Internal Validation Study Revised April 24, 2013

Transcription

1 A Comparative Evaluation of the Invisible Sentinel Veriflow Method for the Detection of in Select Foods and Environmental Surfaces Internal Validation Study Revised April 24, 2013 Erin Crowley, Patrick Bird, Kiel Fisher, Jonathan Flannery, Katie Goetz, M. Joseph Benzinger Jr., Will Judd, Paige Bedinghaus, James Agin, David Goins Q Laboratories, Inc., 1400 Harrison Ave, Cincinnati, OH A comparative validation study of the Invisible Sentinel Veriflow (LM) method for the detection of was conducted at Q Laboratories, Inc. The method was compared to the USDA/ FSIS MLG 8.08 (2012) Isolation and Identification of from Red Meat, Poultry and Egg Products, and Environmental Samples for hot dogs, deli turkey and for stainless steel, concrete, plastic and ceramic tile environmental surfaces and to the AOAC Official Method in Milk and Dairy Products (1996) for 2% pasteurized milk. Twenty (20) replicates of each matrix were analyzed at a low inoculum level of CFU/test potion, 5 replicates were analyzed at a high inoculum level of 2-5 CFU/test portion and 5 control replicates were also analyzed at 0 CFU/test portion (un-inoculated). Results were analyzed by the Probability of Detection (POD) statistical model. There was no statistically significant difference in the number of positive samples detected by the method and reference methods for all 7 matrices analyzed. The method accurately detected L. as a rapid and easy to use alternative method for the detection L. in select foods and environmental surfaces. This report presents the analytical results for comparison of Invisible Sentinel Veriflow (LM) method to the USDA/ FSIS MLG 8.08 (2012) method: Isolation and Identification of from Red Meat, Poultry and Egg Products, and Environmental Samples and to the AOAC Official Method in Milk and Dairy Products (1996). All analyses were conducted at Q Laboratories, Inc. (Cincinnati, OH). All test kits and proprietary media unique to the method were provided by Invisible Sentinel Inc. (Philadelphia, PA). MATERIALS AND METHODS The methodology for this study was followed as outlined in the AOAC Research Institute s Performance Tested Methods Program Protocol: Comparative Evaluation of the Invisible Sentinel Veriflow Method. The study consisted of evaluating a total of 30 samples for each of 3 food matrices and 4 environmental surfaces. Within each sample set, there were 5 un-inoculated samples, 20 low-level inoculated samples and 5 high-level inoculated samples. The target levels of each strain of used for challenging each matrix were as follows: colony forming units (CFU) /test portion for the low-level inoculation, 2-5 CFU/test portion for the high-level inoculation and 0 CFU/test portion for the un-inoculated control samples. The results of the low-level inoculation were designed to produce fractionally positive results, those in which ether the reference or candidate method yields 5 to 15 confirmed positive results out of the 20 replicates. Page 1 of 25

2 A background screen of each food matrix was conducted prior to inoculation and no natural contamination of the target organism was detected in the test matrices. For the method comparison, each matrix was inoculated with commonly associated strains of as indicated in Table A. For each food matrix, a diluted 24 hour broth culture inoculum was added to a bulk sample of each food matrix and mixed thoroughly. The deli turkey, hot dogs and 2 % milk were inoculated with an organism that had been heat stressed for 10 minutes at 50 o C in a water bath. The degree of injury of the culture was estimated by plating an aliquot of diluted culture onto Modified Oxford agar (MOX) and Tryptic Soy agar (TSA). The agars were incubated at 35 ± 1 C for 48 ± 2 hours and the colonies were counted. The degree of injury was estimated as nselect ( 1 ) x100 nnonselect where n select = number of colonies on selective agar and n nonselect = number of colonies on nonselective agar. Results of the percent injury are presented in Table 3 of the Appendix. Following inoculation, all test portions were mixed thoroughly, and were held at 2-5 C for hours prior to analysis to allow time for the organisms to equilibrate within the sample. Table A: Test Matrices and Inoculating Organisms Matrix Inoculating Organism Matrix Inoculating Organism Deli Turkey Hot Dogs 2% Pasteurized Milk Stainless Steel ATCC ATCC 7644 ATCC ATCC 7644 E. coli ATCC 10x level Page 2 of 25 Ceramic Tile Concrete Plastic ATCC ATCC ATCC For hot dogs and 2% pasteurized milk, the level of was determined by Most Probable Number (MPN) on the day of analysis by analyzing 5 x 100 g, 20 x 25 g (reference method test portions) and 5 x 10 g inoculated test samples. For the deli turkey, 5 x 100 g, 5 x 25 g (from the bulk lot of inoculated test product) and 5 x 10 g test portions were analyzed. Each test portion was enriched with the appropriate reference method enrichment broth and analyzed by the reference method procedure. The number of positives from the 3 test levels was used to calculate the MPN using the AOAC MPN calculator. ( Environmental surfaces were inoculated at a level expected to yield fractional recovery. For stainless steel and concrete surfaces, 4 x 4 areas were inoculated with 0.25 ml of diluted culture and for ceramic tile and plastic surfaces, 1 x 1 areas were inoculated with 0.10 ml of diluted cultures. In addition to the culture, the stainless steel surfaces were co-inoculated with a competitor organism, E. coli, at approximately 10x the inoculation level of the culture. Each surface dried for hours at room temperature (24 ±2 o C). To determine the inoculation level for the environmental surfaces, aliquots of each inoculum were plated onto Tryptic Soy agar (TSA) in triplicate. Sampling sponges, pre-moistened with Dey-Engley neutralizing broth, were used to sample each test area using horizontal and vertical motions. The sponges were held at room temperature for at least 2 hours prior to analysis. Results of the MPN analysis of the food matrices and inoculum plates for the environmental surfaces are presented in Table 4A and 4B, respectively

3 USDA/FSIS MLG 8.08 Reference Method For hot dog samples, 25 g test portions were enriched with 225 ± 5 ml of University of Vermont Medium broth (UVM), homogenized for two minutes, and incubated at 30 ± 2 C for hours. For deli turkey samples, 125 g test portions were enriched with 1125 ± 25 ml of UVM broth, homogenized for 2 minutes and incubated at 30 ± 2 C for hours. For environmental samples, sponges were enriched with 225 ml of UVM and homogenized by hand and swabs were enriched with 10 ml of UVM and mixed by vortexing. Sponges and swabs were incubated for hours at 30 ± 2 C. After incubation of all test portions, 0.1 ml of the sample enrichment was transferred to 10 ml of Fraser Broth (FB) containing 0.1 ml of 5% ferric ammonium citrate and incubated at 35 ± 2 C for 26 ± 2 hours. A loopful of the sample enrichment was also streaked to MOX and incubated at 35 ± 2 C for 26 ± 2 hours. After 26 ± 2 hours, FB was examined for any degree of darkening due to esculin hydrolysis. Any FB that displayed darkening was streaked to a MOX plate. If no darkening occurred, FB was re-incubated at 35 ± 2 C for a total of 48 ± 2 hours and re-examined for evidence of darkening. If darkening occurred, the FB was streaked to a MOX plate, if no darkening occurred, samples were considered negative. All FB streaked MOX plates were incubated at 35 C for 26 ± 2 hours. After incubation for 26 ± 2 hours, all MOX plates, either UVM or FB streaked, were examined for the presence of typical colonies. If no suspect colonies were present on MOX, the plates were re-incubated for an additional 26 ± 2 hours at 35 ± 2 C. If suspect colonies were present on the UVM streaked MOX, these suspect colonies were confirmed via confirmatory methods outlined by the USDA/FSIS MLG 8.08 method which included catalase test, β-hemolysis examination and biochemical tests by VITEK 2 GP, AOAC OMA AOAC in Milk and Dairy Products For 2% pasteurized milk samples analyzed, 25 g were enriched with 225 ml of selective enrichment media and incubated at 30 C for 48 hours. At 48 hours the enriched samples were streaked to Oxford agar (OXA) and incubated at 37 C for 48 hours. At 48 hours the OXA plates were examined for suspect colonies. Up to five suspect colonies were streaked to TSA with yeast additive (TSA/ye) to obtain well isolated, pure colonies. Isolates from TSA/ye were confirmed via catalase reaction, β-hemolysis examination and biochemical tests by VITEK 2 GP, AOAC OMA Invisible Sentinel Veriflow Method For environmental sponges analyzed by the method, 100 ml of IS Enhanced Broth was added to the sampling bag and the sponge was homogenized for 30 seconds. For environmental swabs, 20 ml of IS Enhanced Broth was added to the swab and homogenized by vortex for 30 seconds. Both sponges and swabs were incubated for 24 hours at 35 o C. For deli turkey test portions, each 125 g sample was enriched with 375 ml of IS Enhanced Broth and homogenized for 2 minutes. For 2% pasteurized milk and hot dog test portions, each 25 g sample was enriched with 225 ml of IS Enhanced Broth and homogenized for 2 minutes. All food matrices were incubated at 35 o C for hours. After incubation, 1 ml of enrichment was transferred to a 1.5 ml tube containing three drops of Buffer A. The sample tubes boiled in a water bath for 10 minutes, and then sample tubes were removed and allowed to cool to room temperature (20-25 o C). From the cooled sample, a 5 µl aliquot was transferred to a thawed PCR reagent tube containing 45 µl of Master Mix. The sample PCR reagent tubes were cycled according to the parameters specified in the package Page 3 of 25

4 insert. The cycling parameters for the amplification procedure are presented in Table B. Following the amplification procedure, 4 drops of Buffer B were added to each sample PCR tube. The entire contents of the PCR tube were transferred to the sample window of a Flowthrough cassette. The test was allowed to develop for 2 minutes, and then 4 more drops of Buffer B were added to the sample window of the Flowthrough cassette and allowed to develop for 1 minute. The Flowthrough cassette lever was retracted and the test result was visually interpreted. Regardless of the presumptive results, each sample was confirmed according to procedures outlined in the USDA/FSIS-MLG 8.08 or AOAC reference methods. Final biochemical confirmation was conducted according to VITEK 2 GP (AOAC OMA ). Table B: PCR Amplification Parameters Stage 1 Description Temperature Time Number of Cycles Denaturing 94 C 2 minutes 1 Stage 2 Description Temperature Time Number of Cycles Denaturing 94 C 15 seconds Annealing 55 C 15 seconds 40 Extension 70 C 30 seconds Stage Description Temperature Time Number of Cycles Extension 70 C 1 minute 1 Cooling 4 C 1 minute Inclusivity and Exclusivity Evaluation For the inclusivity evaluation, 50 strains of were cultured in IS Enriched Broth and incubated at 30 o C for 24 hours. The cultures were diluted to 10x the Limit of Detection (2.5 x 10 5 ) and prepared for analysis with the method. For the exclusivity portion of the validation, 30 closely related non- cultures were enriched in BHI and incubated at temperatures and conditions specific for organism growth. The non-target strains were undiluted and prepared for analysis with the method. All inclusivity and exclusivity strains were randomized and blind-coded prior to analysis by the method. RESULTS Method Comparison For the method comparison, the probability of detection (POD) was calculated as the number of positive outcomes divided by the total number of trials. The POD was calculated for the candidate presumptive results, POD CP, the candidate confirmatory results, POD CC, the difference in the candidate presumptive and confirmatory results, dpod CP, the candidate method, POD C, the Page 4 of 25

5 reference method, POD R, and the difference in the confirmed candidate and reference methods, dpod C. A summary of results are presented in Table C and a detailed summary of individual sample results are presented in Tables 5-11 of the Appendix. Table C: Summary of Results Matrix Deli Turkey Hot Dogs USDA/FSIS Inoculation Level 24 Hour Enrichment MLG 24 Hour Enrichment Presumptive Confirmed Method Presumptive Confirmed Page 5 of 25 USDA/FSIS MLG Method Un-inoculated 0/5 0/5 0/5 0/5 0/5 0/5 Low 9/20 9/20 7/20 5/20 5/20 8/20 High 5/5 5/5 5/5 5/5 5/5 5/5 Matrix 2% Milk Stainless Steel Inoculation Level AOAC 24 Hour Enrichment 24 Hour Enrichment Method Presumptive Confirmed Presumptive Confirmed USDA/FSIS MLG Method Un-inoculated 0/5 0/5 0/5 0/5 0/5 0/5 Low 11/20 11/20 9/20 6/20 6/20 7/20 High 5/5 5/5 5/5 5/5 5/5 5/5 Matrix Concrete Ceramic Tile Inoculation Level 24 Hour Enrichment USDA/FSIS MLG 24 Hour Enrichment Presumptive Confirmed Method Presumptive Confirmed USDA/FSIS MLG Method Un-inoculated 0/5 0/5 0/5 0/5 0/5 0/5 Low 7/20 7/20 9/20 12/20 13/20 10/20 High 5/5 5/5 5/5 5/5 5/5 5/5 Matrix Inoculation Level Plastic 24 Hour Enrichment Presumptive Confirmed USDA/FSIS MLG Method Un-inoculated 0/5 0/5 0/5 Low 12/20 12/20 7/20 High 5/5 5/5 5/5 Deli Turkey Results For low-level inoculum deli turkey test portions, there were 9 confirmed positives for the method and 7 confirmed positives for the USDA/FSIS-MLG method. All highlevel inoculum test portions were positive and all un-inoculated control samples were negative for both methods. No false positives or negatives were observed. For the low-level inoculum, a POD CP and POD CC value with 95% confidence intervals of 0.45 (0.26, 0.66) was obtained. A dpod CP value of 0 (-0.28, 0.28) was obtained, indicating no significant difference in presumptive and confirmed results. A POD C and POD R values with 95% confidence intervals of 0.45 (0.26, 0.66) and 0.35 (0.18, 0.57) were obtained. A dpod C value of 0.10 (-0.13, 0.33) was obtained, indicating no significant difference between the method and the reference method confirmed results. For the high-level inoculum, a POD CP and POD CC value with 95% confidence intervals of 1.00 (0.57, 1.00) was obtained. A dpod CP value of 0 (-0.43, 0.43) was obtained, indicating no

6 confidence intervals of 1.00 (0.57, 1.00) was obtained. A dpod C value of 0.00 (-0.43, 0.43) was obtained, indicating no significant difference between the method and the reference method confirmed results. For the un-inoculated control level, a POD CP and POD CC value with 95% confidence intervals of 0.00 (0.00, 0.43) was obtained. A dpod CP value of 0 (-0.43, 0.43) was obtained, indicating no confidence intervals of 0.00 (0.00, 0.43) was obtained. A dpod C value of 0.00 (-0.43, 0.43) was obtained, indicating no significant difference between the method and the reference method confirmed results. Detailed results of the POD analysis are presented in Tables 12 and 13 of the Appendix. Hot Dogs For low-level inoculum hot dog test portions, there were 5 confirmed positives for the method and 8 confirmed positives for the USDA/FSIS-MLG method. All highlevel inoculum test portions were positive and all un-inoculated control samples were negative for both methods. No false positives or negatives were observed. For the low-level inoculum, a POD CP and POD CC value with 95% confidence intervals of 0.25 (0.11, 0.47) was obtained. A dpod CP value of 0 (-0.26, 0.26) was obtained, indicating no significant difference in presumptive and confirmed results. A POD C and POD R values with 95% confidence intervals of 0.25 (0.11, 0.47) and 0.40 (0.22, 0.61) were obtained. A dpod C value of (-0.40, 0.13) was obtained, indicating no significant difference between the method and the reference method confirmed results. For the high-level inoculum, a POD CP and POD CC value with 95% confidence intervals of 1.00 (0.57, 1.00) was obtained. A dpod CP value of 0 (-0.43, 0.43) was obtained, indicating no confidence intervals of 1.00 (0.57, 1.00) was obtained. A dpod C value of 0.00 (-0.43, 0.43) was obtained, indicating no significant difference between the method and the reference method confirmed results. For the un-inoculated control level, a POD CP and POD CC value with 95% confidence intervals of 0.00 (0.00, 0.43) was obtained. A dpod CP value of 0 (-0.43, 0.43) was obtained, indicating no confidence intervals of 0.00 (0.00, 0.43) was obtained. A dpod C value of 0.00 (-0.43, 0.43) was obtained, indicating no significant difference between the method and the reference method confirmed results. Detailed results of the POD analysis are presented in Tables 12 and 13 of the Appendix. 2% Pasteurized Milk For low-level inoculum milk test portions, there were 11 confirmed positives for the Veriflow LM method and 9 confirmed positives for the USDA/FSIS-MLG method. All high-level inoculum test portions were positive and all un-inoculated control samples were negative for both methods. No false positives or negatives were observed. For the low-level inoculum, a POD CP and POD CC value with 95% confidence intervals of 0.55 (0.34, 0.74) was obtained. A dpod CP value of 0 (-0.28, 0.28) was obtained, indicating no significant difference in presumptive and confirmed results. A POD C and POD R values with 95% confidence intervals of 0.55 (0.34, 0.74) and 0.45 (0.26, 0.66) were obtained. A dpod C value of 0.10 (-0.19, 0.37) was obtained, indicating no significant difference between the method and the reference method confirmed results. For the high-level inoculum, a POD CP and POD CC value with 95% confidence intervals of 1.00 (0.57, 1.00) was obtained. A dpod CP value of 0 (-0.43, 0.43) was obtained, indicating no Page 6 of 25

7 confidence intervals of 1.00 (0.57, 1.00) was obtained. A dpod C value of 0.00 (-0.43, 0.43) was obtained, indicating no significant difference between the method and the reference method confirmed results. For the un-inoculated control level, a POD CP and POD CC value with 95% confidence intervals of 0.00 (0.00, 0.43) was obtained. A dpod CP value of 0 (-0.43, 0.43) was obtained, indicating no confidence intervals of 0.00 (0.00, 0.43) was obtained. A dpod C value of 0.00 (-0.43, 0.43) was obtained, indicating no significant difference between the method and the reference method confirmed results. Detailed results of the POD analysis are presented in Tables 12 and 13 of the Appendix. Stainless Steel For low-level inoculum stainless steel test portions, there were 6 confirmed positives for the method and 7 confirmed positives for the USDA/FSIS-MLG method. All highlevel inoculum test portions were positive and all un-inoculated control samples were negative for both methods. No false positives or negatives were observed. For the low-level inoculum, a POD CP and POD CC value with 95% confidence intervals of 0.30 (0.15, 0.52) was obtained. A dpod CP value of 0 (-0.27, 0.27) was obtained, indicating no significant difference in presumptive and confirmed results. A POD C and POD R values with 95% confidence intervals of 0.30 (0.15, 0.52) and 0.35 (0.18, 0.57) were obtained. A dpod C value of (-0.32, 0.23) was obtained, indicating no significant difference between the method and the reference method confirmed results. For the high-level inoculum, a POD CP and POD CC value with 95% confidence intervals of 1.00 (0.57, 1.00) was obtained. A dpod CP value of 0 (-0.43, 0.43) was obtained, indicating no confidence intervals of 1.00 (0.57, 1.00) was obtained. A dpod C value of 0.00 (-0.43, 0.43) was obtained, indicating no significant difference between the method and the reference method confirmed results. For the un-inoculated control level, a POD CP and POD CC value with 95% confidence intervals of 0.00 (0.00, 0.43) was obtained. A dpod CP value of 0 (-0.43, 0.43) was obtained, indicating no confidence intervals of 0.00 (0.00, 0.43) was obtained. A dpod C value of 0.00 (-0.43, 0.43) was obtained, indicating no significant difference between the method and the reference method confirmed results. Detailed results of the POD analysis are presented in Tables 14 and 15 of the Appendix. Concrete For low-level inoculum concrete test portions, there were 7 confirmed positives for the method and 9 confirmed positives for the USDA/FSIS-MLG method. All uninoculated control samples were negative and all high inoculum samples were positive for both methods. No false positives or false negatives were observed. For the low-level inoculum, a POD CP and POD CC value with 95% confidence intervals of 0.35 (0.18, 0.57) was obtained. A dpod CP value of 0.00 (-0.28, 0.28) was obtained, indicating no significant difference in presumptive and confirmed results. A POD C and POD R values with 95% confidence intervals of 0.35 (0.18, 0.57) and 0.45 (0.26, 0.66) were obtained. A dpod C value of (-0.37, 0.19) was obtained, indicating no significant difference between the method and the reference method confirmed results. For the high-level inoculum, a POD CP and POD CC value with 95% confidence intervals of 1.00 (0.57, 1.00) was obtained. A dpod CP value of 0 (-0.43, 0.43) was obtained, indicating no Page 7 of 25

8 confidence intervals of 1.00 (0.57, 1.00) was obtained. A dpod C value of 0.00 (-0.43, 0.43) was obtained, indicating no significant difference between the method and the reference method confirmed results. For the un-inoculated control level, a POD CP and POD CC value with 95% confidence intervals of 0.00 (0.00, 0.43) was obtained. A dpod CP value of 0 (-0.43, 0.43) was obtained, indicating no confidence intervals of 0.00 (0.00, 0.43) was obtained. A dpod C value of 0.00 (-0.43, 0.43) was obtained, indicating no significant difference between the method and the reference method confirmed results. Detailed results of the POD analysis are presented in Tables 14 and 15 of the Appendix. Ceramic Tile For low-level inoculum ceramic tile test portions, there were 12 presumptive positives and 13 confirmed positives for the method and 10 confirmed positives for the USDA/FSIS-MLG method. All high-level inoculum test portions were positive and all uninoculated control samples were negative for both methods. One (1) false negative result was obtained and no false negatives were observed. For the low-level inoculum, a POD CP and POD CC values with 95% confidence intervals of 0.60 (0.39, 0.78) and 0.65 (0.43, 0.82) were obtained. A dpod CP value of (-0.32, 0.23) was obtained, indicating no significant difference in presumptive and confirmed results. A POD C and POD R values with 95% confidence intervals of 0.60 (0.39, 0.78) and 0.50 (0.30, 0.70) were obtained. A dpod C value of 0.10 (-0.19, 0.37) was obtained, indicating no significant difference between the method and the reference method confirmed results. For the high-level inoculum, a POD CP and POD CC value with 95% confidence intervals of 1.00 (0.57, 1.00) was obtained. A dpod CP value of 0 (-0.43, 0.43) was obtained, indicating no confidence intervals of 1.00 (0.57, 1.00) was obtained. A dpod C value of 0.00 (-0.43, 0.43) was obtained, indicating no significant difference between the method and the reference method confirmed results. For the un-inoculated control level, a POD CP and POD CC value with 95% confidence intervals of 0.00 (0.00, 0.43) was obtained. A dpod CP value of 0 (-0.43, 0.43) was obtained, indicating no confidence intervals of 0.00 (0.00, 0.43) was obtained. A dpod C value of 0.00 (-0.43, 0.43) was obtained, indicating no significant difference between the method and the reference method confirmed results. Detailed results of the POD analysis are presented in Tables 14 and 15 of the Appendix. Plastic For low-level inoculum plastic test portions, there were 12 confirmed positives for the method and 7 confirmed positives for the USDA/FSIS-MLG method. All highlevel inoculum test portions were positive and all un-inoculated control samples were negative for both methods. No false positives or negatives were observed. For the low-level inoculum, a POD CP and POD CC value with 95% confidence intervals of 0.60 (0.39, 0.78) was obtained. A dpod CP value of 0 (-0.28, 0.28) was obtained, indicating no significant difference in presumptive and confirmed results. A POD C and POD R values with 95% confidence intervals of 0.60 (0.39, 0.78) and 0.35 (0.18, 0.57) were obtained. A dpod C value of 0.25 (-0.05, 0.50) was obtained, indicating no significant difference between the method and the reference method confirmed results. For the high-level inoculum, a POD CP and POD CC value with 95% confidence intervals of 1.00 (0.57, 1.00) was obtained. A dpod CP value of 0 (-0.43, 0.43) was obtained, indicating no confidence intervals of 1.00 (0.57, 1.00) was obtained. A dpod C value of 0.00 (-0.43, 0.43) was Page 8 of 25

9 obtained, indicating no significant difference between the method and the reference method confirmed results. For the un-inoculated control level, a POD CP and POD CC value with 95% confidence intervals of 0.00 (0.00, 0.43) was obtained. A dpod CP value of 0 (-0.43, 0.43) was obtained, indicating no confidence intervals of 0.00 (0.00, 0.43) was obtained. A dpod C value of 0.00 (-0.43, 0.43) was obtained, indicating no significant difference between the method and the reference method confirmed results. Detailed results of the POD analysis are presented in Tables 14 and 15 of the Appendix. Inclusivity and Exclusivity Results Of the 50 inclusivity strains analyzed, 50 were correctly detected as. Of the 30 exclusivity strains analyzed, all 30 strains were not detected by the method. Results of the inclusivity and exclusivity evaluation are presented in Tables 1 and 2. DISCUSSION The results of this study demonstrated the reliability of the method when compared to the traditional reference methods for the detection of on environmental surfaces and in select food matrices. The method offers a significant savings in time compared to the USDA/ FSIS MLG 8.08 or AOAC reference methods by producing accurate presumptive results in approximately 27 hours compared to 72 to 96 hours for the reference methods. The method protocol is easy to follow and simplifies the process for the rapid detection of. Page 9 of 25

10 Appendix Page 10 of 25

11 Table 1: Inclusivity Summary of Results Organism/ Reference Number/ Serotype CWD /2C CWD /2A Source Not Available Carlisle 1981 Organism/ Reference Number/ Serotype CWD /2B CWD B Source Veriflow LM Not Available Not Available Organism/ Reference Number/ Serotype NCIMB# B NCTC# Source Veriflow LM Not Available Human feces CWD /2A CWD B CWD B CWD B CWD B CWD B CWD B CWD B CWD /2B CWD B CWD /2B CWD /2B CWD B CWD B Carlisle 1981 Not Available Not Available Mother and Baby Placenta Lausanna 1987 LA Outbreak 1985 Not Available Halifax 1983 Not Available Not Available Not Available LA San Francisco Not Available CWD B CWD /2B CWD /2A CWD /2A CWD /2A CWD /2A CWD /2A CWD /2A CWD /2A CWD /2B CWD /2B CWD /2A CWD /2A 2 FSL J C 2 Not Available Not Available CWD B FSL J AB 1 Cultures were obtained from the University of Vermont, 2 Cultures were obtained from Cornell University, 3 Culture obtained from Invisible Sentinel Culture Collection Page 11 of 25 Not Available LA Turkey Frank Factory Turkey Frank Factory Turkey Frank Factory Turkey Frank Factory Oklahoma Turkey Frank Factory Turkey Frank Factory Oklahoma Turkey Frank Factory Mother/Baby Oklahoma Turkey Frank Factory Turkey Frank Factory Not Available BAA-751 1/2B B D E B /2A /2A A /2B A 3 BEI# Li2 4B Not Available Poultry, England Spinal Fluid of Man, Scotland Human Isolate Sheep, USA Chicken, England Spinal Fluid of Child Rabbit, England Not Available Dairy Products Belgium Cheese Cheese Not Available

12 Table 2: Exclusivity Summary of Results Organism/ Reference Number/ Serotype Alcaligenes faecalis ATCC 8750 Bacillus cereus ATCC Bacillus subtilis ATCC 6051 Campylobacter jejuni ATCC Campylobacter lari ATCC BAA-1060 Candida albicans ATCC Citrobacter freundii 8090 Edwardsiella tarda Enterobacter aerogenes Enterobacter cloacae Enterococcus faecium Source Bovine Feces - Herring gull cloacal swab, Larus argintatus Nail Infection - Feces, Human - Sputum - Organism/ Reference Number/ Serotype Escherichia coli 1 BEI NR-4356 Escherichia coli 1 BEI NR-12 Hafnia alvei Klebsiella pneumoniae Escherichia coli Clinical Morganella morganii Isolate Culture obtained from Invisible Sentinel Culture Collection, 2 Culture Obtained from the University of Pennsylvania - Kocuria rhizophila 9341 Lactobacillus acidophilus 314 Lactobacillus kefiri Lactobacillus lactis 4797 innocua marthii BAA-1595 welshimeri Source Veriflow LM Soil - Kefir - Cow Brain - Soil - Human - Organism/ Reference Number/ Serotype Proteus mirabilis 7002 Proteus vulgaris 6380 Pseudomonas aeruginosa Salmonella enterica subsp. enterica serovar Abaetuba Salmonella enterica subsp. enterica serovar Dublin Source Veriflow LM Urine - Clinical Isolate - Clinical Isolate - Creek Water, Zaiman, Argentina Dublin - Salmonella enterica subsp. enterica serovar SaintPaul 9712 Shigella sonnei Staphylococcus aureus Staphylococcus epidermidis Staphylococcus haemolyticus Staphylococcus hominis Human Skin - Human Skin - - Page 12 of 25

13 Table 3: Heat Stress Results Matrix Agar CFU Percent Injury TSA 1 1.2x10 9 CFU Deli Turkey MOX 2 6.1x10 8 CFU 50.8% Hot Dogs TSA 1 MOX 2 1.1x10 9 CFU 7.2x10 8 CFU 65.5% 1 TSA: Trypticase soy agar 2 MOX: Modified Oxford Agar Page 13 of 25

14 Inoculum Level Table 2A: MPN Summary Table for Food Matrices Deli Turkey* ATCC Low-Level Inoculum (0.2-2 CFU/Test Portion) Hot Dogs** ATCC % Pasteruized Milk** ATCC A B C D E A B C D E A B C D E 100g Reference Samples /20 9/20 10g MPN/Test portion Low Conf. Limit MPN/Test Portion High Conf. Limit MPN/Test Portion High-Level Inoculum (2-10 CFU/Test Portion) A B C D E A B C D E A B C D E 100g 5 Reference Samples - 5/5 5/5 10g MPN/Test portion Low Conf. Limit/Test Portion High Conf. Limit/Test Portion * MPN was calculated for Deli Turkey using five 100g, five 25g and five 10g and the AOAC LCF MPN Calculator ** MPN was calculated for Hot Dogs and 2% Milk using five 100g, five 10g and the 20 reference method samples and the AOAC LCF MPN Calculator Table 2B: Inoculum Summary Table for Environmental Surfaces Matrix Stainless Steel Ceramic Tile Concrete Plastic Inoculating Organisms ATCC #7644 & Escherichia coli ATCC #25922 ATCC #51778 ATCC #19115 ATCC #13932 Low-Inoculum Level CFU/Test Area* 36 & High-Inoculum Level CFU/Test Area* 190 & *Test area was 4 x 4 for Stainless Steel and Concrete and 1 x 1 for Ceramic Tile and Plastic Page 14 of 25

15 Table 5: Results of the method vs. the USDA/FSIS-MLG Method for Deli Turkey Sample # ATCC #13932 Presumptive Low-Level 0.46CFU/Test Portion Confirmed USDA/FSIS-MLG Total 9/20 9/20 7/20 Un-inoculated 0 CFU/Test Portion Total 0/5 0/5 0/5 High-Level 1.88 CFU/Test Portion Total 5/5 5/5 5/5 Page 15 of 25

16 Table 6: Results of the method vs. the USDA/FSIS-MLG Method for Hot Dogs Sample # ATCC #7644 Presumptive Low-Level 0.53CFU/Test Portion Confirmed USDA/FSIS-MLG Total 5/20 5/20 8/20 Un-inoculated 0 CFU/Test Portion Total 0/5 0/5 0/5 High-Level 4.38 CFU/Test Portion Total 5/5 5/5 5/5 Page 16 of 25

17 Table 7: Results of the method vs. the AOAC Method for 2% Pasteurized Milk Sample # ATCC #19115 Presumptive Low-Level 0.48 CFU/Test Portion Confirmed AOAC Total 11/20 11/20 9/20 Un-inoculated 0 CFU/Test Portion Total 0/5 0/5 0/5 High-Level 4.38 CFU/Test Portion Total 5/5 5/5 5/5 Page 17 of 25

18 Table 8: Results of the method vs. the USDA/FSIS-MLG Method for Stainless Steel Sample # ATCC #7644 Presumptive Low-Level 36 CFU/4 x 4 Confirmed USDA/FSIS-MLG Total 6/20 6/20 7/20 Un-inoculated 0 CFU/4 x Total 0/5 0/5 0/5 High-Level 217 CFU/4 x Total 5/5 5/5 5/5 Page 18 of 25

19 Table 9: Results of the method vs. the USDA/FSIS-MLG Method for Ceramic Tile Sample # ATCC #51778 Presumptive Low-Level 17 CFU/ 1 x 1 Confirmed USDA/FSIS-MLG Total 12/20 13/20 10/20 Un-inoculated 0 CFU/ 1 x Total 0/5 0/5 0/5 High-Level 189 CFU/ 1 x Total 5/5 5/5 5/5 Page 19 of 25

20 Table 10: Results of the method vs. the USDA/FSIS-MLG Method for Concrete Sample # ATCC #19115 Presumptive Low-Level 95 CFU/ 4 x 4 Confirmed USDA/FSIS-MLG Total 7/20 7/20 9/20 Un-inoculated 0 CFU/ 4 x Total 0/5 0/5 0/5 High-Level 560 CFU/ 4 x Total 5/5 5/5 5/5 Page 20 of 25

21 Table 11: Results of the method vs. the USDA/FSIS-MLG Method for Plastic Sample # ATCC #13932 Presumptive Low-Level 33 CFU/ 1 x 1 Confirmed USDA/FSIS-MLG Total 12/20 12/20 7/20 Un-inoculated 0 CFU/ 1 x Total 0/5 0/5 0/5 High-Level 280 CFU/ 1 x Total 5/5 5/5 5/5 Page 21 of 25

22 Table 12: Food Matrices, Presumptive vs. Confirmed POD Results Matrix Deli Turkey Hot Dogs 2% Pasteurized Milk Strain MPN a /Test Portion N b Presumptive Confirmed x c d e POD CP 95% CI x POD CC 95% CI dpod CP f 95% CI g , , , , , , , , , , , , , , , , , , 0.43 a MPN = Most Probable Number is based on the POD of reference method test portions across labs using the AOAC MPN calculator, with 95% confidence interval b N = Number of test potions c x = Number of positive test portions d POD CP = Candidate method presumptive positive outcomes divided by the total number of trials e POD CC = Candidate method confirmed positive outcomes divided by the total number of trials f dpod CP = Difference between the candidate method presumptive result and candidate method confirmed result POD values g 95% CI = If the confidence interval of a dpod does not contain zero, then the difference is statistically significant at the 5% level Page 22 of 25

23 Table 13: Food Matrices, Candidate vs. Reference POD Results Matrix Strain MPN a /25g N b Candidate Reference x c d e POD C 95% CI x POD R 95% CI Deli Turkey Hot Dogs 2% Pasteurized Milk dpod C f 95% CI g , , , , , , , , , , , , , , , , , , 0.43 a MPN = Most Probable Number is based on the POD of reference method test portions across labs using the AOAC MPN calculator, with 95% confidence interval b N = Number of test potions c x = Number of positive test portions d POD C = Candidate method confirmed positive outcomes divided by the total number of trials e POD R = Reference method confirmed positive outcomes divided by the total number of trials f dpod C = Difference between the candidate method confirmed results and candidate method confirmed result POD values g 95% CI = If the confidence interval of a dpod does not contain zero, then the difference is statistically significant at the 5% level Page 23 of 25

24 Table 14: Environmental Surfaces, Presumptive vs. Confirmed POD Results Matrix Strain CFU/Test Area N b Presumptive Confirmed x c d e POD CP 95% CI x POD CC 95% CI Stainless Steel 7644 dpod CP f 95% CI g , , , , , , 0.43 Ceramic Tile Concrete , , , , , , , , , , , , 0.43 Plastic , , , , , , 0.43 a CFU/Test Area = Results of the CFU/Test area were determined by plating the inoculum for each matrix in triplicate b N = Number of test potions c x = Number of positive test portions d POD CP = Candidate method presumptive positive outcomes divided by the total number of trials e POD CC = Candidate method confirmed positive outcomes divided by the total number of trials f dpod CP = Difference between the candidate method presumptive result and candidate method confirmed result POD values g 95% CI = If the confidence interval of a dpod does not contain zero, then the difference is statistically significant at the 5% level Page 24 of 25

25 Table 15: Environmental Surfaces, Candidate vs. Reference POD Results Matrix Strain CFU/Test Area N b Candidate Reference x c d e POD C 95% CI x POD R 95% CI f dpod C 95% CI g Stainless , , , 0.23 Steel , , , 0.43 Ceramic Tile Concrete , , , , , , , , , , , , 0.43 Plastic , , , , , , 0.43 a CFU/Test Area = Results of the CFU/Test area were determined by plating the inoculum for each matrix in triplicate b N = Number of test potions c x = Number of positive test portions d POD C = Candidate method confirmed positive outcomes divided by the total number of trials e POD R = Reference method confirmed positive outcomes divided by the total number of trials f dpod C = Difference between the candidate method confirmed results and candidate method confirmed result POD values g 95% CI = If the confidence interval of a dpod does not contain zero, then the difference is statistically significant at the 5% level Page 25 of 25