Laurence Parker PhD, Presenter: Nothing to Disclose

Andrea J. Frangos MPH, Abstract Co-Author: Nothing to Disclose

David C. Levin MD, Abstract Co-Author: Consultant, HealthHelp Board of Directors, Outpatient Imaging Affiliates, LLC

Vijay Madan Rao MD, Abstract Co-Author: Nothing to Disclose

This study examines geographic variation in magnetic resonance imaging (MRI). Geographic variation in medical practice is well established in the literature. MRI is a high-technology, high-cost modality whose utilization is of current concern. However, little is known about geographic variations in its usage.

The data source for the study was the Center for Medicare and Medicaid Services (CMS) 2004 Limited Dataset (LDS). The dataset is a 5% radom sample of charges for the 37.7 M fee-for-service Medicare beneficiaries. From the physician file, we extracted all allowed MRI claims. The file provides a variable that identifies the state, territory, or foreign residence of the beneficiary of each service. We calculated state fee-for-service denominators using the 2004 State County Files, which report Medicare enrollees and HMO market share for each county. We then calculated rates and 95% confidence intervals (CI) for each state and overall, for head, spine, musculoskeletal (MSK), body, cardiovascular (CV), and total MRI.

Overall MRI utilization was 155/1000 beneficiaries. The five highest states were Delaware, 223; Florida, 205; Alabama, 195; Texas, 185; and, Maryland, 183. The five lowest were Montana, 122; Iowa, 119; North Dakota, 112; Vermont, 106; and Hawaii, 106. Because the denominators are so large, the 95% CI for the total sample is 0.2/1000, the largest state 0.4, and the smallest, 2.8. Ratio of highest to lowest state MRI utilization is: all, 2.2; head, 2.6; Spine, 2.9; MSK, 3.0; Body, 4.6; and CV, 9.4. We included two territories in our analysis, the Virgin Islands and Puerto Rico, and the rates, 102 and 80, are lower than any states.

There is substantial statewide geographic variation in Medicare MRI utilization, more than two-fold overall, and more than four-fold and nine-fold in still-growing areas of MRI, Body and CV. This variation deserves further study, both for over- and under-utilization issues.

Does not apply.

Parker, L, Frangos, A, Levin, D, Rao, V, Variation across States in Magnetic Resonance Imaging Utilization in the Medicare Population: 2004.  Radiological Society of North America 2007 Scientific Assembly and Annual Meeting, November 25 - November 30, 2007 ,Chicago IL. http://archive.rsna.org/2007/5002821.html