Background: Navigation allows for determination of the mechanical axis of the lower extremity. We evaluated the intra- and inter-observer reliability with an image-free navigation system and determined the accuracy of the navigation system to monitor changes in lower limb alignment as compared to alignment measured with a novel 3D CT method.

Methods: A total of 13 cadaver legs were used to evaluate the intra- and inter-observer registration reliability by three observers. Navigated HTOs were then performed on all legs and pre/postoperative values of the varus-valgus angles were recorded. Data were compared to equivalent measures obtained by 3D CT using intra-class correlation coefficients (ICCs).

Results: The ICCs for intra-observer varus-valgus reliability ranged from 0.756 to 0.922, inter-observer reliability was 0.644. ICCs for navigation-CT comparison were 0.784 for varus-valgus angle (pre-op), 0.846 (postop) and 0.873 (delta). Maximum differences in navigation-CT measurements in varus-valgus angle (delta) were 4.5° for all trials. There was poor reliability and accuracy in the axial plane (tibial rotation) as well as fair reliability and accuracy in the sagittal plane (tibial slope).

Conclusion: Image-free navigation is reliable for dynamic monitoring of coronal leg alignment but shows relevant limitations in determination of sagittal and axial plane alignment.