Reliability and convergent validity of the five-step test in people with chronic stroke

Objectives: (i) To estimate the intra-rater, inter-rater and test-retest reliabilities of the Five-Step Test (FST), as well as the minimum detectable change in FST completion times in people with stroke. (ii) To estimate the convergent validity of the FST with other measures of stroke-specific impairments. (iii) To identify the best cut-off times for distinguishing FST performance in people with stroke from that of healthy older adults. Design: A cross-sectional study. Setting: University-based rehabilitation centre. Participants: Forty-eight people with stroke and 39 healthy controls. Interventions: None. Main outcome measures: The FST, along with (for the stroke survivors only) scores on the Fugl-Meyer Lower Extremity Assessment (FMA-LE), the Berg Balance Scale (BBS), Limits of Stability (LOS) tests, and Activities-specific Balance Confidence (ABC) scale were tested. Results: The FST showed excellent intra-rater (intraclass correlation coefficient; ICC = 0.866-0.905), inter- rater (ICC = 0.998), and test-retest (ICC = 0.838- 0.842) reliabilities. A minimum detectable change of 9.16 s was found for the FST in people with stroke. The FST correlated significantly with the FMA-LE, BBS, and LOS results in the forward and sideways directions (r = -0.411 to -0.716, p < 0.004). The FST completion time of 13.35 s was shown to discriminate reliably between people with stroke and healthy older adults. Conclusion: The FST is a reliable, easy-to-administer clinical test for assessing stroke survivors' ability to negotiate steps and stairs.