Study faults self-referring cardiologists for MPI over-utilization
Given that myocardial perfusion imaging (MPI) is a mature technology and that there is no evidence of an increased incidence of coronary artery disease, the sharp rise in MPI utilization by cardiologists is likely due to in-office self-referral, according to a presentation by David C. Levin, MD, at the Radiological Society of North America (RSNA) conference last week in Chicago.

Levin from the Center for Research on Utilization of Imaging Services at Thomas Jefferson University in Philadelphia noted there is ongoing controversy between radiologists and cardiologists over who should be performing MPI.

He said that there is particular “concern among payers and policymakers about self-referral in imaging.”

Levin said that MPI is a good model to study because it is a high cost, high technology; commonly performed by both specialties; and it is most widely performed by non-radiologist physicians in self-referral settings.

In 2005, Levin et al published a study in the Journal of the American College of Radiology on utilization trends from 1998-2002. At RSNA, he presented their updated analysis, which extends to 2006. The researchers also reviewed utilization trends in related studies, like stress echo and diagnostic cardiac catheterization.

The investigators calculated the utilization rates per 1000 Medicare beneficiaries and studied trends.

Since 1998, Levin said that there has been a sharp increase in MPI utilization among cardiologists, while there has been a small decline among radiologists.

He reported that the vast bulk of the growth has been in cardiologists’ private offices; and while radiologists still do more MPI in hospitals, the gap appears to be closing. They also found that MPI by cardiologists is not substituting for catheterization “to any substantial degree.”

In conclusion, Levin referenced a Mayo Clinic study of the appropriateness of stress echoes and stress SPECT MPI, using the American College of Cardiology (ACC)/American Society of Nuclear Cardiology (ASNC) appropriateness criteria, (JACC 2008; 51:1290), which found:

 SPECT MPIstress echo
Number of patients284298

In the accompanying JACC editorial comment,  Robert O. Bonow, MD, wrote that the “increase in cardiovascular imaging procedures is not related to changes in disease prevalence or severity, and it has been difficult to tie the increase in imaging to improved health outcomes.”