When physicians who aren’t radiologists refer patients to imaging facilities they own or lease—known as self-referral—their patients don’t always benefit. In fact, these self-referrals lead to overuse of services, escalate spending, and rarely shorten the duration of illness, according to a series of studies in the December issue of Health Affairs. The findings challenge what proponents argue are the merits of having non-radiologist physicians acquire imaging equipment. Funding for the publication of these articles is made possible by a grant from Blue Shield of California Foundation.
The December Health Affairs issue is titled “Battling Chronic Disease Worldwide.” In addition to the cluster of articles on imaging self-referral, the volume contains a second series of pieces finding that worldwide disease prevention programs and greater use of primary care reduce deaths, rates of illness, and costs associated with chronic illnessess. It also includes a study finding that private insurance spending in McAllen and El Paso does not mirror the huge variation in Medicare spending between the two Texas cities that was publicized by Atul Gawande in the New Yorker.
To test the argument that self-referrals lead to a shorter turnaround time for imaging, and is thus more convenient for patients, American College of Radiology senior researchers Jonathan Sunshine and Mythreyi Bhargavan examined how often an office visit to a physician and a self-referred imaging procedure take place on the same day. In a review of Medicare data from 2006 and 2007, they found one-fourth of self-referred imaging was for chest and musculoskeletal x-rays and almost three-quarters of these patients had their imaging on the same day. But in only 15 percent of cases did Medicare patients get same-day service for high-tech imaging such as nuclear medicine, computed tomography (CT) scanning, and magnetic resonance imaging (MRI). Authors Sunshine and Bhargavan point out that higher tech imaging equipment is more expensive to buy or lease, so it makes financial sense that physician owners schedule appointments ahead of time to maximize use rather than have equipment sit idle and available for patients on a walk-in basis. However, the authors note that one of the main justifications for physician to self-refer is to provide same-day convenience to patients—a benefit they generally don’t seem to be offering.
Medicare now bans self-referral when the physician has a financial interest but it does allow self-referral for designated services, including imaging, if the service takes place in the doctor’s office. If policymakers want to curb unnecessary use of imaging, which has been a rapidly growing component of physician services, the authors recommend that Medicare allow referrals to a physician’s own office only for x-rays.
Increases in Unnecessary Use and Spending
A second study in the journal, by Laurence Baker, chief of health services research at Stanford University, shows that when non-radiologists acquire MRI equipment, use of services and overall spending rise. Baker examined a six-year period of imaging use and spending trends by patients of orthopedists and neurologists, who began billing for MRI scans on equipment they owned. Once they began billing for MRIs, both specialties ordered significantly more scans. The number of MRI procedures used by orthopedists increased 38 percent within 30 days of a first visit by a patient. Neurologists also prescribed substantially more MRI services once they began to bill.
“These changes in use were largely driven by a discrete jump that occurred when physicians first billed for MRI,” says Baker, adding that these increases persisted over time. Developing ways to understand the value of new technologies and evaluate their cost effectiveness would help target efforts to curb unnecessary use to areas where the greatest concerns about efficiency are apparent, he says.
Highlights of other Health Affairs studies on imaging in the December issue:
- Danny Hughes and colleagues at the American College of Radiology looked at the effects of self-referral on costs and the duration of an illness. They found that self-referrals do not offer some of the advantages frequently advanced by orthopedists, neurologists, and others who have invested in high-tech imaging. In fact, they show that the practice is usually associated with higher costs and only rarely with shorter episodes of illness. The only time there was a drop in length of illness for some of the 20 conditions they studied was when doctors self-referred patients for x-rays.
- University of Virginia professor Bruce Hillman and Health Futures Inc. analyst Jeff Goldsmith note that although limits are increasingly being placed on self-referrals, the incentives for doctors owning or leasing high-tech imaging equipment still remain strong because of the “sizable potential economic reward.” Their article traces the history of curbing self-referrals for imaging, including the effects of the Stark II Rule, the Deficit Reduction Act, the Affordable Care Act, and other measures. The authors say it will take years to assess the effect on self-referral of the payment reductions imposed by this legislation, as well as the disclosure requirements within the Affordable Care Act. In the meantime, policymakers should do their utmost to curb the abuses of self-referral and assure that Medicare patients receive only medically necessary imaging services.
- Laurence Baker and colleagues from Stanford and Harvard Medical School offer a case study to understand the complexities of developing policies around technology diffusion. They examine the increased use of a new imaging technology— computed tomographic angiography—which is safer and more cost effective than an older test for diagnosing carotid artery disease, the leading cause of stroke in adults. Although growth in the use of this test expanded the number of patients getting tested, the authors found no evidence that it also increased the use of treatments for carotid artery disease. The authors note that the use of new technologies often spreads beyond situations in which value is clear. In these cases, effectiveness or cost effectiveness is not well demonstrated. They warn policymakers not to base policies only on clinical situations, in which value is demonstrated, a move that could miss the many other uses of a new service and, they say, “Could pose a strong risk of fostering inefficiency.”