Background/Purpose: Previous studies evaluated performance of the 2010 American College of Rheumatology (ACR)/European League against Rheumatism (EULAR) criteria for classification of rheumatoid arthritis (RA) in clinical settings. However, information about performance of these criteria in population-based studies is scarce. We aimed to estimate RA incidence during 2010-2014 using 1987 ACR and the 2010 ACR/EULAR criteria and compare the performance of these criteria sets.

Methods: We evaluated RA incidence in a population-based inception cohort of individuals 18 years of age and older based on fulfillment of the 1987 ACR and the 2010 ACR/EULAR criteria between 1/1/2010 and 12/31/2014. Both rheumatoid factor (RF) and cyclic citrullinated peptide antibody (anti-CCP) were considered for criteria fulfillment. Seropositive was defined as RF and/or anti-CCP positive, and seronegative was defined as neither positive. Incidence rates were estimated and were age-and sex-adjusted to the white population in the US in 2010.

Results: There were 221 patients who first fulfilled 1987 or 2010 criteria for RA in 2010-2014 (216 met 1987 criteria and 175 met 2010 criteria). Table shows RA incidence rates per 100,000 population for the 2010-2014 time period overall and by seropositivity. The estimates for the overall age- and sex-adjusted annual RA incidence were higher when 1987 ACR criteria were applied as compared to the 2010 ACR/EULAR criteria due to lower estimates for seronegative patients identified with the 2010 ACR/EULAR criteria. There were a total of 36 seronegative patients who met 1987 ACR criteria but not 2010 ACR/EULAR criteria in 2010-2014. When these patients were compared with the other 47 seronegative patients who met both 1987 ACR and 2010 ACR/EULAR criteria in 2010-2014, patients fulfilling both criteria had higher joint counts (p< 0.001) and were more likely to be ever smokers (p=0.02). The rest of characteristics of these two groups of seronegative patients were similar, including age at RA incidence, sex, race, body mass index, proportion of patients with prolonged morning stiffness, arthritis in 3 or more joints, arthritis in hand joints, rheumatoid nodules, erosions, abnormal erythrocyte sedimentation rate or C-reactive protein, disease duration >6 weeks. There were no differences between these groups in the proportion of patients started on methotrexate, hydroxychloroquine, sulfasalazine, or other disease-modifying antirheumatic drug (DMARD) as their first DMARD.

Conclusion: The incidence of RA and particularly seronegative RA in population-based studies may be underestimated by 2010 ACR/EULAR criteria as compared to the 1987 ACR criteria for classification of RA. Seronegative RA patients who fulfill 1987 ACR criteria but not 2010 ACR/EULAR criteria have lower joint counts but have otherwise similar RA disease characteristics as patients who fulfilled both sets of criteria.

« Back to 2019 ACR/ARP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/performance-of-2010-acr-eular-and-1987-acr-criteria-for-classification-of-rheumatoid-arthritis-in-a-population-based-incidence-cohort-2010-2014/