December 2012
Feature Story

Anne Paxton

Cautious language may be the hallmark of most government publications. But a new report from the Government Accountability Office, an agency of Congress, has a title that doesn’t mince words: “Higher Use of Advanced Imaging Services by Providers Who Self-Refer Costing Medicare Millions.”

Self-referral occurs when providers refer their patients to entities—such as themselves or a group practice—with which they or immediate family members have a financial relationship. And the report, released Sept. 19, confirms what many other studies have found: Self-referral contributes to the runaway growth in Medicare part B expenditures. Analyzing Medicare payments for magnetic resonance imaging and computed tomography services, the GAO found that between 2004 and 2010, after accounting for practice size, provider specialty, geography, and patient characteristics, the number of self-referred MRI services increased by more than 80 percent, compared with a 12 percent increase for non-self-referred MRI services. For CT, self-referred services more than doubled, while non-self-referred CT services increased by 30 percent.

In 2010, the GAO estimates, self-referring providers made 400,000 more referrals for CT and MRI procedures than they would have if they were not self-referring—not only adding costs of more than $109 million that year to Medicare, but also unnecessarily exposing patients to hazardous radiation. Congressman Pete Stark (D-Calif.), whose name is attached to the two-decade-old Stark I and II laws barring physician self-referrals, said the report “should serve as a wakeup call to Congress that this is an arena where we can’t afford to sit idly by and allow providers to continue these practices.” The GAO believes it is imperative that the Centers for Medicare and Medicaid Services develop policies to address the effect of self-referral on the use of and expenditures for advanced imaging services.

“Simply put, what the GAO study shows is that when people own an imaging device, they end up using it,” says Richard Friedberg, MD, PhD, chair of the CAP Council on Government and Professional Affairs. “GAO’s comments on self-referral and imaging were fairly clear—and scathing.” The CAP, along with the American College of Radiology and six other groups in diagnostics, imaging, and physical therapy which have formed the Alliance for Integrity in Medicine, applauds the study and is encouraging the CMS to take steps to address the problems self-referral creates by ending the IOAS exception.

In the wake of the report, the Center for American Progress has called for expanding the Stark law—extending the self-referral prohibitions beyond Medicare to the private sector—and closing the existing self-referral loopholes for pathology, advanced imaging, and radiation therapy. The center estimates a health care savings of $1.5 billion over 10 years. “The statement by the center—a legitimate, independent think tank—is evidence of further support for our views,” says Dr. Friedberg, chair of the Department of Pathology, Baystate Health, Springfield, Mass.

Using a claims-based methodology for identifying self-referred service, the GAO found more than just a strong correlation between self-referring providers and greater volume of diagnostic imaging services ordered as of the date they began to self-refer. Looking at providers who in 2009 installed their own imaging devices or joined a practice that already self-referred, the GAO compared their usage levels with their previous referral levels. Called “switchers,” these providers referred an average of 25.1 MRI services in 2008, but that average rose to 42 in 2010. This 67.3 percent increase contrasted with a 6.8 percent decline in average referred MRI services by non-self-referrers.

The implications of the IOAS exception may be just as great in other areas of diagnostics as in radiology, Dr. Friedberg says, noting that a report on physician self-referral of anatomic pathology services will likely be issued by the GAO within the next few months. “Pathology is in a similar boat, as far as providers wanting to perform it in their offices in order to generate revenue. So it is a logical follow-on to look at pathology as well. We expect that that report, when it comes out, will have similar comments about the use of other diagnostic services.”

Dr. Friedberg sees the GAO study as a confirmation of well-documented behavior patterns when there are incentives at play. “There are certainly times when it’s appropriate to have in-office diagnostic services. We have always thought if it benefits patients on that visit, then it clearly can be in the patient’s interest. But if you are getting results days later, rather than at the time of the visit, then it is difficult to justify on the grounds of patient convenience or diagnostic necessity, and likely the only reason to do it is a financial one.”

For pathologists, self-referral is a problem of long standing. “It’s been a challenge for our members,” says Dr. Friedberg. The College recently helped fund a study by Georgetown University researcher Jean Mitchell, PhD, who found that urologists who self-refer prostate biopsies order those biopsies at much higher rates than non-self-referring urologists. “The Mitchell study confirmed the large volume of literature demonstrating the psychological effect of incentives on behavior. The GAO study on imaging essentially does the same,” Dr. Friedberg says.

The Stark rules were designed to prohibit physicians from referring to any entity in which they had a financial interest, he notes. “But what happened as a result is that providers would set up their own labs in their offices, always under the guise of good patient care. And often, it is good patient care—but it gets beyond that.”

The College has taken a multi-pronged approach to bringing self-referral in check, by exploring ways to control the excesses through the Medicare coding process or through possible legislation, Dr. Friedberg says. Last year, in addition to discussions with the CMS, the CAP arranged congressional fly-ins for members from around the country to raise the issue with their representatives in Congress.

But a key confounding factor is that group practices can vary greatly in size, making it difficult to sort out where self-referral begins and ends. “We have everything from a handful of doctors in a group who work together in a multispecialty practice, to large academic group practices such as the Cleveland Clinic. Where you have one part of the group referring to another part of the group, is that self referral, or not? In essence, how do you distinguish the legitimate referrals from the ‘gaming-the-system’ self-referrals? At what size of practice do the ethics really change? If there’s a group of 100 doctors that has a pathology group as part of it, is that self-referral or not?”

In addition, some might even point to the routine practice of reflex staining in anatomic pathology. “Some will argue that we also ‘self-refer’ for the reflex staining we do on our biopsies. We can counter that: We don’t determine the number of specimens; the surgeon determines that. And our testing of additional stains based on an initial read is analogous to reflex testing in the clinical lab, so it’s not the same kind of self-referral.”

Though there are those who understand the CAP’s position on self-referral, there have been no commitments to support the cause—in large part, Dr. Friedberg says, because much of the opinion within the house of medicine goes the other way. “The people who have similar issues to ours are the radiologists, although in many ways they are at greater risk, for the simple reason that radiology is a much bigger dollar item than pathology. And many more doctors think they can read an x-ray or CT than a biopsy.”

Recent discussions with the CMS have not been promising either, Dr. Friedberg says. “We’ve suggested a time-based approach, where only those services that benefit patients and inform treatment or diagnosis during the same visit may warrant exception under the Stark laws. CMS wasn’t willing to move forward on that. But we’re a relatively small group of people with a relatively small problem. And we’re probably more likely to get some solutions by piggybacking onto imaging as a bigger problem.”

Perplexing to many, the response of the Department of Health and Human Services to the GAO report has been fairly neutral, despite the report’s strong findings. The Department of HHS does not support insertion of a self-referral flag in Medicare part B claims, for example, though it said the CMS would consider other ways of ensuring appropriateness of advanced medical imaging. HHS believes that other payment reforms, such as accountable care organizations and value-based purchasing programs, will better address overutilization.

“Most of HHS simply sees this self-referral issue as a failure of fee-for-service,” not wanting to acknowledge clear winners and losers, Dr. Friedberg says. “The solution to a lot of what we’re complaining about, as they see it, is to move forward with health care reform, and that’s where we’re going right now. Health care reform is going to change the equation on this one.”

Dr. Friedberg encourages pathologists to look at the GAO study and its conclusions. “Self-referral of imaging is analogous to our problem, and it’s further evidence about the behavior of individuals—not just physicians—when incentives are there. It’s a strong study, and it’s fairly damning evidence that self-referral drives up utilization and costs.”