Parallel reliability of the functional independence measure and the Barthel ADL index.

Abstract

The aim of this study was to evaluate the concordance between the assessments of ADL according to the Functional Independence Measure (FIM) and the Barthel ADL index (BI) by means of a rank-invariant statistical method. The construct validity, also called parallel reliability, of FIM and BI was assessed on the item level. Two different approaches to condensing the FIM assessments into 2-4 scale steps on the item level in order to calibrate the two ADL instruments were compared. One was determined by the theoretical operational definitions and the other defined by the empirical definitions. The 204 assessments of elderly persons were made three months after stroke by means of interviews. The parallel reliability of the FIM and the BI on the item level was strong, both according to the theoretical cut-off levels defined by the operational definitions and the empirical cut-off levels defined by the marginal distributions. The concordance between FIM and BI was high. There was a slight difference in favour of the operational definitions. The clinical key elements are to have a critical attitude towards ADL instruments; how they are constructed and operationalized. These are important elements in the quality assurance in the everyday work.