Comparison of the 1987 ACR and 2010 ACR/EULAR classification criteria for rheumatoid arthritis in clinical practice: a prospective cohort study

Objective: To compare application of the 1987 American College of Rheumatology (ACR) and 2010 ACR/European League Against Rheumatism (EULAR) classification criteria for diagnosing rheumatoid arthritis (RA) in clinical practice. Method: The medical records of patients with early arthritis attending the Rheumatology Department, Umea University Hospital (n = 1026) were analysed. Patients with synovitis in at least one joint, no diagnosis other than RA being better for explaining the synovitis, and duration of symptoms less than 1 year at first visit, and at least 1 year of follow-up were included consecutively. Fulfilment of the 1987 and 2010 criteria at baseline was evaluated. Sensitivity and specificity for each criterion set, where estimated by using the outcome measures: initiation of methotrexate (MTX) therapy during the first year, and a clinical diagnosis of RA at the 1-year follow-up. Radiographs of hands and feet were evaluated using the Larsen score. Results: The study included 313 patients, of whom 56% fulfilled the 1987 ACR criteria, 74% the 2010 ACR/EULAR criteria, and 53% both sets of criteria at baseline. The sensitivity/specificity for the 1987 and 2010 criteria with MTX within the first year as the outcome measure was 0.68/0.79 and 0.84/0.54, respectively, and with a diagnosis of RA at follow-up 0.72/0.83 and 0.91/0.65, respectively. Older patients (i.e. >= 60 years) more often fulfilled the 2010 criteria. Patients who fulfilled the 2010 ACR/EULAR but not the 1987 ACR criteria had a lower Larsen score at inclusion and after 2 years. Conclusions: Compared with the 1987 ACR criteria, the 2010 ACR/EULAR criteria have higher sensitivity but lower specificity, especially in patients aged >= 60 years. The 1987 ACR criteria are suggested to predict a more erosive disease.