Ostomy as a risk factor for posttraumatic infection in penetrating colonic injuries: univariate and multivariate analyses.

Background: Primary repair for penetrating colonic injury is an acceptable practice in uncomplicated injuries, but it is still viewed with trepidation in high risk patients.

Methods: The records of 350 patients evaluated at an urban Level I trauma center for penetrating colonic injuries over an 8-year period (1989-1997) were reviewed. These included 33 stab and 317 gunshot wounds. Thirty-nine patients died within 48 hours. Of the remaining 311 patients, 78 (25%) developed 152 infections. These infections were classified as traumatic or nosocomial in nature. Traumatic infections (46%) included abdominal abscesses or peritonitis (28), wound infections (30), missile tract infections (8), and fistulas (4), whereas nosocomial infections (54%) included pneumonia (25), bacteremia (25), urinary tract infections (17), miscellaneous (8), empyema (4), and sinusitis (3). Significance for analyses was set at p < 0.05.

Results: Univariate analysis was performed to identify risk factors for the development of infections. The five most significant risk factors, using all infections as an outcome, were as follows: penetrating abdominal trauma index (PATI) greater than 30, presence of an ostomy, multiple transfusions, Injury Severity Score (ISS) of 16 or greater, and Revised Trauma Score less than 7.8. All were highly significant (p < 0.0001). Multivariate analysis with all infections as an outcome revealed that four of the five risk factors had independent effects, with the following significance: PATI greater than 30, ISS of 16 or greater, ostomy, and multiple transfusions. Multivariate analysis for traumatic infections revealed only two of the above to be independent risk factors: presence of an ostomy (p = 0.004) and a PATI greater than 30 (p = 0.039), both of which can be considered local factors. Conversely, multivariate analysis of nosocomial infections revealed independent risk for the two other factors, both of which can be considered systemic factors: multiple transfusions (p = 0.011) and ISS of 16 or greater (p = 0.026).

Conclusion: Although most of the above factors are beyond the control of the trauma surgeon, the creation of an ostomy is a clinical decision. The creation of an ostomy in high-risk patients does not protect them from septic complications and, indeed, may independently contribute to local abdominal infections.