Inter- and intra-observer reliability of contrast-enhanced magnetic resonance imaging parameters in children with suspected juvenile idiopathic arthritis of the hip

Previous work at our institution demonstrated discrepancies between radiologists in interpretation of contrast-enhanced magnetic resonance imaging (MRI) in suspected hip arthritis. To assess inter- and intra-observer reliability of selected MRI parameters (effusion, marrow oedema and synovial thickness and enhancement) used in the diagnosis of juvenile idiopathic arthritis. A retrospective cohort study was conducted of patients with confirmed or suspected juvenile idiopathic arthritis who underwent hip contrast-enhanced MRI between January 2011 and September 2014. Three pediatric musculoskeletal radiologists independently assessed all scans for effusion, marrow oedema, measurement of synovial thickness, synovial enhancement and subjective assessment of synovium. Categorical variables were analysed using the Cohen κ, and measurement using Bland-Altman plots. Eighty patients were included. Interobserver reliability was moderate for effusion (κ=0.5-0.7), marrow oedema (κ=0.6), subjective synovial assessment (κ=0.4-0.5) and synovial enhancement (κ=0.1-0.5). Intra-observer reliability was highest for marrow oedema (κ=0.6-0.8) and lowest for effusion (κ=0.4-0.7). Intra-observer reliability for synovial enhancement (κ=-0.7-0.8) and subjective synovial assessment (κ=0.4-1.0) ranged from poor to excellent. For synovial thickness, intra- and interobserver Bland-Altman plots were well clustered around the mean suggesting good agreement. There were large differences across variables and only moderate agreement between observers. The most reliable parameters were presence of joint effusion and bone marrow oedema and subjective assessment of synovium. (orig.)