A prospective study was conducted to evaluate the use of magnetic resonance
imaging (MRI) by orthopedic surgeons and residents versus the use of MRI b
y non-orthopedically trained health care providers in diagnosing knee patho
logy. Fifty patients initially evaluated by members of one of these groups
who underwent subsequent knee MRI evaluation were selected to participate.
Two orthopedic examiners individually examined all patients, recording clin
ical diagnosis and the merit of MRI evaluation in each case. Clinical accur
acy, sensitivity, and specificity were compared between groups based on MRI
findings. Diagnostic accuracy was similar; however, the orthopedic group d
isplayed greater sensitivity, suggesting better clinical assessment. The st
udy examiners observed both groups using MRI equally inappropriately and fo
und 62% of the imaging studies unjustified. We conclude that knee MRI is us
ed inappropriately in the current military setting. An algorithm is propose
d to guide the future use of MRI in the diagnosis and management of knee pa
thology.