THU0269 CLINICAL COMPARISON OF NEW CRITERIA FOR SYSTEMIC LUPUS ERYTHEMATOSUS IN A COLOMBIAN COHORT

Abstract

Background:Due to heterogeneity of the disease, there has been several classification criteria for Systemic Lupus Erythematosus (SLE). These have considered the knowledge obtained through the years and have strived for increased sensibility and specificity. Recently, both EULAR and ACR have proposed new criteria for disease classification that mandate a positive ANA result to apply the criteria.Objectives:To compare the 2019 EULAR/ACR classification criteria (1) with the Systemic Lupus International Collaborating Clinics (SLICC) 2012 classification criteria (2) and the American College of Rheumatology (ACR) 1997 classification criteria in a Colombian cohort (3).Methods:A cross-section retrospective study was done with data collected between 2014 and 2018 from a population diagnosed with SLE by a group of rheumatology in an autoimmunity referral centre and followed for one year. The new 2019 EULAR/ACR classification criteria were applied to the information collected from the clinical records. Three sets of criteria were compared using Cohen´s kappa coefficient and concordance was evaluatedResults:We obtained information for 480 patients, in this analysis were mostly females (96%). Anti-nuclear antibody (ANA) results were available for 95% of the patients. According to SLICC classification criteria the diagnosis of SLE was definite in 92% of patients, 81% by ACR 1997 and 89% using ACR/EULAR 2019. The sensibility was 93% and 97% for ACR/EULAR 2019 and SLICC 2012, and the specificity was 67% and 48% respectively. The concordance analysis between the two sets of criteria showed agreement of 92% (kappa 0.52 p <0.001) in the whole group.Conclusion:We found good agreement between SLICC 2012 criteria and EULAR/ACR 2019 classification criteria. In contrast with previous studies, where the new criteria had a sensitivity of 96.1% and specificity of 93.4%, in our cohort the sensitivity was maintained in 93% but the specificity decreased to 67%. A possible explanation could be the ANA negativity that was seen in 5% of the patients and would force to discard patients with false negative results. Despite this, the agreement of the criteria is good and should continue to be applied in our population, without abandoning the expert’s clinical criteria.