gms | German Medical Science

Objective: Aim of this study was to evaluate accuracy, revision and complication rate, technical obstacles and learning curve of 3D-fluoroscopy navigated pedicle screw placement in the thoracolumbar spine.

Methods: Between January and November 2011 685 pedicle screws (74 thoracic / 611 lumbar; 107 patients; 50 m / 57 f; 1 motion segment 37 pts., 2 / 27 pts., 3 / 29 pts., >4 / 14 pts.) were placed by use of intraoperative 3D-fluoroscopy navigation (Brainlab) for degenerative spine, spondylolisthesis, tumor and infection by 6 experienced board certified neurosurgeons. Accuracy in postoperative CT-scan, mean navigation/screw placement time, intraoperative revisions/resurgeries and complications were analysed and technical problems assessed separately for the first, second and last third of the 11 months period.

Results: Mean 3D-fluoro-navigation time was 11.7 min (15.9 min 1^st third / 11.9 min 2^nd third / 10.2 min 3^rd third). Mean screw placement time (time/screw) was 4.9 min (5 min 1^st and 2^nd third / 3.9 min 3^rd third). Overall accuracy of screw placement was 95.6% (81.4% thoracic / 97.1% lumbar) in postoperative CT-scan. Divided in the 1^st, 2^nd and 3^rd third: accuracy for thoracic screws was 78% / 89% / 77% and for lumbar screws 96.2% / 96.7% / 97.9%. 36 screws (9 thoracic / 27 lumbar) were revised intraoperatively. 5 lumbar screws in 3 pts. were repositioned in revision surgeries. In 2 cases conversion to freehand screw placement were necessary (software failure/extreme obesity).

Conclusions: Intraoperative image-guided spinal 3D-fluoro-navigation achieves good accuracy of screw placement. We could demonstrate a learning curve by reduction of mean navigation and screw placement time and by increase of lumbar screw placement accuracy. The accuracy for thoracic screws was lower compared to the lumbar spine and there was no improvement over the observed time period.