Abstract
OBJECTIVE: To evaluate whether clinical cultures are an appropriate surrogate for surveillance cultures to measure the effect of interventions on the incidence of MRSA and VRE in the hospital. DESIGN: Cross-sectional and quasi-experimental, retrospective analysis. SETTING AND POPULATION: Convenience sample of patients admitted between January 1, 2002, and June 31, 2011, to the medical intensive care unit (MICU) and surgical intensive care unit (SICU) of an acute-care hospital in the United States. INTERVENTIONS: Asynchronously in the MICU and SICU, we introduced (1) universal glove and gown use, (2) bundled intervention to prevent central-line-associated bloodstream infection, and (3) daily chlorhexidine gluconate bathing. RESULTS: We observed a statistically significant correlation between surveillance and clinical culture-based incidence rates of MRSA in the MICU (0.32; P < .001) and the SICU (0.37; P < .001) but not for VRE in either the MICU (0.16, P = .11) or the SICU (0.15; P = .12). For VRE, but not for MRSA, incidence density rates based on surveillance cultures were 2- to 4-fold higher than for clinical cultures. When evaluating the impacts of the interventions, different effect estimates were noted for universal glove and gown use on MRSA acquisition in MICU, and for VRE acquisition in both the MICU and the SICU based on surveillance versus clinical cultures. CONCLUSIONS: For multidrug-resistant organism acquisition, surveillance cultures should be used when feasible because clinical cultures may not be an appropriate surrogate. Clinical or surveillance-based end points for infection control interventions should reflect the conceptual model from colonization to infection and where an intervention might have an effect, rather than considering them interchangeable.