Purpose: Despite the evolution in the management of traumatic colorectal injuries in both civilian and military settings during the previous few decades, they continue to be a source of significant morbidity and mortality. The purpose of this study was to analyze management and clinical outcomes from a cohort of patients suffering colorectal injuries.
Methods: This was a retrospective analysis of prospectively collected data from all patients injured and treated at the 31st Combat Support Hospital during Operation Iraqi Freedom from September 2003 to December 2004.
Results: From the 3,442 patients treated, 175 (5.1 percent) had colorectal injuries. Patients were predominately male (95 percent), suffered penetrating injuries (96 percent), and had a mean age of 29 (range, 4-70) years. Ninety-one percent of patients had associated injuries. Initial management included primary repair (34 percent), stoma (33 percent), resection with anastomosis (19 percent), and damage control only (14 percent). By injury location, stomas were placed more frequently with rectal or sphincter injuries 65 percent (25/40) vs. other sites (right, 19 percent (8/42); transverse, 25 percent (8/32); left, 36 percent (20/55); P < 0.01). Thirteen percent of patients eventually received stomas for failure of initial in-continuity management. Patients with colorectal injuries had a significantly increased mortality rate than those without (18 percent (31/175) vs. 8 percent (269/3267); P < 0.001) but not the subset without colorectal injuries undergoing celiotomy (18 vs.14.4 percent; P = 0.41). Rectal (odds radio, 22; P = 0.03) and transverse colon (odds radio, 17; P = 0.04) injuries were independently associated with increased mortality in multivariate regression analysis. Initial placement of stoma had an independent association with lower leak rates (odds radio, 0.06; P = 0.04).
Conclusions: Injury to the rectum or transverse colon is an independent predictor of mortality. The use of a diverting stoma varied by injury site and was associated with a decreased leak rate but demonstrated no impact on the incidence of sepsis or mortality.