Methodology of the development of the EULAR Antiphospholipid Syndrome Disease Activity Score (EAPSDAS)

To describe the methodology of the development of a EULAR disease activity score in antiphospholipid syndrome (EAPSDAS).

In phase I/item generation, we used a systematic literature review (SLR) and 2 open-ended surveys among Task Force members, experts in antiphospholipid syndrome (APS)/collaborators, patients with primary APS, and health professionals in rheumatology (HPRs). In phase II/item reduction, similar items were eliminated, participants rated items on their importance to be included in EAPSDAS, and the Delphi methodology/voting was used. In phase III/Item scoring, we used real-world clinical vignettes to develop scores for included items. We applied linear regression analysis with physician global assessment (PhysGA) as the standard.

SLR retrieved 3834 articles; 47 were analyzed. Of 170 items generated from the SLR and initial surveys, 140 deduplicated candidate items were rated by 74 physicians, 65 patients, and 21 HPRs and voted on by Task Force members. Twenty-four items were included in the EAPSDAS thrombotic/microvascular/nonthrombotic (TMN) scale, and 6 in the obstetric scale. Three versions of 60 real-world clinical vignettes were randomly assigned to participating physicians, split into 3 groups. The items' score was calculated as the mean PhysGA adjusted for physician characteristics. Pairwise ranking comparisons analysis supported the mean vignette rankings. Analysis of ratings of the combinations of single manifestations showed no significant enhancement in scores except for 6 of 26 combinations of manifestations, supporting a 'bump-up' in these scores. TMN manifestations that should be scored if they remained stable or improved were determined by a survey.

EAPSDAS was developed using data-driven and consensus methodology.