Background: Series of patients with pancreas cancer from single high-volume institutions or surgeons have demonstrated improvements in morbidity and mortality of pancreatic resection in recent decades. The experience of these single institutions or surgeons may not, however, reflect the results achieved by a cross-section of surgeons or hospitals. This article examines the resection outcome for a large unselected group of university hospitals and surgeons.
Methods: Pancreas cancer resection morbidity and mortality were examined using a multi-institution data base of discharge coding data from 26 American university hospitals. The data were analyzed for relationships of morbidity and mortality with the type of resection, patient age, hospital volume, and individual surgeon case load.
Results: Two hundred twenty-three resections were performed in 1989-1990 (pancreaticoduodenectomy, 168 patients; total pancreatectomy, 11; distal pancreatectomy, 30; and islet tumor resection, 14). The mortality rate was 6% (13 of 223) with major complications in 21%. Patient age did not correlate with complications or death. The surgeon case load ranged from 1-15 cases (median, 1) over the 2-year period. The mortality rate did not correlate with the case load. Surgeons performing one to three resections had significantly more complications than those performing four or more resections (P = 0.011).
Conclusions: Pancreas resection is performed by an unselected cross-section of surgeons in American university centers with acceptable morbidity and mortality rates.