Postoperative pancreatic fistulas are not equivalent after proximal, distal, and central pancreatectomy

It is uncertain whether postoperative pancreatic fistulas after distal and central pancreatectomies behave similarly to those after pancreaticoduodenectomy. To date, this concept has not been validated either clinically or economically. Overall, 256 consecutive pancreatic resections from October 2001 to February 2006 (184 pancreaticoduodenectomics, 66 distal pancreatectomies, and 6 central pancreatectomies) were evaluated according to the International Study Group of Pancreatic Fistula classification scheme. Pancreatic fistula was defined as any measurable drainage on or after postoperative day 3, with amylase content greater than three times the normal serum value. Outcomes were divided into four grades: (1) no fistula, (2) grade A: biochemical fistula without clinical sequelae, (3) grade B: fistula requiring any therapeutic intervention, or (4) grade C: fistula with severe clinical sequelae. Grades B and C are considered clinically relevant fistulas based on worsening morbidity, increased length of stay, frequent hospital readmission, and increased costs/resource utilization. Clinical and economic outcomes were compared-grade for grade-across the three resection types. Fistulas of any extent (Grades A-C) occurred in one third of all patients; two thirds had no fistula. Overall, there were 16 readmissions (6%), six reoperations (2%), and no deaths attributable to pancreatic fistula. Outcomes between no fistula and grade A patients were identical across resection types, though grade A fistula was more common in distal pancreatectomy. For each resection type, length of stay and costs progressively increased with grades B and C. However, the negative impact of these clinically relevant fistulas varied between resection types. Rates for intensive care unit admission and rehabilitation placement were higher among pancreaticoduodenectomy patients. Total parenteral nutrition and antibiotic use were similar, but percutaneous drainage was used more often for distal pancreatectomy. Grade B fistula was more severe after distal pancreatectomy, as indicated by increased length of stay, readmissions, and total cost. Although reoperation rates for grade C fistulas were equivalent, intervals to reoperation were substantially longer after distal and central pancreatectomies. When classified according to International Study Group of Pancreatic Fistula criteria, clinically relevant pancreatic fistulas behaved differently depending on type of pancreatectomy. This translates into variable severity that guides management decisions, which ultimately dictate clinical outcomes and economic impact.