OBJECTIVES : Timely referral and early start of treatment are necessary in rheumatoid arthritis (RA) to ascertain rapid improvement of functional disability. This might reduce the financial burden on individual RA patients, health services research and society. In 2010 new ACR/EULAR classification criteria were introduced to support the early initiation of treatment in RA patients. These classification criteria are widely adopted for use as aids in diagnosing RA in daily practice. This change in approach could have influenced the RA incidence in the rheumatology clinic. For socio-economic purposes it is necessary to estimate this time effect.

METHODS : A systematic review was conducted using Embase, Medline Ovid, Cochrane Central and Web of Science from database inception to April 2020. Included were only articles that addressed incidence rates in adults from rheumatology clinics. Three reviewers independently screened articles and performed data extraction. A meta-analysis was performed using a random effects model to assess the time effect.

RESULTS : In total 230 publications were found eligible for full-text review, after which nine studies were included for final analysis. Pooled incidence for RA was 6% (95% CI 3-8%) before 2010 and 15% (95% CI 8-23%) after 2010. Results show a significant increase in reported incidence of 9% for RA (p=0.017) after the introduction of new classification criteria. Overall high heterogeneity (I²=97.93) was present that was of clinical and methodological origin.

CONCLUSIONS : Rheumatoid arthritis is the most frequent inflammatory arthritis diagnosis in newly referred patients, of which over time the incidence increased up to 15%. Due to the heterogeneity between studies and differences in study design, it cannot be stated that the increasing trend is fully attributable to the introduction of the 2010 ACR/EULAR classification criteria.