THU0552 SCORING ULTRASOUND SYNOVITIS IN JUVENILE IDIOPATHIC ARTHRITIS: RESULTS OF A RELIABILITY EXERCISE USING OMERACT US PAEDIATRIC DEFINITIONS

Abstract

Background: Despite the increasing use of musculoskeletal ultrasound (MSKUS) in assessing joints in juvenile idiopathic arthritis (JIA) in daily practice, there is still a lack of data on its reliability in the evaluation of synovitis in children. Lately, the OMERACT ultrasound paediatric task force has developed preliminary definitions and scoring system for US synovitis in children (1). Objectives: To evaluate the reliability of the OMERACT paediatric definitions and scoring system of synovitis through a web reading images exercise. Methods: Thirteen sonographers with at least 2 years’ experience in paediatric MSKUS JIA were invited to participate in the reliability exercise. After a training session, participants were asked to analyze 75 images for the presence/absence of elementary lesions and for the grading of synovitis (2), synovial hypertrophy, effusion and Doppler signal. US images of elbow, wrist, second MCP, knee and ankle were collected using standardized scanning by 2 sonographers experienced (LR and SB), from JIA patients with different age in different disease stages. The intra-reader reliability (evaluated 4 weeks after the first assessment) was calculated by the Cohen’s kappa coefficient and by the quadratic weighted kappa and the inter-reader reliability by the Light’s kappa. Results: The intra-reader reliability was good for the detection of the presence/absence of any elementary lesion (median Cohen’s kappa 0.62, range 0.47-0.75), and for scoring synovitis and synovial hypertrophy and almost perfect for the scoring of Doppler signal (median quadratic weighted kappa 0.77, 0.66-0.86, 0.76, 0.61-0.84 and 0.87, 0.77-0.94 respectively), while it was moderate and rather variable for scoring effusion (0.55, 0.24-0.76). The inter-reader reliability was good for scoring synovitis and synovial hypertrophy (Light’s kappa 0.68, CI 95% 0.61-0.75 and 0.63, CI 95% 0.54-0.71 respectively), almost perfect for scoring Doppler signal (0.85, CI 95% 0.77-0.90) and moderate for the detection of any elementary lesion and for scoring effusion (0.48, CI 95% 0.36-0.64 and 0.49, CI 95% 0.40-0.60 respectively). We found a trend for better results in the evaluation of all parameters in older children (9-16 vs 0-8 years : light’s kappa 0.53 vs 0.44 for any lesion, 0.71 vs 0.64 for synovitis, 0.53 vs 0.43 for effusion, 0.71 vs 0.64 for synovial hypertrophy and 0.78 vs 0.90 for Doppler). Concerning the five joints evaluated the best results were obtained for the knee for all parameters (median Lights kappa 0.71, range 0.54-0.85) excepted for Doppler whose reliability was higher for second MCP (Lights’ kappa 0.87). Conclusion: Excepted for effusion, whose evaluation is limited on static images, the reliability of the OMERACT paediatric US definitions and scoring system for synovitis in JIA was good among a large group of sonographers. These results suggest their potential applicability in clinical practice and in multicenter studies. Reference: [1] Preliminary definitions for the sonographic features of synovitis in children. Roth J, Ravagnani V, Backhaus M, et al. Arthritis Care Res (Hoboken). 2017Aug;69(8):1217-1223. Disclosure of Interests: Linda Rossi-Semerano Grant/research support from: Roche, Sylvain Breton: None declared, Marouane Boubaya: None declared, Haykanush Ohanyan: None declared, Valerie Devauchelle-Pensec Grant/research support from: Roche-Chugai, Speakers bureau: MSD, BMS, UCB, Roche, Sandrine Jousse-Joulin: None declared