MAGNETIC-RESONANCE-IMAGING OF THE KNEE - DIAGNOSTIC PERFORMANCE STATISTICS

Objectives: To review and reassess the published diagnostic performanc e statistics for MRI of the menisci and cruciate ligaments, To illustr ate the potential sources and effects of bias in the evaluation of thi s widely accepted diagnostic technique, Methods: Published evaluations of knee MRI were identified from the literature, Criteria for inclusi on in the review were a total sample size greater than or equal to 35, arthroscopic correlation of MRI findings and presentation of complete results. Diagnostic performance statistics were then recalculated for each published study, Results: Twenty-two studies were identified wit h sample sizes between 35 and 1014. The overall sensitivity for MRI of the menisci and cruciates was 0.88 (95% confidence interval 0.86-0.90 ). The overall specificity was 0.94 (0.93-0.94). Sampling error varied widely amongst studies and was rarely quantified, Conclusions: Diagno stic performance statistics are widely used. It is still not well appr eciated that these are subject to sampling error, Such errors make mea ningful comparisons between published studies more difficult. Neverthe less, the results for meniscal and cruciate lesions are consistently h igh and support the use of MRI for these common problems, The diagnost ic performance of other applications of MRI should be subjected to sim ilar critical review.