Food Protection Trends - May/June 2023 - 237

an option for establishments to abide by enforced final
product standards for select frozen and breaded raw chicken
products (on the basis of testing five 25-g samples for any
Salmonella spp. per lot) in lieu of the options of having
validated processing controls or testing raw chicken mixture
inputs. These Canadian standards specify that Salmonellapositive
product lots of select frozen and breaded raw chicken
products cannot enter the market in the current form (5).
Risk models will require data to estimate the efficacy of
possible regulatory standards in improving public health,
considering factors such as the portion of the industry
able to meet the standard, verification efficacy, regulatory
consequences for failing standards, and ancillary benefits
associated with standards. For example, a regulatory
consequence of the unenforceable performance standards
in the United States was that USDA posted on its website
the identities of poultry processing establishments that
failed the standards. An ancillary benefit was that in the
2 years following the agency's first publication of these
establishment-specific listings in 2006, detected Salmonella
rates in poultry dropped in half (34), presumably, in part,
because establishments face financial consequences from
buyers that prefer to purchase from compliant firms.
Regulatory standards that have previously been implemented
could have associated data sets to assist in model
development. For example, the European Union collects
Salmonella poultry industry data and human illness data from
every member country yearly (10, 11). Recently, developed
standards may have data sets allowing for a time interrupted
analysis, better allowing for parameters to be estimated when
evaluating similar standards.
Modeling the public health impact of regulatory standards
at different stages of poultry production could require different
considerations. For all standards, food supply impact, industry
impact, cost, and feasibility should be evaluated. For processing
standards, the baseline scenario should be the current
U.S. system of performance standards for processing establishments.
Modeling modifications to these standards should
include simulations of the public heath impact of risk-based
classification of Salmonella serovar or subtypes, as described by
Cohn et al. (8), as well as quantitative standards, as described
by Lambertini et al. (22). These standards should also be
modeled as both enforced and unenforced. Models should also
include the ability to assess the public health impact of final
product standards for individual products or categories, such as
comminuted poultry, parts, or breaded products, considering
both consumption volume of different categories and differences
in exposure associated with different product categories
(e.g., different contamination frequencies and loads). The
effectiveness of different testing methodologies and feasibility
of verifying compliance with these product standards should
also be included in models. Our recommendations for the minimum
regulatory risk management options to evaluate in risk
assessments and models are further outlined in Table 1.
CONCLUSIONS
Despite ongoing efforts to decrease the public health
burden of Salmonella in the United States, salmonellosis
incidence remains high, with poultry implicated as a major
source. To implement a risk-based approach to decrease
the public health burden of Salmonella in poultry, a
multifaceted approach is necessary to properly assess the
risks associated with poultry. In this article, we describe
specific data and model needs that will help to inform future
risk assessments to allow for selection and implementation
of food safety policies that will have a quantifiable positive
public health impact, and more specifically, would help
the United States achieve the Healthy People 2030 goal
regarding human salmonellosis cases. Risk assessments
should assess a wide range of options, and ultimately, risk
managers must determine which options are feasible and
best suited to reduce illness. There is good reason to hope
that these determinations will improve as risk assessments
evolve and meet the data and modeling needs identified in
this article. However, agreement on an acceptable level of
risk (also referred to as residual risk (39)) may be difficult
to achieve, for example, due to the difficulty of comparing
relative public health benefits and unintended consequences
of different interventions; this will likely represent a
continued challenge for the implementation of risk-based
approaches to controlling Salmonella in raw poultry. Indeed,
an acceptable level of risk may differ according to the relevant
legal authority (e.g., Administrative Procedure Act, tort law).
Furthermore, although we describe a wide range of data
and model needs, it is important that risk models and risk
assessments are conducted in a timely fashion and regularly
updated as further data or model improvements become
available and feasible. Finally, ongoing complementary
projects by Food and Agriculture Organization of the United
Nations and World Health Organization (13) and the
National Advisory Committee on Microbiological Criteria
for Foods (36) will produce future reports that will provide
additional insights and consensus opinions on how to
minimize Salmonella transmission through raw poultry.
ACKNOWLEDGMENTS
We thank members of the Coalition for Poultry Safety
Reform for helpful discussions and feedback on this
manuscript.
May/June Food Protection Trends 237
Food Protection Trends - May/June 2023

Table of Contents for the Digital Edition of Food Protection Trends - May/June 2023
