Interobserver and intraobserver error in obtaining transepicondylar axis and whiteside line in total knee arthroplasty

Abstract

Four cadaveric knee joints with an intact soft tissue envelope were approached using the medial parapatellar approach employed in standard conventional incision total knee replacement. The transepicondylar axis and Whiteside line of the distal femur were repeatedly identified by three orthopaedic surgeons for a total of twenty times each. The precision in the identification of these landmarks was compared with the reference transepicondylar axis and Whiteside line established at the end of the experiment by dissecting the soft tissue from the cadaveric knees. It was found that there was significant interobserver error in the identification of both the transepicondylar axis and Whiteside line (p<0.001, ANOVA). The repeatability in the identification of these two axes was relatively low (13° for transepicondylar axis and 24° for Whitesid e line).