The 3-Meter Backward Walk Test (3MBWT): Reliability and validity in individuals with subacute and chronic stroke

Backward walking (BW) is an important gait adaptation and BW speed may be an important indicator of walking function and fall risk. However, the measurement characteristics of a standardized assessment of BW post-stroke have not been fully established.

To determine intra- and interrater reliability, concurrent validity and minimal detectable change (MDC) scores for the 3-Meter Backward Walk Test (3MBWT) post-stroke.

Thirty-four individuals with subacute and 29 individuals with chronic stroke participated. Two trials of comfortable BW was measured over a total distance of 5-meters, while speed was calculated during the middle 3-meters of the walking distance. Intra and interrater reliability were determined by comparing the two trials from one rater and simultaneous assessment of two raters, respectively. Two additional trials were performed and BW speed was calculated using 3MBWT and an instrumented walkway to determine concurrent validity. Intraclass correlation coefficients (ICC) estimated reliability and validity. The MDC was calculated from the standard error of measurement.

Excellent ICC values were obtained for the 3MBWT in the subacute (interrater: ICC_2,1 = 0.99; intrarater: ICC_2,1 = 0.96; validity: ICC_2,1 = 0.96) and chronic (interrater: ICC_2,1 = 0.99; intrarater: ICC_2,1 = 0.94; validity: ICC_2,1 = 0.97) groups. The MDC was 0.07 m/s (subacute) and 0.11 m/s (chronic).

Establishment of the 3MBWT as a reliable and valid measure in assessing BW speed is an important addition to the assessment toolbox for rehabilitation post-stroke.