146 / 453
CONFIDENTIAL INFORMATION
Page 4 of 13
2.5 Analysis of Candidate Method Sub-Portions
After the 14 d stabilization period post-inoculation, the samples were subjected to
the following enrichment procedures and testing on the MDA.
3M™
provided all
3M
TM
consumable supplies involved with the MDA test. In brief, 3M
TM
Buffered
Peptone Water (BPW) was pre-heated to 41.5°C + 1°C. The 25 g sample was
added to 225 mL 3M
TM
BPW, gently agitated and incubated at 41.5°C + 1°C for 18
h. The mixing of the enrichment broth and blueberries was performed gently with
efforts to limit any bursting of blueberries. The enriched samples were then subject
to analysis as outlined by the instructions for use for the 3M
TM
Molecular Detection
Assay 2
–
E. coli
O157 test kit. The sample results obtained after completing the
MDA wer
e considered the ‘screening stage’ result for the candidate method.
Concurrent with the transfer and testing on the MDA, aliquots of the enriched
samples were transferred and subjected to the testing outlined in Section 2.4 for
cultural confirmation. The results obtained after completing this methodology were
considered the ‘confirmation stage’ result for the candidate method.
2.6 Most Probable Number Analysis
As mentioned above in Section 2.3, samples of each portion were inoculated with
two different levels of
E. coli
O157 (i.e. Portion 1, and Portion 2) and used for
performing a five tube-three level MPN analysis to determine the concentration of
E. coli
O157 inoculated into the portion. This analysis was setup after the 14 d
stabilization period in order to mimic all handling and treatment of matrices. From
each 300 g sample of each portion, five 50 g samples and five 10 g samples were
weighed into sterile Whirl-Pak bags and hydrated with 425 mL and 90 mL of
mBPWp, respectively. Each sample was then processed according to the
reference method procedures described above in Section 2.4. Note: for the MPN
test samples, the set of five samples (high inoculum) and twenty samples (low
inoculum) for the reference testing were also used (25 g, respectively). Samples
that confirm positive for this procedure were noted and the LCF MPN Calculator
Version 1.6 (Least Cost Formulations, Ltd., Virginia Beach, VA) was used to
calculate the MPN/g and 95% confidence intervals for each portion using the mass
of each sample and the number of reference method confirmed positives at each
level of the MPN setup.
2.7 Probability of Detection Statistical Analysis
After results from the “screening” and “confirmation” stages
were obtained for the
candidate method and “confirmed” results
were obtained for the reference method,
the data was statistically analyzed according to the POD methods of the 2012
version of the AOAC International Methods Committee Guidelines for Validation of
Microbiological Methods for Food and Environmental Surfaces document (1). The
POD for the candidate method’s “screening stage” (presumptive) results (POD
CP
)
were calculated for each level of inoculation, and were compared with the POD for