The Clinical Application Value Of POSSUM And Its Modified Scoring System In Pancreatoduodenectomy Risk Evaluation

Objiective: To explore the related risk factors for complications and death in pancreatoduodenectomy, and to discuss the clinical application value of POSSUM and its modified scoring system in pancreatoduodenectomy through a retrospective analysis.Methods: Three hundred fifty cases of pancreatoduodenectomy were collected from January 2010 to December 2013 in the fourth Hospital of Hebei Medical University.We recorded preoperative systemic physiological state important viscera function, nutritional status, perioperative preparation, operation situation, conditions of pancreas, tumor size and position, if or no applicationing somatostatin after surgery, and POSSUM score of these cases. All datas were organized to analyze the relevance to the postoperative complications and death. Screened the related risk factors for postoperative complications and death using Logistic regression analysis. POSSUM score was carried out to the datas of 350 cases, and calculated the predicted incidence of complications and mortality through scores, compared with the actual ones, observated the fitness of the actual and estimated values. All data were analyzed by SPSS 21.0 statistical software, single factor analysis using chi-square test, multiple factors analysis using Logistic regression analysis. The difference is statistically significant if P<0.05.Results: There were 160 cases developping postoperative complications in 350 patients. Of these, eleven cases died(3.14%), pancreatic fistula 47 cases(13.43%), bleeding 37 cases(10.57%), delayed gastric emptying 31 cases(8.86%), stress ulcer 22 cases(6.28%), abdominal cavity infection 50 cases(14.28%), pleural effusion 8 cases(2.28%), abdominal cavity effusion 5 cases(1.42%), pneumonia 11 cases(3.14%), biliary fistula 6 cases(1.71%). Logistic regression analysis showed that age, weight, respiratory disease, tumor size, tumor location, if or on applicaing somatostatin postoperative were risk factors influencing complications(P<0.05). Physiological index of 350 patients was 12 ~ 34 points, average score was 18.53±4.28, surgery severity index was 14 to 24 points, average 16.33±1.89, the total POSSUM score was 26 ~ 58 points, average 34.85±4.82. The incidence of complications predicted by POSSUM scoring system was 53.88%, while the actual one was 45.71%, Chi-square test showed statistically significant(P<0.05), POSSUM scoring system was over-predicted the incidence of postoperative complications. The postoperative mortality was 11.72% by POSSUM scoring system, and the actual postoperative mortality was 3.14%, Chi-square test showed statistically difference(P<0.01), POSSUM scoring system predicted highly the postoperative mortality. The mortality predicted by P-POSSUM(The modified POSSUM scoring system for the prediction of mortality) scoring system was 3.23%, Chi-square test showed no statistical difference(P>0.05), which meaned P-POSSUM scoring system could predicted accurately the postoperative mortality.Couclusions: Patient’s age, weight, respiratory disease, tumor size, tumor location, postoperative application of somatostatin or no were risk factors influencing complications and death after pancreatoduodenectomy. Incidence of postoperative complications and mortality predicted by POSSUM scoring system were higher than actual ones, while the modified P-POSSUM scoring system could predicted more accurately postoperative mortality, but not a good predictor for complications rate. POSSUM scoring system had certain clinical application value in predicting the risks of pancreatoduodenectomy, It needed further study and improve.