LAUSR

Background: Walk tests are commonly used to evaluate walking ability in frail older adults with dementia but their psychometric evidence in this population is lacking. Objectives: 1) To examine test‐retest and inter‐rater reliability, construct and known‐group validity, and minimal detectable change at 95% level of confidence (MDC95) of walk tests in frail older adults with dementia, and 2) to examine the feasibility and consistency of a cueing system in facilitating participants in completing walk tests. Design: Psychometric study with repeated measures. Setting: Day care and residential care facilities. Participants: Thirty‐nine frail older adults with a mean age 87.1 and a diagnosis of dementia or Alzheimer's disease who were able to walk independently for at least 15 m. Methods: The participants underwent a 2‐minute walk test (2MWT), 6‐minute walk test (6MWT) and 10‐meter walk test (10MeWT) on 6 separate occasions under 2 independent assessors using a cueing system. Functional status was measured using the Elderly Mobility Scale (EMS), Berg Balance Scale (BBS) and Modified Barthel Index (MBI). Results: Excellent test‐retest (ICC = 0.91–0.98) and inter‐rater reliability (ICC = 0.86–0.96) were shown in the 2MWT, 6MWT and 10MeWT. The walk tests were strongly correlated with each other (&rgr; = 0.85–0.94). The correlations between the walk tests and the functional measures were moderate in general (&rgr; = 0.34–0.55). All the walk tests were able to distinguish between those who could walk outdoor and indoor only (p ≤ .036). The MDC95 were 9.1 m in the 2MWT, 28.1 m in the 6MWT, and.16 m/s in the 10MeWT. The cues provided by the assessors in the walk tests were generally consistent (ICC = 0.62–0.89). Conclusions: The 2MWT, 6MWT and 10MeWT are reliable and valid measures in evaluating walking ability in frail older adults with dementia. The MDC95 of the walk tests has been established. The cueing system is feasible and reliable to facilitate the administration of the walk tests in this population group.