A total of 4 strains of organisms,
i. e., 2 strains of
Staphylococcus, 1 of
E. coli and 1 of
Enterococcus were determined for their susceptibilities to a series of 6 antibiotics at several research laboratories by single-disc method, three-disc method and 2-fold dilution method and the results then tabulated by Dr. KUWAHARA and Dr. FUJil at 13 th East Japan Congress of Chemotherapy. Statistical analysis of the data at the fiducial level of 95% were performed with the results as follows:
1) With the single-disc method, values for MIC on which the criteria for quantitative response (
i. e., -, +
etc. ) can be based are given, whereas, with three-disc method, no MICs are shown. In add ition, the interval between the upper and lower limit of criteria for quantitative response is substantially shorter with three-disc method than with single-disc method. Inasmuch as the two methods employ criteria for quantitative response not in common, it is of no import to attempt to institute a comparison of reproducibility of the data between the two.
2) No significant differences were found between th e mean values for MICs obtained by single-disc method and 2-fold dilution method, in 23 cases of all the 24 different combinations of bacteria with drugs studied. Moreover, in 22 cases of all the 24 combinations, mean values of MICs obtained by the single-disc method were found implicit in the fiducial limit of means of MICs by the 2-fold dilution. method, thereby stressing consistency between the two assay techniques. In one case, with significant difference, however, the discrepancy was assumed to be caused by the experimental errors in the dilution method from the results of the statistical analysis of the more numerous data further added.
3) Within the range of standard variances (theoretically 68% and the 95% confidencial range for assays by the 2-fold dilution method, 68. 5% and 95. 5% of a total of 191 parameters as obtained by the single-disc method were found existent respectively. It follows that the MIC values obtained by the single-disc method are conjectured to share their distribution practically in fairly common with those obtained by the 2-fold dilution method.
4) Analysis of variance revealed th e absence of significant differences between the two assay methods in 23 cases of the total of 24 different bacteria-drug combinations. In one case, variance for assays by the single-disc method was smaller than the one by the dilution method: the single-disc. method in this case had greater reproducibility. The results, therefore, in no way provided evidence in support of incapability of the single-disc method. to estimate MIC values obtained by the 2-fold dilution method. Conversely, the results indicate that advantage is implicit in the ability of the single-disc method with the simple procedure to conjecture fairly accurately the values to be obtained by the 2-fold dilution method which requires rather cumbersome procedure.
View full abstract