Inter- and intralaboratory variability in antibiotic susceptibility tests with Pseudomonas aeruginosa and Enterobacteriaceae.

A two-phase collaboratice study was designed for definition of the extent to which results of disk diffusion tests can be reproduced. Two currently recommended techniques, the Kirby-Bauer method and the agar overlay method, were used. Of special concern were tests of Pseudomonas aeruginosa with moderate susceptibility to carbenicillin; gentamicin, tobramycin, polymyxin B, and colistin were also tested with P. aeruginosa as well as with Klebsiella pneumoniae, Escherichia coli, and Proteus species. The agar overlay method tended to give zones that were somewhat smaller than those given by the Kirby-Bauer method. Both disk methods demonstrated an excellent degree of reproducibility, being at least as reproducible as the standard agar dilution technique for measurement of minimal inhibitory concentrations. Slight changes in carbenicillin disk potency affected the zones of inhibition around P. aeruginosa more than those around susceptible E. coli. This fact suggests that a standard strain of P. aeruginosa would be advantageous for quality control of carbenicillin disk tests.