Factors Influencing On the Discordance Between 2011 ACR/EULAR Criteria and Physician's Clinical Judgment for Remission in Rheumatoid Arthritis Patients

Abstract

Background/Purpose: Remission is a primary end point in the treatment of rheumatoid arthritis (RA). To ensure more uniform reporting of outcome measures, American College of Rheumatology/European League Against Rheumatism (ACR/EULAR) recently developed new definitions of RA remission. The 2011 ACR/EULAR remission criteria provide the stringent definition of remission, but do not always concordant with the state of remission defined by physicians. The purpose of this study was to compare the concordance between new remission criteria and physician’s clinical judgment of remission and to identify influencing factors on discordance between them.

Methods: Total of 1,140 patients with RA were recruited from KORean Observational study Network for Arthritis (KORONA), a database generated by rheumatologist investigators across the South Korea. The frequency of remission was evaluated based on various definitions including the Boolean based ACR/EULAR criteria and physician’s clinical judgment. The agreement between them was estimated by Cohen’s kappa (κ). For the patients in remission according to Boolean based ACR/EULAR criteria and/or physician’s judgment (n=279), we divided them into three groups; Group 1(remission according to both criteria and physician’s judgment), Group 2 (remission only by criteria), and Group 3 (remission only by physician’s judgment). On multinomial logistic regression analysis, we identified influencing factors for both discordant groups compared to concordance group in remission

Results: Remission rates with the Boolean based ACR/EULAR remission criteriam CDAI, and SDAI were 10.5%, 16.4% and 17.2%, respectively. Remission rate according to physician’s clinical judgment was 18.4%, while DAS28 remission rate was 27.4%. The agreement between new criteria and physician’s clinical judgment for remission was low (κ =0.202) and the concordant remission rate was only 4.1% (n=51, group 1), while the prevalence of remissions by only criteria and physician’s judgment were 6.1% (n=69, group 2) and 13.9% (n=159, group 3), respectively. Multinomial logistic regression analysis shows that the pain affected on both discordant groups and sleep disturbance and fatigue were associated with remission only by physician’s clinical judgment. These indicated that patient’s subjective symptoms such as pain, fatigue, and sleep disturbance may influence on the discordance between 2011 ACR/EULAR criteria and physicians’ clinical judgment for remission(Table).

Conclusion: Although the 2011 ACR/EULAR remission criteria are stringent than physician’s clinical judgment, their agreement was low and patients’ subjective symptoms such as pain, fatigue, and sleep disturbance were associated with discordance between them.