175 INTEROBSERVER AGREEMENT IN THE RADIOGRAPHIC DIAGNOSIS OF AVASCULAR NECROSIS OF THE FEMORAL HEAD FOLLOWING SLIPPED CAPITAL FEMORAL EPIPHYSIS.

Introduction Slipped capital femoral epiphysis (SCFE) is the most common pediatric hip disorder. Avascular necrosis (AVN) of the femoral head is a complication of SCFE with devastating consequences, including poor outcome, rapid collapse of the femoral head, early osteoarthritis, long-term morbidity, and the possibility for total hip arthroplasty at a young age. Slipped capital femoral epiphysis can be classified as stable or unstable on the basis of the ability to bear weight. The former is more common; however, the latter has been associated with AVN in up to 47% of cases. This has become the accepted prevalence in the literature; however, some studies have reported AVN in 15% of unstable SCFE. It is the objective of this study to estimate the interobserver agreement between two experienced pediatric orthopedic surgeons for the radiographic diagnosis of AVN following SCFE. Methods A retrospective review of all cases of SCFE treated at a Canadian pediatric referral center between 1995 and 2005 was performed. A total of 103 children were treated for SCFE during the time horizon of the study. Of these, eight were diagnosed, by the treating surgeon, with AVN. For the purpose of estimating the interobserver agreement for the diagnosis of AVN, each of these eight children and a random sample of the remaining children, who were not diagnosed with AVN by the treating surgeon, were included in this study. The most recent anteroposterior and lateral radiographs were digitized and presented to two experienced pediatric orthopedic surgeons in random order. Each surgeon reviewed the radiographs independently and recorded which radiographs they believed to be consistent with AVN. Interobserver reliability was determined by calculating kappa to assess for clinical agreement. Each observer repeated this process 2 weeks after the initial review. Clinical Relevance There is considerable variability in the reported prevalence of AVN following SCFE. The cause for this is unknown but may be related to interobserver agreement. If agreement between experienced pediatric orthopedic surgeons in the radiographic diagnosis is less than acceptable, then this may explain the variation in reported prevalence of AVN following SCFE. Further understanding of the possible causes of this discrepancy will lead to a better understanding of the burden of disease of AVN in SCFE.