Student Johns Hopkins School of Medicine Baltimore, Maryland, United States
Introduction: Robotics has demonstrated increasing utility in spine surgery, most notably in pedicle screw placement. Robot-assisted pedicle screw insertion offers several advantages over traditional fluoroscopy-guided or freehand techniques, including greater screw placement accuracy. However, there is a paucity of research identifying factors associated with robot-assisted screw fixation accuracy. Therefore, the objective of the current study is to determine accuracy rates and risk factors for ExcelsiusGPS®-assisted pedicle screw insertion.
Methods: All patients between October 2017-March 2022 undergoing spinal fusion surgery with ExelsiusGPS®-assisted screw implantation at a single tertiary center were retrospectively identified for inclusion. Patient demographics, preoperative symptoms and operative details were collected. Post-operative CT scans of the spine were used to classify screw placement accuracy according to the Gertzbein and Robbins scale (GRS). Stepwise multivariate logistic regression analysis was conducted to determine independent risk factors for clinically inaccurate screws (GRS C/D/E).
Results: 117 patients were included in the study. Mean age was 60.6 ± 13.2 years, with 57% male, 72% white, and an average BMI of 29.9 ± 6.4 kg/m2. Main indications for surgery were spondylolisthesis (32%), instability/fracture (26%), pseudarthrosis (10%) and spondylosis (10%). A total of 728 screws were placed, predominantly in the thoracic (29.5%) and lumbar (52.6%) regions. Average number of levels fused was 3.2 (range 1-11). Accuracy classification of screws indicated 667 GRS A, 35 GRS B, 22 GRS C, 3 GRS D, and 1 GRS E. The clinically acceptable screw placement rate (GRS A/B) was 96% (702/728). Logistic regression analysis identified revision procedures (OR 2.31, 95% CI 1.13, 4.72, p=0.02) as independently associated with inaccurate screw placement. Of the 23 (20%) reoperations that occurred following robotic-assisted screw placement, only 2 (1.7%) were for screw revision. Of the 728 screws placed, only 3 were discovered to have breached the pedicle upon postoperative imaging, requiring revision.
Conclusion : ExcelsiusGPS®-assisted pedicle screw insertion has high placement accuracy and low revision rates. Only revision surgeries had an independent association with clinically inaccurate screw placement, as determined by postoperative CT scans. Awareness of this association can be used to guide patient selection and surgical practice for robot-assisted spinal fusion procedures.
