Determination of Limit of Detection and Relative Limit of Detection of Salmonella in Raw Pet Food Matrices Using Salmonella Bacteriological Analytical Manual Methods

Abstract

Each year, there are multiple Salmonella recalls due to contaminated pet food products. Routine testing of finished products requires reliable, quick, and sensitive detection methods. In this study we comparatively evaluated the sensitivity of a culture-based reference method and two alternative methods, which are based on LAMP and PCR principles. The evaluation was performed in multiple trials using challenging pet food matrices, which were tested after various storage time intervals to better address the reliability of the methods. Raw freeze-dried treats, kibble, and patties were individually inoculated with Salmonella Typhimurium at fractional and high inoculation levels. Raw freeze-dried treats and kibble were inoculated at 0, 1, 3, 5, 7 and 10 CFU/25g and raw patties were inoculated at 0, 1, 3, 5, 7, 10 and 30 CFU/25g. For the culture method, the level of detection (LOD50) values were 0.9, 1.1, and 3.7 CFU/25g for freeze-dried treats, kibble, and patties respectively. The calculated relative level of detection (RLOD) values were close to 1 CFU/25g, indicating that LAMP and PCR methods have similar LOD50 values to the culture method. It was also confirmed that the alternative LAMP and PCR methods can provide reliable results within 24 hours that match the reference culture method. Faster response times are invaluable when responding to potential pet food product contamination.

Keywords: : Salmonella, pet food, level of detection, relative level of detection, method comparison, Bacteriological Analytical Manual