When using the foot as the marker, establishing reliable gait speed measurement has involved averaging the findings using 6-10 passes on a known walkway. Collecting and analyzing this large number of passes is not practical for the clinical setting. The purpose of this study was to compare reliability of velocity measurement when using the trunk as a marker versus the foot for crossing walkway lines. We hypothesized that changing the marker from the foot to the trunk would result in better reliability and allow for fewer measurements to evaluate walking speed in a clinic setting.
Number of Subjects:
Ten Normal, Healthy Females Aged 22–36 Years Old
were timed and recorded walking 6 passes each for 5 conditions (2 self-selected speed, walking with head turning, carrying a light load, and talking). A 14-foot walkway was used with acceleration and deceleration lanes and a marked 10 foot area in the center. Walking speed data were collected using Expert Vision Advanced High Resolution, OrthoTrak 4.1 Motion Analysis, Inc using the Cleveland Clinic Marker Set. At the same time, walking was recorded using a AG-180 Panasonic VHS Movie Camera placed 17 feet from the walkway center. Time was recorded in seconds and milliseconds using a Sports Timer Stopwatch by playing the tape at its slowest forward speed (3.3% of normal). The stopwatch was depressed when either the foot or the trunk crossed the start and finish lines depending on the measurement being recorded (SMV).
Intra-rater reliability was established with both SMV measurement techniques by measuring the first trial of each of the 5 conditions for 5 subjects for four different times. Intra and inter-rater reliability was greater than 0.81 for all conditions. There were no significant difference in velocities between the SMV measurement technique and the Motion Analysis data. When comparing the two markers, the standard deviation of the walking speed measures of the 6 passes using the foot as the marker ranged from 1.9 to 2.5 times the standard deviation when using the trunk as the marker for the 5 conditions. For all but the normal walking conditions, the standard deviations were marginally or significantly higher when using the foot as the marker (t test P = 0.0014–0.10). Using random effects ANOVA, we calculated the intra-class correlation coefficients. The measurement reliability when using the foot was adequate and ranged from 0.45 to 0.65 for the five conditions, but the walking speed reliability when using the trunk was higher and excellent ranging from 0.84–0.92 for the five conditions. The intra-class correlation remained similar when only using the first 3 passes with the trunk as the marker.
Our hypothesis that fewer measurements are needed for reliable estimates of gait speed when using the trunk as a marker was supported.
Inexpensive reliable methods of measuring gait speed which require fewer passes and less data analysis are of practical importance.