Study design: The location of pedicle screws (n = 42) in four human specimens of the lumbar spine and in 30 patients (n = 131 screws) after lumbar spinal fusion was assessed using computed tomography.
Objectives: To determine the accuracy of pedicle screw placement in lumbar vertebrae and the reproducibility and repeatability of the computed tomography examination.
Summary of background data: Failures in the placement of transpedicular screws for lumbar fusion are reported. The evaluation of such screws using computed tomography examination has not been investigated.
Methods: After surgery, the specimens were dissected in transversal slices to observe macroscopically the location of the pedicle screw and to correlate these observations with the computed tomography images. All patients were examined by one observer. To determine the reproducibility and repeatability of the computed tomography examination, two observers studied computed tomography images of 12 patients (n = 58 screws) twice within 3 months.
Results: In the specimens, 10 screws were observed to penetrate the medial wall of the pedicle. This correlated fully with the images. In the patients' group, 40% of all screws penetrated the cortex of the vertebra. Of all screws, 29% penetrated the medial wall of the pedicle. From the computed tomography images, it appeared that a deviation of more than 6 mm medially was a high risk for nerve root damage. Three months after his first examination, Observer 1 documented a different position in three of 58 screws (kappa = 0.90). Observer 2 found a different position in eight screws (kappa = 0.65). The comparison between the reviews of the two observers showed a different opinion for the first evaluation, four disagreements (2-4 mm) and 17 disagreements (0-2 mm; kappa = 0.34), and for the second evaluation, four disagreements (2-4 mm) and 12 disagreements (0-2 mm; kappa = 0.43).
Conclusions: Correct placement of transpedicular screws for spinal fusion seems to be more difficult than it looks. The computed tomography scanning is useful for differential diagnosis of postoperative radicular syndromes after lumbar transpedicular fixation.