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

The effect of physician financial interest on the utilization and cost of healthcare is an issue of persistent concern in health economics. Physicians order medical procedures for their patients, ideally as economically disinterested agents, but these procedures may yield financial benefits to the ordering physician. This topic is of particular interest in high-tech diagnostic medical imaging because high-tech imaging has been the fastest-growing component of healthcare expenditures in recent years. Much of this growth has been in physician offices, where physicians are likely to have an ownership interest in the imaging equipment. Any image performed on this equipment is likely to result in facility fee revenues (also called "technical component" revenues) for the physician as well as in fees for the physician's own work.

This session brings together four papers on the topic. The first paper, by Jonathan Sunshine, uses Medicare claims data from 2004 and 2005 to show that physician FSI is associated with much higher levels of imaging utilization and costs than those of comparable physicians who do not have this interest. Because FSI can be difficult to identify unambiguously, the study uses four definitions of financial self-interest to test robustness of findings. Patients are the unit of analysis and a wide variety (13 in total) of medical conditions are studied. The study tests multiple often-asserted, but rarely-evaluated defenses of FSI.

The second paper, by Stanford's Prof. Laurence Baker, uses 1998-2005 Medicare data to examine the relationship between trends in physician ownership of MRI equipment and trends in utilization of MRI procedures. Analysis is at the level of hospital referral regions (HRRs). At this aggregated level, a conclusive relationship between equipment ownership and imaging utilization is not found, but there appear to be differential effects according to physician specialty.

The third paper, by Mythreyi Bhargavan, also deals with small geographic areas (rather than individual patients). It explores the extent to which the share of imaging by physicians with FSI in imaging explains small area variation (SAV) in imaging use, for each imaging modality and in total. It thus explores the extent to which SAV in utilization, which has long been shown to be very widespread and very costly, can be explained by physician FSI. Physician FSI has been shown to have large effects at the level of individual physicians, but its effect at the level of geographic areas has not been studied. Preliminary analysis finds an association between the proportion of imaging by physicians with FSI in imaging, and imaging volume in the area.

The fourth paper, by Georgetown's Prof. Jean Mitchell, studies the effects of a payer's attempt to control potentially inappropriate imaging that may have been motivated by financial self-interest. Some payers with concerns regarding increasing imaging costs have been taking actions to curb potentially inappropriate imaging, some of which might have been financially motivated. This fourth study examines a natural experiment in the use of administrative controls. It measures their effect and whether they have most effect in situations in which financial self-interest is greatest.