The development of task-specific metrics for grading the robotic gastrojejunostomy in robotic pancreaticoduodenectomy

Background Minimally invasive surgery affords the opportunity for video review. Procedures are unique and global assessments may not be helpful in identifying points of improvement. Robotic pancreaticoduodenectomy (RPD) is a complex procedure. Trainees typically start with the gastrojejunostomy (GJ) step. In this work, we aim to develop task-specific metrics for robotic GJ based on expert consensus and to evaluate these metric items relative to global grading schemes. Methods This is a retrospective review of prospectively collected GJ RPD videos at two quaternary referral centers from 2011 to 2022. These videos were rated according to overall performance, objective structured assessment of technical skills (OSATs), global evaluative assessment of robotic skills (GEARS) and task-specific metrics (TSM). A second grader rated a subset of videos and intra-class correlation was calculated. The final task-specific metrics were developed using partial least squares regression (PLS) based on overall performance. Results A total of 49 videos were analyzed. These were rated according to OSATs, GEARS and task-specific metrics. Overall performance (Likert scale 1–5), total OSATs and total GEARS correlated poorly with total task scores ( R = 0.55, P < 0.05), ( R = 0.5, P < 0.05) and ( R = 0.54, P < 0.05), respectively. After PLS, total scores for the final task-specific metrics correlated well with overall performance ( R = 0.8, P < 0.05), OSATs ( R = 0.76, P < 0.05) and GEARS ( R = 0.74, P < 0.05). Conclusion Task-specific metrics with good correlation with established rating schemes were derived for Robotic Gastrojejunostomy. These metrics may provide trainees with precise task-specific feedback to improve performance.