Antimicrobial susceptibility testing of enterococci: results of a survey conducted by the United Kingdom National External Quality Assessment Scheme for Microbiology.

Six strains of enterococci were distributed to participants in the United Kingdom National External Quality Assessment Scheme for Microbiology with a request that they be tested for susceptibility to ampicillin/penicillin and vancomycin and for high-level resistance to gentamicin. More than 98% of laboratories correctly reported the resistant strains as resistant to penicillin and ampicillin. However, there was a poorer correlation for the susceptible strains, with only 1-6% and 69-83% of laboratories reporting the strains as susceptible to penicillin and ampicillin respectively. Those laboratories which used 5-25 micrograms ampicillin discs or breakpoint methods reported proportionally more results correctly for the susceptible strains than those using 2 micrograms discs. Participants using Escherichia coli or an enterococcus as a control and those not using a control at all were more likely to report ampicillin-susceptible strains correctly than those using Staphylococcus aureus NCTC 6571 (Oxford strain). For vancomycin-susceptible and high-level vancomycin-resistant strains, 93-96% and 96-99% of reports respectively were correct. For the low-level vancomycin-resistant strains, only 50-54% of reports were correct. Participants using a 5 micrograms vancomycin disc reported proportionally more of the low-level resistant strains correctly than those using 10, 20, 25, 30 or 100 micrograms discs. Participants not using controls were proportionally more likely to report the low-level resistant strains as susceptible than those using an enterococcus or S. aureus NCTC 6571 as a control. For the high-level gentamicin-resistant (HLGR) strains, 96-98% of reports were correct. Depending on the strain, between 12% and 41% of participants also reported high-level resistance for the strains which were not HLGR. Those using low-content discs (< or = 30 micrograms) and breakpoint or MIC methods were more likely to report strains which were not HLGR as HLGR than those using high-content discs. Those using an enterococcus as a control were less likely to report strains which were not HLGR as HLGR than those using either S. aureus NCTC 6571 or E. coli as a control and those not using a control.