Presenter: Isabel DeLaura BS | Duke University Medical Center Background: Robotic pancreaticoduodenectomy (RPD) is a complex operation with a well-studied but inconsistent learning curve of approximately 30-250 cases. The learning curve for laparoscopic pancreaticoduodenectomy (LPD) using a team approach at Duke University has previously been reported and defined as 50 cases to reach proficiency. This study aims to compare the experience building the laparoscopic and robotic pancreaticoduodenectomy programs and to determine whether the transition from LPD to RPD shortened the learning curve for RPD. Methods: LPD and RPD cases performed at Duke University Hospital were retrospectively analyzed. Continuous operative, pathologic, and perioperative metrics from LPD and RPD were compared with independent t-test or Wilcoxon rank sum test, while categorical variables used Fisher’s exact test. Learning curves were defined with respect to OR time using CUSUM analysis. Distinct phases on the CUSUM plots for LPD and RPD were determined at clear transitions in the slope of the CUSUM fitline. Results: Thirty-nine LPD and 69 RPD were included in our analysis. Three LPD and 0 RPD were converted to open. Introduction of LPD was completed first as a hybrid technique including laparoscopic resection and open reconstruction for 23 cases (excluded from this analysis), followed by total LPD. Based on operative time, LPD had an apparent bimodal or inverted learning curve likely accounting for proficiency attained from hybrid LPD and introduction of additional surgeons and trainees later in the experience. The learning curve for RPD had an accelerated early experience phase (case 1-10), a skill consolidation phase (case 11-40), followed by an improvement phase marked by reduction in operative time (case 41-69). Compared to LPD, RPD had a shorter operative time (median 379 vs 406 minutes, p<0.03), improved LN harvest (median 15 vs 19, p<0.02), and similar estimated blood loss (EBL) (median 250ml vs 400ml, p = 0.26) and R0 resection (83% vs 85%, p = 1.0). Additionally, RPD had improved LOS (median 7 days vs 9 days, p<0.002), and lower rates of surgical site infection (10% vs 38% p<0.001), abscess (7% vs 23%, p<0.04), and readmission (13% vs 33%, p<0.03). Conclusion: Adoption of LPD can be safely performed by surgeons with experience in complex laparoscopic procedures. The transition of a competent robotic surgeon from LPD to RPD is facilitated by prior familiarity with MIS approaches to pancreatic resection which result in an abbreviated learning curve. Adoption of robotics can improve surgical quality and perioperative outcomes including operative time, LN harvest, LOS, post-operative infection, and readmission. As exposure to robotics within General Surgery, Hepatopancreatobiliary, and Surgical Oncology training programs increase, surgical quality and patient outcomes will be further improved using RPD over LPD.