TY - JOUR T1 - Interpreting Studies of Diagnostic Accuracy JF - The Journal of Rheumatology JO - J Rheumatol SP - 2340 LP - 2342 DO - 10.3899/jrheum.141109 VL - 41 IS - 12 AU - MICHAEL A. McISAAC Y1 - 2014/12/01 UR - http://www.jrheum.org/content/41/12/2340.abstract N2 - In this issue of The Journal, Payet, et al examine a test for elevated anticyclic citrullinated peptide antibody (anti-CCP) levels and demonstrate its inability to identify rheumatoid arthritis (RA) among anti-CCP-positive patients with rheumatic disorders1. Their results seem to be at odds with previous findings, which have shown the high diagnostic accuracy of anti-CCP tests for differential diagnosis of RA2,3. It is important, therefore, to consider how to appropriately interpret studies of diagnostic accuracy and assess generalizability. Such considerations are important when planning, reporting, or reading studies of diagnostic accuracy.Pepe4 lists the following 6 criteria for identifying settings where diagnostic tests would be useful: (1) the disease should be potentially serious, (2) the disease should be relatively prevalent in the target population, (3) the disease should be treatable, (4) the treatment should be available to those who test positive, (5) the test should not harm the individual, and (6) the test should accurately classify diseased and non-diseased individuals. Given that RA is a chronic disease (with a worldwide prevalence of 1%5) that can lead to severe disability, premature mortality6, and a loss of quality of life7, and given that appropriate therapeutic intervention can greatly enhance clinical outcomes6, it is clear that the first 4 criteria have been met in this setting. Anti-CCP antibody tests satisfy the fifth criterion, so it remains to establish that they can be used to accurately classify diseased and non-diseased individuals, which motivates studies of diagnostic accuracy such as that considered by Payet, et al1. Note that there is evidence that this sixth criterion could be met because anti-CCP tests have been shown to be useful in identifying patients with early-stage RA8 and predicting which patients will … Address correspondence to Dr. M.A. McIsaac, Public Health Sciences, Queen’s University, 99 University Ave., Kingston, Ontario K7L 3N6, Canada. E-mail: mcisaacm{at}queensu.ca ER -