The purpose of this study was to assess the reliability and accuracy of the modified Marcus and Enneking classification of osteonecrosis of the femoral head in patients with systemic lupus erythematosus when applied to the reading of the original radiographs and then applied to interpreting the radiologist's written report. Thirty anteroposterior radiographs of the hip from patients with systemic lupus erythematosus with varying degrees of osteonecrosis were selected and submitted to a panel of two rheumatologists and an orthopaedic surgeon who individually graded the radiographs using the modified Marcus and Enneking method. Agreement was determined by the kappa statistics. Identical staging was assigned in 21 of the 30 radiographs by all three observers, providing a weighted kappa of 0.88. Identical staging was assigned by all three observers in 26 of the 30 reports, giving a kappa of 0.95. There was agreement in the staging of 23 of the 30 radiographs by actual reading and the radiologist's report, with a weighted kappa of 0.86. Discordant gradings resulted primarily because the Marcus and Enneking method does not account for degenerative changes on the femoral side of the hip joint. The authors conclude that the radiologist's report can be used to stage osteonecrosis when interpreted by expert observers and when those reports contain all critical information required for staging. The authors recommend that the Marcus and Enneking classification be modified to include changes on the femoral side of the joint.
University of Toronto Rheumatic Disease Unit, The Toronto Hospital; and University of Toronto Lupus Clinic, Centre for Prognosis Studies in the Rheumatic Diseases, The Toronto Hospital, Toronto, Ontario, Canada
†Division of Orthopaedic Surgery, The Wellesley Hospital, Toronto, Ontario, Canada;
††The Department of Statistics, University College, London, England.