PURPOSE: Acinetobacter bumannii (AB) is a highly-resistant pathogen & is an important cause of nosocomial infection. We noted an outbreak of AB bacteremia at our institution following the onset of Operation Iraqi Freedom (OIF). We sought to describe the epidemiology of AB bacteremia and its differential impact on trauma and non-trauma patients.
METHODS: We retrospectively identified all patients with AB bacteremia seen at our facility from Jan 2001 thru Aug 2004. Mortality represented our primary endpoint. We compared outcomes in injured service members to those seen in non-trauma patients and explored the effect of initially inappropriate antibiotic treatment on survival (defined as prescription of an anti-infective to which the pathogen was resistant based on sensitivity testing).
RESULTS: During the study period, there were 47 eases of AB bacteremia. From 2001 to 2004 the prevalence of AB bacteremia increased from 0.7 cases to 14.0 cases per 100,000 patient-days (p<0.001). Seventy-five percent of subjects were in the ICU when their blood cultures became positive, and 40% of cases represented nosocomial spread to non-active duty patients. Nearly 20% of isolates were carbepenem resistant. Colistin was given in 2 cases. Non-OIF subjects were older, more often immunosuppressed, and more severely ill. Although the overall case fatality rate was 17%, mortality was 15.8 times (95% CI: 1.8-144.2) more likely in non-OIF patients than in injured service members. The one death in a soldier was not infection-related. This differential in survival was present despite the fact that OIF subjects more often received initially inappropriate antimicrobial therapy (50.0% vs. 22.2%).
CONCLUSION: Nosocomial spread of AB bacteremia poses a major threat to non-trauma patients. Despite the extent of their injuries and severity of illness, AB has less of an impact on critically ill trauma patients.
CLINICAL IMPLICATIONS: Infection control must remain a central aspect of any approach to addressing AB. That inappropriate antimicrobial therapy did not correlate with mortality in previously healthy trauma patients suggests that the importance of this factor on outcomes relates to the specific population studied.
DISCLOSURE: Andrew Short, None.
Andrew F. Shorr MD * Alex G. Truesdell MD Laura A. Pacha MD Dennis M. Sarmiento MD John D. Betteridge MD Allesa J. Ewell PhD Washington Hospital Center, Washington, DC
COPYRIGHT 2005 American College of Chest Physicians
COPYRIGHT 2005 Gale Group