Background
A previous randomised controlled trial (RCT) by our group found that robotic assisted unicompartmental knee arthroplasty (UKA) surgery was significantly more accurate than conventional instrumentation. The aim of this study was to determine whether a low-cost novel PSI design could deliver the same level of accuracy as the robot in the same time efficient manner as conventional instruments.

Methods
Thirty patients undergoing medial UKA took part. Tibial component position was planned using a low dose CT-scan, and compared to a day 1 postoperative CT-scan to determine the difference between the planned and achieved positions. Operations were performed by one expert surgeon using PSI (Embody, London, UK).

Results
The mean absolute difference between planned and achieved tibial implant positions using PSI was 2.0° (SD 1.0°) in the coronal plane, 1.8° (SD 1.5) in the sagittal plane, and 4.5° (SD 3.3) in the axial plane. These results were not significantly different to the 13 historical robotic cases (mean difference 0.5°, 0.5°, and 1.7°, p = 0.1907, 0.2867 and 0.1049 respectively). PSI mean operating time was on average 62 min shorter than the robotic group (p < 0.0001) and 40 min shorter than the conventional instrument group (p < 0.0001). No complications were reported.

Conclusions
In conclusion, this clinical trial demonstrates that for tibial component positioning in UKA, a novel design PSI guide in the hands of an expert surgeon, can safely deliver comparable accuracy to a robotic system, whilst being significantly faster than conventional instruments.

NIHR Clinical Research Network Reference: 16100.