Woodburn J et al. – With adaptations, FIS–RA was successfully translated and cross–culturally validated for use in 4 European languages. The two subscales can be used at the individual level for patient assessment and at the group level for research purposes.Methods
- The FIS–RA was translated from English to German, Hungarian and Dutch target languages and administered to 653 rheumatoid arthritis (RA) patients.
- Rasch analysis was undertaken on the impairment/footwear (FIS–RAIF) and activity/participation (FIS–RAAP) subscales for each language version separately and for pooled data.
- Overall fit to the Rasch model, item and person fit, unidimensionality, differential item function (DIF), and local response dependency were tested.
- To meet Rasch model assumptions, item deletion, subtests analysis and item splitting strategies were adopted.
- With the exception of the Hungarian FIS–RAIF subscale, preliminary fit to the Rasch model was unsuccessful for all target languages individually and pooled data.
- Multidimensionality, misfitting items, local dependency and DIF by age, gender, disease duration, and language were observed.
- With adjustment, fit to the Rasch model was satisfactorily achieved for all language versions.
- For the pooled data the Rasch model assumptions for cross–cultural validity were met following item deletion, subtest analysis and item splitting for language DIF.