Development and validation of a Disease Activity index for PSoriatic Arthritis based on 44 joints (DAPSA44).

Abstract

Objectives: This study aims to develop a modified Disease Activity index for PSoriatic Arthritis (DAPSA) score using the 44-joint count instead of the 66 swollen/68 tender joint counts (SJC/TJC) in patients with spondyloarthritis (SpA) (axial SpA [axSpA], peripheral SpA [pSpA], and psoriatic arthritis [PsA]) and evaluate its construct validity.

Methods: Patients with axSpA, pSpA, and PsA included in the Assessment of SpondyloArthritis international Society-PerSpA (PERipheral involvement in SpondyloArthritis) study were randomly allocated 1:2 stratified for diagnosis and country into a derivation and validation cohort. For all patients, the 66SJC/68TJC were available, from which the SJC44/TJC44 were extracted. The derivation cohort was used to calculate the conversion factors for SJC66/TJC68 into SJC44/TJC44 to obtain the DAPSA44 using mixed-effects linear regression models. The validation cohort assessed the DAPSA44 construct validity through the agreement between continuous original-DAPSA/DAPSA44 (intraclass correlation coefficient [ICC]) and DAPSA disease activity states (weighted-kappa), Bland-Altman plot, Spearman correlations between original-DAPSA/DAPSA44 and external constructs, and discrimination between known groups (standardised mean differences [SMDs]) after stratifying patients into active/inactive based on a combination of patient global assessment (≥5/<5) and SJC (≥1/0 or ≥2/<2).

Results: A total of 4121 patients (65% axSpA, 10% pSpA, and 25% PsA) were included. Conversion factors from the derivation cohort (n = 1364) for TJC and SJC were 1.30 and 1.34, respectively. The validation cohort (n = 2757) analyses showed almost-perfect agreement between original-DAPSA and DAPSA44 (ICC 0.98, kappa 0.95). Spearman correlation coefficients between DAPSA44 and external constructs fell within the predefined 0.3-wide range of corresponding original-DAPSA coefficients. DAPSA44 demonstrated excellent discriminatory ability (SMD > 0.8) across all scenarios of active/inactive disease.

Conclusions: These findings support the use of DAPSA44 as a comparable tool to the original DAPSA for assessing disease activity due to peripheral arthritis in SpA, especially in situations where only a reduced joint count is available.