Learning curve for robot-assisted rectal resection using the hinotori™ surgical robot system: a risk-adjusted cumulative sum analysis in a surgical team without prior robotic surgery experience.

Learning curves for robot-assisted rectal resection are mainly reported for the da Vinci platform. We evaluated the learning curve and short-term outcomes after introducing the hinotori™ Surgical Robot System as the first robotic platform for a surgeon with no prior console experience and a team with no prior robotic surgery experience. We retrospectively analyzed 60 consecutive robot-assisted rectal resections performed using the hinotori™ Surgical Robot System at a single center. The primary and secondary outcomes were the console time and robot roll-in–to–console start time, respectively. The case mix was adjusted using the Veenhof technical difficulty score and procedure type, and the change point was identified using risk-adjusted cumulative sum (RA-CUSUM) analysis. The RA-CUSUM curve peaked at case 27, defining the learning (cases 1–27) and stable (cases 28–60) phases. The median console time decreased from 198 min (interquartile range, 137–251) to 156 min (112–193) (p = 0.014) despite increased technical difficulty (median Veenhof score, 12 [9–16] vs. 17 [11–22]; p = 0.040). The median robot roll-in–to–console start time was 9 (8–12) and 9 (7–11) min in the learning and stable phases, respectively (p = 0.817). No surgery was converted to open surgery. In addition, postoperative and key pathological outcomes were comparable between phases. In this first robotic platform setting, our case-mix–adjusted analysis suggested that the console time stabilized after 27 cases. The robot roll-in–to–console start time remained stable throughout, suggesting early standardization of the pre-console workflow.