Thirty-two studies were included, but only the results for 26 were included in the primary analysis (n=3,386). The remaining 6 studies dealt primarily with adolescents, patients with osteoarthritis, features of clinical history, or the post-operative patient.
Five studies favoured the use of MRI before arthroscopy, while seven deemed routine MRI unnecessary. Those favouring MRI reported a sensitivity ranging from 84 to 100%, with a reported specificity ranging from 63 to 90%. For clinical examination, the sensitivity ranged from 77 to 100% and the specificity from 6 to 62%. These studies suffered from a number of methodological limitations, including failure to define criteria for a positive clinical diagnosis.
Results for the 7 studies that deemed MRI unnecessary reported a sensitivity/specificity/accuracy ranging from 55 to 100% for clinical examination and from 35 (lateral) to 100% for MRI examination. The studies that found MRI to be unnecessary generally involved diagnoses made by careful examination by an experienced orthopaedist.
Five studies using McMurray’s, Apley’s, Ege’s or Thessaly’s test at 5 degrees flex indicated high specificity and low sensitivity. By comparison, the Thessaly’s test at 20 degrees flex indicated high sensitivity and specificity. Joint line test results had higher sensitivity but lower specificity (4 studies).
Studies involving patients with acute injuries, patients with recurrent pain after a previous meniscal repair or partial meniscectomy, and studies involving adolescents, post-operative patients, patients with osteoarthritis, and features of clinical history were mentioned in the review.