Comprehensive Learning Curve of Robotic Surgery: Discovery From a Multicenter Prospective Trial of Robotic Gastrectomy.

Abstract

OBJECTIVE\nTo evaluate the complication-based learning curve and identify learning-associated complications of robotic gastrectomy.\n\n\nSUMMARY BACKGROUND DATA\nWith the increased popularity of robotic surgery, a sound understanding of the learning curve in the surgical outcome of robotic surgery has taken on great importance. However, a multicenter prospective study analyzing learning-associated morbidity has never been conducted in robotic gastrectomy.\n\n\nMETHODS\nData on 502 robotic gastrectomy cases were prospectively collected from 5 surgeons. Risk-adjusted cumulative sum analysis was applied to visualize the learning curve of robotic gastrectomy on operation time and complications.\n\n\nRESULTS\nTwenty-five cases, on average, were needed to overcome complications and operation time-learning curve sufficiently to gain proficiency in 3 surgeons. An additional 23 cases were needed to cross the transitional phase to progress from proficiency to mastery. The moderate complication rate (CD ≥ grade II) was 20% in phase 1 (cases 1-25), 10% in phase 2 (cases 26-65), 26.1% in phase 3 (cases 66-88), and 6.4% in phase 4 (cases 89-125) (P < 0.001). Among diverse complications, CD ≥ grade II intra-abdominal bleeding (P < 0.001) and abdominal pain (P = 0.01) were identified as major learning-associated morbidities of robotic gastrectomy. Previous experience on laparoscopic surgery and mode of training influenced progression in the learning curve.\n\n\nCONCLUSIONS\nThis is the first study suggesting that technical immaturity substantially affects the surgical outcomes of robotic gastrectomy and that robotic gastrectomy is a complex procedure with a significant learning curve that has implications for physician training and credentialing.