LAUSR

BACKGROUND This study was undertaken to examine our outcomes after robotic pancreaticoduodenectomy and to compare our outcomes with predicted outcomes using the American College of Surgeons (ACS) NSQIP Surgical Risk Calculator and with outcomes reported through ACS NSQIP. METHODS We prospectively followed 155 patients undergoing robotic pancreaticoduodenectomy. Outcomes were compared with predicted outcomes calculated using the ACS NSQIP Surgical Risk Calculator and with outcomes documented in ACS NSQIP for pancreaticoduodenectomy from 2012 to 2017. Median data are presented. RESULTS Eighty-eight percent of our robotic pancreaticoduodenectomies were performed in 2015 to 2018. Predicted outcomes were like those reported in ACS NSQIP. Actual outcomes were superior to predicted outcomes and outcomes reported in ACS NSQIP for overall complications, serious complications, returned to operating room, surgical site infections, deep vein thrombosis, and length of stay. Seventeen percent had conversions to open operations, generally due to failure to progress or need for major vascular reconstruction; only 3 (3.5%) of the last 80 operations were converted to open. Robotic operations took 423 minutes; estimated blood loss was 200 mL. Biliary fistulas occurred in 5% and pancreatic fistulas occurred in 5%. Six percent of patients died perioperatively; 5 patients died due to cardiac deterioration and 4 (3.1%) patients died after pancreaticoduodenectomy completed robotically. CONCLUSIONS Our patients were not a select group, they were like those reported in ACS NSQIP. Their outcomes after robotic pancreaticoduodenectomy were like or better than predicted outcomes or national data. Our mortality was high because of preoperative ill health (eg renal failure) and cardiac risk. Although we believe our results will continue to improve, our current data document the salutary benefits of minimally invasive robotic pancreaticoduodenectomy.