SAT0620 Validation of The Dutch-Flemish Promis Fatigue Item Bank in Patients with Rheumatoid Arthritis. (15th July 2016)

• Record Type: Journal Article
• Authors: Crins, M.H.; Terwee, C.B.; Westhovens, R.; van Schaardenburg, D.; Smits, N.; Joly, J.; Verschueren, P.; van der Elst, K.; Cella, D.; Cook, K.F.; Dekker, J.; Boers, M.; Roorda, L.D.
• Is Part Of: Annals of the rheumatic diseases. Volume 75(2016)Supplement 2
• Page Start: 894
• Page End: 894
• Publication Date: 2016-07-15
• Subjects: Rheumatism -- Periodicals; 616.723005
• DOI: 10.1136/annrheumdis-2016-eular.2153
• Languages: English
• ISSNs: 0003-4967

Abstract

Abstract : Background: In the assessment of patients with rheumatoid arthritis (RA) it is important to measure fatigue. The National Institutes of Health's Patient-Reported Outcomes Measurement Information System (PROMIS) initiative developed an item bank for measuring fatigue. This item bank was translated into Dutch-Flemish language according to the FACIT methodology. Objectives: The aim of current study was to validate the Dutch-Flemish translation of the PROMIS Fatigue item bank (DF-PROMIS-Fatigue) in Dutch and Flemish RA patients. This is the first validation study of the PROMIS Fatigue item bank in RA patients. Methods: 2030 RA patients (1370 Dutch and 660 Flemish RA patients) completed a paper-and-pencil or web-based survey, including the full DF-PROMIS-Fatigue (95 items, 5-point Likert scale). One-factor confirmatory factor analysis assessed unidimensionality. Item response theory (IRT) models evaluated the item characteristics of the item bank, to facilitate future development of a computer adaptive test (CAT). A graded item response model (GRM) was fitted and construct validity was studied. Ordinal regression models evaluated Differential Item Functioning (DIF) for e.g. language (Dutch-Flemish vs. English and Dutch vs. Flemish), to analyse cross-cultural validity. Results: The analyses support unidimensionality of the DF-PROMIS-Fatigue (CFI=0.996 and TLI=0.996). Only 45 out of 4465 (1%) item pairs were marked as possibly locally dependent. The data of the item bank fit the GRM, and showed good coverage across the fatigue continuum (threshold-parameters ranged from -3.8 to 5.8). The item bank showed good cross-cultural validity: none of the DF-PROMIS-Fatigue items showed DIF between Dutch and Flemish language. Furthermore, the item bank showed good reliability (Cronbach's alpha=0.993). Analyses of construct validity are in progress and will be presented at the conference. Conclusions: The results indicate that the items of the DF-PROMIS-Fatigue fit a GRM. The DF-PROMIS-Fatigue can be used to develop a CAT for measuring fatigue in Dutch and Flemish RA patients. Disclosure of Interest: None declared