Beyond the Learning Curve of Robot-Assisted Navigation Spine Surgery: Refinement of Outcomes With Extended Experience.

Objective This study assessed whether robotic-assisted navigation (RAN) spine surgery outcomes, including operative time and pedicle screw accuracy, continue to improve with extended experience beyond 200 cases. Methods This is a retrospective review of 60 patients who underwent lumbosacral transforaminal interbody fusion using RAN. Patients were segmented into three groups of 20 consecutive cases each. The first group represented a surgical performance baseline leading up to the investigating surgeon's 200th RAN case. The subsequent two groups were selected beyond the 200th case with an average of 15 cases between groups. Pedicle screw accuracy and intraoperative outcomes were assessed. Statistical results were significant if p<0.05. Results Measures of surgical efficiency significantly improved beyond the investigating surgeon's 200th RAN case. As case number increased, the following parameters significantly decreased: registration time (group 1: 16.9±6.5, group 2: 12.9±3.0, group 3: 8.7±1.6 minutes; p<0.05), screw insertion time (group 1: 14.9±3.5, group 2: 10.9±2.0, group 3: 8.4±2.7 minutes; p<0.05), and total operative time significantly decreased from group 1 (175.9±58.2 minutes) to group 2 (135.8±23.9 minutes) (p=0.013) with a non-significant decrease to group 3 (121.5±32.3 minutes). Accuracy (Grade = A) significantly increased across groups (group 1: 87%, group 2: 94%, group 3: 98%; p=0.024). Group 1 had the highest misplacement rate of 3.7% (4/108 screws). The overall misplacement rate was 1.4% (4/290 screws) (Grade C-E). There was a higher rate of lateral screw misplacement compared to medial misplacement. Conclusion Even with a small number of initial cases, RAN spine surgery can consistently be performed with high accuracy and acceptable intraoperative outcomes. However, this study demonstrated refined outcomes with extended robotic experience.