Pancreaticoduodenectomy: minimizing the learning curve.

Background: Pancreaticoduodenectomy outcomes improve as surgeon experience increases. We analyzed the outcomes of pancreaticoduodenectomy for any improvements over time to assess the learning curve.

Methods: A retrospective study of patients undergoing consecutive pancreaticoduodenectomy by a single surgeon at the beginning of practice was performed. Operative factors and 90-day outcomes were examined and trends over the course of the 4-year time period were analyzed.

Results: Between July 2011 and June 2015, 124 patients underwent pancreaticoduodenectomy (including total pancreatectomy, n=17) by open (n=93) or a laparoscopic (n=31) approach. The median operative time was 305 minutes which significantly improved over time. The median blood loss and length of stay were 250 mL and 6 days respectively which did not change over time. The pancreatic fistula rate, total morbidity, major morbidity, and mortality, and readmission rate was 7.5%, 41.1%, 14.5%, 1.6%, and 15.3% respectively and did not change over time. Pancreaticoduodenectomy was performed most commonly for pancreatic adenocarcinoma (51.6%) with a negative margin rate of 91.1% which significantly improved over time.

Conclusions: The performance of pancreaticoduodenectomy improves as surgical experience is gained. However, a learning curve that impacts patient outcomes can be considerably diminished by appropriate training, high-volume practice/institution, proficient mentorship and experienced multidisciplinary team.