Accuracy of Lumbar Pedicle Screw Placement Based on Insertion Techniques in a Single-Center Retrospective Study.

INTRODUCTION: Although many studies have evaluated the accuracy of pedicle screw (PS) placement using various techniques, few have conducted large-scale comparisons of open and percutaneous approaches using CT navigation and fluoroscopy within a single institution. This study aimed to evaluate the accuracy of lumbar PS placement based on four different insertion techniques and to identify factors influencing screw placement accuracy. METHODS: A total of 237 cases (1,081 screws) of PS placement at the L1-S1 level were retrospectively analyzed. Patients were divided into four groups: open CT navigation, percutaneous PS (PPS) CT navigation, PPS fluoroscopy, and open freehand. Screw accuracy was assessed using the Gertzbein-Robbins classification based on postoperative CT or intraoperative O-arm images. The proportions of grade A (no breach) and grade A or B (≤2 mm breach) screws were compared among the groups. Statistical analyses included Fisher's exact test with Bonferroni correction. RESULTS: The proportion of grade A screws was highest in the open CT navigation group (96.4%), followed by PPS fluoroscopy (89.3%), open freehand (88.0%), and PPS CT navigation (81.0%) (p < 0.05). Grade A or B screws exceeded 95% in all groups with no significant differences. Lateral breaches were the most frequent, particularly in the PPS CT navigation (93.3%) and PPS fluoroscopy (82.0%) groups. No grade D or E medial or caudal breaches or neurological complications were observed. Screw repositioning was required in two cases (three screws) due to lateral breaches >4 mm with screw tips located outside the vertebral body. CONCLUSIONS: All four insertion techniques achieved high accuracy, with more than 95% of screws classified as acceptable placements and no neurological complications observed. These findings suggest that lumbar PS placement can be performed safely and reliably across different techniques when applied appropriately.