A comparison of the performance characteristics of classification criteria for the diagnosis of psoriatic arthritis

OBJECTIVE To compare the accuracy of published classification criteria for the diagnosis of psoriatic arthritis (PsA) and to see whether data-derived classification criteria would be more accurate. METHODS Data were abstracted from case-note review and radiographic review of patients identified with PsA or rheumatoid arthritis (RA) from 2 clinical disease registers. Each patient was classified according to 7 criteria sets. The test performance characteristics were compared using conditional logistic regression analysis. In an attempt to overcome the problems of the diagnostic gold standard, latent class analysis also was used to calculate test-performance characteristics. Classification and regression-tree methodology was used to derive new criteria and to indicate the diagnostic importance of particular data items, especially rheumatoid factor (RF). RESULTS Four hundred ninety-nine patients were identified with RA (n = 156) or PsA (n 343). Excluding the criteria of Fournie, which could not be applied in 24% of subjects, 446 cases could be classified by all of the other 6 methods. The most sensitive criteria for the diagnosis of PsA were those of Vasey and Espinoza, McGonagle, and Gladman (99%), whereas the others were significantly less sensitive (between 56% and 94%). The specificity of the criteria was high and statistically similar (between 93% and 99%). The Fournie criteria were the most difficult to use, whereas the Vasey and Espinoza and Moll and Wright criteria were the easiest (98% of subjects were able to be classified). A 2-latent class model found very similar test-performance characteristics. Logistic regression and classification and regression-tree models suggested that negative RF was not necessary for diagnosis in the presence of other characteristic features of PsA. CONCLUSIONS Apart from the Bennett and European Spondyloarthropathy Study Group criteria, which have inadequate sensitivity, the published classification criteria for PsA have similar test-performance characteristics. These data suggest that the criteria proposed by Vasey and Espinoza, Gladman, or McGonagle are the most accurate and feasible in distinguishing between PsA and RA.