Correspondence between patient-reported flare and disease activity score variation in axial spondyloarthritis: A 12-months web-based study.

Abstract

The aims of this study were to determine thresholds of variations of BASDAI and pain associated with patient-reported outbreak or resolution of flare and to test performance of ASAS preliminary definitions of flare. SpA patients registered on the Spondy+platform were invited to fill BASDAI and global pain on numeric rating scales every week during one year and to tell if they experienced flare since last week. Performance of BASDAI and pain variations (ΔBASDAI and Δpain) to detect occurrence and resolution of flare was determined with receiver operator characteristic (ROC) curves. Ninety-one of the 99 axSpA patients included reported at least one episode of flare. Area under the ROC curve was significantly higher for ΔBASDAI than for Δpain to predict outbreak of flare (0.81 vs. 0.77, p<0.05) without statically significant difference to predict flare resolution (0.78 vs. 0.80). Best sensitivity/specificity compromise was obtained for ΔBASDAI of 0.2 and 0.4 points to predict flare outbreak or resolution, respectively. All the ASAS definitions obtained a specificity higher than 95% whereas sensitivity was lower than 40%. ΔBASDAI appeared as a suitable variable to monitor occurrence and resolution of patient-reported flare in axSpA.