Radiological and computer tomographic assessment of pedicle screw placement in the lumbar spine.

Purpose: To determine the accuracy of a standard roentgenogram for the plac ement control of pedicle screws following spinal fusion. Method: From 1995 to 1997 we performed computed tomography (CT) after material removal follow ing lumbar and lumbosacral spinal fusion in 16 patients. We compared the pl acement of the 76 pedicle screws in plain X-rays after spinal fusion with t heir appearance on CI. A correct placement was defined as no penetration of the pedicle cortex, no contact of the lateral OF ventral cortex of the ver tebral body or joint, and a sufficient screw length. Results: We found corr ect placement of 58 screws (76.3 %) on the standard roentgenogram, whereas by CT only 46 (60.5 %) were placed correctly. The results of both radiologi cal examinations correlated (correct/incorrect placement) for only 54 screw s (77.1 %). A penetration of the ventral cortex of the vertebral body in 21 cases (27.6%) was identified in only 11 screws (14.5 %) on the postoperati ve X-rays. The two radiological methods in this instance correlated for 62 screws (81.6%). Conclusion: The value of postoperative standard roentgenogr am for the placement control of pedicle screws following spinal fusion is l ow By this method possible contact with the aorta, vena cava, dura or Verte bral joint cannot sufficiently be excluded.