Purpose of the study: Magnetic resonance imaging (MRI) has the highest sensitivity of all methods for the diagnosis of intra-articular knee injuries. In spite of this, its benefit for the decision-making algorithm is questionable. The aim of this study was to evaluate the real situation in our regional conditions.
Material and methods: The medical records of the patients who underwent knee arthroscopy in 2008 and 2009, and had pre-operative MRI examination, were retrospectively reviewed. The group included 92 patients (46 women and 46 men; average age, 41.7 years) of whom 49 had knee injury in their medical history.
Results: In medial meniscus (MM) injuries, the MRI examination had a sensitivity of 0.92 and a specificity of 0.44, and the congruence of MRI and arthroscopic findings was 0.73. In lateral meniscus (LM) tears, the values were 0.70 for sensitivity, 0.81 for specificity and 0.87 for congruence. In injury to the anterior cruciate ligament (ACL), MRI sensitivity was 0.66, specificity was 0.85 and congruence was 0.79. In evaluation of articular chondral lesions, the values were 0.45 for sensitivity, 0.87 for specificity and 0.60 for congruence.
Discussion: In our examination of knee structures for MM, LM, ACL and cartilage injuries, the diagnostic value of MRI was lowest for cartilage damage, with sensitivity being only 0.45. This was in agreement with the findings of other authors. Although this fact is known, our arthroscopic findings in patients with no MRI evidence of injury were very high: 22 knees with grade III or grade IV chondral lesions. Therefore, MRI examination is not considered to be sensitive enough to replace arthroscopy in the diagnosis of cartilage injuries. MRI examination is most frequently indicated in suspected meniscal damage. Its sensitivity reported in the literature varies; generally, it is about 0.90 in MM injuries, and about 0.75 in LM lesions, and this is in agreement with our results. However, in view of our previous experience, the high sensitivity of MRI in the diagnosis of MM lesions was an unexpected finding. A detailed statistical analysis showed that its high value was at the expense of a relatively high negative positivity (0.56) and a low predictive value of the positive test (0.65). In LM injuries these values were even worse: in addition to low sensitivity (0.70), the predictive value of the positive test was only 0.50. In the diagnosis of ACL injuries, MRI examination is reported to have a high sensitivity ranging from 0.85 to 0.90. This study showed poorer results; sensitivity was 0.66 and the predictive value of the positive test was 0.62. For injuries in which the orthopaedist is sure about the diagnosis and indicates arthroscopy, it is doubtful to indicate also MRI examination. This should be reserved for clinical presentations that are not clear, for post-operative conditions, serious knee injuries or combined injuries. However, patients with chronic problems and unclear clinical presentations should benefit from MRI examination which can make the diagnosis more accurate and thus reduce the number of arthroscopic procedures done entirely for diagnostic purposes.
Conclusions: There is no consensus regarding the role of MRI in the diagnosis of intra-articular lesions of the knee. To a certain extent, its use is related to local conditions. It can be concluded that MRI examination is not currently as important for the diagnosis of knee injuries as expected by both medical and lay communities.