Quality assessment of glycopeptide susceptibility tests: a European collaborative study. European Glycopeptide Resistance Group.

The ability of seventy clinical laboratories in nine European countries to detect glycopeptide resistance in Gram-positive bacteria was investigated. Results of routine tests were compared with those on the same strains by a reference method in national co-ordinating laboratories. In addition, control strains were tested by some of the participants. Errors in reporting susceptibility of Staphylococcus aureus to teicoplanin and vancomycin and coagulase-negative staphylococci to vancomycin were < 1%. With coagulase-negative staphylococci however, 44 (3.4%) teicoplanin susceptible isolates were reported intermediate and six (0.4%) resistant; 18 (58.1%) of 31 teicoplanin intermediate isolates were reported susceptible and five (16.1%) resistant; and six of nine teicoplanin resistant isolates were reported susceptible and two intermediate. All seven isolates of enterococci intermediate to vancomycin were reported susceptible. Distribution of a known vancomycin intermediate strain of E. gallinarum indicated problems with vancomycin susceptibility testing (44.4% reported susceptible, 32.7% intermediate, 32.1% resistant) and identification (only 34.1% correct) of this organism. Two of 28 teicoplanin resistant enterococci and three of 30 vancomycin resistant isolates were reported susceptible. Among other organisms, one resistant Lactobacillus sp. was reported susceptible to teicoplanin and vancomycin. In reporting teicoplanin susceptible organisms, there were fewer errors with comparative/Stokes methods than with most other methods and more errors with the ATB and Sceptor methods than most other methods. None of the methods used were reliable for testing teicoplanin intermediate and resistant coagulase-negative staphylococci or low-level vancomycin resistant enterococci. Alternative methods, such as breakpoint screening, should be considered for detecting glycopeptide resistance.