Risk Factors For Major Complications And Death After Pancreatoduodenectomy

Objective For every surgeon, Pancreatoduodenectomy(PD) is a great challenge. More complications and death happened after PD because of resections of more organs and complicated reconstruction. As major complications with high morbidity, anastomotic fistula, intraabdominal abscess and postoperative hemorrhage, which threaten patients' lives, are leading causes of death after PD. At present, data of risk factors for major complications and mortality remains controversial. The aim of this study was to evaluate these risk factors by retrospectively analysing medical records of 141 patients undergone PD, so as to decrease the rate of morbidity and mortality.Patients and Methods 141 patients have undergone panceatodudenectomy in the First Affiliated Hospital of Zhengzhou University from Dec. 1994 to Nov. 2004. There were 93 males (66.0%) and 48 females (34.0%). Mean age for patients was 52.3 years with a range 10 to 79 years. The operations performed consisted of 88 standard pancreatoduodenectomies (SPD) and 53 pylorus-preserving pancreatoduodenectomies (PPPD).Using designed tables, the medical records of 141 cases undergone PD were retrospectively investigated. The data included clinical factors, laboratory factors,operative factors, pathological diagnosis, and complications. The statistical software SPSS 10.0 was used for the statistical analysis. Statistical analysis were performed using the Student's t test, the chi-square test, and Fisher's exact test of probabilities for univariate analysis. We then selected factors with p values less than 0.10 for inclusion as factors in the mutivariate analysis using stepwise logistic regression. Factors kept in the logistic regression model were considered to be significant independent risk factors. The calculated numeric data were expressed as mean ±standard deviation of the mean(x ±s).Results The rates of pancreatic fistula, biliary fistula, intraabdomial abscess, and postoperative hemorrhage were 9.9%, 9.2%, 12.1%, and 11.3% respectively. The 30-day mortality rate was 9.2%. Independent risk factors for pancreatic fistula were jaundice(OR=13.063), history of upper abdominal surgery (OR=10.678), routine invaginated pancreatojejunostomy (OR=7.279), and the use of octreotide(OR=0.062). Predictive equation of pancreatic fistula was P=1/[1+e-(-4.561+2.570jaundice+2.368upperabdominal surgery+1.985method of pancreatojejunostomy-2.783octreotide) ] . Independent risk factors forbiliary fistula were operative blood loss more than or equal to 1000ml (OR=6.271) and diameter of biliary duct less than 1.5cm (OR=5.940). Predictive equation ofbiliarv fistula was P=1 /[1+e-(-3.286+1.836amount of operative blood loss+1.782diameter of biliary duct) ]Independent risk factors for intraabdomial abscess were postoperative intraabdomial hemorrhage (OR=24.026), operative blood loss more than or equal to 1000ml (OR=7.583), and anastomotic fistula (OR=4.3). Predictive equation of intraabdomialabscess was P=1 / [1+e-(-3.265+3.179intraabdomial hemorrhage+2.206amount of operative bloodloss+1.426anastomotic fistula) ]. Independent risk factors for postoperative hemorrhage were man (OR=5.831), soft consistency of pancreas (OR=4.627), postoperative hemoglobin less than 90g/L (OR=3.893), and large amount of operative blood transfusion (OR= 1.001). Predictive equation of postoperative hemorrhage was P=1 /[1+e-(-5.231+1.763gender+1.532consistency of pancreas +1.359postoperative serum hemoglobin levels+0.001 amount ofoperative blood transfusion) ]. Independent risk factors for mortality were postoperative hemorrhage (OR=24.642), preoperative serum albumin less than 30g/L (OR=8.311),operative blood loss more than or equal to 1000ml (OR=6.548), and preoperative serum bilirubin more than or equal to 171 μ mol/L (OR=5.127). Predictive equation ofmortality was P=1 /[1+e-(-4.707+3.204postoperative hemorrhage+2.118preoperative serum albumin levels +1.879amount of operative blood loss +1.634preoperative serum bilirubin levels) ].Conclusion Jaundice, history of upper abdominal surgery, and routine invaginated pancreatojejunostomy indicated a higher rate of pancreatic fistula. Therefore for a decrease of pancreatic fistula, body condition including disorder of blood coagulation and endotoxaemia, which resulted from jaundice, should be improved and palliative operations should be avoided if PD couldn't be performed successfully, as well as we'd better select one-staged PD and binding pancreatojejunostomy. As a retrospective research, this study couldn't satisfy design of clinical trail to estimate the effect of medicine, so we couldn't draw a conclusion about octreotide.The major risk factors of biliary fistula were operative blood loss more than or equal to 1000ml and diameter of biliary duct less than 1.5cm. So the operation should be cautiously performed and we should do our best to decrease the amount of operative blood loss. For patients with a thin biliary duct, choledoehojejunostomy must be performed uniformly and reassuringly. All these were favorable to lessen biliary fistula,Patients with operative blood loss more than or equal to 1000ml frequently suffered from intraabdomial abscess. Postoperative intraabdominal bleeding and anastomotic fistula always resulted in intraabdominal abscess. To attain a lower morbidity of intraabdominal abscess, large amount of operative blood loss should be avoided and postoperative intraabdominal bleeding and anastomotic fistula should be prevented.The risk factors of postoperative hemorrhage included man, soft consistency of pancreas, postoperative hemoglobin less than 90g/L, and large amount of operative blood transfusion. We should accordingly pay more attention to male patients or ones with soft consistency of panceas and decrease the amount of operative blood loss andblood transfusion as we can. Postoperative hemoglobin levels should maintain more than 90g/L. It was of benefit to postoperative hemorrhage prophylaxis.Postoperative hemorrhage, preoperative serum albumin less than 30g/L, amount of operative blood loss more than or equal to 1000ml, and preoperative serum bilirubin more than or equal to 171 μ mol/L played an important role in operative mortality. Consequently it was helpful for a low operative mortality to actively treat preoperative hypoalbuminemia and hyperbilirubinemia and avoid seriously bleeding during or after operation.