[Validation of the functional independence scale]

To assess the psychometric quality of an instrument designed to measure functional independence (Functional Independence Scale [FIS]) in several activities of daily living domains and to be applied by trained non-health-related interviewers. The study was carried out in the autonomous region of Madrid in community-dwelling elders.

We performed a cross-sectional validation study. In addition to the FIS, Pfeiffer's questionnaire, the Depression Subscale of the Hospital Anxiety and Depression Scale, the Comorbidity Index, the Barthel Index, and EQ-5D were used. These measures were cross-sectionally applied to community-dwelling elders (n=500) and outpatients in a general hospital (n=100) aged 65 years. The following FIS psychometric attributes were analyzed: acceptability, scaling assumptions, internal consistency, construct validity, and precision.

A fully computable FIS total score was obtained in 94.3% of the subjects. A ceiling effect (60.65%), but no floor effect (0.22%) was evident in the community-dwelling elders. No floor or ceiling effects were detected in the hospital sample. Scaling assumptions and internal consistency were satisfactory (item-total correlations: 0.57-0.91; Cronbach's alpha: 0.94). Factor analysis identified three factors that explained 74.3% of the variance. Indexes of convergent, internal, and known-groups validity were satisfactory. The scale's precision, determined by the standard error of measurement (2.49; 95%CI=4.88), was also satisfactory.

The FIS is an easy-to-use instrument with appropriate metric attributes. This scale can be usefully applied in broad samples of non-institutionalized elders by non-health related personnel.