Direct comparison of the BD Phoenix system with the MicroScan WalkAway system for identification and antimicrobial susceptibility testing of Enterobacteriaceae and nonfermentative gram-negative organisms

The Phoenix automated microbiology system (BD Diagnostics, Sparks, MD) is designed for the rapid identification (ID) and antimicrobial susceptibility testing (AST) of clinically significant human bacterial pathogens. We evaluated the performance of the Phoenix instrument in comparison with that of the MicroScan WalkAway system (Dade Behring, West Sacramento, CA) in the ID and AST of gram-negative clinical strains and challenge isolates of Enterobacteriaceae (n = 150) and nonfermentative gram-negative bacilli (NFGNB; 45 clinical isolates and 8 challenge isolates). ID discrepancies were resolved with the API 20E and API 20NE conventional biochemical ID systems (bioMerieux, Durham, NC). The standard disk diffusion method was used to resolve discordant AST results. The overall percentages of agreement between the Phoenix ID results and the MicroScan results at the genus and species levels for clinical isolates of Enterobacteriaceae were 98.7 and 97.7%, respectively; following resolution with conventional biochemical testing, the accuracy of the Phoenix system was determined to be 100%. For NFGNB, the levels of agreement were 100 and 97.7%, respectively. Both systems incorrectly identified the majority of the uncommon nonfermentative nonpseudomonal challenge isolates recovered from cystic fibrosis patients; these isolates are not included in the databases of the respective systems. For AST of Enterobacteriaceae, the rate of complete agreement between the Phoenix results and the MicroScan results was 97%; the rates of very major, major, and minor errors were 0.3, 0.2, and 2.7%, respectively. For NFGNB, the rate of complete agreement between the Phoenix results and the MicroScan results was 89.1%; the rates of very major, major, and minor errors were 0, 0.5, and 7.7%, respectively. Following the confirmatory testing of nine clinical isolates initially screened by the MicroScan system as possible extended-spectrum-beta-lactamase (ESBL)-producing organisms (seven Klebsiella pneumoniae isolates and two Escherichia coli isolates), complete agreement was achieved for eight isolates (one ESBL positive and seven negative); one false positive was obtained with the Phoenix instrument. The MicroScan system correctly detected the 10 ESBL challenge isolates, versus the 6 detected by the Phoenix system. Overall, there was good correlation between the Phoenix instrument and the MicroScan system for the ID and AST of Enterobacteriaceae and common NFGNB. The Phoenix system is a reliable method for the ID and AST of the majority of clinical strains encountered in the clinical microbiology laboratory. Until additional performance data are available, results for all Klebsiella pneumoniae or Klebsiella oxytoca and E. coli isolates screened and confirmed as ESBL producers by any automated system should be confirmed by alternate methods prior to the release of final results.