TEST-RETEST RELIABILITY AND MINIMAL DETECTABLE CHANGE OF GAIT VELOCITY IN INDIVIDUALS UNDERGOING REHABILITATION AFTER STROKE

Abstract

Purpose/Hypothesis: Gait velocity is commonly used to assess walking ability in persons with stroke. Previous research related to the psychometric properties of gait velocity has been conducted primarily with individuals who were able to walk independently and/or were in the later stages of recovery. The purpose of this research was to examine the test-retest reliability and minimal detectable change (MDC90) of gait velocity with individuals with stroke undergoing inpatient rehabilitation who required varying levels of assistance to ambulate. Number of Subjects: Thirty-five individuals participated. The mean age was 67.4 years (std 13.8), mean time since stroke was 34.5 days (std 17.7), and median FIM™ locomotion score was 5 (range 2–7). Materials/Methods: Gait velocity was measured using a 5m walk test at a self selected comfortable pace on two occasions separated by one to three days. Test-retest reliability was assessed using ICC2, 1, and the MDC90 was calculated using the technique described by Stratford. Results: For all the subjects combined the ICC2, 1=0.862 and MDC90=0.30 m/s. For the 13 subjects that required physical assistance to walk the ICC2, 1=0.971 and MDC90=0.07 m/s. For the 22 subjects that could walk without physical assistance the ICC2, 1=0.80 and MDC90=0.36 m/s. Conclusions: Gait velocity is a reliable measure for both individuals who require physical assistance to walk and those who do not after stroke. These results indicate that in order to be certain that a change in gait velocity is not a result of chance variation there needs to be a greater change in individuals who can ambulate without physical assistance compared to those who require physical assistance. Clinical Relevance: Assessing walking ability using gait velocity in patients with stroke undergoing rehabilitation who do not require physical assistance requires a change greater than 0.36 m/s to be certain the change is not a result of measurement variability. For patients who require physical assistance to ambulate the change in gait velocity need only be greater than 0.07 m/s to be certain that the change is not a result of measurement variability.