PROJECT SUMMARY – PROJECT 4 Differences in clinical practice are often driven by factors unrelated to the health needs of patients, such as reimbursement rates for procedures. With up to 10-fold differences in payment rates for identical services across the main funders of health care - private insurers, Medicare, and Medicaid – there is enormous potential for misaligned incentives and inefficiency. This includes both overuse of ineffective but well-compensated treatments, or underuse of poorly-compensated, effective treatments and limited access to care, particularly among Medicare and Medicaid enrollees. In this Project, we propose to test several hypotheses regarding the potential harm of high relative commercial prices (or low Medicaid prices) using the large datasets from Core B including 100% Medicare claims, 100% Medicaid data, and commercial insurance from Optum Labs, the Blue Health Intelligence, and the Health Care Cost Institute (HCCI). We first create new regional comprehensive measures of overall health care utilization and spending that are adjusted for differences across regions in age, sex, poverty, and health status. We then ask whether high commercial reimbursement rates (relative to Medicare or Medicaid) affect quality of care for the elderly and vulnerable – Medicare and Medicaid recipients - with a particular focus on access for vulnerable populations. We hypothesize that high commercial rates lead to less Medicare and Medicaid utilization, and more sorting to lower-quality or to safety net providers. Continuing on this theme that price shifts may adversely affect utilization and quality of care, we also consider discontinuities in incentives for emergency department (ED) patients. When Medicare funds emergency care, hospitals lose money when patients are discharged from the ED, but make money when ED patients are admitted to hospital;; this does not hold for private-pay patients. We will test the hypothesis that these financial incentives lead to unnecessary hospitalizations and adverse outcomes, using both a discontinuity approach, and by considering an exogenous change in ED management practices for some (but not all) hospitals. Finally, we propose to study how changes in Medicare reimbursement rates influence private prices and health care quality using a natural experiment in which federal legislative changes caused approximately 100 hospitals to receive an average of 10% higher Medicare reimbursement rates, with no impact on comparable hospitals. In sum, a better understanding of how the health care delivery environment affects physician behavior and quality of care can provide makers with guidance on the health risks (and not just the financial risks) of either high commercial prices, or low Medicare and Medicaid prices.