Sensitivity of the knee joint kinematics calculation to selection of flexion axes

Various flexion axes have been used in the literature to describe knee joint kinematics. This study measured the passive knee kinematics of six cadaveric human knee specimens using two widely accepted flexion axes; transepicondylar axis and the geometric center axis. These two axes were found to form an angle of 4.0degrees +/- 0.8degrees. The tibial rotation calculated using the transepicondylar axis was significantly different than the rotation obtained using the geometric center axis for the same knee motion. At 90degrees of flexion, the tibial rotation obtained using the transepicondylar axis was 4.8degrees +/- 9.4degrees whereas the rotation recorded using the geometric center axis at the same flexion angle was 13.8degrees +/- 10.2degrees. At 150degrees of knee flexion, the rotations obtained from the transepicondylar and the geometric center axes were 7.2degrees +/- 5.7degrees and 19.9degrees +/- 6.9degrees, respectively. The data suggest that a clear definition of the flexion axis is necessary when reporting knee joint kinematics. (C) 2004 Elsevier Ltd. All rights reserved.