Clinical microbiology and infection vol:vol. 12 issue:S4 pages:1-1
European Congress of Clinical Microbiology and Infectious Diseases edition:16th location:Nice date:April 1-4th 2006
Objective: We evaluated identification of Extended-Spectrum Beta-Lactamases (ESBLs) with the VITEK-2 system (bioMe´rieux, Marcy l’Etoile, France) by comparing it to ESBL detection using Etest (AB-Biodisk, Solna, Sweden) in clinical isolates of Intensive Care Unit (ICU). Furthermore we evaluated the possible impact of these data on the antibiotic policy in our hospital.
Methods: We have collected prospectively 100 consecutive Enterobacteriaceae strains (no duplicates) from patients hospitalised in ICU for more than four days. The following samples were included: blood, sputum, broncho-alveolar lavage fluid or tracheal aspirates and pus. Identification as well as susceptibility-profile were determined with the VITEK-2 system. Escherichia coli and Klebsiella pneumoniae growing on screening plates containing ceftazidime 0.5 lg/ mL or cefotaxime 0.5 lg/mL were further tested with following Etest ESBLs strips: ceftazidime ± clavulanate (TZ/TZL) and cefotaxime ± clavulanate (CT/CTL). Klebsiella oxytoca and all other AmpC producers were tested using Etest ceftazidime. Only strains with MIC’s for TZ > = 4 lg/ mL were further tested: Etest ESBLs strips (TZ/TZL) for K. oxytoca and cefepime ± clavulanate (PM/PML) for AmpC producers.
Results: The VITEK-2 expert system identified 16 possible ESBLs. Out of those 16 strains, the expert system reported 10 strains as ‘ESBL or high level cephalosporinase’. For three strains VITEK-2 suggested a number of resistance mechanisms (‘high level natural penicillinase’, ‘impermeability’ and ‘acquired penicillinase’) and three strains were determined as ‘ESBL’ by
VITEK-2. Only three strains were confirmed by Etest as ESBL: two of the isolates were identified by VITEK-2 as ‘ESBL’ and one as ‘ESBL and impermeability’. Thus one of the three strains determined by VITEK-2 as ‘ESBL’ wasn’t confirmed by our Etest.
Conclusion: VITEK-2 overestimates the number of ESBLs among Enterobacteriaceae. The VITEK-2 report of ‘high level cephalosporinase or ESBL’ was systematically false-positive for ESBL. Automatic validation of VITEK-2 results may lead to unnecessary use of carbapenems. Rechecking of possible ESBL positive identifications on VITEK-2 could lead to alternative therapeutic options for these strains and have a substantial impact on antibiotic policy.