Summary
This study aimed to demonstrate the learning curve of robot-assisted minimally invasive esophagectomy (RAMIE). A retrospective analysis of the first 124 consecutive patients who underwent RAMIE with intrathoracic anastomosis (Ivor Lewis) by a single surgeon between May 2015 and August 2020 was performed. An risk-adjusted cumulative sum (RA-CUSUM) analysis was applied to generate a learning curve of RAMIE considering the major complication rate, which reflected the technical proficiency. The overall 30-day morbidity rate was 38.7%, while the major complication rate was 25.8%. The learning curve was divided into two phases based on the RA-CUSUM analysis: phase I, the initial learning phase (cases 1–51) and phase II, the proficiency phase (cases 52–124). As we compared the proficiency phase with the initial learning phase, significantly decreased trends were observed in relation to the major complication rate (37.3% vs. 18.7%, P = 0.017), total operation time (330.9 ± 55.6 vs. 267.3 ± 39.1 minutes, P < 0.001), and length of hospitalization (10 [IQR, 9–14] days vs. 9 [IQR, 8–11] days, P = 0.034). In conclusion, the learning curve of RAMIE consisted of two phases, and at least 51 cases were required to gain technical proficiency.
