Background The classic SELENA SLEDAI flare index (cSFI) has been extensively used for assessment of flares in SLE with the purpose of differentiating severe flares from those falling in the mild/moderate range, and is composed of both clinical and medication criteria.

Objectives To examine the role of medication criteria in determining flare severity by cSFI.

Methods 182 encounters (91 patients, 2 visits each) were assessed. Flares were determined by the cSFI and compared to a physician-enhanced version (eSFI). eSFI differed from cSFI in that when criteria for mild/moderate flare were met, the rater distinguished between mild and moderate by a global opinion. The medication criteria were also eliminated in eSFI. Disease activity (PGA, SLEDAI, composite BILAG), number of BILAG-defined organs involved regardless of flare status and number of organs flaring based on revised SELENA flare index (SFI-R) domains were compared between the different groups.

Conclusions When medication criteria were removed from the classic SELENA SLEDAI flare index, severe flares had higher disease activity and number of organs flaring than when the medication criteria were used. Frequently medication-defined flares were clinically assessed as mild, moderate or not present, and many patients had medication changes due to restarting existing treatments or treating persistent (not flaring) disease. These data suggest that medication criteria falsely upgrade flares to severe.

Disclosure of Interest None Declared