AGREEMENT BETWEEN SEMIQUANTITATIVE AND QUANTITATIVE DOPPLER SCORING SYSTEMS FOR THE ASSESSMENT OF SYNOVIAL PATHOLOGICAL VASCULARIZATION IN RHEUMATOID ARTHRITIS

E. Naredo, I.M. Rutigliano, I. Janta, C. Scirocco, L. Valor, F. Figus, L. Carreño, J. Garrido, A. IagnoccoDepartment of Rheumatology, Hospital General Universitario Gregorio Marañόn Department of Rheumatology, Hospital Universitario Fundaciόn Jiménez Díaz. Autonoma University, Madrid, Spain Dipartimento Medicina Interna e Specialità Mediche, Sapienza Università di Roma, Rome Dipartimento Reumatologia, Università degli Studi di Cagliari, Cagliari, Italy Department of Social Psychology and Methodology, Faculty of Psychology, Autonoma University of Madrid, Madrid, Spain Dipartimento Scienze Cliniche e Biologiche, Università degli Studi di Torino, Turin, ItalyObjectives: To compare power Doppler (PD) vs colour Doppler (CD) semiquantitative and quantitative scoring of synovial vascularization (RA) and to evaluate the relationship between semiquantitative and quantitative scores in patients with rheumatoid arthritis (RA). Methods: One hundred RA patients underwent B-mode, PD, and CD assessments of 12 joints at two European centres. Each joint with synovial hypertrophy detected on B-mode was semiquantitatively scored (0–3) for PD (SPD score) and CD (SCD score) synovial signal. PD and CD synovial signal were also quantitatively scored (0–100%) (QPD and QCD scores, respectively) using a software for counting the colour fraction. Results: We found SH in 184 joints. SPD and SCD agreed in 92.3% (95% CI: 88.4; 96.2%) of paired scores, with Kendall rank correlation coefficient tau-b=0.95. Significant differences between marginal distributions of SPD and SCD were not found (p=0.565). QPD and QCD scores were highly correlated (Pearson's coefficient=0.70) but Blamd-Altman plot showed insufficient agreement, being the QCD scores systematically slightly higher than the QPD scores. The distribution of QPD and QCD values between SPD and SCD scores, respectively, showed significant differences between grade 0 and grade 1 (p<0.001), and grade 2 and grade 3 (p=0.042 and p=0.007, respectively) but not between grade 1 and 2 (p=0.154 and p=0.150, respectively). Conclusions: SPD and SCD scores were concordant and QPD and QCD scores highly correlated although were not concordant. There was consistency between SPD and SCD moderate and severe scores and QPD and QCD scores. There was an overlapping between SPD and SCD mild and moderate scores regarding QPD and QCD scores. Disclosure of Interest: None declared DOI: 10.1136/annrheumdis-2017-eular.3251Citation: Annals of the Rheumatic Diseases, volume 76, supplement 2, year 2017, page 1009Session: Diagnostics and imaging procedures (Poster Presentations )