Accuracy of robot-assisted cortical bone trajectory pedicle screw placement: CT-based retrospective cohort study

Abstract

Background and study aims: Pedicle screw placement accuracy is critical for safe spinal fixation, and robot- assisted techniques aim to improve precision. This study evaluated the accuracy of robot-assisted posterior pedicle screw placement in the lower thoracic and lumbar spine using a CBT trajectory as well as determine what factors are associated with accuracy of screw placement.

Materials and methods: A retrospective single-centre cohort study included 100 consecutive patients undergoing Aposterior spinal instrumentation with pedicle screws between August 2022 and August 2023. Preoperative CT- based planning and robotic guidance (ExcelsiusGPS) were used. Postoperative CT assessed 511 screws using the Gertzbein-Robbins classification, including breach direction. Clinically acceptable placement was defined as grades or B. Interobserver agreement was evaluated using Cohen’s kappa and Gwet’s AC1. Multivariable analyses were performed using generalized estimating equations and mixed-effects models.

Results: Clinically acceptable screw placement exceeded 96% at all spinal levels. Perfect placement ranged from 86.7% to 100%. No clinically relevant (Grade C-E) medial, superior, or inferior breaches were observed; lateral breaches occurred in 1.4%. Interobserver agreement was high (AC1 0.953-0.994). For perfect placement (Grade A), L5 showed higher odds compared to L4.

Conclusion: Robot-assisted pedicle screw placement using CBT demonstrates high accuracy and safety, with absence of clinically relevant medial breaches. These findings support the use of robotic guidance as an effective adjunct in spinal instrumentation.