Background Ultrasound (US) was found to have face and content validity for detecting synovitis in juvenile idiopathic arthritis (JIA) with higher sensitivity than clinical examination. In order to test validity and improve the applicability of US in JIA, the OMERACT US pediatric subtask force recently published preliminary definitions for the sonographic features of synovitis in children.
Objectives Aim of this study was to confirm and improve B-mode and color power/Doppler (PD) US definitions for synovial components and grading in children, by using an image and patient based exercise.
Methods The definitions were confirmed and modified in a multi-step process. In the 1st step, definitions were developed in multi-round Delphi web based consensus process were ≥80% of participants would need to reach ≥80% of agreement on a Likert scale from 1–5 (1 strongly disagree, 2 disagree, 3 neutral, 4 agree, 5 strongly agree). In the 2nd step, in a face to face meeting, a subgroup of these experts revised the definitions for final wording and performed intra- and inter-observer reliability exercise study in JIA patients as the final 3rd phase of the process. The definitions were tested in four joints (wrist, 2nd MCP, knee and ankle) of JIA patients divided in four age groups following standardized image acquisition and machine setting protocol. Statistics program R (version 3.3.0) was used for the statistical analyses. For intra-rater agreement Cohen kappa and for inter-rater agreement prevalence and bias adjusted kappa (PABAK) were calculated if needed.
Results Reliability exercise included 20 JIA patients (distributed in equal numbers by age groups), 14 observers, 4 joints/observer, 3 observers/joint, 360 intra- and 360 inter-observer tests. A 0–3 semi-quantitative B-mode and color power/Doppler US definitions for synovial components and grading, were agreed (presented in Figure 1).
Conclusions The proposed synovitis grading for children showed to be reliable why the next step should be to test sensitivity to change in order to possibly be used as an outcome tool in JIA.
Disclosure of Interest None declared