Evaluation of Vitek system for susceptibility testing of Enterococcus faecium isolates.

Although imipenem is not a first-line drug for treating enterococcal infection, it could well become a useful drug for treating mixed infections that include enterococci. However, there is no NCCLS guideline for susceptibility testing of imipenem versus enterococci. Moreover, there are no statements to indicate that in vitro susceptibility results for other antimicrobial agents can be used to predict the in vitro activity of imipenem against enterococci. In this study, 52 Enterococcus faecium isolates were collected from patients hospitalized at Kangnam St. Mary's Hospital between March 2002 and December 2002. The sources of the isolates were mainly urine specimens and wounds. For ampicillin, the "major" and "very major" error rates observed with the Vitek system were 0% and 2.0%, respectively. For penicillin, the major and very major error rates observed with the Vitek system were both 0%. For imipenem, the major and very major error rates observed with the Vitek system were 0% and 36.5%, respectively. The MICs of ampicillin and penicillin obtained using the Vitek system were reliable, but that of imipenem was unreliable. In the 52 E. faecium isolates, the in vitro activity of penicillin and ampicillin versus enterococci accurately predicted that of imipenem. Therefore, the MIC of imipenem obtained with the Vitek system must be retested by the agar dilution method, when it disagrees with those of penicillin and ampicillin.