Microorganisms responsible for intravascular catheter-related bloodstream infection according to the catheter site

OBJECTIVE: Current guidelines for the management of intravascular catheter-related bloodstream infection (IVC-RBSI) recommend that empirical antimicrobial therapy must have activity against Gram-positive bacteria, but additional empirical coverage for Gram-negative bacteria may be needed for severely ill or immunocompromised patients, and antifungal therapy may be needed in some situations. We hypothesized that the spectrum of etiological microorganisms responsible for IVC-RBSI and, in relation to that, the choice of empirical antimicrobial therapy depends on the catheter insertion site. We therefore compared the proportion of IVC-RBSI due to Gram-negative bacteria and yeasts according to catheter site. DESIGN: Prospective cohort study from May 1, 2000, to April 30, 2004. SETTING: A 24-bed medical-surgical intensive care unit in a 650-bed tertiary hospital. PATIENTS: Patients requiring a central venous or arterial catheter. MEASUREMENTS AND MAIN RESULTS: We diagnosed 88 IVC-RBSIs, comprising 36 femoral catheter sites (26 femoral venous and ten femoral arterial sites) and 52 other catheter sites (36 jugular venous, 11 subclavian venous, and five radial arterial sites). No differences were found between IVC-RBSI of femoral vs. other catheter sites for age, sex, Acute Physiology and Chronic Health Evaluation II, diagnosis at admission, use of antimicrobials, the time the catheter responsible for IVC-RBSI had been in place, or the duration of intensive care unit stay before IVC-RBSI. The proportion of IVC-RBSIs due to Gram-negative bacteria was higher in femoral, 14 of 36 (38.89%), than in the other catheter sites, 4 of 52 (7.69%) (odds ratio, 7.48; 95% confidence interval, 2.19-25.54; p = .001). Also, the proportion of IVC-RBSIs due to yeasts was higher in femoral, 6 of 36 (16.67%), than in the other catheter sites, 1 of 52 (1.92%) (odds ratio, 10.20; 95% confidence interval, 1.17-88.85; p = .035). CONCLUSIONS: Empirical antifungal therapy would seem to be indicated in patients with suspected femoral catheter-related bloodstream infection.