Performance of the QUALIBAXTM System (the BAX Method), a rapid PCR-based qualitative method for detecting pathogens in food, was compared to the Official Method in Japan for detecting low levels of Salmonella in feed ingredients and formula feed. A total of 78 samples, which included 17 naturally contaminated lots, was tested. A pre-enrichment step, where 25 g samples were grown in 250 ml of buffered peptone water with 0.6% Tween, allowed for BAX Method results in 32 hours that were consistent with the Official Method. Analysis determined 97.4% agreement between the two methods, with a false positive rate of 1.3% and an indeterminate rate of 1.3% for the BAX Method. The detection limit was determined by MPN to be approximately 2 cfu/100 g sample. Additional samples were taken from an Official Method selective enrichment in either selenite cystine broth or Hajna tetrathionate broth instead of BAX Method secondary enrichment in brain heart infusion broth. When tested with the BAX Method, the time to result extended beyond 32 hours but all of the BAX Method results agreed with the Official Method, demonstrating advanced precision. These studies showed that the QUALIBAXTM System can be used by feed manufacturers as a fast and accurate method for detecting Salmonella.
In order to improve the throughput of microbiological testing of food samples in the food industry, systems that reduce the number of manual steps and labor required in the testing process are needed. TEMPO is an automated enumeration system comprising automated devices and a set of reagents for microbiological testing based on the most probable number (MPN) enumeration method. In this study, we compared TEMPO with the conventional culture methods for tests of total viable counts, coliforms and Escherichia coli using 5 food samples artificially contaminated with E. coli. Other 343 food samples without spiking for the comparison between both methods. Good correlation was observed for the measured parameters between TEMPO and each conventional culture method. The present method also greatly reduced the labor requirement and reduced volume of waste generated, while providing accountability in the entire testing process. Tests of 5 food samples spiked with E. coli showed correlation coefficients between TEMPO and conventional culture methods that were higher than 0.98. On the tests using non-spiked 343 food samples, number of samples whose difference between log value of each cfu/g of TEMPO and those of conventional culture method were within ±1 were checked. On TVC, 142 samples (94%) were in the range and 130 samples (95.6%) and 19 samples (100%) were in the range on TC and EC, respectively. For tests of total viable count, the labor requirement for TEMPO was about one-third that of the conventional culture method, and the amount of waste generated using TEMPO was about one-forth that of the conventional culture method.