IJCA - Volume 3 - Flipbook - Page 64
62 The International Journal of Conformity Assessment
STEP 4: (FOOD) ITEM VERIFICATION USING eBIAS
(Food) item veri昀椀cation for quantitative methods
is achieved by determining the estimated Bias
(eBias), which is another unique de昀椀nition within this
standard. An accurate determination of the bias (as
conducted in validation studies) is not feasible due to
the small number of samples tested in the veri昀椀cation
study. Therefore, the term eBias is utilized for the
veri昀椀cation study.
To conduct (food) item veri昀椀cation, the laboratory
will use the three challenging (food) items and one
non-food item (swabs with Letheen broth) chosen
from the review of the categories in the scope of the
laboratory application.
Each (food) item will be arti昀椀cially contaminated at
three inoculation levels that cover the range of use of
the method as it is routinely used by the laboratory
(for example: 30-300, 300-3,000 and 3,000-30,000
cfu/g). Each of these three levels will be performed in
duplicate.
A culture of Enterobacteriaceae is grown overnight,
and then serial dilutions are made and plated to
determine the correct inoculation levels. When
diluting the inoculation suspension to prepare for
inoculation in the (food) item duplicates, consider
including additional dilutions that may be needed
to achieve counts within the countable range of the
method for each of the three levels. As illustrated
in Figure 7, the inoculum requires further dilution to
achieve the correct dilution/counting levels for the
high, intermediate, and low levels. The inoculum,
when mixed with the volume of the (food) item,
requires less dilution.
To calculate the eBias, enumeration is conducted and
recorded for:
• The inoculum at three levels.
• The (food) item with inoculum at three levels.
• The (food) item without inoculum in duplicate (as
a negative control) to determine the background
microbiota level (if any) in the (food) item.
For each of the three levels, the counts of the
(food) item duplicates are averaged and a log10
transformation is done to determine log10 cfu/g
for each level. These results are then expressed in
log10 cfu/test portion for each level and compared
to the log transformation on the count of that same
inoculum level determined for each level without the
(food) item. The eBias is the absolute difference in
results between the inoculated (food) item and the
inoculum.
To meet the Acceptability Limits for eBias per the
standard (found in Table 16), the absolute difference
for each level must be ≤ 0.5 log10/ml. Again, the
Excel® -based program (workbook) is also available on
the ISO TC34/SC9 website for you to insert your data
and help complete all calculations.
Figure 7. Quantitative (food) item veri昀椀cation protocol with three levels of inoculum
