Enter Your Email Address to Receive a
Copy of MedicExhange Member Demograhpics

Medicare should stop paying doctors who "self-refer" for imaging on equipment they own or lease -- excluding payments for x-rays -- because in most cases self-referral neither shortens the length of an illness nor provides cost-saving benefits, said researchers in the field.

An analysis of a random sampling of Medicare data for fee-for-service claims over a three-year period found that, with the exception of some x-rays, imaging self-referral not only failed to decrease the duration of an illness but also cost up to 10% more per total episode (from diagnosis to treatment) -- raising the per imaging cost by as much as 40% -- according to Danny R. Hughes, PhD, and colleagues at the American College of Radiology in Reston, Va.

The lone exception was for self-referred x-rays to diagnose chest pain, which shortened the duration of illness but did not make any significant difference in total costs, thus providing an overall advantage, Hughes and co-authors wrote in the December 2010 issue of Health Affairs.

The study contradicts arguments by supporters of self-referral -- where patients are sent for an x-ray, computed tomography ( CT ) scan, or other imaging using physician-owned or leased equipment -- that the practice leads to earlier imaging, quicker and more accurate diagnosis, faster treatment, and shorter duration of illness.

Critics argue that self-referral increases the use of imaging, raises costs, and increases patients' radiation exposure.

Medicare and Medicaid generally prohibit physicians from referring patients to facilities in which they have an ownership stake -- with the exception of referrals for ancillary services, including imaging -- as long as those services are delivered in the same office.

The Affordable Care Act of 2010 requires physicians who self-refer for imaging to inform their patients that they can receive imaging services elsewhere, according to background provided by the authors. The bulk of previous research of self-referral for imaging has largely focused on utilization, but hasn't examined the effect of the practice on illness duration and medical costs.

To study those factors, Hughes and colleagues culled Medicare claims data from a random 5% sample of Medicare fee-for-service beneficiaries from 2004 to 2007. The researchers relied mostly on "evaluation and management" claims for care delivered in nonhospital settings.

For the 733,459 medical episodes evaluated for the study, researchers calculated total costs as well as imaging and nonimaging costs, and determined that self-referral was associated with a higher mean total cost in 14 of the 20 medical conditions/imaging types they studied.

Hughes and co-authors found the largest difference in cost between self-referral and non-self referral was for the use of ultrasound to diagnose heart disease.

Compared with referrals to outside facilities, total average costs for entire episodes of self-referred care were 10% more for CT scans; 9% more for magnetic resonance imaging; 10.5% more for nuclear medicine; 7% more for ultrasound; and 4.5% more for x-ray.

However, x-rays were the one area where self-referral appeared to offer a benefit: Self-referral was associated with an average of 5% shortened duration of illness for x-rays to diagnose chest pain, respiratory illness, and sinusitis (P<0.01).

However, x-rays done for chest pain did not significantly affect costs, in contrast to those for respiratory illness and sinusitis. "Therefore, if only episode duration and cost are considered, self-referral in this case has an overall advantage because it combines a health benefit -- shorter duration -- with no cost disadvantage," the authors wrote.

Self-referral for x-rays to diagnose extremity dislocations and fractures resulted in cost savings compared with sending the patient elsewhere.

The study authors said the link between x-rays and shorter duration of illness is likely because x-rays are the only type of self-referred imaging that takes place on the same day as an office visit.

"Based on our findings, for types of imaging other than x-rays, we reject the hypothesis that self-referral for imaging in nonhospital settings is associated with shorter episodes of illness than occur in the absence of self-referral," Hughes and colleagues commented.

The authors point out that the Medicare Payment Advisory Commission (MedPAC) has suggested that the self-referral problem could be solved by reducing payments for in-office imaging. A 2005 law did in fact implement major reductions in payment for in-office imaging starting in 2007, and additional cuts are expected.

But the authors question that approach, adding that the rule of thumb is generally that cutting payments leads to increases in service volume.

"Based on our findings, we conclude that the Medicare exemption for self-referred imaging should be narrowed so that it includes only x-rays, not other forms of imaging," Hughes and colleagues wrote. "To the extent that state laws or private payers permit self-referral for imaging, they would also do well to follow this policy."

The authors point out that their study has a number of limitations. For example, they didn't investigate all possible benefits of self-referral, such as the convenience of having imaging performed the same time and place as the initial visit.

Source: Health Affairs