Duke 2008 | ASHE | Equity and Efficiency in Health and Healthcare

Existing research has established wide variations in care, both across physicians in a given market and across geographic markets. The papers in this session consider how physicians' information sources and financial incentives influence clinical decisionmaking, particularly as they relate to promoting or decreasing the extent of variation across physicians. In the first paper, Lucarelli and Nicholson consider how physicians learn about newly introduced oncology drugs. Specifically, they consider learning from clinical trials and product characteristics, which would promote standardization of decisionmaking; learning from their peers, which would contribute to geographic variations; learning from their own experiences, which would create variations across physicians; and incorporation of patient-specific characteristics, which would generate within-physician variations across patients.

In the second paper, Ketcham and Léger examine how gainsharing, a recent innovation in physician reimbursement, promotes standardization of cardiologists' choices of drugs and devices. Gainsharing relies on group-based financial incentives integrated with benchmarking information. This both lowers the cost and raises the benefits of sharing information and monitoring each others' choices. The paper presents a detailed, multi-stage model of gainsharing that establishes the mechanisms by which gainsharing can reduce spending on drugs and devices, including standardization of physicians' choices. It also presents empirical results from analysis of a novel dataset that covers all of the completed and ongoing gainsharing programs in cardiology. The results indicate the extent to which gainsharing promoted standardization and which types of physicians and groups were most responsive to its incentives.

In the third paper, Léger, Allard, and Jelovac develop a model that incorporates dynamic changes in patient health into primary care physicians' decisions about whether to treat or refer to a specialist. The model relies on differences in primary care physicians' altruism, diagnostic ability, and financial incentives to understand variations in physicians' referral decisions. Because primary care, specialty care, and hospital care can be substitutes in the health production function, the model determines under what conditions fee-for-service, capitation, and fundholding minimize the likelihood that the patient's health will worsen over time. This model represents an important extension of the existing static models. Together, these papers offer a greater understanding of how incentives and information affect patients' access to care, including new drugs, new devices, and specialists, and subsequent health care spending and health outcomes.