Empiric Therapy and Mortality among Patients with Staphylococcus aureus Bacteremia

Abstract

Background: Staphylococcus aureus bloodstream infections are associated with considerable costs, morbidity and mortality. Appropriate empiric antibiotic therapy, defined as antibiotic therapy with in vitro activity to bacteria given prior to known culture results, may improve patient outcomes. Studies are conflicted as to whether receipt of appropriate empiric therapy is associated with decreased mortality. Objectives: The objectives of this dissertation were to examine whether documentation of prior methicillin-resistant S. aureus (MRSA) is associated with the receipt of appropriate empiric therapy, as well as outcomes associated with appropriate empiric therapy in patients with suspected S. aureus bacteremia while controlling for patient and pathogen characteristics. Methods: These objectives were carried out using two retrospective cohort studies. The first study used a administrative database of patient hospitalization records to assemble a cohort of patients who had a blood culture collected and received empiric antibiotic therapy in order to measure the utility of documentation of prior MRSA colonization or infection in optimizing empiric therapy. The second study was conducted using a central data repository, chart review, and laboratory analysis to assess the independent association between appropriate empiric therapy and mortality among patients with S. aureus bacteremia, after controlling for patient and pathogen characteristics. Results: The first cohort of 25,378 admissions to University of Maryland Medical Center (UMMC) from January 2001 to December 2005 found that documentation of prior MRSA colonization or infection was the strongest independent predictor of both receipt of MRSA-directed empiric therapy (adjusted odds ratio (OR)=2.79; 95% confidence interval (CI): 2.52, 3.09) and subsequent MRSA positive blood culture (adjusted OR: 4.05; 95% CI: 3.02, 5.44). The second study, composed of 814 admissions to UMMC from January 2003 to June 2007, concluded that appropriate empiric therapy for S. aureus bacteremia was not significantly associated with 30-day in-hospital mortality (adjusted hazard ratio (HR):1.17; 95% CI: 0.76, 1.81) or increased post-culture length of stay (HR: 1.04; 95% CI: 0.88, 1.23). Conclusions: Documentation of prior MRSA colonization or infection has value beyond its original infection control purposes. Appropriate empiric therapy for S. aureus bacteremia is not protective against mortality.