Using VRE screening tests to predict vancomycin resistance in enterococcal bacteremia

BACKGROUND AND OBJECTIVE: Enterococcus causes clinically significant bloodstream infections (BSIs). In centers with a higher prevalence of vancomycin resistant enterococcus (VRE) colonization, a common clinical question is whether empiric treatment directed against VRE should be initiated in the setting of a suspected enterococcal BSI. Unfortunately, VRE treatment options are limited, and relatively expensive, and subject patients to the risk of adverse reactions. We hypothesized that the results of VRE colonization screening could predict vancomycin resistance in enterococcal BSI. METHODS: We reviewed 370 consecutive cases of enterococcal BSI over a 7-year period at 2 tertiary-care hospitals to determine whether vancomycin-resistant BSIs could be predicted based on known colonization status (ie, patients with swabs performed within 30 days, more remotely, or never tested). We calculated sensitivity and specificity, and we plotted negative predictives values (NPVs) and positive predictive values (PPVs) as a function of prevalence. RESULTS: A negative screening swab within 30 days of infection yielded NPVs of 90% and 95% in settings where <27.0% and 15.0% of enterococcal BSI are resistant to vancomycin, respectively. In patients with known VRE colonization, the PPV for VRE in enterococcal BSI was >50% at any prevalence exceeding 25%. CONCLUSIONS: The results of a negative VRE screening test result performed within 30 days can help eliminate unnecessary empiric therapy in patients with suspected enterococcal BSI. Conversely, patients with positive VRE screening swabs require careful consideration of empiric VRE-directed therapy when enterococcal BSI appears likely.