Objective: To preliminarily summarize the experience in robotic-assisted pancreaticoduodenectomy (RPD) by the DaVinci system.
Methods: The clinical data of 18 patients undergoing RPD between November 2015 and January 2018 were retrospectively analyzed.
Results: In the 8 patients, the operative time was (450±30) min, intraoperative blood loss was (525±125) mL, with no requirement of blood transfusion; 1 case (5.6%) was converted to open surgery; the time for postoperative intestinal function recovery was (4.0±1.0) d, and the length of postoperative hospital stay was (16±4) d; postoperative complications occurred in 7 cases, including pancreatic fistula in 4 cases (22.2%) (of whom all were grade B pancreatic fistula), biliary fistula in 1 case, and intra-abdominal hemorrhage in 2 cases; no reoperation was required, and the postoperative mortality rate was 5%. Postoperative pathological results showed 3 cases of highly differentiated adenocarcinoma of the pancreas, 1 case of poorly differentiated ductal adenocarcinoma, 
3 cases of pancreatic serous cystadenoma, 2 cases of pancreatic mucinous cystadenoma, and 1 case of pancreatic ductal papillary myxoma; 2 cases of well-differentiated adenocarcinoma of the duodenum, 2 cases of duodenal villus tubular adenoma; 1 case of poorly differentiated adenocarcinoma of lower common bile duct, 2 cases of well differentiated adenocarcinoma of lower common bile duct, and 1 case of chronic pancreatitis. Of the 9 patients with malignant tumors, R0 resection was achieved in 8 cases and R1 resection was obtained in 1 case; the number of dissected lymph nodes was 16±4.
Conclusion: RPD is clinically safe and feasible, without increase of procedure-associated complications, and meanwhile, it can accelerate the postoperative recovery of the patients.