Proficiency of Italian clinical laboratories in detecting reduced glycopeptide susceptibility in Enterococcus and Staphylococcus spp. using routine laboratory methodologies.

OBJECTIVE: To assess the ability of 59 clinical microbiology laboratories distributed throughout Italy to correctly identify and detect reduced susceptibility to glycopeptides in staphylococci and VanA-, VanB- or VanC-mediated glycopeptide resistance in enterococci. METHODS: Eight test strains comprising three staphylococci (S. aureus ATCC 29212 and two vancomycin-intermediate S. haemolyticus [11105301, 10030683Y]) and five enterococci (E. faecalis ATCC 29212, E. faecalis ATCC 51299 VanB, E. faecium AIB40 VanA, E. faecalis V583 VanB and E. gallinarum AIB39 VanC1) were distributed to 59 Italian clinical microbiology laboratories. Each isolate was blind-coded, and laboratories were instructed to identify the strains and test isolates for susceptibility to teicoplanin and vancomycin using their standard methods. Results were assessed against consensus test results obtained by a reference laboratory. In addition, to complement data interpretation, laboratories were asked to provide retrospective routine test results from their respective hospitals. RESULTS: All 59 laboratories participating in the study completed the susceptibility testing and provided data for analysis. A total of 53 laboratories provided retrospective routine data. Overall, laboratories were able to identify isolates to the genus level successfully. E. gallinarum and S. haemolyticus posed problems for species identification, with only 40.6 and 71.2%, respectively, of results reported correctly; most incorrect results were reported as 'other species'. For enterococcal test strains, VanA phenotypes were detected correctly by 96.6% of laboratories; VanB by 30.5% (E. faecalis ATCC 51299) and 88.1% (E. faecalis V583); and VanC1 by 67.8%. For staphylococcal test strains, 28.8% (S. haemolyticus 11105301) and 23.7% (S. haemolyticus 10030683Y) of the laboratories were able to detect reduced susceptibility to vancomycin. Errors in detecting vancomycin resistance in VanB and VanC1 enterococci were made with all methods, most noticeably by disk diffusion users. For staphylococci, most errors in reporting vancomycin-intermediate resistance occurred with disk diffusion and Vitek (software version 5.04) users. Overall, considerably fewer errors occurred with the detection of teicoplanin resistance, especially for staphylococci. For 1999, routine results show that 41/1749 (2.4%) of E. faecium, 220/11 180 (2.0%) of E. faecalis, 29/24 927 (0.12%) of S. aureus and 54/22 102 (0.24%) of coagulase-negative staphylococci were reported as resistant to vancomycin. CONCLUSION: Italian laboratories are able to identify staphylococci and enterococci adequately, although all methodologies used have problems in identifying E. gallinarum and coagulase-negative staphylococci to the species level. While VanA phenotypes were efficiently detected, problems were experienced in detecting VanB and VanC phenotypes. The majority of laboratories were unable to detect reduced vancomycin susceptibility in staphylococci adequately, especially with disk diffusion and older Vitek systems. Teicoplanin appeared useful as a marker for detecting vancomycin resistance, particularly with disk diffusion. Should enterococcal VanB or staphylococcal glycopeptide-intermediate phenotypes become prevalent in Italy, it is likely that they would be under-detected. New systems under development, such as Vitek2, should improve this situation.