BackgroundVague spinal anatomical landmarks in patients with ankylosing spondylitis (AS) make intraoperative insertion of pedicle screws difficult under direct vision. Currently, the clinical outcome is significantly improved with robot guidance. This study aims to explore the efficacy of robot-assisted pedicle screw insertion in treating AS combined with spinal fractures. Methods40 patients (341 screws) who underwent pedicle screw insertion with AS complicated with spinal fractures were included. According to different surgical methods, 16 patients (135 screws) were classified into the robot group and 24 (206 screws) into the free-hand group. Intraoperative blood loss, operative duration, and adverse events were compared between the two groups. Gertzbein and Robbins classification was used to classify the accuracy of screw position. Clinical outcomes were evaluated by Visual Analog Scale, Japanese Orthopedic Association, and Oswestry Disability Index. ResultsNo statistically significant differences between baseline data of the groups. The difference in the blood loss between groups wasn’t significant, nor was the operative duration. No severe adverse events related to pedicle screw insertion were reported in either group. Notably, the accuracy of screw insertion was significantly higher in the robot group (129/135) than in the free-hand group (182/206). The lateral perforation prevalence didn’t differ among groups. VAS in the third month postoperatively was lower in the robot group than in the free-hand group, with a significant difference. ConclusionsThe study demonstrates statistically superior accuracy and surgical outcome of robot-assisted pedicle screw insertion in the treatment of AS complicated with spinal fractures compared with the traditional free-hand operation.