12 / 129
B
ird
et al
.:
J
ournal of
AOAC I
nternational
V
ol
.
96, N
o
. 4, 2013
813
the SPR. The test value is calculated by the instrument and is
equal to the difference between the background reading and
the final reading. The calculation appears on the result sheet. A
negative result has a test value less than the threshold (0.25) and
indicates that the sample does not contain
Salmonella
spp. or
contains
Salmonella
spp. at a concentration below the detection
limit. A positive result has a test value equal to or greater than
the threshold (≥0.25) and indicates that the sample may be
contaminated with
Salmonella
spp. If the background reading
is above a predetermined cutoff, then the result is reported as
invalid (Table
2012.01D
).
G. Confirmation
All positive VIDAS SPT results must be culturally
confirmed. Confirmation should be performed using the non-
heated enrichment broth stored between 2 and 8°C, and should
be initiated within 72 h after the end of incubation at 42 ± 1°C.
Presumptive positive results may be confirmed by isolating
on selective agar plates such as IBISA or ASAP, or on the
appropriate reference method selective agar plates. Typical or
suspect colonies from each plate are confirmed as described
in AOAC
Official Method
967.27
. As an alternative to the
conventional tube system for
Salmonella
, any AOAC-approved
commercial biochemical kits may be used for presumptive
generic identification of foodborne
Salmonella
as described in
AOAC
Official Methods
978.24
,
989.12
,
991.13
, and
2011.17
.
Results of Collaborative Study
In this collaborative study, the VIDAS SPT method was
compared to the to the USDA/FSIS-MLG reference method for
one food product, raw ground beef, at two different test portion
sizes, 25 and 375 g. A total of 15 laboratories throughout the
United States participated in this study, with 14 submitting
data for each matrix, as presented in Table 1. Each laboratory
analyzed 36 test portions for each method: 12 inoculated with
a high level of
Salmonella
, 12 inoculated with a low level, and
12 uninoculated controls. For each test portion size, the actual
level of
Salmonella
was determined by MPN determination on
the day of initiation of analysis. Individual laboratory and sample
results are presented in Tables 2–5. Tables
2013.01A
and
B
summarize the interlaboratory results for all foods tested,
including POD statistical analysis (6). Detailed results for each
laboratory are presented in Tables A–F of the Appendix.
Raw Ground Beef (25 g Test Portions)
Raw ground beef test portions were inoculated at a low and
high level, and analyzed (Tables 2 and 3) for the detection of
Salmonella
spp. Uninoculated controls were included in each
analysis. Fourteen laboratories participated in the analysis of
this matrix, and the results of 12 were included in the statistical
analysis. Laboratory 8 reported that it was unable to confirm
samples via serological testing and indicated that it did not
conduct the alternative confirmation of the VIDAS SPT samples.
Therefore, its results were not included in statistical analysis.
Laboratory 12 produced a low-level presumptive positive result
for one of its uninoculated control test portions, which could
not be confirmed positive by the traditional reference method.
Therefore, its results were not included in the statistical analysis.
The MPNs obtained for this matrix, with 95% confidence
intervals, were 1.10 CFU/test portion (0.49, 2.46) for the low
inoculum level and 4.38 CFU/test portion (1.71, 11.20) for the
high inoculum level.
Traditional Confirmation with Xylose-Lysine-Tergitol 4
(XLT4) and Brilliant Green Sulfa (BGS)
For the high inoculum level, all of the 144 test portions were
reported positive by the VIDAS SPT method, with all portions
confirming positive. For the low inoculum level, all 144 test
portions were also reported as positive by the VIDAS SPT
method, with 143 confirming positive, indicating one false
unconfirmed positive result (Laboratory 6). For the uninoculated
controls, none of the 144 samples produced a presumptive
positive result by the VIDAS SPT method, and all samples
confirmed negative. For test portions analyzed by the USDA/
FSIS-MLG method, 138 out of 144 high and 84 out of 144 low
inoculum test portions confirmed positive. For the uninoculated
controls, none of the 144 test portions confirmed positive.
For the low-level inoculum, a dLPOD
C
value of 0.41 (+0.32,
+0.49) was obtained between the USDA/FSIS-MLG method
and the VIDAS SPT method. The confidence intervals obtained
for dLPOD
C
indicated a statistically significant difference
between the two methods. However, the VIDAS SPT method
detected more positive samples than the USDA/FSIS-MLG
reference method, indicating a higher level of sensitivity than
the reference method. A dLPOD
CP
of 0.01 (–0.02, +0.04)
was obtained between presumptive and confirmed VIDAS
SPT results for both confirmation procedures. The confidence
intervals obtained for dLPOD
CP
indicated no significant
difference between the presumptive and confirmed results.
For the high-level inoculum, a dLPOD
C
value of 0.04 (+0.01,
+0.09) was obtained between the USDA/FSIS-MLG method
and the VIDAS SPT method. The confidence intervals obtained
for dLPOD
C
indicated a statistically significant difference
between the two methods. However, the VIDAS SPT method
detected more positive samples than the USDA/FSIS-MLG
reference method, indicating a higher level of sensitivity than
the reference method. A dLPOD
CP
of 0.00 (–0.03, +0.03)
was obtained between presumptive and confirmed VIDAS
SPT results. The confidence intervals obtained for dLPOD
CP
indicated no significant difference between the presumptive
and confirmed results. Results of the POD statistical analysis
are presented in Table
2013.01A
, and in appended Table A and
Figure 1A and B.
Traditional Confirmation with IBISA and ASAP
For the high inoculum level, all 144 test portions were reported
as positive by the VIDAS SPT method, with all confirming
positive. For the low inoculum level, all 144 test portions were
also reported as positive by the VIDAS SPT method, with 143
confirming positive. For the uninoculated controls, none of
the 144 samples produced a presumptive positive result by the
VIDAS SPT method with all samples confirming negative. For
test portions analyzed by the USDA/FSIS-MLG method, 138
of the 144 high inoculum test portions and 84 out of 144 low
inoculum test portions confirmed positive. For the uninoculated
controls, none of the 144 test portions confirmed positive.
For the low-level inoculum, a dLPOD
C
value of 0.41 (+0.32,
9