Introduction Pancreatic fistula (PF) remains a significant cause of morbidity and mortality after pancreaticoduodenectomy. The relative merits of external (ES) and internal (IS) pancreatic stents on postoperative outcome have not been clearly defined. The aim of this study was to evaluate the role of pancreatic stents after pancreaticoduodenectomy by meta-analysis.

Methods Randomised and observational studies comparing patients with and without pancreatic duct stents (ES or IS) from January 1990 to May 2011 were included in the analysis. Studies were only included if the incidence of PF was separately reported in each group. Secondary outcome measures included postoperative mortality, overall complications and hospital stay. Subgroup analysis was performed for studies that used external stents or internal stents exclusively. Randomised and observational studies were combined separately using a random effects model, and the overall effect was calculated using a mixed effects model. Outcomes were compared using ORs and weighted mean differences (WMD).

Results Sixteen studies were analysed (four randomised and 12 observational), consisting of 1,846 patients (876 stented, 970 not stented). On analysis of all studies (ES and IS), the incidence of PF (p=0.54), overall complications (p=0.23) and perioperative mortality (p=0.19) were similar in stented and non-stented patients. Hospital stay was shorter in the stented group (WMD (−) 2.5 days, CI (−) 4.4 to (−) 0.6 days, p=0.009). On analysis of ES and IS studies separately, PF and overall complications were significantly reduced by external stents (PF: OR 0.49, CI 0.3 to 0.81, p=0.005; complications: OR 0.62, CI 0.43 to 0.89, p=0.01), but no difference was observed with internal stents (PF: OR 1.59, CI 0.97 to 2.61, p=0.07; complications: OR 0.96, CI 0.64 to 1.46, p=0.86).

Conclusion External pancreatic duct stenting appears to reduce the rate of pancreatic fistula and overall complications following pancreaticoduodenectomy. However, due to the lack of high quality evidence, it remains unclear whether internal stents have a role. A well-conducted, adequately powered randomised trial of internal pancreatic stents in patients at risk of pancreatic fistula should help to clarify this.

Competing interests None declared.