What change in isokinetic knee muscle strength can be detected in men and women with hemiparesis after stroke?

To assess the intra-rater (between occasions) test-retest reliability of isokinetic knee muscle strength measurements in subjects with chronic poststroke hemiparesis and to define limits for the smallest change that indicates real (clinical) improvements for stroke patients.Fifty men and women (mean age 58 +/- 6.4 years) 6-46 months post stroke, able to walk at least 300 m with or without a unilateral assistive device.Maximal concentric knee extension and flexion contractions at 60 degrees/s and 120 degrees/s, and maximal eccentric knee extension contractions at 60 degrees/s, with the paretic and nonparetic limbs, were performed seven days apart using a Biodex dynamometer.Reliability of the maximum peak torque measurements was evaluated with the intraclass correlation coefficient (ICC(2,1)), the Bland and Altman analyses, the standard error of measurement (SEM and SEM%) and the smallest real difference (SRD and SRD%).Test-retest agreements were high (ICC(2,1) 0.89-0.96) with no discernible systematic differences between limbs, angular velocities and modes. The SEM%, representing the smallest change that indicates a real (clinical) improvement for a group of subjects, was relatively small (8-20%). The SRD%, representing the smallest change that indicates a real improvement for a single subject ranged from 26% to 33% for concentric knee extension, from 39% to 55% for concentric knee flexion, and from 22% to 25% for eccentric knee extension.Isokinetic knee muscle strength can be reliably measured and used to detect real improvements following an intervention for single subjects as well as for groups of subjects with chronic mild to moderate hemiparesis after stroke.