Prospective evaluation of agreement and accuracy in the diagnosis of meniscal tears: MR arthrography a short time after injection versus CT arthrography after a moderate delay

OBJECTIVE. The purpose of this study was to compare the performance of direct CT arthrography performed after a moderate injection delay with that of MR arthrography performed soon after injection in the diagnosis of meniscal tears. SUBJECTS AND METHODS. Twenty-five patients underwent direct MR arthrography followed by same-day CT arthrography of the knee. Two blinded musculoskeletal radiologists independently reviewed the MR and CT arthrographic images for the presence of medial or lateral meniscal tears in the anterior horn, body, and posterior horn. Their readings were compared with a reference standard reading, which was based on post-MRI arthroscopic findings (n = 11) or the consensus opinion of two other musculoskeletal radiologists who simultaneously reviewed the paired CT and MR arthrographic examinations using all available clinical, surgical, and imaging information. The individual and combined radiologist agreements with the reference standard were calculated for each modality. RESULTS. Interreader agreement was 91% for MR arthrography and 85% for CT arthrography. The overall combined radiologist agreement with the reference standard was 91% for MR arthrography and 86% for CT arthrography (p = 0.03). For the two readers, the overall accuracy rates for diagnosing a meniscal tear were 82% and 88% with MR arthrography and 74% and 76% with CT arthrography. Only 50% (3/6) of surgically proven tears imbibed gadolinium. CONCLUSION. Although MR arthrography performed soon after the contrast injection had higher interreader agreement and greater accuracy, CT arthrography performed after a mean postinjection delay of 100 minutes was moderately accurate in the diagnosis of meniscal tears and can be used as an alternative procedure when MR arthrography cannot be completed.