A Comprehensive Comparison of the BAX® System Salmonella 2 and Real-time Salmonella PCR Assays to the FDA BAM Reference Methods for the Detection of Salmonella in a Variety of Soy Ingredients

Introduction: There have been several well publicized outbreaks of salmonellosis that resulted from the consumption of contaminated low water activity, ready-to-eat foods with the root cause of some of these outbreaks being traced to the low water activity ingredients.  For example, the Salmonella outbreak associated with contaminated peanut butter ingredient resulted in a class I product recalls from over 360 firms. Following this outbreak, the industry and regulators realized the need to evaluate rapid automated pathogen testing technologies in low moisture products and ingredients.

Purpose: The purpose of the evaluation was to proactively evaluate the performance of the BAX System^® Salmonella 2 and Real-Time Salmonella PCR assays using Buffered Peptone Water (BPW) as a primary enrichment against the FDA BAM or ISO 6579 reference methods on a variety of low moisture, soy ingredients: Isolated Soy Protein (ISP), Soy Fiber (SF), Soy Fluid Lecithin (SFL), Soy Deoiled Lecithin (DL), and Soy Nuggets (SN). 

Methods: For each method, 20 replicates, either 25g or 375g, were analyzed at one inoculum level: 0.2-2 CFU/test portion.  Five control replicates were analyzed at 0 CFU/test portion.  Each matrix was analyzed by the candidate method after 22-24 hours of enrichment in BPW with a 3 hour re-growth step in BHI. 

Results: The results of a McNemar’s Chi-square analysis indicated no statistically significant difference observed between the candidate method and the reference methods (X²< 3.84). There was no observed reduction of sensitivity in larger 375 g composite samples.   

Significance: These studies indicate that the new methods, in combination with the single BPW primary enrichment and subsequent BHI re-growth step, demonstrated equivalent sensitivity and robustness when compared to the reference methods for both 25 g and 375 g composite samples.  The new methods also provide a 30h time to presumptive results as compared to up to 3-4 days for the reference methods.