76 / 453
12
intervals, were 0.29 CFU/test portion(0.09,0.55) for the low inoculum level and 1.92 CFU/test portion (1.00, 3.42)
1
for the high inoculum level.
2
For the low inoculum level, 32 out of 120 test portions (POD
CP
of 0.27) were reported as presumptive positive
3
by the 3M MDA 2 –
E. coli
O157 (including H7)method with 31 test portions (POD
CC
of 0.26) confirming
4
positive. For samples that produced presumptive positive results on the 3M MDA 2 –
E. coli
O157 (including H7)
5
method, 29 samples confirmed positive (POD
C
of 0.24). For test portions evaluated by the USDA/FSIS MLG
6
reference method, 42 out of 120 test portions produced positive results (POD
R
of 0.35). A dLPOD
C
value of -0.11
7
with 95% confidence intervals of (-0.22, 0.01) were obtained between the candidate and reference method,
8
indicating no statistical significant difference between the two methods. A dLPOD
CP
value of 0.01 with 95%
9
confidence intervals of (-0.10, 0.12) were obtained between presumptive and confirmed results indicating no
0
statistically significant difference between the presumptive and confirmed results.
1
For the high inoculum level, 97 out of 120 test portions (POD
CP
of 0.81) were reported as presumptive positive
2
by the 3M MDA 2 –
E. coli O157 (including H7)
method with 96test portions (POD
CC
of 0.80) confirming
3
positive.For samples that produced presumptive positive results on the 3M MDA 2 –
E. coli
O157 (including
4
H7)method, 92 samples confirmed positive (POD
C
of 0.77). For test portions evaluated by the USDA/FSIS MLG
5
reference method, 95 out of 120 test portions produced positive results (POD
R
of 0.79). A dLPOD
C
value of -0.025
6
with 95% confidence intervals of (-0.129, 0.080) wasobtained between the candidate and reference method,
7
indicating no statistical significant difference between the two methods. A dLPOD
CP
value of 0.008 with 95%
8
confidence intervals of (-0.092, 0.109) wasobtained between presumptive and confirmed results indicating no
9
statistically significant difference between the presumptive and confirmed results.
0
For the un-inoculated controls, zero out of 120 samples (POD
CP
of 0.00) produced a presumptive positive result by
1
the 3M MDA 2 -
E. coli
O157 (including H7) method with 2test portions (POD
CC
of 0.02) confirming positive. For
2
samples that produced presumptive positive results on the 3M MDA 2 –
E. coli
O157 (including H7)method,
3
zerosamples confirmed positive (POD
C
of 0.00). For test portions evaluated by the USDA/FSIS MLG reference
4
method, 3 out of 120 test portions produced positive results. Colonies isolated from the three test portions were
5
further characterized utilizing partial sequencing of the 16S rRNA gene; isolates were not identified as
E. coli
6
O157:H7. A dLPOD
C
value of 0.017with 95% confidence intervals of (-0.059, 0.017) wasobtained between the
7
confirmed candidate and reference method, indicating no statistical significant difference between the two
8
methods. A dLPOD
CP
value of 0.00 with 95% confidence intervals of (-0.031, 0.031) wasobtained between
9
presumptive and confirmed results indicating no statistically significant difference between the presumptive and
0
confirmed results. POD statistical analysis for the un-inoculated test samples were conducted on the revised
1
results (See Table 2016.2A) from laboratory 11 and 12 after 16S rRNA sequencing.
2
3
Detailed results of the POD statistical analysis are presented in Table 2016.2A and Figures 1A-1B.
4
5
Discussion
6
7
While three laboratories experienced technical issues during the evaluation, no negative feedback was provided by
8
the collaborating laboratories in regard to the performance of the 3M MDA 2-
E. coli
O157 (including H7)
9
method.During the evaluation, three laboratories culturally confirmed
E. coli
O157:H7in their (blinded) un-
0
inoculated control samples, see Table 2016.2A.The two test portions from Laboratory 5 were unavailable for
1
further identification as the collaborative study plates had been disposed.However, laboratories 11 and 12
2
submitted their samples directly to a third party laboratory capable of performing 16S rRNA sequencing for
3
additional identification.The inoculating strain
E. coli
O157:H7 ATCC 43895 was also analyzed as the control.
4
Theprotocol and results for 16S rRNA sequencing can be found in Supplementary Material. Utilizing the Basic
5
Local Alignment Search Tool (BLAST
®
) for nucleotides [10], the isolates were determined to 1) not be the
6
inoculating strain
E. coli
O157:H7 ATCC 43895 and 2) not
E. coli
. Serological cross reactivity of other members
7
of the microbial community for raw beef products has been documented due to the close relationship among
8
species [11, 12]. This indicates that the 3M Molecular Detection Assay 2 –
E. coli
O157 (including H7) is more
9
OMAMAN-35 A : Collaborative Study Manuscript
For ERP Use Only
January 2017
AOAC Research Institute
Expert Review Pa el Use Only