Abstract
Successful clinical outcomes following unicompartmental knee arthroplasty (UKA) depend on accurate component alignment, which can be difficult to achieve using manual instrumentation. A new technology has been developed using haptic robotics that replaces traditional UKA instrumentation. This study compares the accuracy of UKA component placement with traditional jig-based instrumentation versus robotic guidance.
Forty-four UKAs performed using standard manual instrumentation were compared to 33 performed with a robotically guided implantation system without instrumentation. Each was performed using a minimally invasive surgical approach. The two groups were identical in terms of age (p=0.74), gender (p=0.65) and BMI (p=0.72). The coronal and sagittal alignment of the tibial components were measured on pre- and post-operative AP and lateral radiographs. Postoperative tibial component alignment was compared to the pre-operative plan.
For both techniques, the surgical objective was to match the natural tibial posterior slope. The RMS error of the tibial slope was 3.5° manually compared to 1.4° robotically. In addition, the variance using manual instruments was 2.8 times greater than the robotically guided implantations (p< 0.0001). In the coronal plane, the goal of the manual technique was to implant the tibial component perpendicular to the anatomic tibial axis, while the robotic implantations attempted to match the natural varus of the medial compartment. The average error was 3.3 ± 1.8° more varus using manual instruments compared to 0.1 ± 2.4° when implanted robotically (p< 0.0001).
Tibial component alignment in UKA is significantly more accurate and less variable using robotic guidance compared to manual, jig-based instrumentation. By enhancing component alignment, this novel technique provides a potential method for improving outcomes in UKA patients.
Correspondence should be addressed to ISTA Secretariat, PO Box 6564, Auburn, CA 95604, USA. Tel: 1-916-454-9884, Fax: 1-916-454-9882, Email: ista@pacbell.net