Predictors of morbidity and mortality in patients with traumatic duodenal injuries.

The aim of our study is to determine factors that predict morbidity and mortality in patients with traumatic duodenal injury (DI). A retrospective review from July 1996 to March 2003 identified 52 patients admitted to our trauma center (age 24.4 +/- 2.1 years, ISS = 18.8 +/- 1.76). The mortality rate for patients with duodenal injury was 15.4 per cent (n = 8). The mechanisms of injury were blunt (62%), gun shot wound (GSW) (27%), and stab wound (SW) (11%). There was no difference in mortality based on mechanism of injury. Management was primarily nonoperative [n = 30 (57%)]. Of those with perforation (n = 22), 64 per cent underwent primary repair (n = 14), 23 per cent duodenal resection (n = 5), 9 per cent duodenal exclusion (n = 2), and one patient pancreaticoduodenectomy. The method of initial surgical management was not related to patient outcome. Univariate analysis demonstrated that nonsurvivors were older, more, hypotensive in the emergency department, had a more negative initial base deficit, had a lower initial arterial pH, and had a higher Injury Severity Score. Nonsurvivors were also more likely to have an associated inferior vena cava (IVC) injury. Multivariate regression analysis revealed age, initial lowest pH, and Glasgow Coma Score to be independent predictors of mortality, suggesting that the physiologic presentation of the patient is the most important factor in predicting mortality in patients with traumatic DIs.