Multicenter comparison of Etest, Vitek2 and BD Phoenix to broth microdilution for beta-lactam susceptibility testing of Streptococcus pneumonia

Abstract

Purpose To assess performance of Etest (R), Vitek (R) 2 and BD Phoenix (TM) to determine the susceptibility of Streptococcus pneumoniae strains to penicillin, ampicillin and cefotaxime. Methods Sixty unique S. pneumoniae challenge strains were selected to cover a wide range of penicillin, ampicillin and cefotaxime minimal inhibitory concentrations (MICs). Strains were analyzed in four different Belgian laboratories. Etest (R) benzylpenicillin (BEN), ampicillin/amoxicillin (AMP) and cefotaxime (CTA) (bioM & eacute;rieux), Vitek (R) 2 AST-ST03 (bioM & eacute;rieux) and BD Phoenix (TM) SMIC/ID-11 testing were each performed in two different labs. Results were compared to Sensititre (R) broth microdilution (BMD) (Thermo Fisher Scientific) results. MIC results were interpreted using EUCAST non-meningitis breakpoints (v 13.0). Results Essential agreement (EA) was >= 90% for all methods compared to BMD, except for Etest (R) BEN on Oxoid plate (58.3%), Etest (R) AMP (both on Oxoid (65.8%) and BD BBL plate (84.2%)). Categorical agreement (CA) for penicillin was only >= 90% for Vitek (R) 2, for other methods CA ranged between 74 and 84%. CA for AMP was for all methods < 90% (range 75.8-88.3%) and CA for CTA was between 87 and 90% for all methods except for Etest on Oxoid plate (79.2%). Conclusions Our study indicates that Vitek (R) 2 and BD Phoenix (TM) are reliable for providing accurate pneumococcal susceptibility results for BEN, AMP and CTA. Using Etest BEN or AMP on Oxoid plate carries a risk of underestimating the MIC and should be interpreted with caution, especially when the obtained MIC is 1 or 2 doubling dilutions below the S or R clinical breakpoint.