Methods: All patients from 5 Intermountain Healthcare hospitals with ≥ 1 blood culture positive for S.aureus were included. Data were obtained through manual chart review as well as Enterprise Data Warehouse (EDW) extraction. MIC was evaluated by routine automated methods (Phoenix) as well as E-test performed on stored bacterial isolates per standard protocol. Failure was defined as: all cause 60-day mortality, growth of S. aureus in blood cultures ≥7 days after index culture while on therapy, or recurrence of bacteremia within 30 days of completion of therapy. Deep seated infection was defined as: endocarditis, epidural abscess, pneumonia, psoas abscess, osteomyelitis, septic arthritis, or cardiac device infection. Data were analyzed using multivariate logistic regression with interaction variables.
Results: 328 cases of SAB were identified from 1/2010 to 7/2012, 31% were due to methicillin resistant S. aureus (MRSA). 16% of cases met failure criteria and 53% were complicated by deep seated infection. The following variables were independently associated with failure: age (aOR = 1.03, p<0.003); MRSA (aOR = 2.28, p=0.02); presence of deep seated infection (aOR = 2.97, p=0.04); and number of days with fever (aOR = 1.15, p=0.003). Elevated vancomycin MIC was not predictive of failure (aOR Etest =1.32, p=0.62, aOR Phoenix = 1.01, p=0.99). No evidence of interaction between elevated MIC and deep seated infection was observed with Etest or Phoenix.
Conclusion: We found no evidence of interaction between elevated vancomycin MIC and deep seated infection in relation to failure in SAB. Age, MRSA, presence of deep seated infection, and number of days with fever were independent predictors of failure.
D. Handrahan, None
J. Coombs, None
B. K. Lopansri, None
M. Gazdik, None
E. Stenehjem, None