Comparison between agar dilution and three other methods for determining the susceptibility of 228 clinical isolates of non-fermenting gram-negative rods

Introduction: Susceptibility testing for non-fermenting gram-negative rods (NFGNR) is problematic; valid methods are needed for this purpose. Methods: In this study, 228 NFGNR clinical isolates were evaluated, including 85 Acinetobacter spp., 80 Stenotrophomonas maltophilia, 50 Pseudomonas aeruginosa, and 13 other species (8 Ralstonia pickettii and 5 Burkholderia cepacia). Agar dilution was used as the reference method, and results were compared with those obtained by disk diffusion, Etest. and microdilution performed with the VITEK 2 Compact System. Results: The disk method was unacceptable for S. maltophilia and P. aeruginosa, in the latter organism, mainly because of poor agreement in the colistin results. Nonetheless, the disk method is valid for Acinetobacter spp. and the remaining NFGNRs. The VITEK 2 Compact System yielded poor results for piperacillin-tazobactam, tigecycline, and ceftazidime in S. maltophilia. There were minor discrepancies with the VITEK 2 Compact system and the Etest for tigecycline in Acineobacter spp. and S. maltophilia, likely because of the differing composition of the Muller-Hinton agar lots. Hence, Etest results should be interpreted with caution. Conclusion: The disk diffusion method is inadequated for S. maltophilia. This method is unacceptable for testing colistin in P. aeruginosa. The methods Vitek 2 Compact System and Etest show minor discrepancies for testing S. maltophilia and Acinetobacter spp. for tigecycline. (C) 2008 Elsevier Espana, S.L. All rights reserved.