Appraisal of disk sensitivity tests with mecillinam.

Disk susceptibility tests for mecillinam were carried out on 167 hospital isolates of Enterobacteriaceae from Vancouver. Inhibition zone sizes with eight experimental variations were compared by three-way analysis of variance. The Association of Clinical Pathologists' method gave significantly (p = 0.0002) larger zones than the Kirby Bauer method (24.9 +/- SD 7.5 and 23.4 +/- 7.3 mm, respectively). 10 microgram mecillinam disks from Mast Laboratories gave significantly (p = 0.0004) larger zones than the Baltimore Biological Laboratories (BBL) equivalent (24.9 +/- 7.6 and 23.4 +/- 7.2 mm, respectively). Differences between Oxoid Iso-sensitest agar and BBL Mueller Hinton agar were not quite significant (24.5 +/- 7.9 and 23.8 +/- 6.9 mm, respectively, p = 0.061). Each disk test method was significantly different from at least one other (Duncan multiple range test, alpha = 0.05). A semilogarithmic regression was established between the antibiotic disk inhibition zone and the MIC for each experimental variation. BBL Mueller Hinton and Oxoid Iso-sensitest media did not give significantly different results in MIC determinants by a surface inoculation method. An additional 146 organisms from hospitals in New York provided a wider range of organisms from a different location. Pooled data from both centres were used to identify breakpoints for mecillinam in the disk susceptibility test using the Kirby Bauer method with BBL disks and media. A 16-mm breakpoint with 10 microgram mecillinam disks provides a realistic and conservative prediction of probable response to attainable serum concentrations of this antibiotic.