Proximal Facet Joint Violations in Augmented Reality-Guided versus Freehand Pedicle Screw Placement in the Lumbosacral Spine

To compare the incidence and severity of proximal facet joint violations (FJVs) between augmented reality (AR)-guided and conventional freehand (FH) pedicle screw placement. The secondary objective was to evaluate associations between FJVs and patient-specific or surgical factors. Adult patients who underwent lumbar or sacral pedicle screw placement using either AR-guided or conventional FH techniques were included. Proximal FJVs were evaluated on postoperative computed tomography (CT) and classified using the modified Park and Shah systems. Demographic and surgical variables were recorded, and generalized linear mixed models were used to analyze associations with FJV occurrence. A total of 412 proximal facet joints in 206 patients were analyzed, with 103 patients in each group. The incidence of proximal FJVs was significantly lower in the AR group (7.8%, n = 16) compared to the FH group (17.0%, n = 36) (p < 0.001). In both cohorts, FJVs occurred mostly at the L5 level and were predominantly unilateral. Among all violations, most of these findings were classified as modified Park grade 1 (n = 29, 7.0%) or grade 2 (n = 21, 5.1%). One grade 3 violation (0.2%) was observed in the FH group. Baseline demographic and operative characteristics were comparable between groups. These findings demonstrate that AR-assisted pedicle screw placement was associated with significantly fewer proximal FJVs compared to the FH technique. The highest violation rate was observed at the L5 level in both groups. Further studies are warranted to confirm these findings across diverse populations and surgical settings.