Background and Purpose. The purpose of this study was to assess the intrasession and intersession reliability of measurements of quantitative gait variables at two self-selected walking speeds. Subjects. Forty-one patients with osteoarthritis in one or both knees who were referred for physical therapy participated. Methods. Three measurements were made at 1-week intervals. The quantitative gait variables of walking speed, cadence, and stride length were measured using an 8-m electronic footswitch walkway. At each measurement, subjects were asked to walk five times at a self-selected pace that they considered to be normal and five times at a self-selected pace that they considered to be fast. Results. At the normal walking speed, although intraclass correlations were consistently high for all gait variables, an additive factor within and across the first two measurements was evident even when a mean of several trials was used. At the fast walking speed, the intraclass correlations were again consistently high for all gait variables, but there were no changes within or across the measurements. Conclusion and Discussion. The data suggest that quantitative gait analysis is a practical objective assessment tool for persons with osteoarthritis of the knee. Gait at the fast walking speed, however, will provide the more reliable stable measure on which to evaluate the effect of therapy when compared with gait at the normal walking speed. The comparative responsiveness to change between the two walking speeds still needs to be determined.

Key Words:
  • Received June 7, 1996.
  • Accepted March 5, 1997.