OBJECTIVETo evaluate the pancreaticojejunostomy procedures selection strategy in pancreaticoduodenectomy and to analyze risk factors of pancreatic fistula.

METHODSClinical data of 352 patients who received pancreaticoduodenectomy from September 2009 to September 2012 were retrospectively analyzed. For patients with soft pancreas, binding pancreaticojejunostomy was applied to 153 patients. For patients with hard pancreas, duct-to-mucosa pancreaticojejunostomy (DMPJ) was applied (199 cases). The clinical efficacy and incidence of postoperative complications were compared among 2 groups. Risk factors of pancreatic fistula were screened out from many factors by univariate and multivariate analysis.

RESULTSThe overall incidence of pancreatic leakage was 13.9% (49/352). There were no significant difference in incidences of pancreatic leakage (χ(2) = 0.512), peritoneal bleeding (χ(2) = 0.784), abdominal infection (χ(2) = 1.161), digestive dysfunction rate (χ(2) = 4.753) and mean duration of hospital stay (t = 2.13) among 2 groups (all P > 0.05). The results of multivariate analysis showed pancreatic tube diameter < 3 mm (OR = 5.748), preoperative total bilirubin level > 171 µmol/L (OR = 5.112), duration of preoperative jaundice > 8 weeks (OR = 5.090), preoperative albumin level < 30 g/L (OR = 4.464) were independent risk factors of pancreatic fistula (all P < 0.05).

CONCLUSIONSBunding pancreatojejunostomy was as good as soft pancreatic; for duct diameter ≥ 3 mm suggested using duct-to-mucosa pancreaticojejunostomy. For the risk factors for pancreatic leakage actively cooperate with preoperative nutritional support and timely treatment of jaundice, the incidence of postoperative pancreatic leakage will be further reduced.