Background and purpose: One must understand the potentials and limitations of all tests used to evaluate interventions. The aim of the present study was to clarify the reproducibility, smallest percentage difference needed to be able to detect a genuine change and correlation regarding the 10-m and 30-m timed walks (10TW 30TW) and the 'timed up and go' (TUG) test in people with moderate multiple sclerosis (MS).
Method: A repeated-measures design was used, with randomization into two groups and different time intervals used for testing. The 10TW and 30TW were performed three times and TUG twice at each testing. Self-selected speed was used for 10TW and forced speed (quickly but safely) for 30TW and TUG. Forty-three people were tested on three occasions within one week. Each person was tested at approximately the same time of the day and by the same physiotherapist on each occasion.
Results: The reproducibility was very high. For a single testing occasion, the intraclass correlation was 0.97 for the 10TW and 0.98 for the 30TW and TUG. The smallest percentage difference needed to be able to detect a genuine change in the entire study group was approximately -23% or +31% for either the 1OTW or TUG. It was evident from the 30TW testing results that lower values applied to those with less (-14% to +17%) rather than more (-38% or +60%) disability. The correlation between all tests for the entire study group was 0.85 (0.76-0.91).
Conclusion: It is sufficient to use only one attempt and to choose only one of the tests when evaluating people with moderate MS. In the case of the 30TW greater attention must be paid to the degree of disability when determining the smallest percentage difference needed to establish a genuine change, than