Purpose: Internal validation studies were previously performed by DuPont Nutrition & Health on 24 matrices to demonstrate the reliability of the test method among a wide variety of sample types. Two of these matrices –frankfurters and orange juice– were each evaluated in 14 independent laboratories as part of a collaborative study to demonstrate repeatability of the method independent of the end user.
Methods: Samples were prepared at three inoculation levels - unspiked, low (<2.0 MPN/test portion) and high (~5 MPN/test portion) – then blind-coded and sent to each participating site. Frankfurters were compared to the USDA FSIS reference method in a paired study design, and orange juice was compared to the FDA BAM reference method in an unpaired study with enrichment in proprietary BAX® System MP media.
Results: For low-spike samples, the BAX® System method returned presumptive positive results for 186 of the 336 samples tested, and the reference method confirmed 182 positive results. For high-spike samples, the BAX® System method returned positive results for all 336 samples tested, and the reference method confirmed all 336 positive results. Also, all negative control samples (n = 336 samples total) were presumptively and culturally negative. At each inoculation level and for each matrix, results from the BAX® System method and the reference method were statistically indistinguishable when analyzed using the Probability of Detection (POD) statistical treatment as required by AOAC.
Significance: This study indicates that PCR detection of Salmonella using the BAX® System method is rapid and sensitive. Test kit results demonstrated no significant difference when compared with the reference culture methods.