The high variability of tibial rotational alignment in total knee arthroplasty.

Although various techniques are advocated to establish tibial rotational alignment during total knee arthroplasty, it is unknown which is most repeatable. We evaluated the precision and accuracy of five tibial rotational alignment techniques to determine whether computer-assisted navigation systems can reduce variability of tibial component rotational alignment when compared to traditional instrumentation. Eleven orthopaedic surgeons used four computer-assisted techniques that required identification of anatomical landmarks and one that used traditional extramedullary instrumentation to establish tibial rotational alignment axes on 10 cadaver legs. Two computer-assisted techniques (axes between the most medial and lateral border of the tibial plateau, and between the posterior cruciate ligament [PCL] and the anterior tibial crest) and the traditional technique were least variable, with standard deviations of 9.9 degrees, 10.8 degrees, and 12.1 degrees, respectively. Computer-assisted techniques referencing the tibial tubercle (axes between the PCL and the medial border or medial 1/3 of the tubercle) were most variable, with standard deviations of 27.4 degrees and 28.1 degrees. The axis between the medial border of the tibial tubercle and the PCL was internally rotated compared to the other techniques. None of the techniques consistently established tibial rotational alignment, and navigation systems that establish rotational alignment by identifying anatomic landmarks were not more reliable than traditional instrumentation.