AOAC ERP MICRO AUGUST 2018

96 B ird et al .: J ournal of AOAC I nternational V ol . 100, N o . 1, 2017 method, 85 of 132 test portions confirmed positive (POD C of

strain as the organism that was inoculated, indicating that cross- contamination of the sample had occurred. Due to the fact that cross-contamination had occurred, just cause removal of the data was established and the data generated by laboratory 10 was, therefore, not included in the statistical analysis. Overall, the data generated during this evaluation demonstrated the reproducibility of this new method. For the deli turkey analysis, the POD statistical analysis indicated the difference between the candidate and reference methods was not statistically significant at the 0.05 probability level, and that the difference between presumptive and confirmed candidate methods was not statistically significant at the 0.05 probability level. For raw chicken breast fillet, a statistically significant difference was observed between the reference and alternative methods. dLPOD being significantly greater than zero showed an observed higher proportion of positive results by the candidate method than the reference method. One possible contribution for the higher number of positive results observed with the 3MMDA2– Listeria method was the use of DF broth for the candidate method. This enrichment media formulation may be less selective than the modified UVMmedium used in the USDA reference method and may have contributed to the higher level of recovery observed during the evaluation. A second possible contribution for the higher observed proportion of positive results with the candidate method was the duration of the primary enrichment. Test portions evaluated by the 3M MDA 2– Listeria method were incubated for a minimum of 28 h in the primary enrichment, whereas the USDA reference method had a maximum primary enrichment time of 26 h. No statistically significant difference was observed between the candidate method presumptive and confirmed results for this matrix. It is recommended that the 3M MDA 2– Listeria method be adopted Official First Action status for the detection of Listeria in selected foods: hot dogs (25 and 125 g); salmon (25 g); deli turkey (25 and 125 g); cottage cheese (25 g); vanilla ice cream (25 g); queso fresco (25 g); spinach (25 g); melon (whole); raw chicken leg pieces (25 g); raw chicken fillet (25 g); and concrete, stainless steel, and plastic environmental samples. Recommendations

0.64). For test portions evaluated by the USDA/FSIS MLG 8.09 reference method, 64 of 132 test portions produced positive results. A dLPOD C value of 0.16 with 95% confidence intervals of (0.04, 0.28) was obtained between the candidate and reference methods, indicating a statistically significant difference between the two methods, with an observed higher proportion of positive results by the candidate method than the reference method. A dLPOD CP value of 0.02 with 95% confidence intervals of (–0.10, 0.13) was obtained between presumptive and confirmed results, indicating the difference between presumptive and confirmed results was not statistically significant at the 0.05 probability level. For the high inoculum level, 131 of 132 test portions (POD CP of 0.99) were reported as presumptive positive by the 3M MDA 2– Listeria method with 132 of 132 test portions (POD CC of 1.00) confirming positive. For samples that produced presumptive positive results on the 3M MDA 2– Listeria method, 131 of 132 samples confirmed positive (POD C of 0.99). For test portions evaluated by the USDA/FSIS MLG 8.09 reference method, 132 of 132 test portions produced positive results. A dLPOD C value of –0.01 with 95% confidence intervals of (–0.04, 0.02) was obtained between the candidate and reference methods, indicating the difference between methods was not statistically significant at the 0.05 probability level. A dLPOD CP value of –0.01 with 95% confidence intervals of (–0.04, 0.02) was obtained between presumptive and confirmed results, indicating the difference between presumptive and confirmed results was not statistically significant at the 0.05 probability level. For the uninoculated controls, 2 of 132 samples (POD CP of 0.02) produced a presumptive positive result by the 3M MDA 2– Listeria method with 1 of 132 test portions (POD CC of 0.01) confirming positive. For samples that produced presumptive positive results on the 3M MDA 2– Listeria method, 1 of 132 samples confirmed positive (POD C of 0.01). For test portions evaluated by the USDA/FSIS MLG 8.09 reference method, 0 of 132 test portions produced positive results. A dLPOD C value of 0.01 with 95% confidence intervals of (–0.02, 0.04) was obtained between the candidate and reference methods, indicating the difference between methods was not statistically significant at the 0.05 probability level. A dLPOD CP value of 0.01 with 95% confidence intervals of (–0.03, 0.05) was obtained between presumptive and confirmed results, indicating the difference between presumptive and confirmed results was not statistically significant at the 0.05 probability level. Detailed results of the POD statistical analysis are presented in Table 2016.07D and Figure 2A and B. No negative feedback was provided by the collaborating laboratories regarding the performance of the 3MMDA2– Listeria method. During the evaluation of the raw chicken breast fillet, Laboratory 2 isolated L. innocua from an uninoculated control sample. Because the organism recovered was different from the organism that was inoculated, L. monocytogenes, no just cause for removal of the data was determined, therefore, the data were included. For the raw chicken breast fillet, Laboratory 10 reported isolating L. monocytogenes from two uninoculated control samples. The isolates were sent for further identification and it was determined that they were the same Discussion

Acknowledgments We would like to extend a sincere thank you to the following collaborators for their dedicated participation in this study: Robert Brooks, ATC Microbiology, LLC (North Little Rock, AR) Jaspreet Walia and Francisco Hernandez, Certified Laboratories (Turlock, CA) David Bosco and Grizelda Trevino, Food Safety Net Services (Fresno, CA) Alex Brandt and Chris Lopez, Food Safety Net Services (San Antonio, TX) Elizabeth Sjogren and Manish Shekhawat, Microbac Laboratories, Inc. (Worcester, MA) Li Maria Ma, Chris Timmons, and Claudia Diaz Proano, Oklahoma State University (Stillwater, OK) Alexandra Calle and David Campos, Texas Tech University (Lubbock, TX) ERP Use Only