Pancreaticoduodenectomy performed by surgeons in training and the risk of post-operative fistula: a systematic review and meta-analysis

Abstract

Introduction: The complexity of pancreaticoduodenectomy (PD) and fear of morbidity, particularly post-operative pancreatic fistula (POPF), can be a barrier to surgical trainees gaining vital experience. This meta-analysis sought to establish the POPF rate following PD by trainees or established surgeons. Methods: A systematic review of the literature was performed using PRISMA guidelines and meta-analysis compared complication rates using RevMan software. Results: 3 of 53 studies were included for meta-analysis, all defining POPF using ISGPS 2005 criteria. Some 309 PD (16%) were performed by trainees. The rate of POPF after surgery performed by those who had completed training was not different when surgery was performed by trainees (19.6 vs 23.0%; OR: 0.65; 95%CI 0.36-1.18; p=0.07) or mortality (OR: 1.08; 95%CI 0.30-3.97; p=0.60). Neither soft pancreatic texture (OR: 0.62; 95%CI: 0.19-1.99; p=0.042) nor pancreatic duct width significantly differed between the two groups. Gastrointestinal bleeding, blood loss and operative time were greater when operations were performed by trainees but there was no difference in delayed gastric emptying, intra-abdominal collection or mortality. Conclusions: PD, when performed by trainees, is associated with acceptable outcomes. Evidence of heterogeneity in key variables indicates a need for further studies and it is unclear whether outcomes are similar when trainees perform surgery among patients stratified as low or high risk for POPF using established risk scores. The use of risk adjusted CUSUM as reported recently could be a useful tool to assess trainees performance.