Abstract 13261: Determining the Cost-Benefit Yields of Participating in a Cardiac Rehabilitation Program

Abstract

Background and Objective: Regular participation in cardiac rehabilitation (CR) has been consistently shown to improve survival prognosis for patients suffering from a recent cardiovascular episode. No study has quantified the thresholds of baseline risk/behavioral attrition beyond which the cost-benefits of CR become economically unattractive. We aimed to determine how the cost-benefits associated with CR vary across baseline risk and behavioral attrition patterns amongst an actual population of patients participating in an outpatient CR program. Methods: Data was obtained from 11,998 consecutive patient referrals to Toronto Rehab Cardiac Rehabilitation and Secondary Prevention Program (1995 to 2010). Baseline risk was defined as the probability of death/hospitalization at 2 years following program termination and behavioral attrition as the probability of program drop-out. We assumed a fixed cost of $1500 per program and a fixed program efficacy of 20% reduction in death/hospitalization. Multiple logistic regression models were used to predict the number of deaths/hospitalizations avoided per 1000 patients treated. Results: Regression models for baseline risk and baseline behavioral attrition produced C-statistics of 0.66 and 0.85, respectively. Increasing age, type 2 diabetes were found to be concordant predictors of CR cost-benefit and higher baseline functional capacity as a discordant predictor. The correlation between program drop out risk and baseline risk was poor (r=0.22, p<0.001). There was a 13% increased risk of predicted dropout for each quartile of program cost to yield (RR 0.87, 95% CI: 0.87, 0.89), while for each 10% baseline risk increase, RR was found to be 1.36 (95% CI: 1.34, 1.37). For both risk factors as predictors of improved cost-benefit yield ($1000 cost per adverse-event avoided) there was a 21% higher likelihood of the program being economically attractive for each 10% increase in behavioral attrition risk (RR 1.21, 95% CI: 1.207, 1.213). Conclusion: Baseline risk and behavioral attrition are important determinants of CR effectiveness and underscores the potential utility of risk stratification and patient baseline/behavioral attrition risk thresholds to estimate the programmatic benefit of CR programs.