Methods: We performed a retrospective cohort study of patients with VRE colonization/infection who met criteria for decolonization between 2007-2015 at the NIH Clinical Center. Antibiotic exposure was measured as the percentage of antibiotic days per number of VRE decolonized days and analyzed as a binary variable greater than or less than the median. The relationship of antibiotic exposure and time to VRE re-colonization was assessed via survival analysis using Cox proportional hazards regression.
Results: 350 inpatients were identified as VRE colonized/infected; 72 (21%) met decolonization criteria. 21 (29%) of these 72 patients subsequently became re-colonized with VRE, whereas 51 (71%) remained decolonized. The hazard of VRE re-colonization was 6.7 (95% CI 2.7, 16.3) and 4.6 (95% CI 2.1, 9.9) times higher in patients with a percentage of antibiotic days and anti-anaerobic antibiotic days above the median, respectively, after adjustment for confounding by age, race and gender.
Conclusion: The proportion of antibiotic and anti-anaerobic antibiotic days to the number of VRE decolonized days was significantly related to VRE re-colonization. This finding warrants more careful antimicrobial stewardship and adjustment of targeted surveillance for reappearance of VRE colonization.
H. Y. Hughes,
A. V. Michelin, None
E. S. Snitkin, None
N. Sinaii, None
A. Milstone, Sage Products LLC: Grant Investigator , Grant recipient
D. K. Henderson, None
T. N. Palmore, None