A prospective trial of serial gait speed as a measure of rehabilitation in the elderly.

Walking speed was measured on admission and then weekly during treatment of 125 subjects admitted to a geriatric rehabilitation ward. Walking speed was strongly related to the ability to stand up from a chair without help. Placement decisions at discharge were made without knowledge of gait speed data so that they could be used as a standard against which to compare walking speed as a valid and practical measure of mobility. It was found that a discharge walking speed of 0.15 m/s best separated immobile subjects who required long-term hospital care from those sufficiently mobile to be discharged home alone or to a rest home (54% versus 0% below cut-off, respectively). Relative walking speed (speed/height) was no better at predicting placement or mortality than walking speed. Uncorrected walking speed therefore remains the preferred clinical measure of velocity. A serial record showing improvement in walking speed proved useful in predicting eventual independent mobility of poorly mobile subjects. Thus walking speed is an objective yet inexpensive method of monitoring gait rehabilitation.