Venue: The Fuqua School of Business, Duke University, 1 Towerview Drive, Durham, NC 27708-0120
Presentation
The effect of physician financial self-interest (FSI) on the volume of diagnostic imaging ordered for illnesses, and on the costs of care
Motivation: Medical imaging is a large component of health care costs in the United States, with an estimated annual cost almost $100 billion, and has one of the fastest growth rates (approximately 10% per year) among all medical services. Imaging is usually provided through physician referral, and therefore, may be influenced by financial incentives faced by the referring physician. Although FSI has previously been found associated with 2-4-fold higher levels of imaging than that of physicians without FSI, defenders of the practice assert (among other things) (i) that utilization differences between physicians with and without FSI have largely disappeared in the approximately15 years since previous studies, (ii) that the higher level of imaging arises because patients of physicians with FSI are older and sicker, and (iii) that total episode costs are lower with FSI because intensive imaging leads to faster and more accurate diagnosis.
Method: We analyze the differences in imaging utilization and annual cost of care for an illness between physicians with and without FSI in imaging. To address assertion (ii) we control for patient characteristics and co-morbidities and physician specialty. We also control for region because regional care patterns differ. Because FSI can be difficult to identify unambiguously, we use four different operationalizations of the concept of FSI to ascertain if findings are robust. To assess generalizability of findings, we study utilization separately for 13 medical conditions and a total of 26 combinations of medical condition with imaging modality (CT, plain x-ray, etc.). For the half of the conditions that are chronic, we study the effect of FSI on total annual costs of care for the illness.
We perform regression analysis on claims data from Medicare's Limited Data Set 5% beneficiary sample files for 2004 and 2005. We identify patients with specific medical conditions such as acute upper respiratory tract symptoms, headache, or knee pain. We identify whether the physician treating the patient for the condition had FSI in imaging. Outcomes of interest are, on a condition-specific basis, (a) percent of a physician's patients who undergo imaging, (b) number of images and total imaging costs per patient imaged, and (c) total Medicare costs per patient, with costs adjusted to control for geographic price differences.
Preliminary Results: Taking account of all the controls, 2004 claims show adjusted odds of imaging by a physician with FSI are typically 1.5 to 3 times the odds of imaging by a physician who does not have FSI. The adjusted odds are >1 in all but one of our 104 analyses. (104 analyses =26 (condition) x (imaging modality) combinations times 4 definitions of FSI.) When imaging of a given modality takes place, the number of images typically does not differ importantly between physicians with or without FSI. Total costs of care have not yet been assessed.
Conclusions: When FSI exists, physicians are much more likely to order images, and imaging costs are proportionately higher. The results of this study will assist policy makers and payers in designing effective incentive structures to ensure appropriate utilization.