Background: Magnetic resonance imaging of the knee is expensive and is neither needed nor useful for all patients presenting with knee pain. Our objective was to determine the completeness of evaluation prior to ordering magnetic resonance imaging of the knee correlated to the ordering providers’ postgraduate medical training and the rate of positive findings on the subsequent magnetic resonance imaging.
Methods: Six hundred consecutive knee magnetic resonance images were reviewed, including 200 consecutive knee magnetic resonance imaging examinations from each of three provider types: orthopaedic surgeons, non-surgical physicians with sports medicine training, and primary care providers. Positive findings on magnetic resonance imaging were recorded as well as a history of present illness, a physical examination, and radiographs made prior to ordering magnetic resonance imaging of the knee. Patient and injury factors were recorded. Differences in patient factors, evaluation before magnetic resonance imaging, and positive findings were examined. A modified Poisson regression approach was used to determine predictors of a proper evaluation before magnetic resonance imaging and positive findings on knee magnetic resonance imaging.
Results: Orthopaedists and non-surgical sports physicians were significantly more likely to document a physical examination, to evaluate radiographs made prior to ordering a magnetic resonance image, and to identify positive findings on the magnetic resonance image (all p < 0.001). In multivariate models, orthopaedists were more likely to document a history of present illness (relative risk, 1.05; p = 0.043). Compared with primary care physicians, a physical examination was more likely to be documented by both non-surgical sports medicine physicians (relative risk, 1.61; p < 0.001) and orthopaedists (relative risk, 1.60; p < 0.001) and positive magnetic resonance imaging findings were more likely to be found by non-surgical sports medicine physicians (relative risk, 1.41; p = 0.012) and by orthopaedists (relative risk, 1.44, p = 0.009). Other independent predictors of a magnetic resonance imaging study with positive findings were the presence of an acute injury by history (relative risk, 2.04; p < 0.001) and younger age (relative risk, 0.99; p = 0.021).
Conclusions: Orthopaedists and non-surgical sports physicians are more likely to perform and to document a complete evaluation prior to ordering a knee magnetic resonance image with a positive finding. More musculoskeletal training may be useful to enable primary care physicians to use magnetic resonance imaging of the knee in a more efficient manner.
1Department of Orthopaedic Surgery, University of Utah, 590 Wakara Way, Salt Lake City, UT 84108. E-mail address for J.D. Wylie: James.Wylie@hsc.utah.edu
2Department of Radiology, University of Missouri, One Hospital Drive, DC069.00, Columbia, MO 65212
3Department of Pathology, University of Utah, 30 North 1900 East, Room 5R441, Salt Lake City, UT 84132