Purpose: To compare various methods to detect Salmonella in ground beef and cilantro.
Methods: Five methods were compared: Petrifilm Salmonella Express System (SALX) (3M); Molecular Detection Assay (MDS) (3M); BAX System (Dupont Qualicon); GeneDisc (PALL Corporation); and the gold standard from FDA and USDA (USDA MLG 4.08 for ground beef; Bacteriological Analytical Manual (BAM) for cilantro). Commercially-available cilantro and irradiated ground beef were used. Samples were inoculated with a five-strain Salmonella enterica cocktail, (Enteritidis, Newport, Typhimurium, Heidelberg, and Blockley). Protocols indicated by each method were followed. A total of 92 tests were conducted per method and sample type, divided into four groups of 23 each as follows: 23 un-inoculated samples and three groups of 23 each inoculated with the cocktail at different levels of Salmonella concentration (attachment of ca. 0.6, 1, and 2 log, respectively). Data was analyzed using the Generalized Linear Model (GLM) with the statistical software SAS 9.4.
Results: All methods were able to detect 100% of Salmonella from inoculated ground beef; statistical difference (P=0.0003) was found among methods only with un-inoculated samples, from which MDS, SALX, and BAX detected positives (60.87%, 40.00%, and 8.70%, respectively). Regarding cilantro, all methods detected Salmonella at the two highest levels of inoculation. SALX failed to detect Salmonella in 22% of the lowest inoculation level. No statistical difference (P>0.05) was found among methods. BAM, GENEDISC, MDS, and SALX detected positive samples from un-inoculated controls.
Significance: Comparing the efficacy of Salmonella detection by different tests, provides useful information to the food industry from an independent source.