Medical Spending and Health Outcomes

Should the U.S. be using cost effectiveness analysis with regards to medical expenditures?

Prior research on geographical variations in health care utilization and outcomes has shown a weak – if not negative – association between expenditures and outcomes. This finding has been interpreted to imply that up to 20% of health care costs in the US could be saved without compromising health outcomes. If that were so, then the urgency of using cost-effectiveness analysis to guide health care in the US would be greatly attenuated. However, Prof. Milton Weinstein of the Center for Health Decision Science collaborated with Dr. Jonathan Skinner of The Dartmouth Institute for Health Policy and Clinical Practice to formulate an alternative explanation and to begin to test it empirically. The hypothesis is that weak or negative associations between health spending and outcomes across geographic areas or hospitals need not be causal, but may instead reflect differences in the degree to which areas or hospitals adopt cost-effective (high-value, low-cost) interventions in lieu of cost-ineffective (low-value, high-cost) interventions. To the degree to which that hypothesis is true, it would be impossible to reduce costs and improve health outcomes simultaneously within a health service area unless high-cost areas substitute high-value health services (those with low cost-effectiveness ratios) for low-value services (those with high cost-effectiveness ratios). Empirical studies linked to the Dartmouth Atlas Project and Medicare claims data are investigating this hypothesis in the context of health care services following hospital admissions for acute coronary syndrome, and in the surgical treatment of low back pain.