Evaluation of the BD Phoenix M50 Automated Microbiology System for Antimicrobial Susceptibility Testing with Clinical...

The objectives of this study were to evaluate the performance of the BD Phoenix™ M50 system with two antimicrobial susceptibility testing (AST) panels against clinical isolates in South Korea and the accuracy of determining carbapenem and colistin susceptibility compared with reference methods. A total of 825 nonduplicated clinical isolates were included in this study. Bacterial identification was performed using Bruker Biotyper and 16S rDNA sequencing. Antimicrobial susceptibilities were tested by disk diffusion, broth microdilution, and agar dilution methods. AST with the Phoenix system was performed following the manufacturer's instructions. The categorical agreement (CA), very major error (VME), major error (ME), and minor error (mE) rates were calculated for each antibiotic. CA rates between the results of the Phoenix system and reference methods were more than 90% for most antibiotics except for ciprofloxacin in enterococci (82.7%, 163/197) and cefepime in Acinetobacter species (88.9%, 88/99). VME and ME rates were less than 3% for all the antibiotics tested in this study. Minimum inhibitory concentration (MIC) values for carbapenem and colistin determined by the Phoenix system were highly correlated with those of dilution methods, exhibiting 99.2% (384/387), 96.7% (374/387), and 98.5% (129/131) of the agreement rate within onefold dilution difference for imipenem, meropenem, and colistin, respectively. The BD Phoenix M50™ system showed reliable performance for AST in clinical microbiology laboratories and for detecting carbapenem and colistin resistance in Gram-negative clinical isolates.