Variation in MRI/CT utilization among FAMILY physicians and general internists in a multi-specialty group practice

Med Sci Monit. 2005 Mar;11(3):MT19-25.


Background: The objective of this study was to examine variations in MRI/CT utilization between family physicians (FPs) and general internists (IMs) within a multi-specialty group practice.

Material/methods: Using administrative data, we computed ordering rates of MRI/CT per 1,000 outpatient clinic visits and per 1,000 unique patients and rate ratios (RR) to compare rates between 34 FPs and 24 IMs practicing in 7 clinics located within 50 miles of the radiology facility. We also assessed intra-specialty variation. Sources and degree of variation were determined separately for FPs and IMs through multivariate linear regression modeling.

Results: The IMs ordered MRI/CTs at twice the rate of FPs (29.6 vs. 14.8 per 1,000). Although the absolute ranges by specialty were statistically similar, the variance was significantly higher for IMs (86.1 vs. 52.3; p<0.0001). FPs' ordering rates ranged from 2.8 to 35.2 (SD=7.23), while IMs' ranged from 16.0 to 47.9 (SD=9.27). Female physicians ordered the tests at a higher rate (RR=1.38; 95% CI=1.17-1.53). After controlling for physician gender, years of practice, and patient panel size, distances from their clinics to the radiology facility site and patient severity index were the only variables that were significantly associated with MRI/CT ordering among FPs, explaining 39% of the total variance.

Conclusions: Although IMs ordered MRI/CTs at a higher rate, as were females from both specialties, there was a higher ratio between high and low FP utilizers. The variation in FPs can be partly explained by their clinic distance to the radiology facility site and their patient severity index.

Publication types

  • Comparative Study
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Ambulatory Care / statistics & numerical data
  • Data Collection
  • Female
  • Group Practice / statistics & numerical data*
  • Health Services / statistics & numerical data
  • Humans
  • Internal Medicine / statistics & numerical data*
  • Linear Models
  • Magnetic Resonance Imaging / statistics & numerical data*
  • Physicians, Family / statistics & numerical data*
  • Practice Patterns, Physicians' / statistics & numerical data
  • Professional Practice / statistics & numerical data
  • Regression Analysis
  • Severity of Illness Index
  • Texas
  • Tomography, X-Ray Computed / statistics & numerical data*