Sensitivity and Specificity of Radiographic Scoring Instruments for Detecting Change in Axial Psoriatic Arthritis

Abstract

There is no widely recognized method used to assess axial disease in psoriatic arthritis (PsA). We aimed to determine the sensitivity to change of the Bath Ankylosing Spondylitis Radiology Index for the spine (BASRI-s), the modified Stoke Ankylosing Spondylitis Spine Score (mSASSS), the Radiographic Ankylosing Spondylitis Spine Score (RASSS), and the PsA Spondylitis Radiology Index (PASRI) in axial PsA. Radiographs of 105 patients with axial PsA were retrieved for 2 time points at least 2 years apart and subsequently anonymized. All radiographs were scored by 3 rheumatologists blinded to name and order of examination using an electronic application that allowed recording of disease manifestations specific to axial PsA and automatically calculated the BASRI-s, mSASSS, RASSS, and PASRI scores. An independent expert determined whether there was true radiographic progression from an overall impression after viewing the radiographs with knowledge of chronologic order. The sensitivity, specificity, and odds ratios for every 1-unit increase in the scores were determined to identify true change. Of the patients studied, 25 (24%) showed progression, as determined by the independent expert. The respective sensitivity and specificity values for an increase in score to detect true change were as follows: 0.48 and 0.78 (BASRI-s), 0.52 and 0.84 (mSASSS), 0.44 and 0.84 (RASSS), and 0.52 and 0.74 (PASRI). Logistic regression analyses showed that an increase of 1 point in the respective scores was associated with the following odds ratios for identifying true progression: BASRI-s 3.0, mSASSS 5.27, RASSS 3.70, and PASRI 3.06. Available scoring systems for quantifying radiographic axial PsA have moderate sensitivity but high specificity for detecting true change.