OBJECTIVE: To analyse cross-culture validity of the Functional Independence Measure (FIM) in patients with a spinal cord injury using a modern psychometric approach. SETTINGS: A total of 19 rehabilitation facilities from four countries in Europe. PARTICIPANTS: A total of 647 patients at admission, median age 46 years, 69% male. METHODS: Data from the FIM, collected on inpatient admission, was fitted to the Rasch model. A detailed analysis of scoring functions of the seven categories of the FIM items was undertaken before to testing fit to the model. Categories were rescored where necessary. Fit to the model was assessed initially within country, and then in the pooled data. Analysis of differential item functioning (DIF) was undertaken in the pooled data for each of the FIM motor and social cognitive scales, respectively. Final fit to the model was tested for breach of local independence by principle components analysis (PCA). RESULTS: The present scoring system for the FIM motor and cognitive scales, that is a seven category scale, was found to be invalid, necessitating extensive rescoring. Following this, DIF was found in a number of items within the motor scale, requiring a complex solution of splitting items by country to allow for the valid pooling of data. Five country-specific items could not be retained within this solution. The FIM cognitive scale fitted the Rasch model after rescoring, but there was a substantial ceiling effect. CONCLUSIONS: Data from the FIM motor scale for patients with spinal cord injury should not be pooled in its raw form, or compared from country to country. Only after fit to the Rasch model and necessary adjustments could such a comparison be made, but with a loss of clinical important items. The FIM cognitive scale works well following rescoring, and data may be pooled, but many patients were at the maximum score.
OBJECTIVE: To analyse cross-culture validity of the Functional Independence Measure (FIM) in patients with a spinal cord injury using a modern psychometric approach. SETTINGS: A total of 19 rehabilitation facilities from four countries in Europe. PARTICIPANTS: A total of 647 patients at admission, median age 46 years, 69% male. METHODS: Data from the FIM, collected on inpatient admission, was fitted to the Rasch model. A detailed analysis of scoring functions of the seven categories of the FIM items was undertaken before to testing fit to the model. Categories were rescored where necessary. Fit to the model was assessed initially within country, and then in the pooled data. Analysis of differential item functioning (DIF) was undertaken in the pooled data for each of the FIM motor and social cognitive scales, respectively. Final fit to the model was tested for breach of local independence by principle components analysis (PCA). RESULTS: The present scoring system for the FIM motor and cognitive scales, that is a seven category scale, was found to be invalid, necessitating extensive rescoring. Following this, DIF was found in a number of items within the motor scale, requiring a complex solution of splitting items by country to allow for the valid pooling of data. Five country-specific items could not be retained within this solution. The FIM cognitive scale fitted the Rasch model after rescoring, but there was a substantial ceiling effect. CONCLUSIONS: Data from the FIM motor scale for patients with spinal cord injury should not be pooled in its raw form, or compared from country to country. Only after fit to the Rasch model and necessary adjustments could such a comparison be made, but with a loss of clinical important items. The FIM cognitive scale works well following rescoring, and data may be pooled, but many patients were at the maximum score.
Authors: Kim Anderson; Sergio Aito; Michal Atkins; Fin Biering-Sørensen; Susan Charlifue; Armin Curt; John Ditunno; Clive Glass; Ralph Marino; Ruth Marshall; Mary Jane Mulcahey; Marcel Post; Gordana Savic; Giorgio Scivoletto; Amiram Catz Journal: J Spinal Cord Med Date: 2008 Impact factor: 1.985