Comparison of Recent Utilization Trends in Radionuclide Myocardial Perfusion Imaging Among Radiologists and Cardiologists

Abstract

To study recent practice patterns in radionuclide myocardial perfusion imaging (MPI) and related procedures among radiologists, cardiologists, and other physicians. The nationwide Medicare administrative Part B claims summary databases from 1998 and 2002 were used to assess utilization rate changes in the 4 primary procedure codes for MPI, the seven codes for diagnostic cardiac catheterization with coronary angiography, and the single code for stress echocardiography. Rate changes among radiologists, cardiologists, and other physicians were determined for the total Medicare population, as well as for the 3 primary places of service at which imaging is formed: hospital inpatient facilities, hospital outpatient facilities, and private offices. Ratios of the use of the 2 supplementary codes for left ventricular (LV) wall motion and ejection fraction to that of the primary MPI codes were calculated for 2002. The utilization rate per 1000 Medicare beneficiaries of MPI rose among radiologists from 19.8 in 1998 to 20.1 in 2002, a 2% increase. The rate among cardiologists rose from 22.9 in 1998 to 40.7 in 2002, a 78% increase. Most of this growth occurred in cardiologists' offices, where the utilization rate increased 101% over the 4 years. In 2002, the ratios of the use of the supplementary LV wall motion and ejection fraction codes to that of the primary MPI codes were 1.73 for cardiologists and 1.46 among radiologists. Between 1998 and 2002, the utilization of diagnostic cardiac catheterization among cardiologists increased by 19%, and their utilization of stress echocardiography increased by 21%. In recent years within the Medicare population, the rate of utilization of MPI among radiologists has remained relatively stable, whereas it has risen sharply among cardiologists. The greatest growth was seen in cardiologists' private offices. This raises concerns about possible inappropriate utilization of MPI and also about the potential effect self-referral has on this utilization trend. The increased use of MPI by cardiologists did not result in reduction in their use of either cardiac catheterization with coronary angiography or stress echocardiography.