gms | German Medical Science

Objective: Robotic-guidance is a recent technology for placement of pedicle screws. We evaluate a single-center’s experience of 3 years of robotic-guided surgery with regards to specific strong sides and limitations of this technique.

Method: The medical records of consecutive patients undergoing pedicle screw placement in lumbar and thoracic spine in our department over 3 years including emergency cases were analyzed for screw placement accuracy, complications and need for re-operations. Special focus was on cases not operated by robotic guidance for various reasons within this period and reasons of undesired results in robot-guided cases.

Results: In total, 289 patients received instrumentation with lumbar and/or thoracic pedicle screws from October 2011-October 2014. Robotic guidance was attempted in 278 including emergency cases and succeeded in 272 patients (receiving 1363 screws). All of these patients except 1 were operated using a percutaneous technique. Main reason for failure (N=6) of roboted technique were registration problems mainly due to obesity (BMI>35) and/or poor bone quality in osteopenic patients. Hard ware problems occurred in 1 case (0.4%). Robotic surgery was not attempted in 11 cases, because of preexisting instrumentation to be revised or extended (N=9) and surgeons decision (N=2). Robotic guided surgery resulted in optimal screw position in 96% and misplacement in 4 % (according grading by Wiesner et al.). Detailed failure analysis revealed that most common reasons were problems in suboptimal planning of the screws entry point yielding a higher risk of skiving of the drill guide. Unnoticed relative movements of patient and/or robot after registration were found in about 0.3%. Intraoperative operating error of the robot software was found in 0.3%. The overall revision rate of implants was 6.6%, due to misplacement 0.36% of screws. 2.7% cases necessitated second surgery revision due to loosening and 2.7% due to dislocation. Complications related to screw insertion were hemorrhage in 1 case and wound infections in 9.3%, which in 7% required a surgical wound revision.

Conclusions: In our series of 289 patients we found that robotic-guided percutaneous surgery was feasible in the vast majority of cases (including emergency cases). Only 4% of the patients were operated via a conventional open midline approach. Robotic guidance was not attempted in 3.8% (preexisting instrumentation, surgeon's decision) and failed in 2.1% mainly because of poor X-ray quality.