Assessing performance-related disability in patients with knee osteoarthritis.

This investigation describes the development and validation of a test battery for evaluating physical activity restrictions (PAR) in patients with knee osteoarthritis (OA). The tasks on the final version of the PAR include (a) a 6-min walk; (b) a stair climb; (c) a lifting and carrying task; and (d) getting into and out of a car. Data from patients with knee OA revealed that the four tasks loaded highly on a single unrotated principal component yielding an alpha internal consistency reliability of 0.92. These data suggest that investigators may choose to use an aggregate score and/or responses from individual tasks. Two week test-retest reliabilities for the four tasks were all in excess of 0.85 and there was support for their concurrent and convergent validity. Specifically, performance on the tasks correlated with time on treadmill, difficulty with self-reported ADLs, and ratings of difficulty following the performance of each task. Additional research is needed on the predictive validity of the measure and its sensitivity to change.