The sensitivity and specificity of seven methods (agar dilution, broth microdilution, Etest at 0.5 and 2.0 McFarland (McF) inocula, two agar screening methods, and population studies [PS]) were evaluated in a double-blind study involving 284 methicillin-resistant Staphylococcus aureus (MRSA) strains and 45 Staphylococcus strains with reduced susceptibilities to vancomycin (SRSV). The results were compared to the population analysis profile-area under the curve ratio method (PAP-AUC ratio compared to that of Mu3) as described by Wootton et al. The agar screening method using brain heart infusion agar (6 g of vancomycin per ml) gave a sensitivity of 22% and a specificity of 97%. A similar method using Mueller-Hinton agar (5 g of vancomycin per ml) gave a sensitivity of 20% and a specificity of 99%. The PS method detected 34 false positives (12%) and gave a sensitivity of 71% and a specificity of 88%. Etest using 0.5 and 2.0 McF inocula gave sensitivities and specificities of 82 and 93% and of 96 and 97%, respectively. The best Etest interpretative criteria for the 2.0 McF inoculum was >8 mg of vancomycin per liter and >8 g teicoplanin per ml or >12 g of teicoplanin per ml. The direct colony suspension inoculum for this method was found to be equally accurate in detecting (hetero-) glycopeptide-intermediate S. aureus compared to the overnight broth inoculum preparation method. Agar dilution and broth microdilution using the NCCLS breakpoint criteria for vancomycin gave sensitivities and specificities of 20 and 100% and of 11 and 100%, respectively. Using the Etest with a 2.0 McF inoculum, six different media were assessed against a selection of SRSV (n ‫؍‬ 48) and MRSA (n ‫؍‬ 12). Brain heart infusion agar yielded the highest sensitivity and specificity values: 88 and 88%, respectively.

No abstract

The Etest MBL (AB BIODISK, Solna, Sweden) correctly differentiated all 57 isolates of Acinetobacter spp. and Pseudomonas aeruginosa with the bla IMP-1 allele and 135 of 137 (98.5%) Acinetobacter spp. and Pseudomonas spp. isolates with the bla VIM-2 allele. The Etest MBL was reliable for detecting the IMP-1-and VIM-2-producing Pseudomonas and Acinetobacter isolates.The IMP-and VIM-type and other acquired class B metallo-␤-lactamase (MBL)-producing gram-negative bacilli have been increasingly isolated from clinical specimens (10, 15). Among the imipenem-nonsusceptible isolates, 62.1% of the 58 Pseudomonas aeruginosa isolates had a VIM-type MBL in a Greek study (3), while 11.4% of 387 Pseudomonas spp. isolates and 14.2% of 267 Acinetobacter spp. isolates had either an IMP-or VIM-type MBL in a Korean study (7).MBL can hydrolyze ␤-lactams from all classes except the monobactams (1). Higher mortality has been reported in patients infected with the IMP-1-producing strains (5). Therefore, the reliable detection of the MBL-producing strains is essential for the optimal treatment of infected patients and to control the nosocomial spread of resistance. However, the current NCCLS document (12) does not contain any method for detecting MBL.MBL-producing strains can be detected with double-disk synergy tests (8, 13) and a microdilution method (11). While the former method is only qualitative, the latter is not simple for routine testing. Walsh et al. (16) reported that the Etest MBL strip (AB BIODISK, Solna, Sweden) had 100% sensitivity and specificity, but their findings were based on testing only 12 IMP-1-type strains and did not include the prevalent VIM-2-type strains. The performance of Etest MBL was influenced by the choice of the Mueller-Hinton agar brand (16) but may have also been influenced by the level of imipenem resistance of the MBL-producing strains (17, 18) and possibly by other testing conditions.In this study, the performance of Etest MBL was evaluated with 199 bla VIM-2 or bla IMP-1 allele-positive isolates, and the factors that may influence the routine use of the test were also investigated.( Strains. The strains were isolated from the clinical specimens at Korean hospitals between 1995 and 2003, and imipenem susceptibility was tested with the disk diffusion method (12) using 10-g disks and Mueller-Hinton agar (Becton Dickinson, Cockeysville, Md.). MBL screening was tested with the Hodge test and a double-disk synergy test (6,8). The bla IMP-1 and bla VIM-2 alleles were detected by PCR (9). The isolates, which were imipenem resistant but MBL negative by the screening tests and PCR, were used as the negative controls. The test strains were stored at Ϫ76°C until used for the study.Evaluation of Etest MBL. The evaluations were performed at two laboratories, one in Korea and the other in Sweden. The Swedish laboratory used the strains that were transported in semisolid cystine trypticase agar (Becton Dickinson) vials at ambient temperature. Several colonies from a 24-h culture plate were used to prepare the ...

Compared to the North American component of this study, there was substantially less vancomycin resistance among E. faecium isolates (Europe 3.0% vs. North America 63.4%). While the occurrence of penicillin-resistant S. pneumoniae in Europe and North America was similar (25.1% vs. 29.7%), the recovery of high-level penicillin-resistant strains was nearly three-fold higher in North America (4.9% vs. 13.2%). Only linezolid was universally active against all the tested Gram-positive isolates at