Abstracts
2002 Digestive Disease Week

# 105103 Abstract ID: 105103 Outcome after Pancreaticoduodenectomy for Periampullary Cancer: An Analysis from the Veterans Affairs National Surgical Quality Improvement Program
Kevin G Billingsley, Kwan Hur, Jennifer Daley, William Henderson, Shukri F Khuri, Richard H Bell, Seattle, WA; Hines, IL; Boston, MA; Chicago, IL

Introduction: Pancreaticoduodenectomy (PD) remains a formidable technical procedure that is accompanied by significant morbidity and mortality. The objective of this study is to define the clinical risk factors that may predict adverse outcomes after PD for periampullary cancer in the VA healthcare system. Methods: The VA National Surgical Quality Improvement Program (NSQIP) collects clinical data on patients undergoing major surgery at 128 VAMCs. All patients undergoing PD with a diagnosis of periampullary malignancy from 10/90 to 09/00 who were registered in this database were included in the study group. Independent variables examined included 68 preoperative and 12 intraoperative clinical risk factors. The primary dependent variable was 30-day postoperative mortality. Logistic regression analysis was used to identify preoperative risk factors that are associated with 30-day mortality. Results: The study group was comprised of 495 patients undergoing PD. The group was 99% male and the mean age was 65.3+ 9.7 years. The patients were 72% white, 18% black, and 10% other. The 30-day operative mortality was 10.5%(52/495). Preoperative biliary tract instrumentation was coded for 120 patients. Multivariable logistic regression identified four preoperative predictors of 30-day operative mortality. Predictors included: serum albumin, presence of disseminated cancer, American Society of Anesthesiologists (ASA) classification, and impaired sensorium. Additional modeling examined the significance of process of care variables. Operative time was a significant predictor of mortality, but preoperative biliary tract instrumentation was not. Conclusions: These data indicate that there are several patient specific factors that are associated with the risk of postoperative death following PD. Many of these factors are related to underlying severity of illness. These findings underscore the importance of risk adjustment for severity of illness when comparing postoperative outcomes between different health care systems. These factors may provide parameters to guide risk-adjusted analyses of PD outcomes.