A study by the Medicare Payment Advisory Commission has confirmed what critics of in-office self-referred imaging have long claimed. Physicians who have a financial interest in medical imaging equipment are more likely to refer patients to use it, and they incur higher costs generally than physicians who do not have similar financial incentives.
The study also determined that imaging performed in an in-office environment is more expensive than such services provided elsewhere. In fact, tests and treatments that follow an initial self-referred imaging study tend to be more expensive than care provided after imaging that was not self-referred, the study said.
The study was published in Report to Congress: Improving Incentives in the Medicare Program. It appeared on the MedPAC website June 16.
Dr. David Levin, a professor emeritus of radiology at Thomas Jefferson University, said the MedPAC study reinforces what he and other radiologists have found repeatedly during 20 years of self-referral research.
"It shows that self-referral inevitably leads to high utilization," he said.
The study breaks new political ground as well, according to Levin. Until now, self-referral supporters could dismiss most of the self-referral literature because the research was performed by radiologists who themselves have a financial interest in who controls imaging services.
"MedPAC obviously is not beholden to radiologists. It is as neutral as an organization can get," Levin said. "Yet it has come out with this really damning evidence."
Levin was frustrated, however, that MedPAC reported the results without a recommendation to Congress to change Medicare self-referral policy.
But the American College of Radiology is comfortable with MedPAC's decision to examine the matter further, said Maurine Spillman-Dennis, senior director of economics and health policy.
"We believe there are many drivers that increase utilization. Self-referral definitely contributes to that," she said.
In-office self-referral has been implicated in the rapid growth medical imaging since the early 2000s. The cost of outpatient medical imaging covered by Medicare grew at nearly twice the rate of all other physician services from 2002 to 2007, according to MedPAC. The growth rate has since slowed.
In 2008, the Government Accountability Office reported that physician offices accounted for 64% of imaging spending under the Medical physician fee schedule in 2006, compared with 58% in 2000.
Proponents of in-office imaging say it improves patient access and convenience. They believe that quick access to imaging allows the referring physician to develop a treatment plan more quickly than cases referred to outside imaging services and that the quality of imaging is easier to supervise when it is performed in a physician's office.
Federal Stark law has generally prohibited physician self-referral for imaging since the mid-1990s, but an exemption allows physicians to own and operate imaging equipment in their offices.
In its most recent study, MedPAC investigators tracked in-office utilization from 2005 Medicare claim data, evaluating the experience of all Medicare beneficiaries in Boston; Miami; Orange County, CA; Greensville, NC; Minneapolis; and Phoenix.
Symmetry Episode Treatment Groups (ETGs), a classification system from Ingenix in Eden Prairie, MN, was used to adjust the data for differences in illness type and severity, physician specialty, and geographic region. Thirteen ETGs, such as ischemic heart disease, kidney stone, and knee joint derangement, were examined.
Based on 493,000 episodes of care, analysts found the physicians involved in in-office self-referring were more likely than non-referring physicians to order at least one imaging study for each of the 13 ETGs.
The size of the difference ranged from two to 23 percentage points. The variation was greatest for nuclear imaging of ischemic heart disease where 38% of the patients received at least one imaging procedure when they were attended by a self-referring physician, and 19% were imaged at least once when they visited a non-self-referring physician.
For MRI of spinal trauma, 37% of patients in a self-referred situation received imaging at least once compared with 22% in non-self-referred environments. For standard imaging of the knee and lower leg, the percentages for self-referrers and non-self-referrers were 58% and 35%, respectively. The differences for all but one of the treatment groups were statistically significant.
Episodes of care with a self-referring physician involved higher-than-expected spending on imaging as well. When controlling for illness type, severity, physician specialty, and geographic market, the difference ranged from 5% to 104%, depending on the ETG and type of imaging.
The study also found a correlation between in-office imaging self-referral and relatively high treatment costs, compared with therapies for patients who visited a physician who did not self-refer.
Based on 509,000 episodes of care across 13 ETGs, analysts determined that more patients who received self-referredimaging went on to receive care that was slightly more expensive than the follow-up for patients outside a self-referred situation. For cerebral vascular accidents, spinal trauma, ischemic heart disease, and congestive heart failure, the relative treatment costs were substantially higher.