Acinetobacter baumannii Transmission: Role of Vancomycin Use as an Intermediate Risk Factor

Background: During December 2008-January 2009 (epidemic period), surgical intensive care unit (SICU) personnel observed increased numbers of patients with Acinetobacter baumannii bloodstream infections (ABSI).  Despite reported adherence to recommended infection control practices and procedures, the number of ABSI continued to increase.  SICU personnel thought the underlying risk factor for ABSI was bronchoscopy.  Thus, we initiated an epidemiologic investigation after confirming that ABSI rates during the epidemic period was significantly higher than ABSI rates during the pre-epidemic period (January-November 2008).

Objective: (i) to identify risk factors for acquisition of ABSI; and (ii) control transmission of A. baumannii and implement appropriate preventive measures.

Methods: A retrospective cohort study and observational studies of SICU infection control practices and procedures were conducted.  A case was defined as any adult SICU patient who acquired ABSI during the epidemic period. A non-case was defined as any SICU patient admitted to the SICU during the study period but did not acquire an ABSI.  Cases were ascertained through review of medical records and microbiology line listings. Data were recorded in a standardized questionnaire and included demography, clinical, epidemiologic, and microbiological information, device usage, and outcomes. Relative risks (RR) and 95% confidence interval (CI) were determined. Independent risk factors were identified through logistic regression analysis.

Results: Of 32 patients admitted to the SICU during the epidemic period, 9 (28%) met the case definition; 24 (75%) were male. Case characteristics were as follows: median age: 47 (range: 25-67) years; sex: 8 (75%) male. Case- and non-case-patients were similar for age, sex, surgeon, co-morbidities, device usage, exposure to individual healthcare professionals, duration of device use, or survival. On univariate analysis, cases were significantly more likely than non-case-patients to have received a prior antimicrobial (RR: 2.7, CI: 1.0-7.4); linezolid (RR: 5.4, CI: 2.1-14.3); vancomycin (2/2 vs. 6/30, p=0.06); or to have had a previous positive culture for A. baumannii (2/2 vs. 7/30, p=0.07).  On multivariate analysis, the single independent risk factor for acquiring an ABSI was vancomycin use (adjusted odds ratio: 6.6, CI: 2.3-32).

Conclusions: Vancomycin use was the single major risk factor associated with acquisition of ABSI among SICU patients. Bronchoscopy was not a risk factor as was thought by SICU personnel.  These data are consistent with published data that suggest vancomycin eliminates enteric microorganisms that normally inhibit A. baumannii growth through production of a specific peptide factor.  Vancomycin likely facilitated overgrowth and deep invasion among SICU patients already colonized with A. baumannii and other gram-positive and gram-negative pathogens.