Abstract
BACKGROUND: Vancomycin-resistant enterococcus (VRE) causes 4% of all healthcare-associated infections in the USA. The process by which the local hospital milieu contributes to VRE acquisition is not fully understood. AIM: To determine the importance of specific factors within the local hospital environment for healthcare-associated VRE acquisition. METHODS: This retrospective cohort study included patients admitted to six intensive care units at an academic medical centre from January 2012 to December 2016 with negative rectal VRE cultures on admission. VRE acquisition was defined as a positive surveillance swab performed at any time after the initial negative swab during the index hospitalization. The exposures of interest were VRE colonization pressure, VRE importation pressure, and use of vancomycin. Multivariable Cox proportional hazards modelling was performed, with patients followed until VRE acquisition, death, or for up to 30 days. FINDINGS: Of 8485 patients who were initially VRE negative, 161 patients acquired VRE. On univariate analysis, patients with VRE acquisition were more likely to have received vancomycin, to have had a neighbouring patient who received vancomycin, to have high VRE importation pressure, or to have high VRE colonization pressure. On multivariable analysis, only high VRE colonization pressure was an independent predictor of VRE acquisition (adjusted hazard ratio: 1.79; 95% confidence interval: 1.19-2.70). CONCLUSION: VRE colonization pressure was the most important risk factor for healthcare-associated VRE acquisition, regardless of VRE importation pressure. Interventions seeking to reduce VRE acquisition should focus on minimizing transmission between patients with known VRE and the local hospital environment.