Food Protection Trends - May/June 2020 - 208

levels (commonly used in Europe and gaining use in the U.S.)
can be more telling than testing for coliform or generic E. coli.
Wilger noted that although all these tests are good indicators
of hygiene, Enterobacteriaceae testing illuminates the potential
for Salmonella spp. to be present more than do other indicators.
Meeting attendees were interested in whether companies
tested for Listeria spp. or Listeria monocytogenes when conducting environmental monitoring in their manufacturing
facilities that are at risk for Listeria monocytogenes. The FDA

Figure 5. Enterobacteriaceae testing provides information about enteric
pathogens that coliform testing does not (reprinted from Environmental
Microbiology, 3rd Edition; Gerber, C. P.; Indicator Organisms; Figure
23.1; Copyright 2015; with permission from Elsevier).

draft guidance "Control of Listeria monocytogenes in readyto-eat foods" recommends testing for Listeria spp. because
such testing will find L. monocytogenes as well as other, more
common, non-pathogenic species of Listeria. If Listeria species
are found, it indicates the potential for L. monocytogenes to be
present because conditions are suitable for its survival and/
or growth. However, one meeting participant indicated that
their company policy is to test for Listeria monocytogenes when
testing food contact surfaces to facilitate rapid response actions
if a positive is found.
What microbiological test should you use?
When conducting microbiological testing for verification of
PCs, an appropriate, validated method should be used, according to meeting speakers Akins-Lewenthal and Thomas Hammack (FDA). Just because an AOAC method exists doesn't
mean it will work in all food matrices, cautioned Hammack.
The analytical method used should be fit for purpose, which
means it is validated for the sample size and for the microorganism within the food or other matrix that is being tested. It is

208

Food Protection Trends May/June

also important to verify that the method works in the hands of
qualified analysts.
Hammack discussed validation requirements for microbiological methods that might be used as part of verification
activities. Method validation is required for new methods and
whenever a significant modification to an existing method
(such as a new food matrix or larger sample size) is made. If an
assay is modified without proper validation, a negative result
can be questioned. For example, if an assay was previously
validated to detect L. monocytogenes in milk but is then used
(without additional validation) to detect the same organism in
yogurt, there may be a difference between the milk and yogurt
matrices that prevents L. monocytogenes from being detected
in the yogurt. If a negative result is obtained in yogurt, it is not
clear whether L. monocytogenes is truly absent or something
present in the yogurt interfered with the assay. Validation is
done to ensure that negative results truly are negative and that
new methods are equivalent to previously validated reference
methods, to demonstrate that a method is truly fit for purpose,
and to comply with regulatory requirements. In contrast,
verification of microbiological methods involves demonstration that a laboratory can properly perform validated methods
and achieve adequate performance through training records,
proficiency testing, etc.
How should you perform sampling?
Jenny Scott commented that the more hazardous a pathogen and the more likely that it may grow in an RTE food, the
more samples are needed for a testing program and the more
important a positive result becomes. The sampling process
in a microbiological testing program is therefore critical.
However, in contrast to analytical detection methods, which
are well studied, sampling methods are underappreciated and
poorly understood. Nevertheless, according to consultant
and meeting speaker Nancy Thiex, better sampling is often
more effective at reducing overall error than a more expensive
analytical instrument.
Robert Buchanan, professor emeritus at the University of
Maryland, provided a primer on microbiological sampling
plans. Two general classes of sampling plans exist: attribute
(which categorize the test result into "presence" or "absence"
or within defined numerical ranges) and variables (which use
quantitative data directly). Both classes of sampling plans can
be used with either culture-based or "omics"-based testing.
Buchanan also explained the concept of defect rate, and
illustrated how defect rates below 2% mean more samples need
to be tested to avoid missing the defect. Lower defect rates
decrease the cost effectiveness of testing programs and increase
the likelihood that a contaminated lot will be accepted. To
circumvent low limits of detection, it may be necessary to
increase the size of the analytical unit, increase the number of
samples, or enrich/concentrate samples. Regardless, the defect
rates in many foods are well below the practical limits of sampling. That does not mean that microbiological testing is not

Food Protection Trends - May/June 2020

Table of Contents for the Digital Edition of Food Protection Trends - May/June 2020