Objective Performance Indicators Differ Between Expert Versus Novice Surgeons During Robotic Proctectomy

Objective performance indicators (OPIs), machine learning-enabled metrics calculated from robotic systems data, have potential to offer objective insights into surgical skill. There are limited studies investigating OPIs in colorectal surgery. In this study, we aim to identify OPIs that differ expert and novice surgeons during robotic proctectomy (RP).

Endoscopic videos synchronized to robotic system data were annotated to delineate individual surgical tasks during 30 RPs. We analyzed 541 critical steps: inferior mesenteric artery dissection (78), sigmoid mobilization (167), posterior rectal mobilization (88), and mesorectal dissection (208). OPIs from each step were compared across two expert (>500 robotic procedures) and eleven novice surgeons (<50 procedures).

OPIs analyses revealed significant differences between expert and novice surgeons. Across all steps, experts exhibited greater velocity and acceleration for camera, dominant and nondominant arms. During mesenteric artery dissection, experts utilized fewer arm swaps and longer energy activation time. During sigmoid mobilization, experts used fewer arm swaps, shorter camera moving time, fewer camera movements, and less dominant wrist pitch. During rectal mobilization and mesorectal dissection, experts exhibited greater dominant wrist articulation (roll, pitch, and yaw), and longer path length for camera, dominant and nondominant arms. These results identify specific OPIs that designate surgeon expertise during RP.

Step-specific OPIs significantly differ between expert and novice surgeons during RP. With further validation, these metrics may serve as objective tools for benchmarking performance, guiding surgical training curricula, and informing credentialing processes.