Reliability and validity of Alternate Step Test times in subjects with chronic stroke.

OBJECTIVE (i) To investigate the intra-rater, inter-rater and test-retest reliability and minimal detectable change of the Alternate Step Test (AST) when assessing people with chronic stroke. (ii) To quantify the correlation between AST times and stroke-specific impairments. DESIGN Cross-sectional study. SETTING University-based rehabilitation centre. PARTICIPANTS A convenience sample of 86 participants: 45 with chronic stroke, and 41 healthy elderly subjects. METHODS The AST was administered along with the Fugl-Meyer Lower Extremity Assessment (FMA-LE), the Five Times Sit-To-Stand Test (FTSTS), limits of stability (LOS) measurements, Berg Balance Scale (BBS) scores, Chinese-translated Activities-specific Balance Confidence Scale (ABC-C) ratings, and the Timed "Up and Go" test (TUG). RESULTS Excellent intra-rater, inter-rater and test-retest reliability were found, with a minimal detectable change of 3.26 s. AST times were significantly associated with FMA-LE assessment, FTSTS times, LOS in the forward and backward directions and to the affected side, BBS ratings and TUG times. CONCLUSION AST time is a reliable assessment tool that correlates with different stroke-specific impairments in people with chronic stroke.