AOAC RI ERP Final Action Recommendations eBook

Dupont BAX Salmonella Pre- Collaborative Study For Expert Review Panel Use Only Do Not Distribute

H. Confirmation 1 All test method samples, regardless of presumptive result, were confirmed by culture and the 2 biochemical and serological protocols described in U.S. FDA Bacteriological Analytical Manual (FDA- 3 BAM) Chapter 5, USDA-FSIS Microbiology Laboratory Guidebook (MLG) Chapter 4.05, and/or Health 4 Canada Compendium of Analytical Methods MFHPB-20, using the appropriate confirmation media, B(o). 5 6 Note: All results are described for samples taken at the longest enrichment time point. Unless otherwise 7 noted, each matrix returned identical results at each time point tested (data provided upon request). 8 The results for meat and seafood samples are summarized in Tables 4 and 5. For each sample type, the 9 BAX® System method and the reference methods demonstrated no significant statistical difference as 10 indicated by POD analysis (the 95% confidence interval of the dPOD included 0 in all cases). 11 The results for poultry and egg samples are summarized in Tables 6 and 7. For each sample type, the 12 BAX® System method and the reference methods demonstrated no significant statistical difference as 13 indicated by POD analysis (the 95% confidence interval of the dPOD included 0 in all cases). 14 The results for produce samples are summarized in Tables 8 and 9. For each sample type, the BAX® 15 System method and the reference methods demonstrated no significant statistical difference as 16 indicated by POD analysis (the 95% confidence interval of the dPOD included 0 in all cases). 17 The results for dairy product samples are summarized in Tables 10 and 11. For each sample type, the 18 BAX® System method and the reference methods demonstrated no significant statistical difference as 19 indicated by POD analysis (the 95% confidence interval of the dPOD included 0 in all cases). 20 The results for environmental samples are summarized in Tables 12 and 13. For each sample type, the 21 BAX® System method and the reference methods demonstrated no significant statistical difference as 22 indicated by POD analysis (the 95% confidence interval of the dPOD included 0 in all cases). 23 The results for the remaining sample types evaluated are summarized in Tables 14 and 15. For each 24 sample type, the BAX® System method and the reference methods demonstrated no significant 25 statistical difference as indicated by POD analysis (the 95% confidence interval of the dPOD included 0 in 26 all cases). One finding from this study is that Salmonella will reliably grow from cocoa enrichments to 27 relatively high cell densities, which allow for a dilution to remove PCR inhibitors without a three-hour 28 regrowth time. Though this may not be applicable to all chocolate matrix variants, this option allows for 29 a reduced time to result for this matrix. 30 Overall, the DuPont™ BAX® System Real-time PCR Assay for Salmonella demonstrates accuracy and 31 reliability statistically equivalent to the reference culture methods used in this study, with a number of 32 additional benefits. Due to the greater robustness of this new PCR assay in the presence of PCR 33 inhibitors, the test method no longer requires a three-hour re-growth step for many matrices (with the 34 exception of some plant-based food items). For many matrices, re-growth was validated as an optional 35 step to allow for validated methods in cases where PCR inhibitors may be present in a variant of a matrix 36 Results and Discussion

AOAC INTERNATIONAL