Gait speed is a responsive measure of physical performance for patients undergoing short-term rehabilitation.

Based on its ability to predict consequential outcomes, gait speed has been recommended as a "vital sign" for older individuals. Recommendations notwithstanding, the value of comfortable gait speed as a measure is limited unless clinicians can use it to identify real changes in performance over time. The purpose of this study, therefore, was to broadly describe the responsiveness of comfortable gait speed measured from patients admitted to short-term rehabilitation. For 136 consecutive patients who met inclusion and exclusion criteria (mean age=73.6 years), information regarding assistance, device use, and speed was obtained. Speed was measured as patients walked at a comfortable speed over 5.2m. The admission gait speed was a mean (SD) .40 (.18)m/s. The discharge speed was .59 (.20)m/s. The change in speed from admission to discharge (mean 15.1 days) was .20 (.16)m/s. The minimal detectable change (95%) was .13 m/s. The effect size was 1.11; the standardized response mean was 1.25. Based on receiver operating characteristic curve analysis, the minimal clinically important difference associated with 4 anchors (1 - subjective change in overall walking, 2 - change in assistance required, 3 - change in assistive device used, and 4 - change in speed based classification) ranged from .10 to .18 m/s. On the basis of multiple distribution- and anchor-based indicators, comfortable gait speed can be considered a responsive measure for patients undergoing short-term rehabilitation.