VALIDATION DATA

Because each muscle group is scored individually and not included in a subscale, internal consistency analysis is not relevant for the FI-2

The measurement error for each task varies between 5 and 16%

The FI-2 demonstrated good to excellent intrarater reliability for all tasks, with intraclass correlation coefficients (ICCs) for the 7 tasks varying between 0.75 and 0.99.

To establish content validity, repeated administrations of the FI from patients with adult PM/DM were analyzed for floor and ceiling effects as well as for internal redundancy and consistency. No tasks were redundant, but grip strength, neck flexion, and trunk flexion (sit-up) showed poor internal consistency with other upper extremity tasks. These results were discussed with a group of health professionals and patients, and hip abduction, transfers, and peak expiratory flow were removed due to ceiling affects and lower relevance. Despite ceiling effects and poor intra- and interrater reliability, the neck flexion was considered relevant and remains in the tool.

The shoulder flexion task correlated moderately with the shoulder flexion isokinetic muscle endurance test (Spearman’s r 0.58) and less with other measures, confirming that the FI-2 assesses muscle endurance in patients with adult PM/DM. The knee extension task of the FI-2 (step test) correlated moderately with maximal isokinetic strength of the knee extensors (Spearman’s r 0.42), less with other constructs, and not at all with the isokinetic knee extension endurance test. This lack of correlation could be because the step test is performed in a closed-chain movement that also stresses the cardiovascular system, whereas the isokinetic test is open chained.