In the United Kingdom, patients with a suspected ligament or meniscal damage are often seen in the accident and emergency department or peripheral clinic or the general practitioner in the first instance. A symptomatic treatment in the form of a knee support device or physiotherapy is offered until seen by a specialist and a definitive treatment is planned. This approach may reduce the pain and make subsequent clinical examination easier and more conclusive. On rare occasions the patient is seen directly by the concerned specialist.
The demographics of the population focused in our study were comparable and more than 50% were in the 4th and 5th decade. With increasing life expectancy and activity levels, we believe this age group will be a major subset of population seen in orthopaedic clinics in the UK.
A good history with particular reference to the nature of injury and a well-performed clinical examination will in most situations indicate the underlying problem. This is improved by experience, and arthroscopy may be justified on clinical grounds alone . Though the accuracy of clinical diagnosis of meniscal and ligament injuries has been varied in the literature [2, 3], a thorough clinical examination carried out by an experienced examiner in most situations will indicate the nature of the intra-articular injury. Clinical examination is as accurate as MRI and MRI should be reserved for confusing and special cases .
The decision to use an expensive investigative tool like MRI should be based on the criteria that the test will confirm or expand the diagnosis or change the diagnosis in such a way that this is going to alter the proposed treatment. It should supplement to formulate a therapeutic decision as well . This entirely rests on the treating physician. In unclear situations, the clinician requests an MRI for additional information to aid plan the operation and to predict the prognosis. This is compounded by high patient expectations, high degree of awareness amongst the public and availability of MRI in most district general hospitals in the UK. A wait period for an MRI and a definitive arthroscopy thereafter is inevitable considering the load in the National Health Service (NHS).
In knees with multiple ligament injuries, the diagnostic specificity of MRI for ligament tears decreases, as does the sensitivity for medial meniscus tears . MRI added valuable information in 4 clinically confirmed patients which helped the surgeon for better planning. MRI is useful but should be reserved for situations in which an experienced clinician requires further information before arriving at a diagnosis . Our observations agree with the above findings.
Though MRI has been recommended as a clarifying diagnostic tool , as in other studies we found MRI added little information to an already established clinical diagnosis . Interestingly in our study, patients in whom all the modalities fully agreed consisted of younger patients. Those with highly suggestive symptoms but with negative clinical tests had arthritic changes on plain radiographs, which were confirmed at arthroscopy. An accurate examination may be difficult even for an experienced examiner in this situation and it may be that an arthritic knee may not allow a complete examination. A conclusive diagnosis was therefore not possible. This may account for the low sensitivity of clinical tests in our study. In these situations, the value of MRI is heightened and invariably is requested for confirming the diagnosis.
In the middle aged and elderly patients a lower threshold of suspicion is warranted for meniscal tears as they follow minor trauma  and MR signal alterations are significantly higher in older population . MRI accuracy depends to a large extent on the structure studied, technical factors including imaging parameters, coil strength, surface coil use and planes of image . Partial tears of ACL may be identified as an altered signal alone and imaging may not be accurate due to the overlying synovial reaction . Further, the sensitivity of MRI for medial and lateral menisci being different there would be many lateral meniscal tears being missed and medial meniscal tears being over diagnosed . A high reliability on the MRI for a diagnosis and additional information will in these situations be a futile attempt . We agree with the above findings. A sound clinical judgment and experience is therefore required in the presence of a normal MRI. However the decision to do an arthroscopy was already made in these patients considering the clinical picture and MRI scans in these patients would have misled the surgeon into not doing an arthroscopy.
Cartilage lesions have not been addressed in the present study. Earlier studies suggested that MRI has a doubtful value in cartilage lesions . Even though un-enhanced MRI using a 1.5-Tesla magnet with conventional sequences (proton density-weighted, T1-weighted, and T2-weighted) is most accurate at revealing deeper lesions and defects at the patellae, a considerable number of lesions will remain undetected until arthroscopy . MRI scans with 3-Tesla field strength however improves the visualisation of hyaline cartilage with comparatively good diagnostic values but the positive predictive values remains low for all grades of lesions. . In our study, there were no traumatic cartilage lesions and most of the cartilage tears were degenerate and superficial, though we did not attempt to classify the tears as it was beyond the scope of the present study. MRI scans with 1 tesla field strength as in our study failed to highlight these tears in most of our patients accounting for a low sensitivity and specificity, which would perhaps been picked up by a higher field strength MRI scan. High quality MRI films may therefore still be useful in delineating the anatomical location and the geometry of the tear, as treatment options differ. This would thus help the surgeon in better planning but may not completely avoid an arthroscopy procedure. We presume that the plicae were symptomatic in a few patients as the symptoms resolved following removal.
Reports from radiology literature have highlighted the importance of quality reporting by experienced musculoskeletal radiologists [14–16]. To be of value, MRI of the knees should follow a specific protocol and should be performed and reported by experienced musculoskeletal radiologists . For practical reasons, it may not be possible to have a specialised musculoskeletal radiologist in all district general hospitals in the UK. With these subjective and inherent factors influencing the outcome of MRI report, it would seem unrealistic to base the decision to deny an arthroscopy on a negative MRI alone. As in other studies a negative MRI did not prevent us from doing an arthroscopy .
We recognise the limitations of this study in terms of the small numbers but believe that the groups studied are representative of the population normally attending the orthopaedic clinics.