Assessment of the Phoenix™ automated system and EUCAST breakpoints for antimicrobial susceptibility testing against isolates expressing clinically relevant resistance mechanisms

EUCAST breakpoint criteria are being adopted by automatic antimicrobial susceptibility testing systems. The accuracy of the Phoenix Automated System in combination with 2012 EUCAST breakpoints against recent clinical isolates was evaluated. A total of 697 isolates (349 Enterobacteriaceae, 113 Pseudomonas spp., 25 Acinetobacter baumannii, 11 Stenotrophomonas maltophilia, 95 Staphylococcus aureus, 6 coagulase negative staphylococci, 77 enterococci and 21 Streptococcus pneumoniae) with defined resistance phenotypes and well-characterized resistance mechanisms recovered in Spain (n = 343) and Italy (n = 354) were tested. Comparator antimicrobial susceptibility testing data were obtained following CLSI guidelines. Experimental agreement (EA), defined as MIC agreement ±1 log2 dilution, category agreement (CA) and relative discrepancies (minor (mD), major (MD) and very major discrepancies (VMD)) were determined. The overall EA and CA for all organism-antimicrobial agent combinations (n = 6.294) were 97.3% and 95.2%, respectively. mD, MD and VMD were 4.7%, 1.3% and 2.7%, all of them in agreement with the ISO (ISO20776–2:2007) acceptance criteria for assessment of susceptibility testing devices. VMD were mainly observed in amoxicillin-clavulanate and cefuroxime in Enterobacteriaceae and gentamicin in Pseudomonas aeruginosa, whereas MD were mainly observed in amoxicillin-clavulante in Enterobacteriaceae. mD were mainly observed in Enterobacteriaceae but distributed in different antimicrobials. For S. aureus and enterococci relative discrepancies were low. The Phoenix system showed accuracy assessment in accordance with the ISO standards when using EUCAST breakpoints. Inclusion of EUCAST criteria in automatic antimicrobial susceptibility testing systems will facilitate the implementation of EUCAST breakpoints in clinical microbiology laboratories.