Background:

While several studies report on accuracy rates of pedicle screws, risk factors associated with inaccurate pedicle screw positioning in patients with thoracolumbar fractures are reported rarely. CT scan as a routine postoperative control is advocated by various authors, however its necessity remains unclear.

Methods:

Two hundred forty-five patients were included in this retrospective study. Percutaneous dorsal instrumentation was most commonly performed (n = 201). Classification of Zdichavsky et al. and Rao et al. were used to classify screw misplacement and anterior perforation was further evaluated according to the extent of perforation (< 2 mm; > 2 mm). Multivariate analysis was performed to identify risk factors for misplacement of screws.

Results:

One thousand sixty-eight pedicle screws were inserted in 245 patients. Misplacement was found in 51 screws (4.8%) in 42 patients (17.1%) according to the classification of Zdichavsky et al. and in 75 screws (7.0%) in 64 patients (26.1%) according to the classification of Rao et al.. An anterior perforation of the vertebral cortex was found in 56 screws (5.2%). Multivariate analysis showed fracture location in the upper thoracic (p = 0.048) and lumbar spine (p = 0.013) to be the only independent predictors for screw misplacement. In addition a significant correlation between pedicle diameter and the occurrence of screw malposition was found (p = 0.003). No consequences were drawn from postoperative routine CT in asymptomatic patients.

Conclusion:

An overall low rate of screw misplacement was found with fracture location in the upper thoracic and lumbar spine being the only factors independently associated with the risk of screw misplacement. No consequences were drawn from postoperative routine CT in asymptomatic patients. Therefore its use has to be discussed critically.