Non-image based navigation has catalysed an ongoing revolution in total knee arthroplasty technique. Few systems, however, have been designed and subsequently analysed in UKA. A new system designed for use with the Oxford UKA prosthesis offers two ways to determine the ideal tibial resection depth for insertion of the tibial component of the prosthesis. Tibial resection may be calculated using femoral or tibial referencing. Clinical experience suggests that femoral referencing calculates tibial resection depth incorrectly. We set out to confirm whether femoral referencing is, in fact, less accurate than tibial referencing in calculating tibial resection depth in this new navigation system, and to determine factors responsible.

The navigated procedure for UKA was carried out on an artificial limb to the point of tibial resection using the latest Vector Vision-Biomet-BrainLAB navigation system. The depth of resection achieved was measured and the procedure repeated in multiple series using femoral and tibial referencing. Intraoperative flexion angle and joint space were altered to assess their effect on accuracy. Analysis of variance revealed that femoral referencing is less accurate than tibial referencing (p = 0.001). Though intraoperative hyperflexion and increased joint space width exacerbated the inaccuracy, their contribution did not achieve significance (p = 0.078 and p = 0.02, respectively).

We propose that femoral referencing be omitted as an option in the determination of tibial resection depth in this system. This study has revealed an inaccuracy in a commercially available navigation system that can result in a prolonged and more complicated procedure, suggesting that more stringent pilot assessment of orthopaedic products is needed.