Abstract
BACKGROUND: Robot-assisted spine surgery aims to improve the accuracy of screw placement. We compared the accuracy and safety between a novel robot and free hand in thoracolumbar pedicle screw placement. METHODS: Eighty patients scheduled to undergo robot-assisted (40 patients) and free-hand (40 patients) pedicle screw placement were included. The patients' demographic characteristics, radiographic accuracy, and perioperative outcomes were compared. The accuracy of screw placement was based on cortical violation and screw deviation. Safety outcomes mainly included operative time, blood loss, revision, and complications. RESULTS: A total of 178 and 172 screws were placed in the robot-assisted and free-hand groups, respectively. The rate of perfect screw position (grade A) was higher in the robot-assisted group than in the free-hand group (91.0% vs. 75.6%; P < 0.001). The rate of clinically acceptable screw position (grades A and B) was also higher in the robot-assisted group than in the free-hand group (99.4% vs. 90.1%; P < 0.001). The robot-assisted group had significantly lower screw deviation than the free-hand group [1.46 (0.94, 1.95) mm vs. 2.48 (1.09, 3.74) mm, P < 0.001]. There was no robot abandonment in the robot-assisted group. No revision was required in any of the groups. CONCLUSIONS: Robot-assisted pedicle screw placement is more accurate than free-hand placement. The second-generation TiRobot-assisted thoracolumbar pedicle screw placement is an accurate and safe procedure. Trial registration retrospectively registered.