Should RA Diagnostic Cut-off Points Be Lowered?

Do We Need to Lower the Cut Point of the 2010 ACR/EULAR Classification Criteria for Diagnosing Rheumatoid Arthritis?

van der Ven M, Alves C, Luime JJ, et al
Rheumatology (Oxford). 2016;55:636-639

Study Summary

Using a cohort of 557 patients from an early arthritis clinic in Rotterdam, The Netherlands, van der Ven and colleagues evaluated the effect of lowering the established cut-off from 6 points to 5 points for diagnosing rheumatoid arthritis (RA) using the 2010 American College of Rheumatology (ACR)/European League Against Rheumatism (EULAR) criteria.

After 1 year of follow-up, 253 (45%) of the patients were classified as "cases," defined as having been put on methotrexate for RA. At the baseline visit, the sensitivity for a cut-off point of a score of ≥6 was 61%, with a specificity of 76%. With a cut point of ≥5, the sensitivity increased to 76%, and the specificity decreased to 68%.

The authors concluded that lowering the cut-off of the 2010 criteria from 6 points to 5 points identified 15% more patients with RA, at the expense of including 8% false-positive patients.

Viewpoint

As wonderfully discussed in a recent review by Aggarwal and other members of the ACR's Subcommittee on Classification and Response Criteria,^[1] the use of "classification" and "diagnostic" criteria in rheumatic diseases is complex. Classification criteria are primarily designed for research in order to ensure that the patients studied are homogenous. As such, classification criteria may be somewhat arbitrary, capturing many important aspects of disease but perhaps not all elements of a particular patient's presentation. On the other hand, diagnostic criteria are designed to guide clinical care for individual patients but are actually quite uncommon in rheumatic diseases.

As van der Ven and colleagues pointed out, given the established and emerging data that indicate that treating RA early can lead to improved long-term outcomes, it may be that ultimately using an even lower cut-off to define RA and begin therapy may lead to improved long-term outcomes. One wonders if in real-life rheumatology practices and not in research studies this may already be occurring. In particular, it would be of interest to know how many practitioners are treating patients with inflammatory arthritis with methotrexate or other disease-modifying antirheumatic drugs when RA is suspected even if the classification criteria are not met—especially if the patient is rheumatoid factor and/or anti-citrullinated protein antibody positive.

More research is needed to help us understand when and whether to treat patients with inflammatory arthritis that appears to be RA. In particular, we need to determine what adverse effects could result from "overtreatment" of patients, although such events may be quite variable and based on patient-specific factors as well as the specific medications used. Overall, it may turn out that it is very reasonable and beneficial to have a low cut-off point to start RA treatment as long as discontinuation of medications over time is considered if the disease is controlled.

Abstract