Reliability of E-Tests and the Phoenix Automated Method in Assessing Susceptibility to IV Fosfomycin—Comparative Studies Relative to the Reference Method

The agar dilution method (ADM) recommended for IV fosfomycin (IV FOS) is complex and labor-intensive. Keeping in mind the reality of everyday laboratory work, we have evaluated the agreement of IV FOS susceptibility results obtained using the E-test and the Phoenix system with the results obtained using the ADM. Materials and methods: The tests were performed on 860 strains. To evaluate susceptibility to IV FOS, BioMerieux E-tests (bioMerieux, Warsaw, Poland), BD Phoenix panels (BD Phoenix, Sparks, MD, USA), and the ADM were used. Clinical interpretation was performed in accordance with EUCAST Guidance (v12.0, 2021). The significance of the E-test and the Phoenix was analyzed in relation to the ADM by defining categorical agreement (CA), major error (ME), and very major error (VME). Essential agreement (EA) has also been defined for the E-test. A method was considered reliable, in accordance with ISO 20776-2:2007, when CA and EA were above 89.9% and VME was <3%. Results: A categorical agreement of >98.9% was demonstrated between the E-test and the ADM for overall strains and for Echerichia coli, ESBL-producing Enterobacterales, and Staphylococcus aureus, while between the Phoenix and the ADM, a CA of >98.9% was shown only for Escherichia coli, Staphylococcus aureus, and Proteus spp. A very major error rate of <3% was obtained only for Staphylococcus aureus and MBL-producing Pseudomonas evaluated by both the E-test and the Phoenix. An essential agreement of >98.9% between the E-test and the ADM has not been demonstrated for any of the tested groups of strains. The Phoenix yielded more VMEs than the E-test (50 and 46, respectively). The highest VME rate was demonstrated using the Phoenix method for Enterobacter spp. (53.83%). Conclusions: Both the E-test and the Phoenix have turned out to be reliable in assessing IV FOS susceptibility only for Staphylococcus aureus (CA > 89.9% and VME < 3%). For the remaining tested groups of strains and genera, the simultaneous high CA rate and low VME rate required by ISO were not achieved. Both methods fared particularly badly in detecting strains resistant to IV.