Do we need to lower the cut point of the 2010 ACR/EULAR classification criteria for diagnosing rheumatoid arthritis?

OBJECTIVE

In this study we aimed to evaluate the effect of lowering the cut point of the 2010 criteria to identify more patients with RA among early inflammatory arthritis patients.

METHODS

We included early arthritis patients from the Rotterdam Early Arthritis Cohort with at least one joint with clinical synovitis and symptoms for <1 year, with no other explanation for their symptoms. The demographic and clinical characteristics of each patient were recorded at baseline. Patients were classified as case or non-case at the 1-year follow-up by the definition used in the development of the 2010 criteria (MTX initiation). To assess the diagnostic performance of the 2010 criteria, the sensitivity and specificity at each cut point were determined.

RESULTS

We included 557 patients in our analysis. At the 1-year follow-up, 253 patients (45%) were classified as case (MTX use). In the group of patients who scored 0-5 points (n = 328), 98 patients (30%) were classified as case (MTX use). The sensitivity and specificity of the 2010 criteria using the cut point of 6 were 61% and 76%, respectively. With the cut point of 5, the sensitivity would increase to 76% and the specificity would decrease to 68%.

CONCLUSION

By lowering the cut point of the 2010 criteria from 6 to 5 points, we were able to identify 15% more RA patients at the cost of 8% more false-positive patients.