Evaluation of the ISO 7932 standard for the enumeration of Bacillus cereus in foods

Twenty, mostly European laboratories took part in a collaborative study to validate the general 1993 ISO 7932 standard for the enumeration of Bacillus cereus in foods (Anonymous, ISO 7932, Microbiology – General Guidance for the Enumeration of Bacillus cereus – Colony-count Technique at 30°C. International Organisation for Standardization, Geneva, Switzerland, 1993). The objective was to determine the precision data in terms of repeatability ( r) and reproducibility ( R) of the method using three different food types at various inoculum levels. The results are intended for publication in Comité Européen de Normalisation (CEN) and ISO standards. The method was challenged with three types of food product: fresh cheese, minced beef and potato powder. Each participant received eight samples for each food type: blind duplicates at four inoculum levels (target values of B. cereus of 0, 10 3, 10 4 and 10 5–10 6 colony forming particles (cfp)/g). In addition, two reference materials (RMs, capsules containing milk powder, artificially contaminated with B. cereus) were included in the study. All test materials were tested extensively for homogeneity and stability prior to the collaborative trial. In addition to determining the precision parameters of the ISO method, polymyxin pyruvate egg yolk bromothymol blue agar (PEMBA, incubation at 37°C) was included in the study to evaluate possible differences in performance compared to mannitol egg yolk polymyxin agar (MEYP, incubation at 30°C) which is prescribed in the ISO standard. In this study no difference in performance between MEYP and PEMBA medium was observed. Results from the glucose fermentation confirmation test indicated that generally 48 h was needed to obtain a yellow colour throughout the whole test tube. A high number of false negative results with the Voges Proskauer (VP) reaction, even after 48 h of incubation of the tubes, was observed in some laboratories. Values for r and R were therefore calculated without VP test results. Further studies have been initiated by the ISO technical committee in order to improve the performance of the confirmation tests. The precision of the test method appeared to be unaffected by the type of food or the concentration of B. cereus present in the test sample. The overall repeatability value found for the food samples was 0.29 log 10 units and the overall reproducibility value was 0.42 log 10 units. For the reference material, the repeatability was 0.11 log 10 units and th