Incidence and risk factors for the misplacement of pedicle screws in scoliosis surgery assisted by O-arm navigation-analysis of a large series of one thousand, one hundred and forty five screws

Purpose To assess the accuracy of O-arm-navigation-based pedicle screw placement in scoliosis surgery and identify the potential risk factors for the misplacement of pedicle screws. Methods One hundred forty four scoliosis patients treated with O-arm-navigation-based pedicle screw instrumentation were enrolled, and 1145 pedicle screws implanted in the apical region of the curves were retrospectively reviewed for accuracy according to post-operative CT images. The potential risk factors and independent predictive factor(s) for the misplaced screws were identified statistically. Results The overall malpositioning rate of pedicle screw was 9.8%; 54.5% of which were misplaced laterally. Univariate and multivariate logistic regression analysis of clinical and surgical treatment variables indicated that patients with congenital scoliosis (CS) [OR: 1.489 (95% CI: 1.002-2.213; P = 0.035)] and neurofibromatosis type I (NF-1) [OR: 1.785 (95% CI: 1267-2.045; P = 0.026)], middle-thoracic spine [OR: 1.661 (95% CI: 1.107-2.481; P = 0.021)], the concave pedicles [OR: 1.527 (95% CI: 1.020-2.285; P = 0.019)], and the segments three levels away from the tracker [OR: 3.522 (95% CI: 2.357-5.263; P = 0.001)] were independently associated with pedicle screw misplacement. Conclusion O-arm-assisted navigation does improve the accuracy and safety of pedicle screw placement in scoliosis surgery. However, unavoidable screw malpositioning remained, which occurred significantly more often in patients with CS and NF-1, in middle-thoracic spine, in the concave pedicles, and in the segments three levels away from the tracker.