Objectives In a prospective cohort study of patients with undifferentiated arthritis (UA) to investigate the ability of ACR/EULAR 2010 criteria with or without different MRI variables to diagnose RA, considering ACR 1987 criteria as gold standard reference
Methods 119 pts (≥2 tender/swollen joints in MCP, PIP, wrist or MTP joints for >6 wks but <24 mths) without a rheumatological diagnosis, underwent clinical (68SJC/68TJC), biochemical, X-ray and MRI (unilateral wrist, MCP, PIP and MTP, scored by OMERACT RAMRIS (MTP not scored for synovitis (syn))) examinations and were followed for >12 mths. Sensitivity, specificity, accuracy, positive/negative predictive value (PPV/NPV) for diagnosing RA by 1) using ACR/EULAR 2010 criteria with and without including MRI syn, 2) substituting 68SJC/68TJC with MRI syn or 3) incorporating MRI bone oedema (BE) in the criteria are shown in Table.
Results ACR/EULAR criteria with MRI syn variables included, gave a slight increase in sensitivity (Table, row 2-4) compared to without MRI (row 1). BE increased the sensitivity even more, highest with BE in wrist and/or MTP joint (row 5-8). Highest sensitivity, specificity, agreement and PPV were demonstrated when 68SJC/68TJC was replaced with counting only joints with MRI syn (row 9-11). Using syn ≥1 as cut-off (row 9) the sens, spec, accuracy, PPV and NPV were all higher than using the criteria based on clinical examination (row 1).
Conclusions When fulfilment of the ACR 1987 criteria at follow-up is considered the gold standard reference, substituting IJC with MRI synovitis in joints of one hand, markedly increases the sensitivity and specificity of the ACR/EULAR 2010 criteria in UA.
Disclosure of Interest None Declared
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.