Background The 20-item BRAF-MDQ was developed in the UK to measure RA fatigue, with a total score and 4 distinct subscales reflecting different fatigue concepts; there are 35 translations. If the translations adequately capture these concepts, then their internal factor structure should be similar; if not this would indicate important cultural differences in fatigue perception, requiring country-specific scoring. Criterion and construct validity have not yet been examined outside the UK.
Objectives To test the internal structure, criterion and construct validity of the BRAF-MDQ in France (F), Germany (G), Netherlands (N), Spain (Sp), Sweden (Sw) and UK.
Methods 100-200 RA patients in each country completed the BRAF-MDQ (sufficient for reliable factor analysis by country), the RA Impact of Disease (RAID) and the SF-36. BRAF-MDQ internal factor structure was tested by confirmatory factor analysis of the original 4 factors; internal consistency by Cronbach's alpha; criterion validity by Spearman's correlation with RAID and SF-36 vitality subscale; and construct validity by correlation with the remaining 7 SF-36 domains.
Results 1276 patients participated (F 206, G 216, N 317, Sp 157, Sw 170, UK 210). Overall, 76% were female; 10% <40 yrs old and 50% ≥60 yrs; 25.5% disease duration <5 yrs, with no major differences between countries. Overall mean BRAF-MDQ (0-70) was 26.04 (SD 16.18, range 0-70), country range 22.02 (G) to 34.19 (UK).
The original 4-factor structure was upheld in each country with all items loading >0.7: Physical fatigue factor 0.749-0.910, Living with fatigue 0.705-0.899, Cognitive fatigue 0.797-0.959 and Emotional fatigue 0.793-0.919. The factor structure was upheld in 476/480 further tests comprising the 4 factors tested in 20 randomly selected subsets of 50% of patients per country (bootstrapping).
Internal consistency was strong in all 6 countries for the total BRAF-MDQ: Cronbach's alpha 0.932-0.954. Subscales were also internally consistent in all 6 countries: Physical 0.75-0.791; Living 0.918-0.943; Cognition 0.912-0.955, Emotion 0.852-0.893.
Criterion validity was shown in the toal sample and all 6 countries by high correlations amongst the BRAF subscales (0.367-0.986); high correlations of the total BRAF with the BRAF subscales (r=0.815–0.928); and of the total BRAF and subscales with both the RAID fatigue item (0.512-0.878) and SF-36 Vitality item (-0.501 to -0.830).
Construct validity was shown in all 6 countries and in the total sample by appropriate correlations between the total BRAF, subscales and the remaining 7 SF-36 health domains: r=-0.223 (Sw, Emotion with Physical Function) to -0.760.
Conclusions The BRAF-MDQ has an internal factor structure that holds across 6 EU countries, supporting its conceptual validity internationally. It has evidence for internal consistency, criterion and construct validity in 6 countries, further supporting its increasingly widespread use.
Disclosure of Interest None declared