Abstract
OBJECTIVES: To evaluate whether rheumatologists experienced in psoriatic arthritis (PsA) assess peripheral and axial involvement in the same way and to consider core clinical measurements that should be included in clinical trials in PsA. METHODS: Ten patients with PsA, representing a broad range of joint inflammation, joint damage, and spinal involvement, were selected for the study. Each patient was examined by each of 10 rheumatologists, members of the Spondyloarthritis Research Consortium of Canada, according to a Latin Square design. Assessments included scoring actively inflamed joints and damaged joints, dactylitis, enthesitis, and spinal measurements. Variance components analyses were conducted for continuous measurements based on models with observer, patient, and order effects. Estimates of intraclass correlation coefficients and associated 95% confidence intervals were obtained. RESULTS: There was substantial reliability in the assessment of the number of actively inflamed joints and excellent agreement in the number of damaged joints. Only moderate agreement was found for the number of digits with dactylitis. There was excellent agreement among observers in the intermalleolar distance measurements, but there was not as good agreement in the other measurements of spinal mobility. There was good agreement among the observers in detecting plantar fasciitis, however, the other entheses did not fare as well. CONCLUSION: In this first multicenter study of the assessment of clinical evaluation of patients with PsA we found that the assessment of peripheral joint disease is reliable although training should be performed prior to initiation of drug trials or comparative studies in this disease. The assessment of back measurements in PsA and other spondyloarthritis requires further study.