Evaluation of the performance of extreme patient-reported outcomes as surrogate markers for fibromyalgia in axial spondyloarthritis.

Abstract

The aim of the study is to evaluate the performance of extreme patient-reported outcomes (PRO) against definitions of fibromyalgia (FM) in patients with axial spondyloarthritis (axSpA). Ancillary analysis of the Predict-SpA trial, an observational study of axSpA patients receiving TNF-α inhibitor, was performed. 'Extreme PRO' was defined as a score ≥ 8 on three out of the first five Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) questions (scored 0-10). FM was defined by the American College of Rheumatology (ACR) 1990 criteria and the Fibromyalgia Rapid Screening Test (FiRST). Performances of 'extreme PRO' for FM were evaluated by the sensitivity, specificity and positive likelihood ratio using the 1990 ACR criteria as gold standard. As secondary analysis, the FiRST was used as the external standard. The prevalence of 'extreme PRO' in this population was 28.8% at baseline and decreased to 9.9% at 12 weeks. 'Extreme PRO' had low sensitivity 12 weeks after TNF initiation (0.18, 95% confidence interval [CI] 0.10-0.27 vs 0.60, 95% CI 0.50-0.71, at baseline), but high specificity (0.92, 95% CI 0.89-0.94 vs 0.78, 95% 0.74-0.82, at baseline), using ACR 1990 criteria as gold standard. Performances when tested against FiRST at 12 weeks showed higher sensitivity (0.27, 95% CI 0.20-0.35) and specificity (0.96, 95% CI 0.94-0.98). The proposed extreme PRO definition showed great specificity for FM recognition in patients with axSpA, suggesting it could be used in observational studies when specific items for FM classification are not available.