Active knee joint velocity replication measures are stable and accurate in healthy individuals

The purpose of this study was to determine the stability and accuracy of active knee joint velocity replication methods in healthy subjects. We used a repeated measures design with 14 healthy volunteers. Measures of velocity replication were performed in two ranges of knee joint flexion (0 degrees -30 degrees and 60 degrees -90 degrees ), across four testing velocities (5, 10, 15, and 30 degrees /s) in two movement directions (flexion and extension). Statistical analysis included intraclass correlation coefficients (ICCs; 2, k) and associated standard error of the measures calculated between day 1 and 2. We performed z-tests between all possible combinations of ICC pairs using Fisher's Z transformations to determine if any significant differences existed between observed ICCs. We also calculated correlation ratios (eta2) to explain the source of variability in the calculated ICCs. To assess measurement accuracy, we calculated constant error and absolute error between criterion and replication velocities. Results on ICCs and standard error of the measurements (SEMs) ranged from r = -0.44 +/- 7.00 to 0.88 +/- 0.72 degrees /s. Calculated z-tests indicated six paired ICCs were significantly different ( p < 0.1). In all six pairs, the faster test velocity had a lower ICC magnitude. The eta2 calculations demonstrated that inconsistent performance between day 1 and 2 caused the low ICC magnitudes observed with faster testing velocities. Significantly more absolute error occurred at 30 and 15 degrees /s compared with 5 degrees /s. Significantly less constant error was observed for 30 degrees /s compared with 15 degrees /s. A significant direction by range of motion interaction indicated less constant error for flexion movements in the 60 degrees -90 degrees range of motion (ROM) as compared with extension movements in either ROM. Healthy individuals could actively replicate slower criterion velocities in the mid and end ranges of knee joint motion in both movement directions with an acceptable amount of consistency and accuracy. The data support the use of velocity replication in future investigations on proprioceptive function.