Classification of Robotic-Assisted Pedicle Screw Errors

Introduction: Although the reported accuracy of robotic-assisted pedicle screws is very high, the current literature does not capture near misses or incidental procedural errors that may have been accounted for during surgery or did not alter treatment. Here, we characterize the types of errors encountered during robotic-assisted pedicle screw insertion and suggest best practices to minimize their occurrence.

Methods: We performed a retrospective review of all cases at our institution in which robotic-guided pedicle screw placement was used. Intraoperative imaging was analyzed to look for evidence of robotic-related errors. Cases where errors occurred were critically appraised to determine the mechanism of complication. We then performed a systematic review to determine the incidence and types of errors in the current literature.

Results: We critically reviewed 75 consecutive surgeries using robotic guidance at the University of Utah. We identified 15 cases where screw placement complications occurred, of which 14 were near misses with no harm to the patient and 1 resulted in neurological injury. Patterns of screw placement complications were classified as errors in: 1) registration, 2) screw planning, 3) procedural interference. Our systematic review yielded 5 relevant citations reporting robotic-assisted screw errors. Four studies reported registration errors, accounting for an average of 48.6% of screw failures. Three studies highlighted screw placement errors, contributing to an average of 45.23% of failures, and interference errors were reported in three studies as well, making up an average of 35.3% of failed screws.

Conclusion : The characterization of errors commonly encountered during robotic spine surgery is not commonly reported. We describe a classification scheme to standardize the reporting of robotic guidance-related screw complications. Awareness of how errors occur may increase the safety of this technology.