Quality assessment of the FIM (Functional Independence Measure) ratings through Rasch analysis

The Functional Independence Measure (FIM(sm)) is an 18-item, 7-level ordinal scale, widely accepted as a standard measure of disability. The instrument is used mainly for analyses of effectiveness/efficiency of inpatient rehabilitation. National data banks are available, providing normative admission/discharge scores, thus allowing quality control of care. Besides the analysis of the overall score, the pattern of response to the various items can be used as a quality check of the data. The Rasch analysis is used to check the internal consistency of the FIM scores. This is a model of a more general item-response theory. On a common true interval scale, it allows the estimation of the 'ability' of the patients and the 'difficulty' of the items, on the basis of the raw interval scores. Measures of ability are independent of the specific items adopted, and measures of difficulty are independent of the specific sample of subjects tested. The Rasch analysis also allows identification of the 'misfitting' subjects, i.e., those who pass too difficult items and/or do not pass too easy items compared to their overall ability (e.g., due to clinical peculiarity, rating errors, rater's tricks). The average difficulty of the FIM items in the various impairments is then calculated in the large national sample. Within the samples sent by each facility to the data bank, cases misfitting with respect to the national response pattern can be easily singled out. This keeps the data bank free from ambiguous scores. In the meanwhile, these cases are brought to the attention of the submitting facility which can then opt for further examination or, when needed, for redirection of the treatment plan.