Navigation versus experience: providing training in accurate lumbar pedicle screw positioning.

PURPOSE: Accurate placement of spinal pedicle screws (PS) is mandatory for good primary segmental stabilization allowing consequent osseous fusion, requiring judgmental experience developed during a long training process. Computer navigation offers permanent visual control during screw manipulation and has been shown to significantly lower the risk of pedicle perforation. This study aims to evaluate whether safety, accuracy, and judgmental skills in screw placement, comparable to an experienced surgeon, can be developed during training using computer navigation.

METHODS: Lumbosacral PS were placed in 18 patients in a prospective setting, in one segment side with conventional fluoroscopy by a senior spine-surgeon, and computer navigated on the other side by a trainee without prior experience in the technique. At the beginning and at the end of the study, PS were placed freehand in solid foam models by the trainee. PS placement time, intraoperative placement revisions, PS placement accuracy on postoperative CT scans, and postoperative complications were assessed.

RESULTS: Significant improvement of trainee's PS placement accuracy (Sclafani score 8.2-8.83; p = 0.006) and time (13.3-6.8 min per screw; p = 0.005) to a similar level as the experienced surgeon state (5.2-4.1 min per screw; p = 0.39) was explored; similar improvement was explored in the foam models. The number of intraoperative placement revisions kept on a low level for surgeon (3.3-0.0%) and trainee (5.1-2.6%) during the whole study, no postoperative complications occurred.

CONCLUSION: Navigated PS insertion allows safe teaching from the early beginning of surgical training, due to steady intraoperative control on PS placement. Adequacy of PS placement is similar to screws placed by an experienced surgeon. Progress in judgmental skills in screw placement can be gained rapidly by the trainee, which can also be transferred to non-computer navigated PS placement.