RESUMO
Meniscal tears account for approximately 15% of all knee injuries and almost 25% of them require surgical procedures. Magnetic Resonance Imaging (MRI) is widely used for noninvasive assessment of the knee joint and is considered reliable and a powerful tool for the detection of soft tissue injuries of the knee. The aim of the study was to evaluate the sensitivity, specificity, and accuracy of magnetic resonance imaging (MRI) to predict the meniscal tears repair in sports practitioners. 104 incoming consecutive patients who underwent knee joint ligament reconstruction and/or arthroscopy for the treatment of meniscal injury at knee joint were imaged using a 1.5-T MRI scanner prior to arthroscopy. MRI images were evaluated for anterior cruciate ligament (ACL), articular cartilage, and meniscal injury. Images were correlated with arthroscopic findings, used as the gold standard. The sensitivity, specificity, and accuracy of MRI in predicting meniscal repair were 61.1%, 65.94%, and 64.58%, respectively. The agreement between MRI and arthroscopy yielded a kappa index of 0.236, indicating fair agreement. When the menisci were evaluated separately, 65.85% sensitivity, 45.45% specificity, and 54.16% accuracy were found for the medial meniscus, while 46.15%, 79.51%, and 75.0% for the lateral meniscus, respectively. The accuracy was 62.09% in whose patients that arthroscopy was performed up to 3 months after MRI and 67.18% in those that this time frame was more than 3 months before surgery. The 54 meniscal injuries occurred more frequently in the posterior horn; most injuries had a longitudinal pattern and were located in the red-red (vascular) zone. We suggest that magnetic resonance imaging is only moderately accurate for the prediction of meniscus reparability.