The precision and robustness of published protocols for disc diffusion assays of antimicrobial agent susceptibility: an inter-laboratory study

Abstract

The precision of the disc diffusion protocols previously published by Alderman and Smith (Aquaculture 196 (2002) 211) was analysed in a seven-laboratory trial using Escherichia coli ATCC 25922 as the test strain. Discs containing seven antimicrobial agents were employed and 2899 zone size measurements were generated. The total data generated in the trial was used to quantify the intra- and inter-laboratory precisions. The study design also facilitated the investigation of the influence of the source of media and the source of discs on zone sizes. A smaller two-laboratory trial was employed to investigate the influence of incubation time of zone size. The intra-laboratory precision was relatively high with the mean of the coefficients of variation calculated for each laboratory and each agent being 4.7%. In contrast, the inter-laboratory precision was very much lower with the mean of the values for each agent being 11.1%. Significant influences on zone size were detected for all three parameters of the protocol. Media source effects were particularly notable with respect to oxytetracycline and oxolinic acid discs, disc source effects with respect to ampicillin and sulphamethoxazole/trimethoprim discs and incubation times with the ampicillin and amoxycillin discs. ANOVA analysis of the total data set confirmed that inter-laboratory variation was the major factor influencing the low precision of the protocol. The overall precision of the protocols used here was found to be significantly lower than that implied by the control limits associated with the same bacterium in other validated disc diffusion protocols. The implications of these results, for the further development of the protocols under investigation, are discussed.