• 30 July 2002
    • journal article
    • research article
Abstract
This study was designed to evaluate risk factors influencing pancreatic leakage and pancreatic leakage-related mortality in a medium-volume hospital. We retrospectively reviewed the clinical records of 107 patients who underwent pancreaticoduodenectomy at the Kobe University Hospital. Fourteen predictive factors for pancreatic leakage and the pancreatic leakage-related mortality were evaluated using univariate and multivariate logistic regression models. In univariate analysis, the degree of pancreatic fibrosis, type of resection (PD/PPPD), anastomosis techniques (invagination or duct-to-mucosa anastomosis), anastomosis sites (jejunum/stomach), and the presence of congestion in anastomosis sites significantly influenced pancreatic leakage, and the degree of pancreatic fibrosis influenced pancreatic leakage-related mortality. Multivariate logistic regression analysis revealed that congestion in anastomosis sites was the strongest parameter for pancreatic leakage. Univariate analysis of the patients with normal/mild fibrosing pancreas revealed that pancreatic leakage was influenced by type of resection, anastomosis techniques, anastomosis sites, congestion in anastomosis sites and the management of pancreas parenchyma. In a medium-volume hospital, reconstruction after pancreaticoduodenectomy should be performed with careful attention to pancreas and anastomosis sites. In the patients with normal/mild fibrosing pancreas, duct-to-mucosa anastomosis without suturing the pancreas parenchyma may be a useful technique for reconstruction.