How reproducible are clinical measurements in robotic knee surgery?

Robotic-assisted surgery has been recently introduced to improve biomechanical restoration, and thus better clinical and functional outcomes, after knee joint arthroplasty operations. Robotic-assisted uni-compartmental knee arthroplasty (UKA) aims indeed to improve surgical bone resection and alignment accuracy, optimized component positioning and knee balancing, relying on a series of calibration measurements performed during the surgery. These advantages focus therefore on improving the reproducibility of UKA surgeries, reducing (if not eliminating) eventual differences among high- and low-volume surgeons. The purpose of this study is to investigate and quantify the reproducibility of in-vivo measurements performed with a robotic system: the intra- and inter-observer variability of a series of measurements was therefore analyzed and compared among differently experienced operators.

Five patients were analyzed and underwent robotic-assisted UKA using a semi-active robotic system. Three different observers with different experience levels were involved to independently perform the measurements of two parameters of the preoperative knee (Hip-Knee-Ankle angle [HKAa], Internal-External Rotation) at different degrees of knee flexion. Inter-observer and intra-observer comparisons were performed.

The average variability in the measurements obtained from the intra-observer and inter-observer comparisons were always < 0.68° for HKAa and < 2.59° for internal-external rotation, and the ICCs showed excellent agreement (> 0.75) for most cases and good agreement (> 0.60) in the remaining ones.

This study demonstrated high reproducibility of the measurements obtainable in clinical environment with the robotic system. The inter-observer results furthermore showed that the level of confidence with the robotic system is not significantly influencing the measurement.