Pedicle screw placement assisted by 3D imaging (O-arm system with StealthStation® software) versus free-hand technique for multilevel posterior thoracolumbar fusion

Abstract

Background: Comparison of pedicle screw placement accuracy and risk of perioperative complications using O-arm/Stealth Navigation system (Medtronic, Memphis, TN) compared with the free-hand technique for multilevel posterior instrumented thoracolumbar fusion is an actual problem. Methods: A retrospective, single-institution case series study was performed. Patients who underwent multilevel instrumented thoracolumbar fusion between 2011 and 2012 were included; age: >20 yr; gender: male and female. Patients with pedicle screw placement assisted by the O-arm navigation were assigned to the navigated group, while patients operated using a free-hand technique to the nonnavigated group. Breaches were defined by CT images and classified by screw extension beyond the cortex: mild (<2 mm), moderate (2-4 mm), and severe (>4 mm). Risk was defined by odds ratio (OR), and P-value by the χ2-test. Bayesian analysis was applied to take into consideration previous findings. Results: Placement of 597 pedicle screws (401 navigated and 196 nonnavigated) was defined in 70 patients: 41 (12 male) navigated and 29 (10 male) nonnavigated. The mean age was 62 yr (SD, 9.3). Risk of breaches was less in the navigated group, OR=0.36, P=0.002; in particular, the risk of mild breaches, OR=0.33 (P=0.003), and the risk of moderate and severe breaches, OR=0.48 (P=0.25); more significantly at thoracic levels, OR=0.04 (P=0.005); and specifically, medial and lateral breaches (P<0.03). The total risk of perioperative complications was less in the navigated group, OR=0.68 (P=0.3). The obtained results well complement previously published data. Conclusions: Pedicle screw placement using the O-arm is more accurate than a standard free-hand technique, specifically at the thoracic spine.