Axial Convergence in Percutaneous Robotic Pedicle Screw Placement: Mixed-Effects Analysis of 1006 Screws

Study Design: Retrospective single-center cohort study. Objective: To evaluate screw accuracy, axial trajectory characteristics relative to the anatomic pedicle axis (APA), and perioperative safety of robotic-assisted (RA) pedicle screw placement, and to determine whether percutaneous instrumentation (PI) enables reproducible medial trajectory modulation without increasing cortical breach rates. Summary of Background Data: RA pedicle screw placement achieves high categorical accuracy. However, most studies focus on breach rates and do not quantify systematic variation in axial screw angulation relative to the APA. Whether robotic workflows enable controlled trajectory modulation beyond simple intrapedicular containment remains unclear. Methods: A consecutive cohort of 146 adult patients undergoing RA pedicle screw placement using the CIRQ robotic arm (1006 screws) was analyzed. Accuracy was graded using the Gertzbein-Robbins classification (GR). Axial angulation and deviation from the APA were measured on postoperative computed tomography. Screw-level outcomes were assessed using multivariable mixed-effects regression accounting for case-level clustering. Sensitivity analyses addressed chronological case era and selection bias using inverse probability weighting. Results: Clinically acceptable placement (GR A/B) was achieved in 97.9% of screws; major malposition (GR D/E) occurred in 0.4%. Cervical location (OR: 7.10, P <0.001) and open instrumentation (OI) (OR: 8.50, P <0.001) were independently associated with increased odds of cortical breach. Imaging modality was not independently associated with accuracy. PI demonstrated significantly greater medial convergence relative to the APA (interaction P <0.001) without increased breach risk. Three screws (0.3%) required revision. No neurological deficits attributable to screw malposition were observed (mean follow-up 5.9 mo). Conclusions: RA pedicle screw placement using CIRQ achieved high accuracy and consistent safety independent of imaging modality and surgeon experience level. PI was associated with reproducible medial trajectory modulation without increased cortical breach. These findings suggest that RA permits controlled intrapedicular trajectory modulation within safe anatomic boundaries.