Objective: To assess the 2 most recently developed sets of spondylarthropathy (SpA) diagnostic and classification criteria [Amor and the European Spondyloarthropathy Study Group (ESSG)] in a survey involving a Spanish population with rheumatic disease.
Methods: The survey involved a cross sectional study of 1549 patients with rheumatic disease, who were examined over a week by 36 expert rheumatologists at 28 Spanish rheumatological centers. The head researcher at each participating center was to diagnose patients as having definite SpA, possible SpA, or as having other rheumatic diseases (definite controls) based on the physician's experience and assessment with no reference to the criteria under study.
Results: Overall a total of 218 patients were classified as having definite SpA, 1242 as definite controls and 89 as having possible SpA. An analysis of the patients diagnosed as definite showed 90.8 and 83.5% sensitivity, 96.2 and 95.2% specificity, 80.8 and 75.5% positive predictive value, and 98.4 and 97.0% negative predictive value for the Amor and ESSG criteria sets, respectively.
Conclusions: Both the Amor and ESSG criteria had excellent intrinsic (sensitivity and specificity) and extrinsic (positive and negative predictive value) performance, with no substantial differences between the 2. The results support use of the criteria for classifying SpA in daily rheumatological practice. Although not intended as diagnostic criteria, they make useful tools for the early identification of initial, atypical, or undifferentiated forms, which fail to strictly meet the diagnostic criteria of the entities that make up the SpA group.