Background: Many studies have reported an association between antibiotic exposure and MRSA or VRE acquisition in hospitals. However, these studies have not accounted for confounding by indication, whereby empiric antibiotics are given for symptoms ultimately due to an MRSA or VRE infection.
Methods: We previously reported several predictors of MRSA and VRE acquisition in a case control study of intensive care unit (ICU) patients admitted to a tertiary care center between Sep 2003 and Apr 2005 who had an initial negative nares or rectal screening culture followed by either a subsequent negative screening culture (controls) or positive screening or clinical culture (cases).1 Cases and controls were selected at a 1:1 ratio. Within and prior to this eligible interval for acquisition, detailed data had been collected, including demographics, comorbidities, and daily device and antibiotic utilization. We now conduct a secondary analysis in which antibiotics given for symptoms ultimately attributed to MRSA or VRE infection are removed. Remaining antibiotic exposures are classified into the following periods: during the eligible interval, 2 weeks prior to the interval, or 2 months prior to the interval. Generalized linear mixed models are used to assess variables associated with MRSA or VRE acquisition, accounting for clustering by ICU.
Results: Among 244 cases and 248 controls for MRSA acquisition, we find that exclusion of empiric antibiotics used for MRSA infection affects 113 cases resulting in different antibiotic predictors of acquisition (Table). In contrast, among 227 cases and 248 controls for VRE, similar exclusion affects 5 cases and has no impact on predictors. Antibiotic exposure in the prior 2 weeks appears most associated with MRSA acquisition, whereas exposure in the prior 2 months appears most associated with VRE acquisition.
Conclusion: Failure to account for treatment indication when assessing antibiotic exposures may cause common empiric antibiotics to falsely appear predictive of MRSA acquisition. This effect is not seen for VRE, likely because VRE infection (and thus empiric therapy) is less common. Timing of antibiotic exposure may also be important.
1Huang et al. Critical Care 2011, 15:R210
S. Rifas-Shiman, None
H. Placzek, None
J. Lankiewicz, None
R. Platt, None
S. S. Huang, None