Health-care costs and exercise capacity

Background: While the beneficial effect of exercise capacity on mortality is well-accepted, its effect on health-care costs remains uncertain. This study investigates the relationship between exercise capacity and health-care costs. Methods: The Veterans Affairs Health Care System recently implemented a Decision Support System that provides data on patterns of care, patient outcomes, workload, and costs. Total inpatient and outpatient costs were derived from existing administrative and clinical data systems, were adjusted for relative value units, and were expressed in relative cost units. We used univariable and multivariable analyses to evaluate the 1-year total costs in the year following a standard exercise test. Costs were compared with exercise capacity estimated in metabolic equivalents (METs), other test results, and clinical variables for 881 consecutive patients who were referred for clinical reasons for treadmill testing at the Palo Alto Veterans Affairs Health Care System facility between October 1, 1998, and September 30, 2000. Results: The patients had a mean age of 59 years, 95% were men, and 74% were white. Eight patients (< 1%) died during the year of follow-up. Exercise testing showed an average maximum heart rate of 138 beats/min, 8.2 METs, and a peak Borg scale of 17. In unadjusted analysis, costs were incrementally lower by an average of 5.4% per MET increase (p < 0.001). In a multivariable analysis adjusting for demographic variables, treadmill test performance and results, and clinical history, METs were found to be the most significant predictor of cost (F-statistic, 21.8; p < 0.001). Conclusion: These findings are consistent with the hypothesis that exercise capacity is inversely associated with health-care costs.