CMS decision didn’t come  out of the blue

March 4, 2008

Restrictions on coronary CT angiographypayments respond to self referral, utilization

On Dec. 13, 2007, the Centers for Medicare and Medicaid Services issued a proposed national coverage decision on cardiac CT. Among its provisions was a decision not to reimburse for coronary CT angiography except for two indications, and those only when done within approved clinical trials.

The national coverage decision (NCD) supersedes all local coverage decisions by regional Medicare carriers. This decision, among the most restrictive ever issued by CMS, is at once surprising yet understandable. CMS finds itself caught in a dilemma, between technological advancement and fiscal responsibility.  

We have been performing CCTA in our practice on a strictly referral basis for three years. Our radiologists find the test to be accurate, demonstrating coronary anatomy extremely well. Over these past three years, the literature has corroborated the benefits of the examination. Nevertheless, our referrals are scant, because the cardiologists in our community are loath to adopt CCTA, as they are all imaging modalities that they do not control and for which they do not bill.

The literature has also been clear in another aspect of which CMS is surely aware: Economic self-interest frequently dictates referrals for examinations, corroborating our practice’s experience. As has been the case with other cardiac imaging studies, CCTA is especially susceptible to self-referral and its attendant overutilization. If this examination is inappropriately utilized, the economic cost could be catastrophic to our fragile Medicare Trust Fund.

Over the past year, the body of data demonstrating the effects of self-referral has been supplemented by several published research studies and comments to CMS regarding other proposed rules.

Research by economist Jean Mitchell, a professor of public policy at Georgetown University, on specialty hospitals and imaging showed that investment in hospitals and imaging affects practice patterns. Different physicians in the same region practice differently depending on whether they have ownership interest in the process. The same physicians’ practice patterns changed coincident with their acquisition of ownership interest in specialty hospitals and imaging equipment.

A New York Times article on Dec. 1, 2006, “Profit and questions on prostate cancer therapy,” described how urologists who get a $40,000-plus “machine fee” for prostate radiation therapy have abruptly changed their patterns of practice. The article quotes Leslie Norwalk, Medicare’s chief administrator, “. . . because of the potential conflicts, urologist-owned in tensity-modulated radiation the rapy is the type of arrangement that Medicare should be watching.”


Comments submitted to CMS on the differential use of certain tests suggest that their referral is often tied to the referring physicians’ ownership of the technology. That is, if they don’t own it, they don’t use it. On the other hand, frequently only doctors who own the equipment order the test. Following is one example:

The first situation involves an independent diagnostic testing facility providing cardiac PET/CT. This IDTF immediately embarked on a dual business model, leasing time to cardiology groups who would provide services as if the machine belonged to them and the rest of the time acting as an independent provider of the service. Data from three large private payers in the region show the following over the 18-month period through June 2007:

•Two hundred eighteen cardiac PET examinations were done in the entire state by three payers.

•Two of the sites are the two largest hospitals in the state, which own and bill for the examinations done on their scanners.

•A majority-134 of the 218 (63%)-were done at the site where the two cardiology practices  provide services on leased time. These are the only two cardiology practices that have the ability to bill for PET, by virtue of being in the same building. These two practices make up less than 20% of the cardiologists in the stated service area of the IDTF, yet virtually all the examinations in the service area were referred by them, and none was referred anywhere by any of the other cardiologists. Even though they are a small percentage of the cardiologists in the state, they perform nearly two-thirds of the total studies.

•All patients examined at the IDTF’s address were performed by the physicians in those two practices as the providers.

•None of the examinations was billed by the IDTF entity on patients referred by the general medical community. This raises serious questions as to the need for this examination if the other cardiologists never order it from the IDTF.

•One hundred eighteeen of the 134 (96%) were referred by physicians with same tax ID or practice address. This means only six of 134 examinations might have been in non-self-referred category. Even those might have been referred at the suggestion of the cardiologists.

Thus, we see the only demand comes from self-referring practices. No other cardiology practices in the area ever order the test, which is readily available at an independent facility.


Anecdotes reveal changes in patterns of care before and after acquisition of in-office imaging equipment: If owned, it is used by practitioners more than it was before ownership. An example of this is shown in Figure 1.

Research by Drs. David Levin and Vijay Rao in the Journal of the American College of Radiology  points out that noncardiac thoracic imaging has shown very little growth in recent years. The authors conclude the reason is that nonradiologists haven’t gotten very involved in doing it (JACR 2007;4[12]:886-889).

Another paper by the same authors reports the shift among radiologists from the invasive and expensive diagnostic catheter angiography to safer and lower cost MR angiography and CT angiography. At the same time, the utilization of diagnostic catheter angiography by nonradiologists has increased markedly, thereby allowing them to bill for the expensive, outdated procedure (The effect of the introduction of MR and CT angiography on the utilization of catheter angiography for peripheral arterial disease. JACR 2007;4[7]:457-460).

These examples show how economic interest not only increases utilization but also leads to changes in patterns of care based on the economic interest of the referring physicians.

The differential growth of imaging by radiologists and nonradiologists is well known to CMS. In addition, a 2007 Office of Inspector General report on the growth of imaging by the various specialties clearly shows that certain specialties are exponentially increasing their billing for imaging procedures. This growth cannot be sustained.


CCTA is a test whose dose of radiation to the patient can vary considerably, with higher doses in the range being among the highest of any examination we perform. Responsible physicians, organized medicine, and regulatory bodies must be vigilant about the unfettered proliferation of its use. A recent article in The New England Journal of Medicine  highlighted concerns about the potential dangers associated with medical radiation from diagnostic tests. While the methodology of the research is in some question, radiologists have always had a healthy respect for the tests we oversee, and we strive for lowest dose and appropriate exams.

In reaction to the NEJM  paper, nonradiologist Dr. Timothy Johnson, the well-known ABC medical commentator, railed against the added utilization that is causing patients to be exposed to potentially damaging radiation by physicians who are not qualified to oversee this machinery and who overorder because they have an economic interest.

We frequently see reports of situations in which financial investments by physicians affect their judgment with respect to research and patient care. In January 2008, three such situations were documented in the media. The Office of Inspector General reported that many physician-owned specialty hospitals do not meet Medicare guidelines, with inadequate staffing and substandard emergency facilities and processes. An NBC news report on overuse of coronary intervention was entitled “Conveyor-belt cardiology puts profits first.” On Jan. 30, The New York Time s reported on how spine surgeons participate in clinical studies on the effectiveness of devices in which they have investments.  

Such is the environment in which CMS must determine coverage for a new and expensive test, and one with a strong likelihood of being billed by the same physicians who order it. Thus, CMS’s decision that coronary CT angiography not be covered at this time is both unfortunate and understandable. The clinical data clearly support its validity. Those of us who know the potential benefits of the test believe that the decision will restrict availability of the procedure  and at the same time retard further refinement of the technique and technological enhancements.

On the other hand, those of us who are aware of the situations described, who have seen self-referral thwart patient care and cause undue cost and radiation exposure, understand the underlying dilemma at CMS. That this may be the beginning of an unfortunately restrictive stance on new technology is, perhaps, the most distressing aspect of the NCD.

CMS, along with responsible physicians of every specialty, must recognize that we cannot continue to allow the in-office ancillary services exception in the Stark laws, whose huge costs have such an overbearing effect on the advancement of patient care. The cost of in-office imaging is driving decisions on coverage that  impede the development and adoption of innovations to the detriment of patients.