israeli Survey national surveys between 2006-2013. A total of 6551 patients discharged with a diagnosis of acute coronary syndrome were included. Results Referral to cardiac rehabilitation following an acute coronary syndrome increased from 38% in 2006 to 57% in 2013 ( p for trend < 0.001). Multivariate modeling identified the following independent predictors for non-referral: 2006 survey, older age, female sex, past stroke, heart or renal failure, prior myocardial infarction, minority group, and lack of in-hospital cardiac rehabilitation center (all p < 0.01). Kaplan-Meier survival analyses showed one-year survival rates of 97% vs 92% in patients referred for cardiac rehabilitation as compared to those not referred (log-rank p < 0.01). Multivariate analysis showed that referral for cardiac rehabilitation was associated with a 27% mortality risk reduction at one-year follow-up ( p = 0.03). Consistently, a 32% lower one-year mortality risk was evident in a propensity score matched group of 3340 patients (95% confidence interval 0.48-0.95, p = 0.02). Conclusions Over the past decade there was a significant increase in cardiac rehabilitation referral following an acute coronary syndrome. However, cardiac rehabilitation is still under-utilized in important high-risk subsets of this population. Patients referred to cardiac rehabilitation have a lower adjusted mortality risk.