Robot-assisted cement-augmented percutaneous pedicle screw fixation is associated with greater screw accuracy and fewer mechanical complications in osteoporotic thoracolumbar spine fractures: a retrospective comparative cohort study

Abstract Purpose To compare screw placement accuracy and mechanical complications between robot-assisted and conventional fluoroscopy-assisted cement-augmented percutaneous fixation for osteoporotic thoracolumbar spine fractures. Methods This retrospective cohort included 128 patients (64 per group) and 658 screws. Modified Gertzbein–Robbins grade A placement was the primary outcome, analysed with patient-clustered generalised estimating equations. Mechanical complications were assessed with logistic models and longitudinal outcomes with linear mixed-effects models. Independent rereading after database lock assessed grade and breach-distance reliability without replacing the primary grades. Results Grade A placement occurred in 285/326 conventional screws (87.4%) and 314/332 robot-assisted screws (94.6%; clustered odds ratio 2.49, 95% confidence interval (CI) 1.43–4.35; P = 0.001). Among 118 patients with evaluable 12-month status, mechanical complications occurred in 17/60 (28.3%) and 5/58 (8.6%), respectively (risk ratio 0.30, 0.12–0.77; P = 0.008); the adjusted odds ratio was 0.21 (0.07–0.64; P = 0.007). At 12 months, Cobb angle was 2.1° lower and anterior vertebral height ratio was 5.3 percentage points higher with robot assistance, whereas pain was similar. Exact grade agreement was 91.8%; quadratic-weighted κ was 0.86 (95% CI 0.79–0.91), and the breach-distance intraclass correlation coefficient was 0.97 (0.94–0.98). Conclusion Robot assistance was associated with more grade A screws, fewer mechanical complications and better maintenance of fracture correction. These associations require confirmation in adequately controlled prospective studies of longer-term safety and cost-effectiveness.