[Total knee arthroplasty: reproducibility of gait speed].
OBJECTIVE: To calculate the inter-session (test-retest) and inter-observers' reproducibility of the maximal gait speed measured in patients recently operated for osteoarthritis with a total knee arthroplasty. METHODS: Evaluation of the maximal gait speed in 20 patients [t72 +/- 6.5 years] on a 10-meter track by two observers using a manual stopwatch and by an automatic photoelectric stopwatch. The maximal gait speed was measured from three trials in two conditions, with and without crutches, with 1 day of interval at the third and the fourth post-surgery week. RESULTS: The inter-session reproducibility (1-day test-retest) is 0.91 [intraclass correlation coefficient] with crutches and 0.9 without crutches at the third post-surgery week, and 0.96 and 0.94 at the fourth post-surgery week with and without crutches respectively. The inter-observers' reproducibility is 0.9 to 0.99. CONCLUSION: When a standardized protocol is used, the reproducibility of maximal gait speed measurements is excellent. These tests evaluate easily and objectively the maximal gait speed in patients recently operated with a total knee arthroplasty.