Comparative results between conventional and computer-assisted pedicle screw installation in the thoracic, lumbar,...

A comparative study on the position of pedicle screws in patients treated surgically with and without computer assistance.

To evaluate the accuracy of computer-assisted pedicle screw installation, and to evaluate its clinical benefit as compared with conventional pedicle screw installation techniques.

In vitro and clinical studies have documented a significant rate of misplaced screws in the thoracolumbar area. Neurologic complications are recognized problems caused by screw misplacement.

Patients treated surgically with computer assistance were compared with a historical control group of patients treated surgically with conventional techniques in the same hospital and by the same surgical team. All screw positions were measured with a postoperative magnetic resonance tomography, and cortical effractions were categorized in 2-mm increments. Patients' charts also were reviewed to assess individual neurologic outcomes.

The control cohort was composed of 100 patients, with 544 screws from T5 to S1. The computer-assisted cohort was composed of 50 patients, with 294 screws from T2 to S1. In the control cohort, 461 of 544 screws (85%) were found completely within their pedicles as compared with 278 of 294 screws (95%) correctly placed in the computer-assisted group (P < 0.0001). All 16 screws incorrectly placed with computer assistance were found 0.1 mm to 2 mm from the pedicle cortex. In the control cohort, 68 screws were found 0.1 mm to 2 mm, 10 screws 2.1 mm to 4 mm, and 5 screws more than 4 mm from the pedicle cortex. Seven patients in the control cohort were surgically retreated because of postoperative neurologic deficits, whereas no patients in the computer-assisted group were surgically retreated.

Computer assistance can decrease the incidence of incorrectly positioned pedicle screws.