Frequency of extended spectrum beta lactamase producing Escherichia coli strains isolated from central venous catheter related infections quantified by three laboratory methods

Abstract number: 1134_01_393

Balotescu M., Lazar V., Larion C., Banu O., Bulai D., Cernat R.

Objectives:  

The purpose of this study was the ESBL screening of clinical isolates of Escherichia coli isolated from patients with central venous catheter related infections, by automatic methods (VITEK system), double-disk synergy test and E-test^® ESBL.

Methods:  

The study was carried out on 51 E. coli strains recovered in 2004 from patients admitted for cardiovascular surgery at the Cardiovascular Disease Institute of Bucharest, Romania. The strains identification was performed using the VITEK system. The strains were subsequently tested for antibiotic susceptibility by standard disk diffusion method and the strains suspected of ESBL production were comparatively tested by double disk diffusion confirmatory test using clavulanic acid and the standard E-test^® strips, containing of two gradients aligned in opposing directions, one side of the strip cefotaxime (CT) MIC range of 0.5–32 mcg/ml and the other a cefotaxime MIC range of 0.125–8 mcg/ml overlaid with a constant level of 4 mcg/ml clavulanic acid (CTL). MIC ratios of CT/CTL (MIC ratio >8) have been used as a discriminative criteria for the presence or absence of ESBLs.

Results:  

Out of the total number of strains, 17.64% were given ESBL positive by VITEK system. The results of antimicrobial resistance testing by standard disk diffusion method showed that 35.29% exhibited beta-lactamase pattern, with high resistance levels to ampicillin and low or high resistance to 1st generation cephalosporins and 17.64% of the tested strains were suspected of ESBL production showing decreased sensitivity to cephtazidime and cefotaxime. ESBL screening of strains by double disk diffusion test showed that 5 strains (9.8%) demonstrated clavulanic acid inhibitory effect in the phenotypic confirmatory test. The E-test results confirmed the presence of ESBL only in 4 strains of the suspected of ESBL production (7.84%), the rest being not determinable, probably due to MICs values exceeding the concentration range, or to the overlapping of other beta-lactamases (e.g. AmpCs and inhibitor-resistant TEMs (IRTs) masking the ESBL pattern.

Conclusion:  

Our results showed that different methods used for the detection of ESBL producing E. coli strains gave variable results and showed different sensitivity levels in detecting ESBL strains. These data suggest that clinical microbiology laboratories should not rely on one single method for the screening ESBL producers, complementary tests being necessary for incidence studies.