Jean-Franc ̧ois Ravaud, Michel Delcey and Alain Yelnik
DOI: 10.1080/16501977313141

The Functional Independence Measure (FIM) is one of the most widely used disability and dependence assessment instruments in rehabilitation medicine. As for other similar scales, the expression of results as a unique score raises an important question. Is it legitimate to consider the object being measured (functional independence) as a unidimensional entity? The answer is of major practical importance in justifying the use of the FIM. Having made a critical analysis of the previous validation procedures, the authors then submitted admission FIM items of 127 consecutive patients admitted in a French rehabilitation unit to different multidimensional statistical methods in order to analyse the structure of the FIM. Their findings demonstrate the multidimensional nature of the phenomenon assessed by the scale. This observation raises the question of the relevant use of the FIM total score, currently too widely applied without sufficient precaution, and suggests that preferably subscores should be used