Volume 32, Issue 3 262 p. 419-428
Article

Risk Prediction for Development of Pancreatic Fistula Using the ISGPF Classification Scheme

Wande B. Pratt

Wande B. Pratt

Department of Surgery, Beth Israel Deaconess Medical Center and Harvard Medical School, 330 Brookline Avenue, ST 9, 02215 Boston, Massachusetts, USA

Search for more papers by this author
Mark P. Callery

Mark P. Callery

Department of Surgery, Beth Israel Deaconess Medical Center and Harvard Medical School, 330 Brookline Avenue, ST 9, 02215 Boston, Massachusetts, USA

Search for more papers by this author
Charles M. Vollmer Jr.

Corresponding Author

Charles M. Vollmer Jr.

Department of Surgery, Beth Israel Deaconess Medical Center and Harvard Medical School, 330 Brookline Avenue, ST 9, 02215 Boston, Massachusetts, USA

[email protected]Search for more papers by this author
First published: 04 January 2008
Citations: 18

Presented at the American Hepato Pancreato Biliary Association, Annual Meeting, Las Vegas, NV; April 21, 2007, and the Pancreas Club Annual Meeting, Washington, DC, May 20, 2007.

Abstract

Background:

The International Study Group on Pancreatic Fistula (ISGPF) classification scheme has become a useful system for characterizing the clinical impact of pancreatic fistula. We sought to identify predictive factors that predispose patients to fistula, specifically those with clinical relevance (grades B/C), and to describe the clinical and economic significance of risk stratification within this framework.

Methods:

Overall, 233 consecutive pancreatoduodenectomies were performed between October 2001 and March 2007 in our institution. Pancreatic fistula is defined according to the ISGPF classification scheme. Logistic regression analysis was performed to identify risk factors for pancreatic fistula development. These features were then analyzed to determine whether additive risk severity equates to worsening clinical and economic impact.

Results:

Fistulas of any extent occurred in 60 patients, but only 31 (14%) were clinically relevant. There are no identifiable risk factors for grade A biochemical fistulas. Multivariate analysis shows that small pancreatic duct size (<3 mm); soft gland texture; ampullary, duodenal, cystic, or islet cell pathology; and increased intraoperative blood loss (> 1,000 ml) are associated with clinically relevant fistulae. An additive effect is further illustrated, in which clinical and economic outcomes progressively worsen as risk profile increases. Each additional risk factor increases the odds of developing a clinically relevant fistula by 52%.

Conclusions:

For pancreatoduodenectomy, small duct size; soft gland texture; ampullary, duodenal, cystic, or islet cell pathology; and increased intraoperative blood loss are convincing risk factors for the development clinically relevant fistulae as judged by ISGPF classification. As risk profile accrues, patients suffer more complications, encounter longer hospital stays, and incur greater hospital costs. These outcomes can be predicted in the operating room through accurate delineation of high-risk glands.

The full text of this article hosted at iucr.org is unavailable due to technical difficulties.