Persistencia de Peligros Biológicos

Adopted: 6 December 2023
DOI: 10.2903/j.efsa.2024.8521
SCIENTIFIC OPINION
Persistence of microbiological hazards in food and feed

production and processing environments
EFSA Panel on Biological Hazards (BIOHAZ) | Konstantinos Koutsoumanis | Ana Allende |

Declan Bolton | Sara Bover-Cid | Marianne Chemaly | Alessandra De Cesare | Lieve Herman |
Friederike Hilbert | Roland Lindqvist | Maarten Nauta | Romolo Nonno | Luisa Peixe |
Giuseppe Ru | Marion Simmons | Panagiotis Skandamis | Elisabetta Suffredini | Edward Fox |
Rebecca (Becky) Gosling | Beatriz Melero Gil | Trond Møretrø | Beatrix Stessl |
Maria Teresa da Silva Felício | Winy Messens | Ancuta Cezara Simon | Avelino Alvarez-Ordóñez
Correspondence: biohaz@efsa.europa.eu
Abstract
Listeria monocytogenes (in the meat, fish and seafood, dairy and fruit and vege-
table sectors), Salmonella enterica (in the feed, meat, egg and low moisture food
sectors) and Cronobacter sakazakii (in the low moisture food sector) were identi-
fied as the bacterial food safety hazards most relevant to public health that are
associated with persistence in the food and feed processing environment (FFPE).
There is a wide range of subtypes of these hazards involved in persistence in the
FFPE. While some specific subtypes are more commonly reported as persistent, it
is currently not possible to identify universal markers (i.e. genetic determinants)
for this trait. Common risk factors for persistence in the FFPE are inadequate zon-
ing and hygiene barriers; lack of hygienic design of equipment and machines; and
inadequate cleaning and disinfection. A well-designed environmental sampling
and testing programme is the most effective strategy to identify contamination
sources and detect potentially persistent hazards. The establishment of hygienic
barriers and measures within the food safety management system, during imple-
mentation of hazard analysis and critical control points, is key to prevent and/or
control bacterial persistence in the FFPE. Once persistence is suspected in a plant,
a ‘seek-and-destroy’ approach is frequently recommended, including intensified
monitoring, the introduction of control measures and the continuation of the
intensified monitoring. Successful actions triggered by persistence of L. mono-
cytogenes are described, as well as interventions with direct bactericidal activity.
These interventions could be efficient if properly validated, correctly applied and
verified under industrial conditions. Perspectives are provided for performing
a risk assessment for relevant combinations of hazard and food sector to assess
the relative public health risk that can be associated with persistence, based on
bottom-up and top-down approaches. Knowledge gaps related to bacterial food
safety hazards associated with persistence in the FFPE and priorities for future re-
search are provided.
KEYWORDS
cleaning and disinfection, Cronobacter sakazakii, food processing, interventions, Listeria
monocytogenes, Persistence, risk factors, Salmonella enterica, subtypes
This is an open access article under the terms of the Creative Commons Attribution-NoDerivs License, which permits use and distribution in any medium, provided the
original work is properly cited and no modifications or adaptations are made.
© 2024 European Food Safety Authority. EFSA Journal published by Wiley-VCH GmbH on behalf of European Food Safety Authority.
EFSA Journal. 2024;22:e8521. efsa.onlinelibrary.wiley.com/journal/1831-4732 | 1 of 114

https://doi.org/10.2903/j.efsa.2024.8521
18314732, 2024, 1, Downloaded from https://efsa.onlinelibrary.wiley.com/doi/10.2903/j.efsa.2024.8521 by CochraneArgentina, Wiley Online Library on [22/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
2 of 114 | PERSISTENCE OF MICROBIOLOGICAL HAZARDS IN FOOD AND FEED ENVIRONMENTS
CO NTE NT S
Abstract��1
Summary��5
1. Introduction��7
1.1. Background and Terms of Reference as provided by the requestor��7
1.2. Interpretation of the Terms of Reference��8
1.3. Additional information�� 10
2. Data and Methodologies�� 10
2.1. Most relevant bacterial food safety hazards associated with persistence in the FFPE (AQ1)�� 10
2.1.1. Most relevant bacterial pathogens of public health relevance in the various food and feed
production and processing sectors in the EU/EEA�� 10
2.1.1.1. Food production and processing sectors�� 10
2.1.1.2. Feed production and processing sectors��11
2.1.2. Persistence of bacterial pathogens of public health relevance in the FFPE of the various sectors��11
2.1.2.1. Food production and processing sectors��11
2.1.2.2. Feed production and processing sectors�� 12
2.2. Main (sub)types of the most relevant bacterial hazards involved in persistence and the main features
responsible for their persistence in the FFPE (AQ2 and 3)�� 12
2.3. Factors at facility level that increase the likelihood of persistence of the most relevant bacterial hazards in the
FFPE (AQ4)�� 13
2.4. Measures for monitoring, preventing and/or controlling the persistence of the most relevant bacterial hazards
in the FFPE (AQ5)�� 13
2.5. Perspectives of integrating the information gathered in risk assessment (AQ6)�� 13
2.6. Knowledge gaps and priorities for future research (AQ7 and 8)�� 14
2.7. Uncertainty analysis�� 14
3. Assessment�� 14
3.1. Most relevant bacterial food safety hazards associated with persistence in the FFPE (AQ1)�� 14
3.1.1. Most relevant bacterial pathogens of public health relevance in the various food and feed
production and processing sectors in the EU/EEA�� 14
3.1.2. Persistence of bacterial pathogens of public health relevance in the FFPE of the various sectors�� 15
3.1.2.1. Pathogens able to persist in the processing environment of the feed for food animal
production sector�� 15
3.1.2.2. Pathogens able to persist in the processing environment of the meat sector�� 16
3.1.2.3. Pathogens able to persist in the processing environment of the fish and seafood sector�� 16
3.1.2.4. Pathogens able to persist in the processing environment of the dairy sector�� 17
3.1.2.5. Pathogens able to persist in the processing environment of the egg sector�� 18
3.1.2.6. Pathogens able to persist in the processing environment of the fruit and vegetable sector�� 18
3.1.2.7. Pathogens able to persist in the processing environment of the LMF sector�� 19
3.1.3. Concluding remarks related to the most relevant bacterial food safety hazards associated with
persistence�� 19
3.2. Main (sub)types of the most relevant bacterial hazards involved in persistence and the main features
responsible for their persistence in the FFPE (AQ2 and 3)�� 20
3.2.1. L. monocytogenes subtypes and features�� 20
3.2.1.1. Subtypes linked to persistence�� 20
3.2.1.2. Features associated with persistence and link to subtypes�� 22
3.2.1.3. Analysis of clusters of related genome sequences in the NCBI Pathogen Detection database�� 25
3.2.2. Salmonella enterica subtypes and features�� 30
PERSISTENCE OF MICROBIOLOGICAL HAZARDS IN FOOD AND FEED ENVIRONMENTS | 3 of 114
3.2.3. C. sakazakii subtypes and features�� 33
3.2.4. Concluding remarks related to (sub)types and features�� 35
3.3. Factors at facility level that increase the likelihood of persistence of the most relevant bacterial hazards in the
FFPE (AQ4)�� 36
3.3.1. L. monocytogenes in FoPE of four sectors�� 36
3.3.1.1. Environmental niches or site of persistent L. monocytogenes strains�� 36
3.3.1.2. Risk factors for persistence of L. monocytogenes��37
3.3.2. Salmonella enterica in the FFPE of four sectors�� 39
3.3.2.1. Environmental niches or location of persistent Salmonella enterica strains�� 39
3.3.2.2. Risk factors for persistence of Salmonella enterica�� 40
3.3.3. C. sakazakii in the FoPE of LMF sector�� 41
3.3.3.1. Environmental niches or location of persistent C. sakazakii strains�� 41
3.3.3.2. Risk factors for persistence of C. sakazakii�� 42
3.3.4. Concluding remarks related to factors at facility level that increase the likelihood of persistence�� 42
3.4. Measures for monitoring, preventing and/or controlling the persistence of the most relevant bacterial hazards
in the FFPE (AQ5)�� 43
3.4.1. Sampling and testing in the FFPE��44
3.4.2. Hygienic measures�� 45
3.4.2.1. PRP infrastructure (building, equipment)�� 45
3.4.2.2. PRP cleaning and disinfection�� 45
3.4.2.3. PRP technical maintenance and calibration��46
3.4.2.4. PRP water and air control��46
3.4.2.5. PRP personnel (hygiene, health status)�� 47
3.4.2.6. PRP working methodology�� 47
3.4.2.7. PRP food safety culture (FSC)�� 47
3.4.3. Actions triggered by persistence�� 47
3.4.3.1. Chemical interventions�� 51
3.4.3.2. Physical interventions�� 51
3.4.3.3. Biological interventions�� 52
3.4.4. Concluding remarks related to measures for monitoring, preventing and/or controlling persistence�� 52
3.5. Perspectives of integrating the information gathered in risk assessment (AQ6)�� 53
3.5.1. Bottom-up approach and data needs�� 53
3.5.2. Top-down approach and data needs�� 57
3.5.3. Concluding remarks related to perspectives of integrating the information gathered in risk
assessment�� 58
3.6. Knowledge gaps and priorities for future research (AQ7 and 8)�� 58
4. Conclusions�� 59
5. Recommendations�� 61
Glossary�� 62
Abbreviations�� 63
Acknowledgements�� 65
Conflict of interest�� 65
Requestor�� 65
Question number�� 65
Panel members�� 65
Copyright for non-EFSA content�� 65

Map disclaimer�� 65
References�� 65
Appendix A�� 83
Appendix B�� 87
Appendix C�� 89
Appendix D��94
Appendix E�� 97
Annex A�� 114
SUM M ARY
The European Food Safety authority (EFSA) asked the Panel on Biological Hazards (BIOHAZ) to deliver a scientific opinion
on the persistence of microbiological hazards in food and feed production and processing environments (FFPE), excluding
primary production. In the scope of this mandate, microbial ‘persistence’ was defined as the ability of a given organism to
be established in niches (or harbourage sites) within the FFPE for a long term, despite the frequent application of clean-
ing and disinfection (C&D). It requires prolonged existence (spanning months or years) usually with multiplication of the
microorganism in the specific FFPE. Feed-producing environments (FePE) were restricted to the environments of facilities
producing and processing feed for food-producing animals. Food-producing environments (FoPE) cover environments
where food of animal or non-animal origin is industrially produced or processed at post-harvest level. The sectors consid-
ered were: (i) feed for food animal production, (ii) meat (incl. slaughterhouses and processing plants), (iii) fish and seafood,
(iv) dairy, (v) egg and egg products, (vi) fruit and vegetable (including herbs) and (vii) low moisture food (LMF).
In Term of Reference 1 (ToR1), EFSA was requested to identify the most relevant microbiological food safety hazards
associated with persistence in the FFPE of these sectors in the EU/EEA. Based on the definition of persistence, viruses and
parasitic protozoa were excluded because, due to their inability to multiply on abiotic surfaces, they cannot become estab-
lished for a long term or constitute a contamination reservoir in the FFPE. Likewise, microbial toxins and other hazardous
microbial metabolites were excluded. The most relevant bacterial food safety hazards were identified as: Salmonella en-
terica in the feed for food animal production sector; Listeria monocytogenes and S. enterica in the meat processing sector;
L. monocytogenes in the fish and seafood processing sector; L. monocytogenes in the dairy sector; S. enterica in the eggs
and egg processing sector; L. monocytogenes in the fruit and vegetables processing sector; and S. enterica and Cronobacter
sakazakii in the LMF sector. Other bacterial hazards were either not of highest public health (PH) relevance in the specific
sector or were not considered as most relevant food safety hazards associated with persistence in the FFPE in the specified
sector based on the available information.
In ToR2, EFSA was requested to identify the main (sub)types of the most relevant hazards involved in persistence and the
main features responsible for their persistence in the FFPE. It was concluded that there is a wide range of subtypes reported
to be involved in persistence in the FFPE for the three most relevant hazards listed above. Some specific subtypes are more
commonly reported as persistent: for L. monocytogenes, especially CC 121, CC8, CC9 from lineage II and CC 5, CC6, CC2 from
lineage I; for S. enterica, S. Typhimurium and S. Agona; and also S. Derby, S. Anatum, S. Infantis, S. Heidelberg, S. Mbandaka
and S. Senftenberg; and for C. sakazakii, CC64, CC1, CC83 and CC4. For L. monocytogenes, some markers have been iden-
tified as possibly associated with persistence: stress survival islets SSI-1 and SSI-2, genomic islands LGI-1 and LGI-2, heavy
metal (cadmium and arsenic) and biocide (bcrABC, qacC, qacH, emrE and emrC) resistance determinants, often located on
mobile genetic elements (mainly plasmids) and bacteriophage regions (comK), globally linked to increased environmental
robustness, tolerance to disinfection and/or biofilm formation. The set of phenotypic and genomic features that have been
investigated for Salmonella and C. sakazakii in relation to persistence in the FFPE is incomplete. As such, it is difficult to de-
duce certain features, that are either indispensable for or may markedly contribute to, persistence, alone or in combination
with other key genotypic and phenotypic elements. For Salmonella, most studies focused on features inherent to most
infectious foodborne hazards, and reported resistance of some strains to one or more antimicrobials, carriage of plasmid-
mediated virulence factors, biofilm formation ability or reduced susceptibility to alkaline disinfectants. Several features
have been associated with the ability of C. sakazakii to survive for long time periods and persist in the dry conditions of
the LMF FoPE, including the ability to form biofilms on a variety of abiotic surfaces; a high heat tolerance and desiccation
resistance; the production of a capsule that aids attachment and provides resistance to biocides and desiccation; and the
production of a yellow carotenoid pigment which stabilises cell membranes and protects against stress. However, none
of these features seem to be specifically linked to particular subtypes frequently associated with persistence. Overall, no
universal markers or features, responsible for persistence have been identified. Although the carriage of different combi-
nations of genetic determinants linked to increased environmental robustness possibly confers the ability to persist on
particular subtypes, persistence is a multifactorial process that also depends on specific environmental conditions and risk
factors.
In ToR3, EFSA was requested to identify the risk factors i.e. those factors at facility level that increase the likelihood of
persistence of the food safety hazards in the FFPE. The main risk factor of the three bacterial hazards listed above in the
FFPE is poor hygienic design of equipment and machines. This leads to niches (or harbourage sites) which are difficult to
clean and disinfect and where food debris and moisture can accumulate, and the hazards can survive and persist. Other
important factors are: (i) inadequate zoning and hygiene barriers, that enables the spread of contamination from contami-
nated to clean areas; (ii) inadequate C&D of the facilities; (iii) introduction of the hazards through raw materials, which may
lead to the colonisation and spread of persistent clones in the processing environment; and (iv) humidity, which favour
persistence. Specifically for hazards of relevance in dry (LMF/feed) processing environments, additional risk factors are
airborne transmission through dust, the limited use of disinfectants due to dry cleaning operations or the presence of
water in the FoPE, whether from wet cleaning, condensation generated through temperature gradients within the facility
or within equipment, or other sources.
In ToR4, EFSA was requested to assess available and enhanced measures or interventions for monitoring, preventing
and/or controlling the persistence of the most relevant microbiological food safety hazards in the FFPE. It was concluded
that a well-designed environmental sampling and testing programme, following a risk-based approach, is the most effec-
tive strategy to identify contamination sources and detect potentially persistent hazards. The establishment of hygienic
barriers and measures within the food safety management systems (FSMS), during implementation of hazard analysis and
critical control point (HACCP), is key to prevent and/or control bacterial persistence in the FFPE through avoiding the entry
of the hazard(s) to the processing plant and/or their spread across the facility. The following prerequisites are of particu-
lar importance: infrastructure (building, equipment), C&D, technical maintenance and calibration, water and air control,
personnel (hygiene, health status), working methodology and food safety culture. The confirmation of the presence of a
persistent strain and identification of its niche within the facility requires the detailed characterisation of isolates of the
specific hazard(s) recovered from positive samples using subtyping methods with enough resolution, preferably whole
genome sequencing (WGS). Once persistence is suspected in a plant, a ‘seek-and-destroy’ approach has been frequently
recommended, which includes: (i) intensified monitoring; (ii) the introduction of measures to control the event; and (iii)
the continuation of the intensified monitoring programme to confirm the efficacy of the measures taken or to identify the
requirement for additional measures. Alternatively, systematic ‘root cause analyses’ can be applied to identify the most
probable factors/sites within the facilities contributing to the problem and define the most appropriate interventions to
eliminate the pathogen from the premises. Successful actions triggered by L. monocytogenes persistence in the FoPE were
identified, for example, the introduction of new or specialised (deep) C&D, the implementation of workflows, the installa-
tion of a new drainage system; the implementation of structural changes and renovations; the control of the contamina-
tion of raw ingredients and the improvement of the compartmentalisation, or the simultaneous implementation of various
corrective actions. In addition, some options of interventions to eliminate the persistent hazard(s) with direct bactericidal
activity and of different nature (i.e. as chemical (e.g. biocides), physical (e.g. heat or novel non-thermal technologies) or
biological (e.g. competitive exclusion, phage)) are described but in some cases these are not yet commercially available
and/or their efficacy is not yet fully validated under industrial conditions.
In ToR5, EFSA was requested to identify knowledge gaps and priorities for future research and develop the perspectives
(or future opportunities) of integrating the information gathered in the previous ToRs in risk assessment. Perspectives are
provided for the use of risk assessment for relevant combinations of hazard and food product to assess the relative PH
risks that can be associated with persistence, based on bottom-up (or forward) and top-down (or backward) approaches.
A basic model for persistence, proposed to be used in bottom-up food chain QMRA, can be used to study the role of
persistence in the PH risk for a specific food production process. The dynamics of persistence and its role in PH risk will
however be very food process specific, and the model proposed may be too simple to capture important biological pro-
cesses, such as biofilm formation. With the currently available data, top-down risk assessment cannot be used to assess
the relative PH risk that can be attributed to persistence. Risk assessment cannot fully exploit the data gathered to support
answering the previous AQs, and the data needs for risk assessment are not well covered. Application of these data would
require better translation of genotypic information of strains into phenotypic characteristics that can be converted into
parameters of risk assessment models, as well as extensive quantitative data to describe the dynamics of transfer, survival
and growth of bacteria in the FoPE.
Nine specific knowledge gaps have been identified and translated into recommendations for research. Most recom-
mendations would involve activities at industry settings, but some of the research activities could be performed using
industrial-like model systems of certain niches, where different strains, environmental conditions and potential interven-
tions can be tested. These research activities would enable to establish the contribution of specific genetic markers and
their link to phenotypes associated with persistence, and to monitor the impact of particular interventions in reducing or
preventing persistence. They would also allow the generation of quantitative data to describe the dynamics of transfer,
survival and growth of bacterial hazards and to define strain-or subtype-specific parameters for QMRA.
It is recommended to apply clear definitions of persistence in all involved research areas (observational, experimental,
epidemiological, etc.) aiming at the same unambiguous definition for all of them. The environmental sampling and testing
programme should be robust and carefully planned by the food business operators and ensure an adequate surveillance
of higher risk niches for target bacterial hazards, and, during outbreak investigation, the sampling strategy should be op-
timised and data reporting of official and industrial sampling improved, in order to strengthen the link between FFPE and
the outbreak cases. It is also recommended to promote the use of interoperable standards to collect and report metadata
associated to WGS data to ensure auditability, to streamline data sharing and to reduce uncertainty. Finally, it is recom-
mended to promote the open access to both WGS data and complete and unambiguous associated metadata related to
the strain isolation, respecting data confidentiality and the interests of different partners in the food chain for investigating
persistence in the FFPE.
1 | I NTRO DUC TIO N
1.1 | Background and Terms of Reference as provided by the requestor
In slaughterhouses and facilities where food and feed are produced and/or processed, persistence of microbiological
hazards in the production environment occurs commonly and often involves repeated occurrence of the same strain for
months or even years at the same premises or equipment (Davies & Wray, 1997; Unnerstad et al., 1996). This represents a
great concern for public health, and food and feed business operators, since the persistent microbiological hazards in the
food and feed processing environment (FFPE) can lead to contamination of processed products, with important associated
health risks for consumers and economic losses for producers.
The ability of some microbiological hazards to persist in food processing environments (FoPE) is well documented, with
Listeria monocytogenes persistence being a major focus of attention (Fagerlund et al., 2021; Townsend et al., 2021). Indeed, the
persistence of L. monocytogenes strains has been described in cheese factories (Fox et al., 2011; Lomonaco et al., 2009),
salmon or crab meat production plants (Elson et al., 2019; Wulff et al., 2006), meat and meat products processing facilities
(Nesbakken et al., 1996; Ojeniyi et al., 2000), and produce packing houses and fresh-cut facilities (Estrada et al., 2020; Sullivan
& Wiedmann, 2020). Nevertheless, other biological hazards also have the ability to persist in FoPE. Larsen et al. (2014) re-
viewed persistence of Campylobacter spp. in food processing facilities and of Cronobacter spp. in processing facilities for pow-
dered foodstuffs, as examples to highlight factors involved in the persistence of microbiological hazards at FoPE. Of note is that
persistence of Campylobacter spp. has been demonstrated to occur for longer periods than expected, given the supposed ‘fra-
gility’ of the organism (Garcia-Sanchez et al., 2017). Persistence of Salmonella in low-moisture and high lipid matrices (such as
chocolate, sesame-based halva, tahini or peanut butter) and their production environments also represents a challenge due to
the pathogen's ability for long-term survival in low-moisture products and dry production environments (Finn et al., 2013).
Salmonella persistence has also been described in pig and poultry slaughterhouses (Hald et al., 2003; Zeng et al., 2021). Moreover,
apart from bacteria, some pathogenic viruses can persist in FoPE,1 as reviewed by Kotwal and Cannon (2014) for enteric viruses.
In recent years several high-profile foodborne outbreaks (FBO) have been associated with strains persistently colonising
FoPE or equipment, even for several years, and some processing facilities have been recurrently linked to FBOs and cases of
infection caused by closely-related genotypes of some microbiological hazards. As an example, a nationwide outbreak of
human listeriosis in Switzerland was traced to persisting environmental contamination of a cheese processing plant with
L. monocytogenes serotype 4b, multi-locus sequence type clonal complex 6 (CC6) (Nüesch-Inderbinen et al., 2021). Luth
et al. (2020) described one of the largest listeriosis outbreaks in Germany, with 83 cases of invasive listeriosis between 2013
and 2018, linked to persistence of the pathogen in a single producer of ready-to-eat (RTE) meat products. The finding of
L. monocytogenes 4b, ST6, matching a multi-country FBO strain in frozen corn and other frozen vegetables produced during
the 2016–2018 production seasons at a freezing plant, led to the suggestion that the outbreak strain could have been persist-
ing in the FoPE of the plant after standard cleaning and disinfection (C&D)2 procedures carried out, in conjunction with periods
of inactivity (EFSA and ECDC, 2018b). Likewise, outbreaks of salmonellosis have been traced back to persistent contamination
of pig slaughterhouse equipment (Bertrand et al., 2010; Kuhn et al., 2013; Schroeder et al., 2016), and a large proportion of the
Salmonella strains associated with food animals are associated with persistence in feed processing plants and compound feed
mills. These examples highlight the challenge of preventing environmental persistence of microbiological hazards in the FFPE.
The factors influencing microbial persistence in the FFPE, and the causes and genetic determinants involved in this, are
a matter of intense debate. The existence of harbourage sites that are difficult to clean and disinfect adequately, or the
special abilities of certain microbial strains to withstand conditions of environmental stress, desiccation, or disinfection, or
to form biofilms in industrial environments, have been mentioned among the most relevant determinants of persistence.
Some review articles have highlighted factors involved in the persistence of microbiological hazards, including L. monocy-
togenes, Cronobacter spp., and Campylobacter spp. in selected food chains (Larsen et al., 2014); they identified, for Listeria,
locations where it is commonly found to persist (i.e., floors, drains, conveyor belts, slicers, and tables), the most common
risk factors at processing facility level (equipment cleanability and lack of hygienic zoning), and interventions for the elim-
ination of persistent strains with variable results (Belias et al., 2022).
Remarkably, evidence from recent years suggests that, within the most relevant microbiological hazards, particular
lineages or genotypes more frequently colonise and persist in processing environments. For example, some L. monocyto-
genes sequence types (e.g., ST121, ST9) have been recurrently identified as persistent in FoPE (Alvarez-Molina et al., 2021;
Ferreira et al., 2014; Schmitz-Esser et al., 2015). However, the molecular mechanisms underlying these frequent associations
are not yet fully elucidated. In 2018, the EFSA BIOHAZ Panel described that some hypovirulent molecular subtypes of
L. monocytogenes, such as ST121, seem to encompass multiple isolates with a proven capability to persist. Whether their
persistence is a result of improper hygiene conditions, or the involvement of strains equipped with an arsenal of spe-
cific genetic determinants is under investigation. A high adaptive capacity against physical–chemical factors as well as
biofilm-forming capacity could partly explain the persistence phenomenon (EFSA BIOHAZ Panel, 2018). Recent advances in
next generation sequencing technologies have promoted a more detailed characterisation of persistent isolates through
whole genome sequencing (WGS) and transcriptomics, and offer opportunities for the study of persistence episodes at
1
This review article deals with transfer and environmental survival of foodborne and waterborne viruses outside the human host. It does not contain evidence on the
long-term establishment of viral hazards in FoPE.
2
Disinfection is used as a synonym of sanitation. The terms have been explained in the glossary.
processing plant level through the characterisation of the FFPE microbiome through metagenomics, which will provide
information on the role of complex microbial communities and the cellular processes and microbial interactions driving
microbial persistence in the FFPE (Fenner et al., 2021; Hu et al., 2022).
The strategies that can be employed to combat persistence by some microbiological hazards have been also reviewed
(for example by Larsen et al. (2014)). Those at the processing environment included hygiene measures; cleaning routines
of facilities and design of equipment; cleaning, disinfection and biofilm removal; and sampling. The EFSA BIOHAZ Panel
provided possible control options that may be implemented by food business operators during the production process of
frozen fruit and vegetables including herbs, blanched during processing. Additional control measures (technologies and
antimicrobial solutions) were identified to reduce or eliminate L. monocytogenes in the environment, mainly on surfaces,
and on the product (EFSA BIOHAZ Panel, 2020). In addition, measures for enhanced environmental monitoring can be im-
plemented, for example through agent-based in silico modelling to simulate the dynamics of foodborne pathogens in the
built environment (surfaces and equipment) of processing facilities, and to reduce the time and cost usually linked to classical
environmental monitoring activities (Sullivan et al., 2021).
EFSA is asked to deliver a scientific opinion on the persistence of microbiological hazards in food and feed production
and processing environments, excluding primary production.
More specifically, EFSA is requested to address the following terms of reference (ToRs):
ToR1. To identify the most relevant microbiological food safety hazards associated with persistence in the FFPE
ToR2. To identify the main (sub)types of the most relevant hazards involved in persistence and the main features re-
sponsible for their persistence in the FFPE
ToR3. To identify the risk factors at facility level responsible for the persistence of the most relevant hazards in the FFPE
ToR4. To assess available and enhanced measures or interventions for monitoring, preventing and/or controlling the
persistence of the most relevant microbiological food safety hazards in the FFPE
ToR5. To identify knowledge gaps and priorities for future research and develop the perspectives of integrating the
information gathered in the previous ToRs in risk assessment
1.2 | Interpretation of the Terms of Reference
In the scope of this mandate, microbial ‘persistence’ was defined as the ability of a given organism to be established in
niches (or harbourage sites) within the FFPE for a long term, despite the frequent application of C&D. It requires prolonged
existence usually with multiplication of the microorganism in the specific FFPE. It is a phenomenon which may lead to
recurrent food contamination events and is normally detected through the repeated isolation from the same premises
or equipment on different dates (spanning months or years) of strains that are subsequently identified as highly related
subtypes (as determined by phenotypic or genotypic methods). Persistence does not include continuous reintroduction
in the facility of the same organism, although in practice it is often not possible to distinguish between both phenomena.
Strains identified as sporadically (not repeatedly) contaminating the FFPE of a processing plant are frequently referred to
as ‘presumed non-persistent’, as a more intensified or a longer sampling campaign could result in their repeated isolation
from the FFPE. This terminology is also used in this assessment.
The ‘microbiological food safety hazards’ include all microorganisms which may adversely affect human health through
food consumption, including bacteria, viruses and parasitic protozoa, and any hazardous substance they may produce.
However, based on the definition of persistence (see above), it was agreed that viruses and parasitic protozoa are excluded
from the assessment because, due to their inability to multiply on abiotic surfaces, they cannot become established for a long
term and then constitute a contamination reservoir in the FFPE. Indeed, their occurrence is mainly linked to contamination
at the primary production level or through human manipulation of food matrices that are not processed afterwards or are
minimally processed, such as raw milk, fresh meat, oysters or berries. Likewise, microbial toxins and other hazardous microbial
metabolites (e.g. histamine) are excluded. Antimicrobial resistance (AMR) determinants are considered as possible features
related to persistence under ToR2, but they are not considered to define the most relevant hazards in ToR1.
The ‘food and feed production and processing environments’ cover slaughterhouses and facilities where food and
feed are produced and/or processed. It excludes primary production. Feed-producing environments are restricted to the
environments of facilities producing and processing feed for food processing animals. FoPE cover all environments where
food of animal or non-animal origin is industrially produced or processed, at post-harvest level (e.g. slaughterhouses, plant
factories, processing facilities). As such, the retail stage is not considered. Although farm environments are not included,
small scale food production facilities linked to a farm are covered (for example artisanal cheese production on a dairy farm).
The focus is on EU food and feed production systems, or, where applicable, in similar production systems from other regions
of the world. The following sectors are considered: (i) feed processing for food-producing animals, (ii) meat processing
(including slaughterhouses and processing plants), (iii) fish and seafood processing, (iv) dairy processing, (v) egg and egg
products processing, (vi) fruit and vegetable processing (including herbs), including Controlled Environment Agriculture
(CEA)/production through indoor hydroponic operations and (vii) low moisture food (LMF) processing. According to ILSI
(2011), a wide range of products fall in this LMF category (with a water activity below 0.85): cereals, chocolate, cocoa powder,
dried fruits and vegetables, egg powder, fermented dry sausages, flour, meal and grits, dried herbs, spices and condiments,
honey, hydrolysed vegetable protein powder, meat powders, dried meat, milk powder, pasta, peanut butter, peanuts and
tree nuts, powdered infant formula, rice and other grains, and seeds.
For ToR1, with ‘relevant’ it is understood those food safety hazards that have public health impact and have been found
to persist in the FFPE of each of the sectors.
For ToR2, (sub)types refers to a grouping of bacteria within a species that share certain characteristics, usually derived
by molecular typing (molecular or genotypic subtype) and/or spectroscopy/spectrometry based phenotypic methods.
Different types (e.g. derived from serotyping, multi-locus sequence typing or MLST, pulsed-field gel electrophoresis or
PFGE, WGS) will be considered as the methods have evolved over time. With ‘features’, it is considered the phenotypic
characteristics (e.g. ability to persist/survive in competition, survive disinfection, form biofilms, etc.) and the genotypic
characteristics (i.e. carriage of genetic determinants of resistance/persistence).
For ToR3, the ‘risk factors’ are interpreted as those factors at facility level that increase the likelihood of persistence of
the food safety hazards in the FFPE.
For ToR4, ‘measures’ or ‘interventions’ are considered synonyms. Strategies and control options cover those that are
already in place (either exceptionally or routinely used), as well as those not yet implemented. Enhanced control options
are considered the latter ones and those used already exceptionally. It was clarified that only measures or interventions
assessed in industry settings are to be considered. Economic or environmental impacts as well as the impact on the human
exposure to the hazards resulting from these mitigation options are outside the scope. Food businesses are obliged to
develop and implement food safety management systems (FSMS) including prerequisite programme (PRP) activities and
hazard analysis and critical control point (HACCP) principles. For this reason, control options will be based on prerequisite
programmes (PRPs; e.g. C&D), operational prerequisite programmes (oPRP) and, if possible, control points (CP) and critical
control points (CCP; i.e. the steps at which control can be applied and is essential to prevent or eliminate a food safety
hazard or reduce it to an acceptable level). PRP are preventive practices and conditions needed prior to and during, the im-
plementation of HACCP and which are essential for food safety. However, some prerequisites, typically linked to a specific
production process, may be identified as essential to control the likelihood of the introduction, survival and/or prolifera-
tion of food safety hazards in the product(s) or in the processing environment. These are referred to as oPRP.
For ToR5, ‘perspectives’ was understood to mean future opportunities. It is expected that limited data are available, and
there are no standard methods available to integrate the information gathered in a risk assessment of the public health
risk related to persistence. Therefore, it was agreed to explore future opportunities to perform such a risk assessment, if
possible, illustrated by some case studies.
The ToRs have been translated into assessment questions (AQs). A conceptual map of the AQs to be addressed in the
current assessment can be found in Figure 1.
FIGURE 1 Conceptual map of the AQs to be addressed in the current assessment.

The AQs are as follows:
• AQ1 (ToR1). What are the most relevant bacterial food safety hazards associated with persistence in the FFPE of the
various food and feed production and processing sectors in the EU/EEA?
• AQ2 (ToR2). Considering the most relevant bacterial hazards (from AQ1), what are the main (sub)types of these hazards
involved in persistence in the specific sector?
• AQ3 (ToR2). What are the main bacterial features responsible for the persistence of the most relevant bacterial hazards/
(sub)types (from AQ2) in the FFPE across sectors?
• AQ4 (ToR3). Considering the most relevant bacterial hazards (from AQ1), what are the factors at facility level that in-
crease the likelihood of persistence in the FFPE?
• AQ5 (ToR4). Considering the most relevant bacterial hazards (from AQ1), what are the available and enhanced measures
for monitoring, preventing and/or controlling their persistence in the FFPE?
• AQ6 (ToR5). What are the perspectives of integrating the information gathered in the previous AQs into a risk assessment?
• AQ7 (ToR5). What are the knowledge gaps related to bacterial food safety hazards associated with persistence in the
FFPE?
• AQ8 (ToR5). What are the priorities for future research related to bacterial food safety hazards associated with per-
sistence in the FFPE?
1.3 | Additional information
The approach to answer the ToR was defined in advance and is described in the protocol (Annex A). It covers both the prob-
lem formulation (i.e. what the assessment aims to address) and which methods will be used for addressing the problem.
The problem formulation (‘what’) includes the clarification of the mandate (see further refined in Section 1.2) and consists
of the steps (1) translation of the mandate into scientifically answerable AQs, (2) definition of the sub-questions (SQs) of
each AQ, if needed, and their relationship (conceptual model) and (3) the selection of the approach for the assessment.
The planning of the methods for conducting the assessment (‘how’) consists of (1) specifying the evidence needs and the
methods for answering each AQ/SQ, including uncertainty analysis and (2) the methods for integrating evidence across
AQs/SQs and addressing the remaining and overall uncertainty. Protocol development followed the draft framework for
protocol development for EFSA's scientific assessments (EFSA, 2020).
2 | DATA AN D M E TH O DO LOG I ES
2.1 | Most relevant bacterial food safety hazards associated with persistence in the FFPE (AQ1)
To identify the most relevant bacterial hazards associated with persistence in the FFPE, first the most relevant bacterial
pathogens of public health (PH) relevance in the various sectors in the EU/EEA were listed. Then, for those most relevant
pathogens, the evidence for their persistence in the FFPE of the corresponding sector was evaluated.
2.1.1 | Most relevant bacterial pathogens of public health relevance in the various food and feed
production and processing sectors in the EU/EEA
2.1.1.1 | Food production and processing sectors

Bacterial pathogens of PH relevance in the various food sectors were identified based on FBO as derived from various
sources and notifications in the Rapid Alert System for Food and Feed (RASFF) database. Some causative agents were ex-
cluded from the extracted data as defined in the clarifications to the ToR (i.e. viruses, parasitic protozoa, microbial toxins
and other hazardous microbial metabolites such as histamine).
Strong evidence FBO at EU/EEA level. Data on ‘strong evidence’ FBO from 2010 to 2020 were extracted from the EFSA
zoonoses database. The available epidemiological evidence was summarised by sector and causative agent, retrieving in-
formation on the number of outbreaks, number of cases, number of hospitalised cases and number of deaths. Thefood
vehicle (i.e. the food or foodstuff, that is suspected of causing human cases) was categorised as follows: meat and meat
products3 (for meat processing), fish and fishery products4 (for fish and seafood processing), fruit and vegetables and
3
Bovine meat and products thereof; Broiler meat (Gallus gallus) and products thereof; Cooked cured (or seasoned) poultry meat; Meat and meat products; Meat from
bovine animals – meat products; Meat from bovine animals – meat products – ready-to-eat; Meat from pig – fresh; Meat from pig – meat products – fresh raw sausages;
Meat from poultry, unspecified – fresh; Meat from poultry, unspecified – meat products – non-ready-to-eat; Meat from wild boar – meat products; Meat from wild boar – meat
products – fresh raw sausages; Meat, mixed meat – meat products – ready-to-eat; Other, mixed or unspecified poultry meat and products thereof; Pig meat and products
thereof; Sheep meat and products thereof; Turkey meat and products thereof.
4
Crustaceans, shellfish, molluscs and products thereof; Fish – smoked; Fish -smoked – hot-smoked; Fish and fish products; Live bivalve molluscs – oysters.
products thereof5 (for fruit and vegetable processing), milk and milk products6 (for dairy processing), egg and egg prod-
ucts7 (for egg and egg products processing) or various other food groups such as bakery products, nuts or sweets (for LMF
processing). Further information from other data fields, such as ‘more food vehicle information’ and ‘contributory factors’,
was consulted, when available. More information about the reporting of FBO can be found in the technical report
(EFSA, 2023).
Multi-country outbreaks in EU described in the Rapid Outbreak Assessments (ROA). Information was extracted
from the ROA reports of the multi-country outbreaks (2012–2023 period) in EU related to the year and date of publication,
incident, causative agent, suspected vehicle, number of cases, countries concerned and duration of the outbreak. Then
they were categorised by sector based on the suspected vehicle and the causative agent was evaluated.
Multi-country outbreaks described in the joint notification summaries (JNS) as included in the annual RASSF
reports. The JNS mentioned in the published annual RASSF reports between 2018 and 2020 were listed (European
Commission, 2019, 2020, 2021). The outbreaks were categorised by sector and the evidence on the causative agent was
evaluated.
RASFF notifications. Data were extracted from the RASFF database, considering the period 1 January 2010 to 11 July
2022, on the notification type ‘food’, the hazard category ‘pathogenic microorganisms’ and for each of the following prod-
uct categories: (i) meat and meat products (other than poultry), poultry meat and poultry meat products (for meat pro-
cessing), (ii) bivalve molluscs and products thereof, cephalopods and products thereof, crustaceans and products thereof,
fish and products thereof (for fish and seafood processing), (iii) milk and milk products, excluding dairy powder and infant
formula (for dairy processing), (iv) eggs and egg products, excluding egg powder (for egg and egg products processing),
(v) fruits and vegetables and herbs and spices, excluding those dried (for fruit and vegetable processing), (vi) cereals and
bakery products, cocoa and cocoa preparations, coffee and tea, confectionery, fats and oils, honey and royal jelly, nuts,
nut products and seeds, prepared dishes and snacks, dried fruits and vegetables, vegetable oil/flour/powder manually
selected from the fruits and vegetables group, dried herbs and spices, egg powder, dairy powder and infant formula (for
LMF processing). Notifications of viruses and parasitic protozoa were excluded.
FBO published on websites from non-EU authorities and agencies and on the scientific literature. The following
websites from authorities and agencies were consulted on 19 July 2022 for information related to foodborne outbreaks:
Centers for Disease Control and Prevention (CDC) of the U.S. Department of Health & Human Services;8 US Food and Drug
Administration (US FDA);9 Food Standards Australia New Zealand (FSANZ);10 Public Health Agency of Canada.11 In addition,
the Eurosurveillance journal12 was consulted for information related to foodborne outbreaks using the search term ‘out-
break’ in ‘All fields’. The search included reports between 1 January 2010 and 3 October 2022. Both surveillance and out-
break reports were considered eligible article types. All outbreaks were categorised by sector and the evidence on the
causative agent was evaluated.
For evidence integration, a list of bacterial pathogens that were found to be associated with FBO for each food sector
was prepared, obtained from these sources of evidence and the most relevant pathogen(s) were identified based on the
available data and the knowledge/expertise of the Working Group and Panel members.
2.1.1.2 | Feed production and processing sectors

For the feed sector, only data extracted from the RASFF database, considering the period 1 January 2010 to 11 July 2022, for
the category feed materials, feed additives, feed premixtures, compound feeds (for feed for food processing animals) were
considered. Also, a former scientific opinion of the BIOHAZ panel on Microbiological Risk Assessment in feeding stuffs for
food-producing animals for both public health and animal health (EFSA BIOHAZ Panel, 2008) was consulted. For evidence
integration, a list of bacterial pathogens that were found to be associated with feed was prepared and the most relevant
pathogen(s) identified based on the available data and knowledge/expertise of the Working Group and Panel members.
2.1.2 | Persistence of bacterial pathogens of public health relevance in the FFPE of the
various sectors
2.1.2.1 | Food production and processing sectors

The evidence for persistence of the most relevant bacterial pathogens in the FoPE of the corresponding sector was evalu-
ated by assessing the evidence (a) of at least one outbreak linked to strains persisting in the FoPE and/or (b) that the bacte-
rial hazard persists in the FoPE of plants from that particular sector.
5
Fruit, berries and juices and other products thereof; Fruits – whole; Herbs and spices; Lentil sprouts; Vegetables; Vegetables – pre-cut; Vegetables and juices and other
products thereof.
6
Cheese; Cheeses made from cows' milk; Cheeses, made from unspecified milk or other animal milk – spreadable; Dairy products (other than cheeses); Milk; Milk,
cows' – pasteurised milk; Milk, cows' – raw milk; Milk, goats' – raw milk; Milk, sheep's; Milk, sheep's – raw milk.
7
Eggs; Eggs – raw material (liquid egg) for egg products; Eggs – table eggs – mixed whole; Eggs – table eggs – shell; Eggs and egg products.
8
https://w ww.cdc.gov/foodsafety/outbreaks/multistate-outbreaks/outbreaks-list.html
9
https://w ww.fda.gov/food/outbreaks-foodborne-illness/investigations-foodborne-illness-outbreaks#date-posted
10
https://w ww.foodstandards.gov.au/industry/FoodIncidents/Pages/default.aspx
11
https://w ww.canada.ca/en/public-health/services/public-health-notices.html
12
https://w ww.eurosurveillance.org/search?value1=&option1=fulltext
For the first (point a), the FBOs in the ROA reports and JNSs and those in the websites from authorities and agen-
cies and the Eurosurveillance journal were screened for evidence of persistence of the outbreak strain in the FoPE. Also,
non-systematic searches to retrieve publications linked to relevant outbreaks were conducted and the knowledge of the
Working Group and BIOHAZ Panel members was considered. Scientific publications of outbreak investigations found
through the literature search (next point (b)) were also assessed.
For the second point (b), a literature search was carried out, restricted to publications post 2010. The search strategy
followed is described in detail in Appendix A. Firstly, review articles including information on outbreaks directly linked to a
hazard persisting in the FoPE, or on primary research studies demonstrating the persistence of hazards in processing
plants, were identified and consulted. Then, experimental studies were considered relevant when these: (i) took place in a
FoPE from one of the sectors considered, with description of the environment/plant;13 (ii) included sampling of specified
surfaces and microbiological testing of samples with reporting of results for each specific hazard; (iii) specifically referred
to the following terms related to persistence: persistence, permanent, residence, recurrence, dispersal or other relevant
terms; (iv) had matching subtypes for at least two sampling events over different time points; (v) used a molecular-based
subtyping method, including either genotypic or spectroscopy/spectrometry based phenotypic methods.14 Experimental
studies were excluded when these (i) were conducted in an environment not relevant to this assessment;15 (ii) described
solely food product sampling (e.g. raw material or finished product), a predictive microbial model in a food matrix or an
intervention in food products; (iii) presented laboratory studies conducted on lab-scale inoculated surfaces; (iv) did not
have matching subtypes for at least two sampling events over different time points; (v) did not use one of the substantive
subtyping methods referred to above. The evidence obtained was used to list the relevant bacterial pathogens identified
per sector.
For selecting the most relevant pathogen(s), the following criteria were considered by expert judgement: available
evidence on persistence and on outbreaks related to persistence, and on the PH relevance of the hazard (e.g. number of
cases or mortality rate).
2.1.2.2 | Feed production and processing sectors

A specific literature search was carried out, following the search strategy described in Appendix A, to retrieve evidence that
the bacterial pathogen persists in the FePE of plants from the feed sector. The most relevant pathogen(s) in the FePE were
selected by expert judgement.
2.2 | Main (sub)types of the most relevant bacterial hazards involved in persistence and the
main features responsible for their persistence in the FFPE (AQ2 and 3)
The literature search as described in Section 2.1.2 was used to identify the subtypes most frequently identified as involved
in persistence and their main features. Reviews and primary research studies were considered. The grey literature was also
consulted, for example reports from agencies. Additionally, information (when available) on the subtypes that have been
found in the outbreaks linked to persistence (see Section 2.1.1) was retrieved.
Data were extracted for each of the primary research studies selected as relevant in relation to persistence of the haz-
ards in the FFPE of processing plants on: the country where sampling took place, the sector and plant within the sector,
the reason for sampling (outbreak related or not), the subtype identified, the location of persistence (i.e. non-food contact
surface (NFCS) and/or food contact surface (FCS)) with further details, if available, the typing method(s) used, the features
of the persistent strain, the factors influencing its persistence, the intervention(s) implemented, the number of sampling
events, the sampling period and the timespan for persistence.
The evidence of relevance for replying to AQ2 and AQ3 was synthesised by listing all those subtypes identified as linked
to persistence in the FFPE for each of the most relevant hazards (from AQ1) and for each sector. For L. monocytogenes and
C. sakazakii (formerly Enterobacter sakazakii), the clonal complex (CC) of strains identified as persistent was recorded and
presented, while for S. enterica it was the serotype of persistent strains.
For those subtypes most frequently involved in persistence, the main features possibly associated with persistence, as
retrieved from the literature search, were identified and listed. Further references obtained through non-systematic spe-
cific literature searches were consulted to provide more insights into the role of some of the genetic markers identified as
possibly linked to persistence.
The subtypes involved in the main clusters of related genome sequences of L. monocytogenes, S. enterica and C. saka-
zakii available in the National Center for Biotechnology Information (NCBI) Pathogen Detection database16 were further
analysed, considering their distribution by source and year of isolation. For L. monocytogenes17 and Salmonella,18 a minimum
13
Small scale production facilities linked to a farm were also considered (e.g. an artisanal cheese production at a dairy farm).
14
For example, AFLP, MLVA, MLVST, MLST, PFGE, RAPD, REA, Ribotyping, WGS-based methods, serotyping, spa typing, panC typing, Fla-t yping, RFLP, FTIR or MALDI-TOF.
15
For example, home kitchen, institutional kitchen, farm or retail.
16
https://w ww.ncbi.nlm.nih.gov/pathogens/
17
Listeria 2023-4 -23; https://f tp.ncbi.nlm.nih.gov/pathogen/Results/Listeria/PDG000000001.3193/
18
Salmonella 2023-7-30; https://f tp.ncbi.nlm.nih.gov/pathogen/Results/Salmonella/PDG000000002.2708/
threshold of at least 100 isolates per cluster was used to select clusters for further interrogation; in the case of Cronobacter,19
the largest cluster contained only 67 isolates, and 24 clusters contained only up to 10 isolates. The NCBI Pathogen Detection
database integrates genomes from numerous ongoing surveillance and research efforts whose sources include food, en-
vironmental sources, such as water or production facilities, and patient samples. The database designates single nucleotide
polymorphism (SNP) clusters using two approaches: firstly, using a reference whole genome multi-locus sequence typing
(wgMLST) scheme using a 25-allele cut-off to cluster related isolates; and secondly, using k-mer distances to first cluster
related isolates, then a 50-SNP single-linkage clustering SNP analysis. For L. monocytogenes, further analyses on the car-
riage of persistence-, biofilm-and virulence-related genetic determinants were undertaken for a selection of clusters from
subtypes more frequently reported as persistent in experimental studies in the literature, together with two clusters from
subtypes that have not been associated with persistence yet. Genetic determinants were screened by BLAST analysis of the
selected genetic markers against all available genome assemblies within the selected clusters. Clusters were extracted for
analysis in June 2023.
2.3 | Factors at facility level that increase the likelihood of persistence of the most relevant
bacterial hazards in the FFPE (AQ4)
The literature search as described in Section 2.1 was used to identify the site (or location) where the persistent strains were
isolated and, if available, the niche (or harbourage site), as well as the risk factors at facility level that increased the likeli-
hood of persistence of the most relevant bacterial hazards in the FFPE. Reviews, primary research studies and the grey
literature were consulted. Additionally, the factors at facility level, when available, that have been found in the FBOs (from
AQ1) were listed.
The evidence retrieved was synthesised, for each of the most relevant hazards (from AQ1), by listing the sites or niches of
persistent strains within the plants. Next, those risk factors at facility level that increase the likelihood of persistence in the
FFPE were described across production sectors and further, when relevant, specific to each of the sectors.
2.4 | Measures for monitoring, preventing and/or controlling the persistence of the most
relevant bacterial hazards in the FFPE (AQ5)
The literature search as described in Section 2.1 was used to summarise the measures for monitoring, preventing and/or
controlling the persistence of the most relevant bacterial hazards in the FFPE. Only experimental studies on measures or
interventions assessed in industry settings were considered. Reviews and primary research studies were consulted. Grey lit-
erature was also consulted, for example reports from agencies, as well as sector-specific guidance documents. Additionally,
the interventions (when that info was available) that have been used to control the problem in the FBOs (from AQ1) were
listed.
In addition, relevant BIOHAZ scientific opinions were reviewed and referred to for the environmental monitoring and
safety control measures in specific commodities (e.g. EFSA BIOHAZ Panel (2017, 2020)).
The evidence retrieved was synthesised by listing the measures for monitoring, preventing and/or controlling per-
sistence in the FFPE. First, the daily measures included in the FSMS, including PRP activities, were addressed. Secondly,
corrective measures triggered when a persistence event was detected were also addressed.
2.5 | Perspectives of integrating the information gathered in risk assessment (AQ6)
Potential risk assessment questions related to persistence in the FFPE were discussed between Working Group members
and the relevant questions were selected, with feedback from Panel members. Approaches to answer these questions in-
cluded the development of quantitative mathematical risk assessment modelling structures, based on existing risk assess-
ment models available from the literature and/or known by the experts. A non-systematic literature review was performed
to identify existing risk assessment models involving persistence in the FFPE. Both ‘top-down’ (or backward) and ‘bottom-
up’ (or forward) approaches were explored, i.e. approaches where risks are assessed based on human disease data and ap-
proaches where risks are assessed based on pathogen occurrence in the food chain and the effect of food chain processes
on hazards, including a dose–response (DR) model. For the ‘bottom-up’ approach, a ‘persistence model’ was proposed and
the perspectives of its usage were studied.
Two case studies were defined, based on information gathered from the previous ToR and on discussion within the
Working Group. For these, the bottom-up approach using the developed model structure was applied. Relevance of the
hazard and of persistence in the FFPE, as well as availability of existing predictive models for growth in the food product
and DR models, required to perform a risk assessment, were important considerations in the selection of the two case stud-
ies. Information gathered from ToR2 ((sub)types and features of persistent strains) was used to develop and test a strain
19
Cronobacter 2023-7-27; https://f tp.ncbi.nlm.nih.gov/pathogen/Results/Cronobacter/PDG000000043.195/
specific risk assessment, where the potential association between for example growth potential, environmental resistance
and virulence is included. Based on the case studies and the modelling structures, more generic conclusions were drawn
on the data needs for risk assessment of microbiological food safety hazards associated with persistence in the FFPE.
Finally, as a last step, the collected data was compared to the data needs that were identified for risk assessment. This
comparison was used to draw conclusions on the perspectives of performing risk assessments considering the role of per-
sistence in the FFPE.
2.6 | Knowledge gaps and priorities for future research (AQ7 and 8)
Uncertainties linked to answering AQ1-6 were listed and used to formulate the knowledge gaps related to bacterial haz-
ards associated with persistence in the FFPE (answering AQ7) based on expert knowledge (Working Group and BIOHAZ
Panel members). Based on the identified knowledge gaps, research needs related to bacterial food safety hazards associ-
ated with persistence in the FFPE were identified and prioritised (answering AQ8), also based on expert knowledge.
2.7 | Uncertainty analysis
As recommended by the EFSA guidance and related principles and methods on uncertainty analysis in scientific assess-
ments (EFSA Scientific Committee, 2018a, 2018b), an uncertainty analysis was implemented. Given the nature and context
of the ToRs of the mandate, the uncertainty analysis was restricted to an overview of the uncertainty sources affecting the
different AQs (Appendix B). They describe the strengths and weaknesses of the collected evidence and served as a source
of information for the discussion on knowledge gaps and research needs.
3 | ASSESSM E NT
3.1 | Most relevant bacterial food safety hazards associated with persistence in the FFPE
(AQ1)
3.1.1 | Most relevant bacterial pathogens of public health relevance in the various food and feed
production and processing sectors in the EU/EEA
Appendix C includes the detailed assessment for each of the food and feed production and processing sectors on the
bacterial pathogens which are of PH relevance in the EU/EEA. A summary is provided here below for each sector based on
reported FBO and/or RASFF notifications and an overview of the pathogens selected as of highest relevance is presented
in Table 1.
Serovars of S. enterica are the major hazard associated with microbial contamination of feed for food processing animals.
S. enterica subsp. enterica (especially serovars Enteritidis and Typhimurium), L. monocytogenes, human pathogenic E. coli,
C. jejuni and C. coli, and Clostridium perfringens and Clostridium botulinum are the most relevant hazards associated with
meat and meat products. Salmonella outbreaks were associated with meat and meat products from all animal origins.
L. monocytogenes was mainly associated with RTE meat products, human pathogenic E. coli with bovine meat (products),
Campylobacter with broiler meat (products) and Clostridium spp. with general meat products. S. enterica caused the highest
number of cases and hospitalisations linked to FBOs; however, L. monocytogenes caused most deaths. Clostridium spp. and
Campylobacter spp. were also associated with a high number of cases, but human pathogenic E. coli caused more hospi-
talisations than these two Clostridium species. Other biological hazards which have been involved in less frequent and/
or severe reported outbreaks linked to meat and meat products are Bacillus cereus sensu lato, Staphylococcus aureus and
Yersinia enterocolitica.
T A B L E 1 Overview of the bacterial pathogens of the highest public health (PH) relevance in the various food and feed production and processing
sectors in the EU/EEA (highlighted in grey).
Pathogens of highest PH relevance in the sector
Bacterial pathogen F M FS D E FV LMF

Bacillus cereus sensu lato
Campylobacter jejuni/coli
Clostridium botulinum/perfringens
Cronobacter sakazakiia
Listeria monocytogenes
Human pathogenic E. coli
TA B L E 1 (Continued)
Pathogens of highest PH relevance in the sector
Staphylococcus aureus
Salmonella enterica
Vibrio parahaemolyticus
Abbreviations: F, feed for food animal production sector; M, meat sector, excluding low moisture food (LMF) products; FS, fish and seafood sector, excluding LMF
products; D, dairy sector, excluding LMF products; E, egg sector, excluding LMF products; FV, fruit and vegetable sector, excluding LMF products; LMF, low moisture food
sector.
a
Formerly Enterobacter sakazakii.
S. enterica and L. monocytogenes are the most relevant bacterial hazards linked to fish and seafood products; these
hazards caused most hospitalisations and deaths linked to FBOs in the EU/EEA since 2010, were also linked to FBOs outside
the EU and were most often notified in the RASFF database. Outbreaks and notifications on S. enterica are associated with
various fish and shellfish products and on L. monocytogenes mostly with cold smoked fish (salmon). V. parahaemolyticus
has also a high PH relevance, as it has been involved in FBOs inside and outside the EU and has been the topic of various
RASFF notifications linked to shellfish. Other bacterial hazards which have been involved in less frequent and/or less severe
outbreaks linked to fish and seafood products are B. cereus s. l., C. botulinum and C. perfringens, human pathogenic E. coli
and Campylobacter.
S. enterica, Campylobacter spp., human pathogenic E. coli, L. monocytogenes and S. aureus are the most important bacte-
rial hazards associated with milk and dairy products. S. enterica, S. aureus, pathogenic E. coli and L. monocytogenes caused
most of the infections associated with cheese consumption, with the latter having the highest mortality. Campylobacter
was the main pathogen in milk-related outbreaks, followed by S. enterica and human pathogenic E. coli. Other biological
hazards associated with the contamination of raw milk, cheese and dairy products that have been linked less frequently to
outbreaks are B. cereus s. l., Brucella melitiensis, Mycobacterium bovis, C. perfringens, Yersinia pseudotuberculosis and Shigella
flexneri.
S. Enteritidis is the most relevant bacterial hazard linked to eggs and egg products, with other serovars of S. enterica
being also causative agents of less frequent outbreaks. S. enterica caused the vast majority of infections, hospitalisations
and notifications associated with eggs and egg products. B. cereus s. l. has been involved in less frequent and/or severe
reported outbreaks linked to eggs and egg products, although detailed information on the food vehicle is in most cases
not available to evaluate whether they can be linked to a LMF such as egg powder.
S. enterica, human pathogenic E. coli and L. monocytogenes are the most relevant bacterial hazards linked to whole
fresh, fresh-cut or frozen fruits and vegetables, berries, juices and products thereof. This is in agreement with the most
relevant hazards identified in the hazard prioritisation conducted based on EU reported outbreaks (2014–2020) in the
scientific opinion on microbiological hazards associated with the use of water in the post-harvest handling and process-
ing operations of fresh and frozen fruits, vegetables and herbs (ffFVHs) – Part 1 (EFSA BIOHAZ Panel, 2023). Despite the
high number of outbreaks and associated cases due to S. enterica, their severity is rather low with no deaths reported, as
opposed to human pathogenic E. coli and L. monocytogenes infections, which, despite the lower number of cases, caused
more deaths. B. cereus s. l., Shigella sonnei, C. botulinum, C. perfringens, Campylobacter spp., Aeromonas hydrophila, S. aureus
and Y. enterocolitica have been involved in less frequent and/or severe reported outbreaks linked to fresh or frozen fruits
and vegetables, berries, juices and products thereof.
S. enterica is the most relevant bacterial hazard linked to LMF, being frequently involved in outbreaks associated with
confectionary products and snacks, chocolate products, nuts and nut products, cereals and grains, dried fruits and vege-
tables, seeds for consumption and powdered foods including infant formula. Other bacterial hazards linked to outbreaks
in LMF and having also a high PH relevance are B. cereus s. l., mainly in cereal-based products, especially cooked rice and
pasta dishes, and spices and dried aromatic herbs; human pathogenic E. coli, mainly in cereal-based products (flours) and
nuts and nut products; and C. sakazakii, in powdered infant formula. Other biological hazards which have been involved
in less frequent and/or severe reported outbreaks linked to LMF are Clostridium (C. perfringens and C. botulinum), S. aureus
and L. monocytogenes.
3.1.2 | Persistence of bacterial pathogens of public health relevance in the FFPE of the various sectors
3.1.2.1 | Pathogens able to persist in the processing environment of the feed for food animal production sector
The presence of Salmonella in animal feed has only rarely been directly linked to human cases of salmonellosis (Bonifait
et al., 2022; Schroeder et al., 2016), possibly due to the number of steps between animal feed production and human con-
sumption of the respective food of animal origin. Nevertheless, there is evidence of Salmonella strains found in feed also
being found in the associated livestock consuming the feed (APHA, 2022). Salmonella strains have been associated with
persistence in feed processing plants and compound feed mills, where some serovars can be repeatedly isolated for many
years. The clones may persist on parts of the factory equipment or environment (Larsen et al., 2014; Nesse et al., 2003;
Vestby et al., 2009). There have also been indications of Salmonella growth in feed mill environments, such as floors, valves
and cyclones (Binter et al., 2011), and in storage containers when moisture had accumulated (Binter et al., 2011).
According to a previous scientific opinion of the BIOHAZ panel (EFSA BIOHAZ Panel, 2008), the repeated and long-term
isolation of certain serotypes in feed ingredients or compound feed has often been found to be the result of persistent
contamination of crushing and feed-producing plants. This opinion highlighted that equipment used for cooling feed may
become persistently contaminated by Salmonella as a result of intake of contaminated cooling air or passage of feed which
has been incompletely heat treated, and that in situations when the post heat treatment feed processing equipment does
become persistently contaminated by Salmonella it is common to find that finished products are more likely to be contam-
inated than the ingredients used to manufacture the feed.
3.1.2.2 | Pathogens able to persist in the processing environment of the meat sector
There are numerous documented outbreaks of human listeriosis linked to meat product contamination, and moreover with
strong evidence of persistence of the L. monocytogenes outbreak strain in the associated FoPE. An outbreak in Germany
lasting 5 years was linked to persistent contamination of the meat processing plant environment (Luth et al., 2020).
Contaminated slicers, in a RTE meat processing plant, were the source of the strain implicated in an outbreak in Canada
(Currie et al., 2015; Weatherill, 2009). Moreover, other listeriosis outbreaks linked to the meat FoPE have been possibly
related with persistence but with no repeated isolation of the outbreak strain (Duranti et al., 2018; ECDC and EFSA, 2019b;
Hachler et al., 2013; Lachmann et al., 2021). For example, a single RTE meat facility was linked to one of the biggest listeriosis
outbreaks reported worldwide, in South Africa in 2017–2018, with the implicated strain being isolated from environmen-
tal samples collected at several facility sections (precooking and post-cooking) in a unique sampling occasion (Thomas
et al., 2020). Collectively, there is strong evidence that L. monocytogenes can persist in the environment of meat products
processing premises.
In the case of Campylobacter outbreaks, strong evidence is lacking regarding association with persistence in the meat
FoPE. Weak evidence for persistence of an outbreak strain in a poultry slaughterhouse was found in an outbreak lasting
several months (Joensen et al., 2021). Moreover, although Campylobacter has frequently been detected in environmen-
tal samples from pork and poultry slaughterhouses and processing plants (Quintana-Hayashi & Thakur, 2012; Torralbo
et al., 2015), evidence of the same subtypes of C. jejuni being repeatedly isolated in poultry processing environments has
been only reported by Melero et al. (2012) and Garcia-Sanchez et al. (2017).
Strong evidence of persistence of S. enterica in the meat processing environment linked to human salmonellosis out-
breaks is lacking. Several outbreaks with weak evidence of environmental persistence have been reported, where no de-
tailed environmental sampling or repeated isolation of the strain involved was reported (Bertrand et al., 2010; Gieraltowski
et al., 2016; Hobbs et al., 2017; Kuhn et al., 2013; Wingstrand et al., 2012). Schroeder et al. (2016) reported the isolation of the
outbreak strain, previously associated with an outbreak related to a feed processing plant, in the environment of a pork
slaughterhouse in several sampling occasions after the C&D procedures. More generally, studies targeting the environ-
ment of poultry slaughterhouses suggest persistence of strains of Salmonella over time (Dantas et al., 2020) with subse-
quent dissemination to retail also noted (Shang et al., 2019), indicating potential routes of transmission of persistent strains
through contaminated meat products to retail and the consumer.
Evidence of outbreaks related to persistent strains of human pathogenic E. coli is lacking. King et al. (2014) traced back an
E. coli O157:H7 outbreak strain to a beef processing facility, however, no strong evidence of persistence was presented as no
environmental sampling was reported. The literature lacks longitudinal sampling studies demonstrating the persistence of
STEC in meat processing environments over prolonged timeframes.
There was no evidence of outbreaks of C. perfringens and C. botulinum linked to persistence. Some studies describe
C. perfringens outbreaks related to episodes occurring after consumption of improperly reheated meat products in restau-
rants (Mellou et al., 2019; Wahl et al., 2013) and a catered lunch (Rinsky et al., 2016). There are limited published studies
tracing these clostridial species in meat processing or identifying persistent contamination; although Jiang et al. (2022)
demonstrated distribution of the same pulsotype across the FoPE and associated meat samples of beef slaughterhouses
on single sampling occasions, persistent contamination of the FoPE over time was not identified.
3.1.2.3 | Pathogens able to persist in the processing environment of the fish and seafood sector
There have been several multi-year outbreaks of human listeriosis linked to cold smoked or gravad fish (mostly salmon) and
two outbreaks linked to crab meat. In one multi-year outbreak in Germany linked to smoked/gravad salmon, the outbreak
strain was repeatedly isolated from the processing environment over 2 years (Lachmann et al., 2022). For other outbreaks, out-
break strains have been found in the processing environments, but evidence for repeated isolation over time is lacking (ECDC
and EFSA, 2019a; Lassen et al., 2016). There are multiple reports of L. monocytogenes persisting in fish processing environments
(Fagerlund et al., 2022; Ferreira et al., 2014). As an example, investigations due to increased listeriosis cases in Finland revealed
two fish production plants with persistent L. monocytogenes contamination (Nakari et al., 2014; Wulff et al., 2006).
No strong evidence for outbreaks of S. enterica linked to persistence in the processing environment of fish and seafood
was found. Two outbreaks with S. Thompson lasting several months associated with raw and cooked seafood and smoked
salmon, respectively, had a possible link to the processing environment, although repeated isolation of the outbreak strains
from the FoPE was not documented20 (Friesema et al., 2014). There is some evidence of persistence of Salmonella in
20
https://w ww.cdc.gov/salmonella/thompson-10-21/details.html
rocessing plants (Wang et al., 2019). Isolation of the same pulsed-field gel electrophoresis (PFGE) types of S. Senftenberg
p
over time was reported in several mussel processing plants (Martinez-Urtaza & Liebana, 2005). It is not entirely clear if this
was due to reintroduction of the bacteria through contaminated salt/brine or semi-processed mussels, or due to per-
sistence of the bacteria in the processing environment. Likewise, persistence (same PFGE types) of S. Stanley and S. Bareilly
has been shown in a tilapia sashimi processing plant in Taiwan (Wang et al., 2019).
Most outbreaks of V. parahaemolyticus are linked to raw or undercooked seafood, usually shellfish (Haque et al., 2023).
Cross-contamination from raw to RTE food is reported but no evidence for persistence in the FoPE was linked to outbreaks.
Re-occurring infections or outbreaks over time with the same strain can be explained by persistence of the strain in marine
habitats, not in the FoPE (Yang et al., 2022).
3.1.2.4 | Pathogens able to persist in the processing environment of the dairy sector
Persistent bacteria in the dairy sector are mainly described in relation to primary production and farmhouse cheesemak-
ers. S. aureus, a contagious mastitis pathogen, has been frequently related to persistence in the bovine udder. Some
S. aureus subtypes are widely spread among dairy cattle (e.g. CC151 and CC97) or have been described in dairy cattle from
particular countries (e.g. CC705, CC398, CC479, CC8) (Campos et al., 2022). S. aureus genotype B (CC8) is highly prevalent in
the alpine dairy industry and has been associated with human intoxications caused by the enterotoxins Sea, Sed and Sej
(Johler et al., 2015). Methicillin resistant S. aureus (MRSA) are rarely reported in dairy animals. Studies indicate persistence
of livestock associated ST1 and ST398 among dairy goats and cattle (Cortimiglia et al., 2016; Schnitt et al., 2020). Some
S. aureus genotypes harbouring particular capsular polysaccharide types (e.g. cap 5) have been associated with a stronger
potential for biofilm production in the udder tissue and food processing environments (Salimena et al., 2016). However,
although S. aureus is described as a biofilm former with potential to persist on abiotic surfaces in the food and medical sec-
tors (Abdallah et al., 2014; Miao et al., 2017), data from FCS or environmental samples in dairy industry settings are lacking.
There is strong evidence from listeriosis outbreaks linked to strains persisting in the processing environment in the dairy
sector. A nationwide outbreak of human listeriosis linked to cheese processing in Switzerland was traced to persistent envi-
ronmental contamination with the hypervirulent L. monocytogenes subgroup CC6 (Nüesch-Inderbinen et al., 2021). In Italy,
Gorgonzola products were linked to human cases caused by L. monocytogenes epidemic hypervirulent clones from CC3
(Bergholz et al., 2016; Filipello et al., 2017). Of interest are studies that retrospectively clarify the cause of outbreaks across
continents such as the study by Acciari et al. (2016), who sought the causes of L. monocytogenes contamination in tradi-
tional Italian cheese associated with US outbreaks in the country of origin. In 2012, a US multistate outbreak of listeriosis
was linked to Ricotta Salata imported from Italy. The follow-up sampling identified the same PFGE type in the Italian cheese
plant suggesting an event of persistence. L. monocytogenes persists preferentially in cheese production with surface rip-
ening, being found in environmental samples and product-associated samples, such as cheese smear, brine, wash water
and smear robot. This has been adequately described in retrospective studies using molecular epidemiological methods
(WGS, PFGE) (Barría et al., 2020; Kaszoni-Rückerl et al., 2020; Melero, Stessl, et al., 2019; Muhterem-Uyar et al., 2018; Nüesch-
Inderbinen et al., 2021; Stessl et al., 2014).
There are some studies assessing the prevalence of Campylobacter in milk. For example, in a Portuguese cattle farm,
C. jejuni ST-21, ST-22, ST-206 and ST-403, all strongly associated with cattle and goat milk according to the PubMLST data-
base,21 were present at a low prevalence (4%; (Barata et al., 2022)). C. jejuni ST-883, associated with an outbreak, persisted
in a milk tank for more than 7 months on a Finnish dairy farm (Jaakkonen et al., 2020). Nevertheless, overall, Campylobacter
is most likely detected in raw milk by faecal contamination during milk collection, and persistence of Campylobacter in milk
and cheese processing environments has not been described.
There are various studies reporting the occurrence of some S. enterica serotypes in primary production and dairy prod-
ucts. S. Dublin, one of the most common serovars in dairy cattle (Holschbach & Peek, 2018) and raw milk cheeses, can per-
sist in some dairy facilities, which were also associated with regional outbreaks in France between 2015 and 2017 (De Sousa
et al., 2022; Ung et al., 2019). In 2020, an outbreak of S. Enteritidis occurred in Central Italy, transmitted by Pecorino cheese.
Cheese, bulk milk, faecal and environmental samples taken at the dairies linked the outbreak to potential short-term per-
sistence of S. Enteritidis in the FoPE. However, the sources of the outbreak were infected sheep and their unpasteurised milk
used for cheese processing (Napoleoni et al., 2021). Overall, there is a data gap regarding the survival and persistence of
S. enterica in the dairy processing environment.
Considering human pathogenic E. coli, during 2013 an E. coli O157:H7 outbreak occurred in five Canadian provinces
transmitted by contaminated Gouda cheese originating from a cheese processing facility in British Columbia. The raw milk
was the primary source of E. coli O157:H7, which persisted during the production and at least during 60 days of ripening.
This outbreak was the third outbreak caused by E. coli O157:H7 attributed to Gouda cheese made from raw milk in North
America within a narrow timeframe. Pathogenic E. coli are introduced via raw milk and are detectable at various stages
of ripening during cheese processing (Dos Santos Rosario et al., 2021; Rios et al., 2020), but there is a clear lack of data on
sampling of pathogenic E. coli in the FoPE.
21
https://pubmlst.org/
3.1.2.5 | Pathogens able to persist in the processing environment of the egg sector
Large salmonellosis outbreaks extended in time linked to consumption of eggs and egg products frequently occur (ECDC
and EFSA, 2020a; EFSA and ECDC, 2017b). Such outbreaks are commonly associated with Salmonella persistence at farm
level. This persistent contamination in laying hen farms has sometimes been reflected in contamination with the same
strain of contact surfaces in the corresponding egg packhouses, which test more frequently positive when working with
high-prevalence farms. This supports an association between the prevalence at the laying environment and at the pack-
house egg contact surfaces (Kim et al., 2015; Kingsbury et al., 2019). However, no strong evidence was found of outbreaks
directly linked to persistence of S. enterica in the processing environment of eggs and egg products processing plants
(excluding LMF).
There is evidence that S. enterica can persist for long periods of time in the processing environment of egg products
processing plants. For example, Jakociune et al. (2014) demonstrated that the continuous contamination of lightly pas-
teurised egg products with S. Tennessee at a large European producer of industrial egg products was caused by persistent
contamination of the production facility. Likewise, Kim et al. (2015) found contamination by S. Bareilly in liquid egg samples
produced by one of the eight egg-breaking plants sampled, due to contamination of the product line. The authors con-
cluded that the contamination could have originated from a specific farm that provided shell eggs to the plant, and that
the contaminating Salmonella could persist at the processing plant for a long period. Indeed, the contamination was not
properly controlled in the product line, even after pasteurisation, which may indicate post-process contamination from the
production environment.
3.1.2.6 | Pathogens able to persist in the processing environment of the fruit and vegetable sector
Outbreaks linked to enteric pathogens in fresh fruits and vegetables are traced to strains present in incoming raw materials
at the processing plants or accumulated in the post-harvest process water and further disseminated in the processing and
packing plant (Fatica & Schneider, 2011; Jung et al., 2014; Wang, 2019). Survival of strains in the FoPE also allows a long-term
presence in the processing plants (Fatica & Schneider, 2011), especially in moist environments (Williamson et al., 2018),
albeit without a clear implication of persistence as the causative link to outbreaks. Listeriosis outbreaks have been com-
monly linked to the presence of strains in dry or wet FCS (brushes, slicers, conveyor belts, etc.) and NFCS (e.g. drains, floors,
coolers) (Chen et al., 2022; Estrada et al., 2020; Truchado et al., 2022).
Considering human pathogenic E. coli, the available evidence on the occurrence of this hazard in the fruit and vege-
table sector suggest, but do not clearly establish, a link between outbreak strains and persistence in the FoPE. Similarly,
there is limited (if any) systematic link of genetically related strains (judged by subtyping) to particular niches in the pro-
cessing environment of the fruit and vegetable sector, that could explain persistence and thus, re-occurrence in the prod-
ucts. STEC O157:H7 lineage I/II has remained the dominant lineage in England since 1980s and recently IIc has become
increasingly associated with consumption of fresh produce and particularly pre-packaged salads (Dallman et al., 2021),
suggesting a potential for persistence in the raw materials and the processing environment. Nevertheless, in general, the
majority of single or multi-country STEC or EPEC outbreaks linked to packaged leafy greens (e.g. rocket, lettuce or mixed
salad), such as those in the UK (165 cases) and Finland (237 cases), are attributed to the raw materials, rather than the pack-
ing or processing plants.
Similarly to human pathogenic E. coli, the available outbreak and literature evidence on S. enterica implies, but it
does not clearly establish a link of outbreak strains with persistence in FoPE. Neither does it demonstrate a clear long-
term establishment of genetically similar strains in the environment of this sector. Most Salmonella outbreaks during
1973–2010 caused by various serovars present in tomatoes were attributed to strains in the packing house or the farm,
yet, without clear source tracing (Bennett et al., 2015). In 2008, a Salmonella outbreak occurred in Finland with 77 cases
linked to serovar Newport and 30 cases to serovar Reading, including one case with a double infection (Lienemann
et al., 2011). However, whether packing house (particularly chopping) environments were the source of contamination
was not clarified. In January 2022, the FDA published the report of the first domestic investigation of a foodborne
salmonellosis outbreak caused by S. Typhimurium associated with leafy greens grown in a CEA indoor hydroponic22
operation. Although the investigation did not result in the identification of the specific source or route of contamina-
tion of the leafy greens, as the outbreak strain was not recovered from inside of the greenhouse, the same strain was
found in a water pond located close to the farm. The final report included an overview of the various factors that po-
tentially contributed to the introduction and spread of pathogens of PH significance into the crop, which could serve
as source of strains that may subsequently be established in the processing environment of fresh-cut products
(FDA, 2022).
The finding of L. monocytogenes 4b, CC6, matching a multi-country FBO strain in frozen corn and other frozen
vegetables produced during the 2016–2018 production seasons, at a freezing plant led to the suggestion that the
outbreak strain could have been persisting in the FoPE of the plant after standard C&D procedures were carried out,
in conjunction with periods of inactivity (EFSA and ECDC, 2018b). WGS revealed that most fresh produce outbreaks
were associated with L. monocytogenes contamination originating from the processing environment and equipment
(Chen, Burall, et al., 2016; Garner & Kathariou, 2016; Truchado et al., 2022). This suggests its ability to establish in various
niches within a processing plant, e.g. brushes, blowers, blades, dryers (Ruiz-Llacsahuanga et al., 2021). Multiple reports
22
A cultivation technique of plants, where plants are growing in the water environment without soil or any aggregate medium.
characterising the prevalence and genotypic diversity of L. monocytogenes strains in the processing environments of
leafy greens, fruit trees and mushrooms have demonstrated the widespread occurrence of this organism in packing
houses for months to years, as the likely cause of fresh produce outbreaks (Chen et al., 2014; Chen, Burall, et al., 2016;
Lake et al., 2021; Pennone et al., 2018; Simonetti et al., 2021; Sullivan & Wiedmann, 2020; Truchado et al., 2022; Viswanath
et al., 2013).
3.1.2.7 | Pathogens able to persist in the processing environment of the LMF sector
There is strong evidence from salmonellosis outbreaks linked to S. enterica strains persisting in the processing environ-
ment of some LMFs. Some examples are a S. Bareilly outbreak (2017–2018) associated with the consumption of a powdered
egg product, where a massive contamination of the equipment of the spray-drying technology, which was confirmed as
the contamination source, was found (Labska et al., 2021). Another example is a S. Poona outbreak (2018–2019), associated
with consumption of rice-based infant formula, where the outbreak isolates were linked by WGS to a 2010–2011 S. Poona
outbreak associated with formula manufactured in the same facility, indicating a persistent contamination source. A dry-
ing tower was identified as the source of contamination (Jones et al., 2019). Also, a S. Mbandaka outbreak (2018), linked
to sweetened puffed wheat cereal, where the thorough investigation of the manufacturing facility identified persistent
environmental contamination with strains closely related genetically to the outbreak strain (Keaton et al., 2022). At last, a
S. Agona outbreak (2008), associated with toasted oats cereal, where the outbreak isolates were linked by PFGE and WGS
to a 1998 S. Agona outbreak associated with cereal manufactured in the same facility, indicating a persistent source of
contamination (Hoffmann et al., 2020; Russo et al., 2013).
No strong evidence was found on outbreaks directly linked to C. sakazakii strains persisting in the LMF processing
environment. However, various studies have reported the isolation of C. sakazakii strains of the same pulsotype or
sequence type from processing equipment and finished product in powdered infant formula processing plants, with
indistinguishable strains being recovered from the production environment for few months up to 4 years (Craven
et al., 2010; Forsythe, 2013; Jacobs et al., 2011; Power et al., 2013). Therefore, there is sufficient evidence that Cronobacter
spp. can persist in dry food processing and preparation environments, particularly in those of infant formula process-
ing plants.
These observations are in agreement with the conclusions of (ILSI, 2011), where several examples of the capacity of
S. enterica and C. sakazakii to survive in dry environments for long periods of time were highlighted, suggesting cross-
contamination and persistence in the manufacturing areas.
Due to its capacity to form endospores B. cereus s. l. species may withstand harsh treatments including evaporation,
drying and disinfection and survive throughout processing. In addition, B. cereus s. l. can produce long-lasting and hard to
remove biofilms in and on equipment. Indeed, some environmental monitoring studies at LMF (mainly powdered infant
formula) processing plants have reported its frequent isolation in FPPE, with high prevalence (in the range of 30%–40%) in
drying and packaging areas (Liu et al., 2018; Zhuang et al., 2019). It is generally acknowledged that B. cereus s. l. and other
spore-forming bacteria are ubiquitous and their presence in raw materials and FPPE appears to be inevitable. However,
despite their capacity to colonise surfaces and equipment, no evidence was found on outbreaks directly linked to B. cereus
s. l. strains persisting in the processing environment of LMF and very limited information was collected from studies sub-
typing strains recovered from FoPE and demonstrating environmental persistence of the hazard for short periods of time
(Liu et al., 2018; Zhuang et al., 2019).
E. coli can be present in the FoPE, where it is commonly regarded as an important indicator of manufacturing hygiene
(Xi et al., 2015). Some studies have revealed the occurrence of human pathogenic E. coli in the processing environment of
some LMF, such as dairy powder factory environments (Duffy et al., 2009). However, no evidence was found on outbreaks
directly linked to human pathogenic E. coli strains persisting in the processing environment of LMF and there is paucity of
studies subtyping E. coli strains recovered along time from particular FoPE.
3.1.3 | Concluding remarks related to the most relevant bacterial food safety hazards associated
with persistence
• With the available evidence, the bacterial hazards of highest PH relevance in the various food and feed production and
processing sectors in the EU/EEA assessed as most relevant for persistence in the respective FFPE of these sectors can be
found in Table 2.
• For example, S. enterica was considered among the most relevant bacterial food safety hazards associated with
persistence in the FFPE of the feed for animal food production sector, the meat sector, the egg sector and the LMF
sector.
T A B L E 2 Overview of bacterial hazards of the highest public health (PH) relevance in the various food and feed production and processing
sectors in the EU/EEA indicating which of those have been assessed as most relevant for persistence in the respective FFPE of these sectors.
Pathogens of highest PH relevance in sector and/or persisting in the FFPE of sector

Bacillus cereus sensu lato
Campylobacter jejuni/coli
Clostridium botulinum/perfringens
Cronobacter sakasakii
Listeria monocytogenes
Pathogenic E. coli
Staphylococcus aureus
Salmonella enterica
Vibrio parahaemolyticus
Notes: Orange cells: bacterial pathogens of highest PH relevance in the specified/specific sector but not considered as most relevant bacterial food safety hazards
associated with persistence in the FFPE in the specified/specific sector; Red cells: bacterial pathogens of highest PH relevance and considered as most relevant bacterial
food safety hazards associated with persistence in the FFPE in the specified/specific sector; blank cells: bacterial pathogens not considered of highest PH relevance in the
specified/specific sector.
Abbreviations: F, feed for food animal production sector; M, meat sector, excluding low moisture food (LMF) products; FS, fish and seafood sector, excluding LMF products;
D, dairy sector, excluding LMF products; E, egg sector, excluding LMF products; FV, fruit and vegetable sector, excluding LMF products; LMF, low moisture food sector.
3.2 | Main (sub)types of the most relevant bacterial hazards involved in persistence and the
main features responsible for their persistence in the FFPE (AQ2 and 3)
A wide range of subtypes of L. monocytogenes, S. enterica and/or C. sakazakii were identified as involved in persistence
in the FFPE from the diversity of experimental studies as retrieved from the literature search. Of note is that the individ-
ual sampling approach used in each retrieved study is a source of uncertainty. Studies showed a variation in sampling
methodology, period of analysis and typing methodology, and used different definitions of persistence, whereas in this
assessment the harmonised criterion used for persistence evidence were matching subtypes of the specific zoonotic
pathogen for at least two sampling events of the FFPE at different time points. Thus, for example, if multiple sources, in-
cluding raw materials, are not included in the longitudinal samplings, the attribution of certain subtypes to persistence
may be uncertain, since some widely distributed subtypes can be repeatedly introduced by raw materials.
Some of the experimental studies reported persistence-related features of the specific hazards or assessed particular
genotypic or phenotypic characteristics of persistent isolates recovered from the FFPE. While the information on features
available for L. monocytogenes was quite vast, more limited evidence was found for S. enterica and C. sakazakii.
Many studies compared a subset of persistent and presumed non-persistent strains in relation to some phenotypes
sometimes with contradictory conclusions. This could be due to the fact that they use strains from different subtypes, that
a presumed non-persistent strain could in some circumstances also have the capability to become persistent, or even to
the different experimental conditions used in different studies.
The following subsections summarise the evidence obtained from the retrieved studies on the main subtypes and
features linked to persistence of the most relevant bacterial hazards identified in AQ1 (i.e. L. monocytogenes, S. enterica,
C. sakazakii). Further studies were only consulted to provide more insights into the role of some of the genetic markers
identified as possibly linked to persistence (identified only in L. monocytogenes).
3.2.1 | L. monocytogenes subtypes and features
3.2.1.1 | Subtypes linked to persistence

L. monocytogenes forms a structured population consisting of four divergent lineages (I– IV). The genetic lineages have dis-
tinct, although at times overlapping, genetic, phenotypic and epidemiological characteristics, with the majority of human
illness being caused by strains in lineages I and II (Painset et al., 2019).
To detect genetic similarities between potentially persistent L. monocytogenes isolates, the subtyping method must be
of high resolution, such as the former gold standard PFGE and currently WGS. The use of different typing methods makes
it difficult to compare persisting subtypes between different studies. An exception is the use of sequence-based typing
methods, where the reporting of persistent CC and/or ST allows to get an overview of the subtypes involved in persistence.
Currently, there are three approaches to compare L. monocytogenes in a food facility using WGS-based subtyping methods.
The first two methods are based on the determination of core genome (cg) and/or whole genome (wg) MLST, where scor-
ing is based on allelic similarity. Strains with less than 10 different alleles in the cgMLST analysis are counted as one genetic
complex and are assigned to a clonal type (CT) (Nüesch-Inderbinen et al., 2021; Stoller et al., 2019). The third approach is
based on reference genome comparison and SNP scoring. For example, isolates with than 25 SNP differences are assigned
to persistent clones in an epidemiologic case (Pasquali et al., 2018).
The literature search confirmed a high number of subtypes being involved in persistence, with a total of 36 persisting
CC types (of the total of 262 CC present in the Institute Pasteur MLST database23), belonging to two lineages. A wide range
of CC types were found persisting in each of the sectors, ranging from 24 CC types in the meat sector to 10 CC types in the
dairy sector and 2 CC types in the LMF sector. Overall, for all sectors, at the lineage level, persistence was mostly reported
for lineage II (129 cases of persistence retrieved in the literature search), followed by lineage I (56 cases) (Figure 2A,B). The
most reported persistent CC types were CC5 (17 cases), CC2 and CC6 (8 cases), of genetic lineage I, and CC121 (22 cases), CC8
(19 cases) and CC9 (18 cases), of lineage II. Twelve individual CC types were found associated with persistence in a single
processing plant.
F I G U R E 2 Overview of the different L. monocytogenes subtypes (clonal complexes; CC) persisting in the FoPE of each sector. (A) Genetic lineage I
CC associated with environmental persistence; (B) Genetic lineage II CC associated with environmental persistence. Note: Based on (Brown et al., 2021;
Burnett et al., 2022; Centorotola et al., 2021; Chase et al., 2017; Chen et al., 2017; Chen et al., 2022; Chen, Burall, et al., 2016; Cherifi et al., 2018; Chiara
et al., 2014; Conficoni et al., 2016; Corcoran et al., 2013; Daeschel et al., 2022; De Cesare et al., 2017; Demaitre et al., 2021; Demaître et al., 2021; Elson
et al., 2019; Fagerlund et al., 2016; Fagerlund et al., 2020; Fei et al., 2015; Gan et al., 2021; Gelbicova et al., 2018; Gelbícová et al., 2019; Guidi et al., 2021;
Guidi et al., 2022; Harrand et al., 2020; Holch et al., 2013; Hurley et al., 2019; Jensen et al., 2016; Kaszoni-Rückerl et al., 2020; Knudsen et al., 2017;
Kovacevic et al., 2016; Lachmann et al., 2021; Lachmann et al., 2022; Lake et al., 2021; Lassen et al., 2016; Lee et al., 2019; Liu et al., 2022; Louha
et al., 2020; Lu et al., 2019; Luth et al., 2020; Madden et al., 2018; Maggio et al., 2021; Maurella et al., 2018; McLauchlin et al., 2020; Melero, Manso,
et al., 2019; Melero, Stessl, et al., 2019; Mohan et al., 2021; Moretro et al., 2017; Muhterem-Uyar et al., 2018; Negrete et al., 2022; Nilsson et al., 2012;
Nowak et al., 2017; Nowak et al., 2021; Nüesch-Inderbinen et al., 2021; Ortiz et al., 2016; Oswaldi et al., 2022; Oxaran et al., 2017; Palaiodimou et al., 2021;
Palma et al., 2017; Palma et al., 2020; Pasquali et al., 2018; Pei et al., 2019; Pérez-Baltar et al., 2021; Ruckerl et al., 2014; Self et al., 2019; Shedleur-
Bourguignon et al., 2021; Smith et al., 2019; Stessl et al., 2020; Stoller et al., 2019; Sun et al., 2021; Thomassen et al., 2021; Tirloni et al., 2020; Truchado
et al., 2022; Veghova et al., 2017; Yan et al., 2015; Yang et al., 2020; Zhang et al., 2021; Zuber et al., 2019). The numbers indicate the cases of persistence
being identified from the studies retrieved through the literature search for each CC and sector.
23
https://bigsdb.pasteur.fr/listeria/, accessed on the 16 October 2023.
Some clonal complexes appear to be widely distributed across several sectors: e.g. CC5, CC8, CC121 and CC204 in five
sectors (Figure 2A,B). CC9 was mainly attributed to the meat sector and was not found persisting in the fish and seafood
sector. CC121, CC2 and CC6 were frequently found persisting in plants from at least three different sectors but were not
reported in the fruit and vegetable sector.
The relative prevalence in humans versus food varies between L. monocytogenes subtypes. In a large European study
with 1143 isolates, a higher relative prevalence of lineage II isolates, especially CC121 and CC9, was reported in foods than in
humans, while the opposite distribution was found for lineage I isolates, especially CC1 and CC4 (Møller Nielsen et al., 2017;
Painset et al., 2019). Similar findings have been reported by others, and it has been suggested that certain lineage I sub-
types, such as CC1, CC4, CC6 are hypervirulent, because they are more common among human isolates, showing high
virulence in in vitro studies, and contain certain specific virulence factors (e.g. internalin genes inlG and inlL, Listeria patho-
genicity islands LIPI-3 and LIPI-4, lapB and vip genes), while certain lineage II subtypes, especially CC121, CC9 and CC31, are
often associated with hypovirulence (truncated inlA gene, lacking specific virulence factors) and are more common among
food isolates (Maury et al., 2019; Muchaamba et al., 2022; Schiavano et al., 2022; Vazquez-Boland et al., 2020). Nevertheless,
L. monocytogenes infection is not solely based on the genetic prerequisites of strains (e.g. the presence of a functional InlA),
but also involves interaction with the immune system. As a result, even hypovirulent strains can cause infection in immu-
nocompromised individuals (Schiavano et al., 2022).
Despite isolates of lineage II are found most often to persist, also lineage I isolates have been shown to persist based on
the literature search (129 vs. 56 reports). CC5 was the single most persisting CC of this lineage, but also subtypes frequently
associated with outbreaks and human illness, such as CC1, CC4 and CC6, have been reported to persist in the food industry.
Examples of FBOs that have been proven to be linked to persistent environmental contamination are two multi-year
outbreaks in England of CC1 and CC2 linked to crab meat (Elson et al., 2019) and a CC19 outbreak linked to salmon in
Germany (Lachmann et al., 2022). In addition, long-lasting (multi-year) outbreaks where the outbreak strain has been found
in the FoPE at a single occasion/sampling (or where it is unclear if found more than once) are also hypothesised to be likely
due to persistence of L. monocytogenes in the FoPE. Examples of such outbreaks are a CC5 outbreak in USA linked to ice
cream (Chen et al., 2017), the large CC6 outbreak associated with meat products in South Africa (Thomas et al., 2020), a CC6
outbreak in the UK linked to meat products (McLauchlin et al., 2020), a CC6 outbreak in Switzerland linked to soft cheese
(Nüesch-Inderbinen et al., 2021), a CC6 outbreak related to frozen corn (McLauchlin et al., 2021; Sarno et al., 2021), two
outbreaks of CC89 (ST391) and CC6 linked to smoked fish in Denmark (Lassen et al., 2016), and two CC8 outbreaks linked
to turkey meat in the Czech republic (Gelbicova et al., 2018) and RTE meat products in Germany (Lachmann et al., 2021),
respectively.
Among the L. monocytogenes subtypes identified as persistent in the retrieved literature (Figure 2), all but four have
been linked to human clinical illness among isolates in the Institute Pasteur database (the exceptions being CC183, CC554,
CC382 and ST1048). Interestingly, CC183 has been associated with recent outbreaks of listeriosis in the US and may ap-
pear to be a potential emerging clonal subgroup of L. monocytogenes of PH importance (Gorski et al., 2022; Kayode &
Okoh, 2022). Previous studies have examined CCs over-represented among human clinical cases of illness, or those with a
higher proportion of clinically derived isolates among the overall isolate population of a given CC group. Considering the
top CCs here identified as persistent from lineage I and lineage II groups, all those from lineage I have been implicated as
important in human clinical listeriosis (i.e. CC2, CC5 and CC6). In the case of the top five CC from lineage II (i.e. CC7, CC8, CC9,
CC121 and CC321), only two of these were among the top 10 human clinical CCs identified in France (CC9 and CC121) (Maury
et al., 2016); in addition to these, CC7 and CC8 were represented among those related to human clinical disease linked to
RTE foods in the EU (EFSA BIOHAZ Panel, 2018). Collectively, this highlights how persistent contamination can have import-
ant PH implications, as these CC most frequently implicated in persistence are clearly important clinical subgroups.
3.2.1.2 | Features associated with persistence and link to subtypes

The persistence of L. monocytogenes in the FoPE is the most studied among foodborne pathogens. Several studies have
focused on finding and understanding the characteristics that allow L. monocytogenes to survive in the FoPE for long peri-
ods of time, mostly related to the adaptability to physico-chemical conditions and to the identification of genetic markers
associated with increased survival capacity (EFSA BIOHAZ Panel, 2018).
The BIOHAZ Panel concluded in 2018 that persistence of L. monocytogenes in FoPE is an often observed and important
phenomenon for the contamination of RTE foods. It was highlighted that some hypovirulent molecular subtypes such as
CC121 seem to encompass multiple isolates with a proven capability to persist too. It was also acknowledged that whether
persistence is a result of improper hygiene conditions or more the effect of strains equipped with an arsenal of genetic deter-
minants is under debate; and that a high adaptive capacity against physical–chemical factors and biofilm-forming capacity
could partly explain the persistence phenomenon. The former EFSA BIOHAZ Panel opinion also identified the following
genetic markers as possibly linked to persistence of L. monocytogenes: a transposon (Tn6188) and the bcrABC cassette, asso-
ciated with tolerance against some disinfectants, and the hypervariable genetic hotspot lmo0443-lmo0449, which appears to
play a role in stress response as it may harbour two independently acting stress survival islets (either SSI-1 or SSI-2).
As a general characteristic of L. monocytogenes, survival or even growth at low temperatures is possible due to an in-
crease in the concentration of unsaturated fatty acids in the cell membrane, that prevents the formation of a gel-like state
that could lead to leakage of cytoplasmic content (Beales, 2004). The expression of cold shock proteins (Csp), which act
as molecular chaperones enabling replication, transcription, translation and protein folding at low temperatures has also
been documented (Eshwar et al., 2017). Additionally, the σB-mediated accumulation of cryoprotectants, such as betaine and
carnitine, confers growth phase-dependent adaptation at low temperatures (Angelidis & Smith, 2003; Becker et al., 2000).
However, the ability to adapt to low temperatures has not been demonstrated to be enhanced in persistent strains, as no
differences in growth rate have been observed between persistent and presumed non-persistent strains at cold tempera-
tures (4°C and 11°C) (Cabrita et al., 2015; Magalhaes et al., 2016).
L. monocytogenes is also able to respond positively to other stresses. Its response to osmotic stress is mainly mediated
by the uptake of potassium cations, glutamate and osmoprotectants, such as glycine-betaine and carnitine, both of which
contribute to maintaining turgidity and help in the stabilisation of protein structure and function (Matereke & Okoh, 2020).
In the absence of osmoprotectans, tolerance to high osmolarity is associated with the expression of the protein Ctc (Gardan
et al., 2003). Although the pathogen presents this general characteristic, its role in persistence is still controversial. (Harrand
et al., 2020) showed that the target strains in a persistent cluster did not grow faster than those included in a non-persistent
cluster at 6.5% of NaCl. Contrary to this, (Magalhaes et al., 2016) reported a shorter lag phase and higher growth rate in a set
of persistent strains compared to presumed non-persistent strains when subjected to 8% NaCl.
Generally, the ability to overcome acidic stress and maintain homeostasis is mediated by the acid tolerance response
(ATR), glutamate decarboxylase (GAD), arginine deiminase (ADI) and F1F0 -ATPase systems (Wiktorczyk-Kapischke et al., 2021).
Albeit a higher growth rate in acidic conditions of persistent strains from cheese processing plants at pH 5 (Magalhaes
et al., 2016) and higher tolerance of persistent strains from meat processing plants at pH 2.4 (Lundén et al., 2008) have
been reported, Harrand et al. (2020) showed a lower growth rate of persistent strains from a salmon processing facility at
pH 5.5 than that of the presumed non-persistent ones. Regarding alkaline stress, L. monocytogenes also has the capacity to
withstand high pH by the increased production of acids and induction of transporters and enzymes responsible for proton
retention and cell surface modifications (Soni et al., 2011).
The ability to adapt to oxidative stress is mediated through the expression of σB, cold and heat shock proteins, proteases
(ClpC, ClpP and GroEL), reactive oxygen species (ROS) detoxification systems, such as catalase (Cat), superoxide dismutase
(Sod) and alkyl hydroperoxidase (AhpCF), and the ferritin-like protein (fri) (Bucur et al., 2018). However, conflicting results
have been obtained regarding the behaviour of persistent and presumed non-persistent strains against oxidative stress.
Manso et al. (2020) reported that a persistent strain (from CC9) isolated in a meat processing plant was more resistant to
cumene hydroperoxide (CHP) at 37°C than a presumed non-persistent strain from the same CC isolated in a cheese pro-
cessing plant, but the opposite occurred with two strains belonging to CC5. Moreover, when the same oxidising agent
was applied at 10°C no differences were found between persistent and presumed non-persistent strains. In line with these
results, Harrand et al. (2020) observed no differences between persistent and presumed non-persistent strains when ex-
posed to 10 mM CHP at 37°C.
Stress survival islet 1 (SSI-1) or stress survival islet 2 (SSI-2) have been detected in persistent L. monocytogenes CC (Tables 3
and 4). SSI-1 (an 8.7-kbp region consisting of five genes: lmo0444, lmo0445, pva, gadD1 and gadT1) has been shown to im-
prove survival at high salt, bile and acid conditions and is widespread among genetic lineage I and II CC (Ryan et al., 2010).
Nevertheless, Harrand et al. (2020) reported that a persistent CC321, carrying the SSI-1 in the genome, was slower to grow
at pH 5.5 and 6.5% NaCl than presumed non-persistent strains lacking this genetic trait. The presence of SSI-2 (comprising
lin0464, coding for a putative transcriptional regulator and lin0465, encoding an intracellular PfpI protease) in L. monocyto-
genes CC121 and CC31 enables them to overcome high pH and oxidative stress conditions (Harter et al., 2017).
Apart from the capacity to withstand a wide range of stresses, L. monocytogenes also has the ability to form biofilms on
materials normally used in the FoPE, such as stainless steel, rubber, polystyrene, glass or polytetrafluorethylene (Nowak
et al., 2021; Sinde & Carballo, 2000). However, depending on the test method used, different results have been observed
(Osek et al., 2022). L. monocytogenes is a relatively weak biofilm former in comparison to many other bacterial species, and
environmental conditions and the microbiota present in the biofilm have a significant impact on biofilm formation by L.
monocytogenes (Rodríguez-López et al., 2018). Interactions with other microorganisms, such as Pseudomonas, Acinetobacter
or Janthinobacterium, can have an important effect on listerial biofilm formation (Finn et al., 2023; Zwirzitz et al., 2021).
The occurrence of persistent L. monocytogenes adapted to sublethal concentrations of biocides in the FoPE has not been
well studied. However, it has been shown that L. monocytogenes cells exposed to biocides can transform into a persistent
state with a higher frequency in mature biofilms (Byun & Kim, 2023). Virulence markers (flaA, ActA, InlA and InlB) and their
regulator operon prfA are important for biofilm development, as shown in deletion mutants that are unable to form bio-
films. In addition, the transcriptional regulator of stress response genes, SigB, is also required in the later stages of biofilm
development. The role of the biofilm associated protein (BapL) is not well understood and is controversially associated
with different serotypes. Truncation of inlA has been associated with increased biofilm production (Franciosa et al., 2009),
whereas inlL mutants showed decreased attachment to surfaces. Myintzaw et al. (2023) identified the presence of bapL to
be specific in CC121, CC14, CC204, CC9 and CC20, while inlL was exclusively present in CC155, CC26, CC37, CC18, CC204, CC20,
CC412 and CC7.
The role of the agrBDCA operon of the signal peptide-based sensing system in biofilm development by L. monocyto-
genes has been questioned, while the LuxS system is important, with mutations in luxS leading to denser biofilms. The DNA
repair and defence protein (RecO) and a putative cell wall binding protein (Lmo2504) were shown to be overexpressed in
biofilms. In addition, prophage insertions into the comK gene are associated with enhanced biofilm production, as demon-
strated in vitro with strains from a 12-year persistent epidemic clone (Colagiorgi et al., 2017; Finn et al., 2023; Orsi et al., 2008;
Verghese et al., 2011).
In the literature, higher biofilm formation by persistent strains, as compared to presumed non-persistent ones, has
been reported in polystyrene at 30°C for 48 h (Nowak et al., 2017) and at 37°C for 24 h, with persistent strains isolated in
a poultry processing plant (Rodríguez-Campos et al., 2019). However, Nilsson et al. (2011) observed no differences in bio-
film formation between persistent and presumed non-persistent strains in polystyrene after 24 h at 10°C, 20°C, 25°C or
37°C. In addition, conflicting results were obtained with persistent and presumed non-persistent strains isolated from a
poultry processing plant in Spain, where the persistent CC121 and CC9 were able to form biofilm in polystyrene as the pre-
sumed non-persistent CC1 and CC87, however the better biofilm former was the presumed non-persistent ST199 (Manso
et al., 2020).
Resistance to heavy metals such as cadmium and arsenic are encoded by genetic markers often located on mobile
genetic elements, mainly plasmids. The cadA1, cadA2, cadA3 and cadA4 genes have been associated with cadmium re-
sistance, with the first two localised on transposon Tn5422 and plasmid pLM80, respectively, and associated with per-
sistence (Hingston et al., 2019; Nelson et al., 2004; Osek et al., 2022). Listeria Genomic Island 2 (LGI2) harbours an arsenic
resistance operon (arsR1D2R2A2B1B2) and Tn544 contains the arsenic resistance cassette arsRDABC (Kuenne et al., 2013;
Nelson et al., 2004) that is correlated with cadmium resistance as it also contains the cadA4 gene. Mixed results have been
observed in the literature regarding the presence of these markers in persistent strains, and more studies are needed for a
clear persistence association. Palaiodimou et al. (2021) found cadmium resistance genes both in persistent and presumed
non-persistent strains, however cadA1, that allows to grow at concentrations of cadmium higher than 140 μg/mL, was more
commonly present in persistent strains. Moreover, LGI2 was present only in two presumed non-persistent strains. Pasquali
et al. (2018) observed differences in the cadmium genetic markers for the persistent ST121 and ST14 recovered from a
rabbit meat processing plant. While none of the ST121 isolates studied harboured LGI2, it was present in 88.89% of ST14
isolates, and moreover, none of the ST14 isolates presented the cadA1C gen while it was present in 87.60% of ST121 isolates.
These authors also confirmed that those isolates carrying the cadA1C gen had higher cadmium chloride MIC.
In L. monocytogenes, several benzalkonium chloride (BC) tolerance genetic determinants have been identified (qacH,
located in Tn6188, bcrABC and emrE), localised on mobile genetic elements and found mainly in lineage II isolates (CC9,
CC13, CC14, CC31 and CC121) (Table 4). There are reports of plasmids carrying emrC, which is particularly identified in CC6
(Table 3), qacA or qacC (Lakicevic et al., 2022). Overexpression of the chromosomal efflux pump MdrL is also capable of
effluxing some antibiotics (such as macrolides or third generation cephalosporins and fluoroquinolones as ciprofloxacin)
and heavy metals (Baquero et al., 2020; Douarre et al., 2022). Although, these genetic markers are well studied, discrepan-
cies in their possible role in persistence have been found in the literature. The efflux cassette bcrABC and emrC were re-
cently detected in both persistent and non-persistent candidates (Palaiodimou et al., 2021). These authors also found that
persistent CC121 isolates were more likely to contain a truncated inlA, SSI-2 and the qacH gene (Palaiodimou et al., 2021).
Recently, Cherifi et al. (2020) showed that the presence of the bcrABC cassette and emrE (present in the LGI1) conferred an
enhanced tolerance to BC disinfectants. Moreover, the presence of CC harbouring the bcrABC cassette was significantly
higher within the group of persistent strains, while two persistent CCs did not present any of these two genes, and only the
persistent CC8 presented the emrE gene. Fox et al. (2011) showed that persistent strains were more tolerant to QAC based
disinfectants than presumed non-persistent strains. However, Magalhaes et al. (2016) reported that the susceptibility to
BC (50 ppm) of persistent and presumed non-persistent pulsotypes from cheese processing environments was similar.
Manso et al. (2020) observed that a persistent strain from CC121, carrying Tn6188, did not show the highest BC MIC, that
was reported by a presumed non-persistent CC1 strain. The authors also observed differences in BC resistance within the
persistent CC9, lacking qacH and the bcrABC cassette, with two isolates showing the highest MIC for both BC disinfectants.
Similarly, Harrand et al. (2020) showed that a presumed non-persistent strain, carrying the bcrABC cassette, had a higher
BC MIC than a persistent strain harbouring the same genetic determinant. Although it is generally unknown whether the
persistent strains characterised in all these studies directly come from settings where QACs have been used and they do
show resistance to QAC concentrations typically used in the industry for disinfection, strains with molecular mechanisms of
resistance to QACs can show a higher tolerance to them than wild type strains, and it is speculated that this could facilitate
survival in some niches through long-term exposure to low doses of disinfectants (e.g. drains).
Some epidemiological studies have tried to identify mobile genetic elements and variable genetic hotspots on bacte-
riophage regions (e.g. comK, tRNA-ArgTCT loci) and plasmid types (e.g. pLM5578, pLM33, pLM80) (Fagerlund et al., 2016;
Kuenne et al., 2010; Schmitz-Esser et al., 2015). A genome comparison of genetic lineage II CC121 strains persisting in dif-
ferent FoPE showed a high degree of conservation among prophages (tRNA-Arg-TCT prophage) and plasmids (pLM6179,
pLM4423 and pLM3253 comparable to pLM5578), suggesting that strong selective pressure has acted on them (Schmitz-
Esser et al., 2015). The same was observed in a genome comparison by Muhterem-Uyar et al. (2018) in genetic lineage I CC5
strains persisting in the processing environment of a cheese plant, which harboured conserved tRNA-Arg-TCT prophages
and a pLM80 prototype plasmid with a bcrABC cassette and genes for heavy metal resistance. Some researchers apply plas-
mid typing, which is a helpful tool to identify the global dissemination of successful plasmids (Anast et al., 2022; Hingston
et al., 2019; Muhterem-Uyar et al., 2018). In a study by Hingston et al. (2019), 26 strains from Canada and Switzerland, cov-
ering many different genotypes, carried an identical plasmid type (pLMG1-7), highlighting the unique conservation of L.
monocytogenes plasmids worldwide. Schmitz-Esser et al. (2021) compared 1037 plasmids from 1921 genomes (54%) and
clearly showed that plasmids were significantly more abundant in L. monocytogenes isolated from food and the FoPE com-
pared to clinical strains, which appears to be a prerequisite for adaptation and dissemination to the environment and food
niche (Tables 3 and 4).
L. monocytogenes has been traditionally considered to be genetically highly conserved. Genetic expansion was not
present to the same extent as in Salmonella or Campylobacter (Buchrieser et al., 2003). However, in epidemiological anal-
yses including food, environmental and animal isolates of L. monocytogenes, it has been recognised that genetic lineage
II is subject to greater selection pressure, with successful CCs and STs having incorporated genetic material from microor-
ganisms in the specific niches (den Bakker et al., 2008; Orsi et al., 2008). Of note, this genetic expansion based on horizontal
gene transfer, mutation or recombination is high among the CCs most frequently associated with persistence in the FoPE.
Thus, according to the Institute Pasteur database (accessed on 16 October 2023), CC5 contains 81 different STs that diverge
in a housekeeping gene, CC2 includes 96 different STs and CC6, which was also involved in the last major Listeria outbreaks,
51 different STs. The genomes of CC9 (n = 94) expanded most frequently in genetic lineage II, followed by CC8 (n = 79 ST)
and CC121 (n = 59).
Certain CC types in genetic lineage II have more stress tolerance genes (Tn6188_qacH transposon, SSI-1 and SSI-2,
plasmid-borne brcABC efflux pump), but have more often attenuated inlA genes due to premature stop codons (PMSC)
than lineage I isolates and none of the Listeria pathogenicity islands 3 (LIPI-3) and 4 (LIPI-4) (Tables 3 and 4). In addition,
the peptidoglycan binding gene (bapL) appears to be specific for genetic lineage II CCs, particularly CC121 and CC9. This
has led to the suggestion that they (e.g. CC9, CC121) can better persist in the FoPE and infect only severely immunocom-
promised individuals (hypovirulence) (Palma et al., 2020). CC121 is the best described L. monocytogenes complex, the first
to be described with a truncation of inlA (hypovirulent) and sublethal adaptation to quaternary ammonium compounds
(QACs) (through genetic traits such as qacH) and appears to be the most prevalent genetic lineage II representative world-
wide. These characteristics have been proposed to be related with persistence in the FoPE (Ortiz et al., 2016; Palaiodimou
et al., 2021).
Nevertheless, despite the fact that some researchers highlight a unique or multiple, genetic markers (e.g. QAC efflux
genes) that may contribute to L. monocytogenes persistence when comparing persistent versus presumed non-persistent
strains (Martínez-Suárez et al., 2016; Mirena et al., 2023), there are findings that support the hypothesis that persistence is
rather an interplay of different genetic markers and the ecological niche (Daeschel et al., 2022; Palaiodimou et al., 2021).
3.2.1.3 | Analysis of clusters of related genome sequences in the NCBI Pathogen Detection database
There were 51 SNP clusters identified for L. monocytogenes with ≥ 100 isolates, as shown in Figure 3. The greatest frequency
was found for CC9, which had five separate SNP clusters, followed by CC7, CC155 and CC554, each with four individual SNP
clusters containing ≥ 100 isolates. Of the top 10 CCs most frequently identified as persistent in the literature (Figure 2), 5
were also among the top 10 largest SNP clusters in the NCBI Pathogen Detection database including strains from both clini-
cal and environmental/other sources (CC6, CC8, CC9, CC121 and CC321).
F I G U R E 3 Number of SNP clusters in the NCBI Pathogen Detection database with at least 100 isolates, by predicted CC identified as persistent in
the literature screening.
An overview of a selection of the largest SNP clusters from the NCBI Pathogen Detection database is shown in Figure 4,
including two clusters related to CCs (CC573 and CC639) that were never reported as persistent in FoPE in any of the re-
trieved studies and six large clusters from CCs more frequently associated with persistence (i.e. CC5, CC6, CC7, CC8, CC9
and CC121). When comparing both groups, the two CCs not associated with persistence events lacked most of the genetic
stress markers described above as possibly related to persistence in the FoPE, including disinfectant tolerance and heavy
metal tolerance markers. In relation to virulence markers, while CC9 and CC121 SNP cluster isolates were associated with
production of a truncated InlA, suggesting reduced virulence, those CCs not associated with persistence (i.e. CC573 and
CC639) carried full length inlA genes. This again suggests that important stress tolerance features of relevance to the FoPE,
and truncations with loss of function of inlA, are more widely represented across persistent subtypes. Similar trends have
been identified previously (Palaiodimou et al., 2021), as already highlighted in Section 3.2.1.2. In relation to the source of
isolation, some clear commodity source associations were seen (e.g. CC9 with meat/pork sources, CC573 with horticulture
products).
However, it is important to note that a SNP cluster will include over-representation of isolates from specific source at-
tribution investigations; for example, if pork meat was implicated in an outbreak related to a given SNP cluster, then this
SNP cluster may be over-represented with isolates from the contaminated pork product, and the FoPE that produced it,
since these would be specifically sampled at higher frequency. Thus, there could be a high number of isolates of a specific
outbreak strain from pork-related sources, represented within that SNP cluster. Of the SNP clusters analysed, the majority
of isolates were collected within the previous 10–15 years; it is notable, however, that some clusters included large propor-
tions of isolates collected over extended timeframes. For example, isolates of the CC121 SNP cluster were collected over a
period of more than 10 years. This suggests that certain clonal subgroups can occur stably and widespread over prolonged
periods.
TA B L E 3 Overview of genetic features of L. monocytogenes lineage I possibly linked to persistence.
CC, STa 1 2 3 4 5 6 59 87 183 (ST382) 217 218 224 288 554

b
Globally spread Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes Yes
Clinical & Yes Yes No Yes Yes Yes No Yes Yes Yes No No No Yes
Environmentc
Hypo/ Hypervirulent Hypervirulent / Hypervirulent / Hypervirulent / Hypervirulent Hypervirulent / / / / /
hypervirulentd
LIPI-hypervirulent LIPI-3 LIPI-3, LIPI-4 LIPI-3 LIPI-3, LIPI-4 / LIPI-2, LIPI-3 LIPI-3 LIPI-3 LIPI-3, LIPI-4 LIPI-3, LIPI-3, LIPI- LIPI-3 LIPI-3
LIPI-4 LIPI-4 3
Stress survival/ SSI-LMOf2365-0481 SSI-LMOf2365-0481 SS-1 SS-1 SS-1 / / / / / / SS-1 / /
responsee
PMSC / PMSC inlA / / PMSC inlA, inlB PMSC actA / / / / / / / /
some
Plasmid groups Group 1, Group 2, Group 1, Group 2, Group 1, Group 4 Group 1, Group 2, Group 1, Group / / / / / / / /
Group 4 Group 1 & 2, Group 1 Group 1 & 2 2, Group 1
Group 4 &2 &2
Plasmidf / pLM33, pLM5578 Group 1 / Group 1 (pLM33 Group 1 Group 2 SG1 / / / pLI100, / / /
(pLMG1-7) (rep25), (pLMST6, (plMG2-13) j1776,
pLMG1-7, QAC efflux pLM33
pLMG1-9, (emrE),
pLMG1-12), resistance to
Group 2 SG2 amoxicillin,
(pLM80, gentamycin)
plMG2-8,
plMG2-10),
LM-F-131
Adaption to LGI 2, arsenic LGI 2, Arsenic LGI 1, QAC efflux pump QAC efflux QAC efflux / LGI 2, arsenic / / cadA and / / /
disinfection resistance resistance, efflux (ladR, lde, (bcrABC, (bcrABC), resistance cadC
(arsA1), QAC efflux pump (bcrABC), mdrL) emrC), efflux efflux pump
efflux (bcrABC), (lde, mdrL) cadA1C1_ pump (clpL, (lde, mdrL),
cadA4C Tn5422 lde, mdrL), cadA, cadC
cadmium
resistance
cass. (cadA1,
cadA2, cadA4),
cadA, cadC
Biofilm markerg comK comK comK comK / / / / / / comK / comK /
Abbreviations: CC, clonal complex; ST, sequence type; PMSC, premature stop codon; SSI, stress survival islet; LIPI, Listeria Pathogenicity Island; QAC, quaternary ammonium compound; LGI, Listeria Genomic Island; /, unknown, not reported.
a
Green cells: major human CC; blue cells: more linked to human; white cells: linked to human and food.
b
Globally spread: a genotype that occurs worldwide in various niches (human, food, environment, animal).
c
Clinical & environment: present in the dataset from the literature search in distiller SR and NCBI genome data comparison related to Figure 4.
d
Hypervirulent, presence of additional Listeria Pathogenicity Islands (LIPIs) and involved in severe infection and documented outbreaks.
e
SSI-1 (stress survival islet-1): adaption to acid and osmotic stress, bile stress in the stomach; single-gene insert LMOf2365_0481 instead of SSI-1; SSI-2 (stress survival islet-2): adaption to alkaline and oxidative stress.
f
Group 1 and group 2 plasmids: typed by plasmid replication protein (RepA sequence). The group 2 plasmids are associated with resistance to heavy metals, other stress conditions and persistence.
g
comK Prophage junction fragments as markers for persistence.
TA B L E 4 Overview of genetic features of L. monocytogenes lineage II possibly linked to persistence.

CC, STa 7 8 9 11 14 20 31 37 101 121 155 199 204 321 403
Globally spreadb Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No
Clinical & Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes No Yes Yes No
Environmentc
Hypo/ / / Hypovirulent / Hypervirulent Hypervirulent Hypovirulent / / Hypovirulent / / / / /

hypervirulentd
LIPI-hypervirulent LIPI-2-inlI LIPI-2, LIPI-3 LIPI-2-inlI / / / LIPI-2-inlI LIPI-2 LIPI-2-inlI LIPI-2-inlI / / / / /
Stress survival/ SS-1 SS-1 SS-1 / / / SS-1 SS-1 SSI-LMOf2365 SS-2, Alkaline and SS-1 SS-1 SS-1 SS-1 /
responsee -0 481 oxidative stress
resistance
Truncated inlA PMSC prfA PMSC actA PMSC inlA, PMSC / PMSC inlA / PMSC inlA / / PMSC inlA PMSC inlA PMSC inlA / PMSC inlA /
prfA, PMSC
actA
Plasmid groups Group 1 Group 2 Group 1, Group 2, Group 1 Group 1, / / / / Group 2 Group 2 / Group 1, Group 2, Group 2 /
Group 1 & 2 Group 2 Group 1 & 2
Plasmidf Group 1 (pLMG1-6, Group 2 SG1 Group 1 (pLMG1-1, Group 1 / / Group / / Group 2, SG1 (plM5578, Group 2 (plmG2-9), Group 1 Group 2 (plmG2-10), Group 2 SG1 Group 1 (pLMST6
pLMST6 (plMG2-1, pLMG1-2, (pLMG1-7, 1(pLM33 plMG2-2), Group plM80 (pLMG1-10) plM80 (plMG2-3) (pLmN12-0935))
(pLmN12-0935)), plMG2-11, pLMG1-4, pLMG1-13) (rep25)) 2 SG2 (pLI100 (pN1-011A/ (pN1-011A/
Group 1 plMG2-12, pLMG1-5, (rep26)), pLM6179 pCFSAN021445/ pCFSAN021445/
(pLMG1-7), Plasmid plMG2-5), pLMG1-7, (pLM6179/ pCFSAN004330/ pCFSAN004330/
pAUSMDU00000235, Group 2 pLMG1-8, pGMI16–004/ pLIS1) pLIS1)
plM80 SG2 (pLI100 pLMG1-11, pCFSAN022990)
(pCFSAN004330/ (rep26), pLM33
pLIS1/ pN1- 011A/ plMG2-4, (rep25)),
pCFSAN021445) plMG2-7) Group 2 SG1
(plMG2-6),
pLM58
Adaption to Tn6188-qaC/lmo2082, LGI 1, QAC efflux LGI 1, LGI 2, Tn554- / / / cadA1C, arsA, / LGI 1, LGI 3, QAC efflux-Tn6188 QAC efflux (bcrABC) QAC efflux LGI 1, LG 2 (TnyfbR- QAC efflux /
disinfection emrC (emrE), Arsenic arsD, QAC (qacH, emrC), (bcrABC) arsA2), QAC (bcrABC),
cadA1C1_ resistance, bCrB and efflux tolerance to BC, efflux (bcrABC), Tn5422
Tn5422, arsA, arsD, bCrC lack (emrC), cadA1C1_Tn5422, TnILP, cadA2C,
Tn6188, QAC efflux cadA, cadA, QAC efflux cadA4C, arsenic
Tn5422, (bcrABC), Tn916 (emrE), efflux resistance
arsD QAC efflux pump (clpL, lde, cassette (arsA-1,
(emrC, emrE), mdrL) arsA-2, arsR,
Tn6188, arsD-1, arsD-2,
Tn5422, acr3)
cadA1C
Biofilm markerg vip-lack, comK comK, vip-lack comK, bapL / bapL bapL vip-lack / comk comK, bapL comK / comK, bapL comK /
Abbreviations: CC, clonal complex; ST, sequence type; PMSC, premature stop codon; SSI, stress survival islet; LIPI, Listeria Pathogenicity Island; QAC, quaternary ammonium compound; LGI, Listeria Genomic Island; /, unknown, not reported.
a
Blue cells: more linked to human; orange cells: more linked to food; white cells: linked to human and food.
b
Globally spread: a genotype that occurs worldwide in various niches (human, food, environment, animal).
c
Clinical & environment: present in the dataset from the literature search in distiller SR and NCBI genome data comparison related to Figure 4.
d
Hypervirulent, presence of additional Listeria Pathogenicity Islands (LIPIs) and involved in severe infection and documented outbreaks.
e
SSI-1 (stress survival islet-1): adaption to acid and osmotic stress, bile stress in the stomach; single-gene insert LMOf2365_0481 instead of SSI-1; SSI-2 (stress survival islet-2): adaption to alkaline and oxidative stress.
f
Group 1 and group 2 plasmids: typed by plasmid replication protein (RepA sequence). The group 2 plasmids are associated with resistance to heavy metals, other stress conditions and persistence. Tn 6188, a novel transposon in L. monocytogenes
conferring tolerance to BC. Tn5422, first-described natural transposon of L. monocytogenes, mediates plasmid-mediated cadmium resistance. Ars, arsenic resistance cassettes are comprised of three (arsRBC) to five (arsRDABC) genes that are
transcribed into a single polycistronic mRNA. Cad, cadmium efflux P-t ype ATPase (cadA) and its repressor cadC, involved in plasmid-mediated cadmium resistance. QacH, transporter, a molecular mechanism leading to increased tolerance to BC. emrE,
gene located on the Listeria Genomic Island 1 encodes for an efflux pump involved in BC tolerance. bcrABC, benzalkonium chloride resistance cassette.
g
comK prophage junction fragments as markers for persistence. BapL, Biofilm-Associated Protein. Vip, is anchored to the Listeria cell wall by sortase A and is required for entry into some mammalian cells.
F I G U R E 4 Overview of genetic features of selected SNP clusters from the NCBI Pathogen Detection database. This includes six SNP clusters related to CC more commonly associated with persistence in food
processing environments (above the black line), and the two SNP clusters present in the database containing at least 100 isolates from CC which have not been identified as persistent in the literature screened for this
scientific opinion (below the black line). ‘inlA*’ indicates isolates harbouring an inlA gene with a premature stop codon, associated with a truncated InlA protein. Presence/absence is indicated by the heatmap, with
numbers relating to the number of isolates in that cluster carrying a given marker. ‘Max SNPs’ refers to the largest SNP difference among isolates within that cluster in the NCBI Pathogen Detection database.
3.2.2 | Salmonella enterica subtypes and features

There are over 2500 different Salmonella serotypes (Grimont & Weill, 2007). Human salmonellosis is caused by a relatively
small number of serotypes, with many serotypes being host specific or unable to cause infection in humans. As in previ-
ous years, the most commonly reported Salmonella serovars in 2021 were S. Enteritidis (54.6%), S. Typhimurium (11.4%) and
monophasic S. Typhimurium (1,4,[5],12:i:-) (8.8%), representing 74.8% of the confirmed human cases. The fourth and fifth
serovars, S. Infantis (2.0%) and S. Derby (0.93%), were at the same levels as in 2020 and 2019, closely followed by S. Coeln
(0.91%) (EFSA and ECDC, 2022).
A wide range of serotypes of Salmonella (n = 36) have been reported to be linked to persistence in the FFPE (see Figure 5).
The greatest number of serotypes of persistent Salmonella were associated with meat processing plants, including sero-
types of human health importance. In the meat processing sector serotypes reported as persisting in multiple plants were
S. Typhimurium, S. Derby, S. Infantis, S. Heidelberg, S. Enteritidis, S. Agona, S. Rissen, S. Kentucky, S. Typhimurium 4,[5],12:i:-
and S. Ohio (Arguello, Carvajal, et al., 2013; Bersot et al., 2021; Bertrand et al., 2010; Boubendir et al., 2021; Bridier et al., 2019;
Charlebois & Horan, 2010; Corcoran et al., 2013; Dantas et al., 2020; Duggan et al., 2010; Gantzhom et al., 2014; Gu et al., 2020;
Kawakami et al., 2019; Kuhn et al., 2013; Medina-Santana et al., 2022; Monte et al., 2020; Monte et al., 2022; Morganti
et al., 2018; Piras et al., 2011; Piras et al., 2015; Prasertsee et al., 2019; Prendergast et al., 2011; Schroeder et al., 2016; Tadee
et al., 2015; van Hoek et al., 2012; Vinueza-Burgos et al., 2019; Wang et al., 2014; Wingstrand et al., 2012; Zeng et al., 2021).
Other serotypes described in single cases included S. Indiana, S. Larochelle, S. Livingstone, S. Muenchen, S. Panama,
S. Paratyphi var Java, S. Saintpaul, S. Schwarzengrund and S. Weltevreden (Boubendir et al., 2021; Gantzhom et al., 2014;
Hiko et al., 2016; Mannion et al., 2012; Ren et al., 2016; Wang et al., 2014; Zeng et al., 2021).
For LMF, S. Agona was reported in two studies (Hoffmann et al., 2020; Russo et al., 2013), and S. Bareilly, S. Bredeney,
S. Mbandaka and S. Poona in single studies (Jones et al., 2019; Keaton et al., 2022; Labska et al., 2021; Viazis et al., 2015). Only
Grasso et al. (2015) reported two serotypes in the same LMF premises, Anatum and Tennessee.
S. Bareilly (Kim et al., 2015), S. Tennessee (Jakociune et al., 2014) and S. Typhimurium (Moffatt et al., 2017) were also re-
ported in individual studies related to persistence in the processing environment of egg and egg products.
In animal FePE 15 different serotypes were reported to persist, and many studies reported the persistence of more
than one serotype. S. Anatum, S. Senftenberg, S. Mbandaka, S. Agona, S. Montevideo and S. Cubana were reported across
multiple studies (Davies et al., 1997; Häggblom, 2012; Löfström et al., 2006; Moretro et al., 2003; Morita et al., 2022; Nesse
et al., 2003; Nesse et al., 2005; Parker et al., 2019; Parker et al., 2022; Pellegrini et al., 2015; Trinetta et al., 2020; Wierup &
Kristoffersen, 2014). Other serotypes reported included S. Agama, S. Liverpool, S. Livingstone, S. Lexington, S. Tennessee,
S. Kedougou, S. Ohio, S. Singapore, S. 4,12,d:-and S. 13,23,i:-(Davies et al., 1997; Eriksson et al., 2005; Gosling et al., 2022;
Parker et al., 2019; Wang et al., 2014; Wierup & Kristoffersen, 2014).
The identification of serotypes of human health importance as being persistent within the meat and egg and egg
products processing environment is not surprising as food categories from these two sectors are the ones more frequently
involved in FBOs of human salmonellosis (EFSA and ECDC, 2022). Interestingly, S. Agona was the only serotype to be found
in persistence events across three different food and feed sectors, namely feed, meat and LMF (Corcoran et al., 2013; Dantas
et al., 2020; Hoffmann et al., 2020; Russo et al., 2013; Wang et al., 2014).
F I G U R E 5 Overview of the various Salmonella enterica subtypes (serotypes) that have been found to persist in in the FFPE of each of the
food and feed production and processing sectors. The numbers indicate the cases of persistence being identified from the experimental studies
retrieved in the literature search for each serotype and sector. Note: Based on (Arguello, Carvajal, et al., 2013; Bersot et al., 2021; Bertrand et al., 2010;
Boubendir et al., 2021; Bridier et al., 2019; Corcoran et al., 2013; Dantas et al., 2020; Davies et al., 1997; Duggan et al., 2010; Eriksson et al., 2005; Friesema
et al., 2014; Gantzhom et al., 2014; Gosling et al., 2022; Grasso et al., 2015; Gu et al., 2020; Häggblom, 2012; Hiko et al., 2016; Hoffmann et al., 2020;
Jakociune et al., 2014; Jones et al., 2019; Kawakami et al., 2019; Keaton et al., 2022; Kim et al., 2015; Kinross et al., 2014; Kuhn et al., 2013; Labska
et al., 2021; Lienemann et al., 2011; Löfström et al., 2006; Mannion et al., 2012; Medina-Santana et al., 2022; Moffatt et al., 2017; Monte et al., 2020;
Monte et al., 2022; Moretro et al., 2003; Morganti et al., 2018; Morita et al., 2022; Nesse et al., 2003; Nesse et al., 2005; Parker et al., 2019; Parker
et al., 2022; Pellegrini et al., 2015; Piras et al., 2011; Piras et al., 2015; Prasertsee et al., 2019; Prendergast et al., 2011; Ren et al., 2016; Russo et al., 2013;
Schroeder et al., 2016; Tadee et al., 2015; Trinetta et al., 2020; van Hoek et al., 2012; Viazis et al., 2015; Vinueza-Burgos et al., 2019; Wang et al., 2014;
Wang et al., 2019; Wierup & Kristoffersen, 2014; Wingstrand et al., 2012; Zeng et al., 2021)

The FFPE poses a variety of challenges to bacteria, including low pH, osmotic and heat stress, starvation, exposure to bio-
cides, microbial competition, etc. In principle, any stress adaptation strategy that assists in survival of hazards in the FFPE,
and/or selects resistant clones that can maintain viability longer than their susceptible isogenic counterparts, could (by
default) be considered as candidate mechanisms of persistence. Such mechanisms are common to most hazards (Begley &
Hill, 2015). Salmonella possesses a broad armoury of stress resistance mechanisms, located in the cell envelope or in the cy-
toplasm (Finn et al., 2013). An indicative (but not exhaustive) list of cell envelope-located mechanisms includes transporters
of potassium (Kdp) or osmoprotectants, such as glycine and betaine (ProU, ProP, OsmU, belonging to the major facilitator
superfamily permeases), and porins (OmpC), all assisting in survival in high salinity environments. Cytoplasm-located stress
response features include the trehalose 6-phosphate synthetase (OtsA) for synthesis of the osmo-protective trehalose (at
low aw), or decarboxylases, e.g. lysine decarboxylase systems (cadB), and chaperones (e.g. GroESL, DnaK), for combating
acid and heat stress, respectively (Begley & Hill, 2015; Finn et al., 2013). However, to identify which of these mechanisms
and how they contribute to persistence, requires a systematic and targeted experimental approach, that combines genetic
and phenotypic traits. In this assessment, an effort was made to capture both the persistence-related features (if any) of
Salmonella and review the features that have been assessed particularly in persistent isolates.
In the available studies where persistent strains of Salmonella were isolated, the genotypic and phenotypic character-
isation (when performed) suggested that isolates had one or more of the following features: (i) AMR and/or resistance to
disinfectants, (ii) ability for biofilm formation, (iii) growth or survival capacity in foods produced in the FFPE where strains
persist, (iv) harbourage of mobile genetic elements, mainly plasmids and (v) carriage of virulence genes and/or a confirmed
cellular invasion phenotype.
It needs to be noted however that these studies describe selected traits of strains without essentially suggesting that
these features are responsible for persistence. Besides, most of these, e.g. virulence, growth/survival and biofilm formation
are typical to most infectious foodborne hazards, including Salmonella. In addition, not all studies isolating persistent
strains evaluated the genotypic and phenotypic features of isolates, nor did they assess a common (and exhaustive) list
of determinants possibly contributing to persistence. Overall, the set of genotypic and phenotypic characteristics of per-
sistent strains assessed in the different literature reports evaluated seems to lack completeness. As such, it is difficult to de-
duce those features that are either indispensable for, or may markedly contribute to, persistence, alone or in combination
with other key genotypic and phenotypic elements.
Strains of the most frequently persistent serotypes in the studies included in this opinion (S. Typhimurium, S. Agona)
were identified as having characteristics enabling antimicrobial resistance, biofilm-forming ability, cell invasion and viru-
lence (for S. Agona only).
There has been diverse evidence about the correlation between resistance or decreased susceptibility to antibiotics and
biocides (disinfectants), suggesting a lack of a systematic mechanism consistent in all resistant isolates. Efflux pumps and
modification of the membrane composition (e.g. increase of short-chain polysaccharide fractions of the LPS) are among
the presumptive mechanisms shared between isolates resistant to alkaline disinfectants and cell wall/membrane-targeting
antibiotics, e.g. β-lactams, cephalosporins and polymyxins (Dubois-Brissonnet, 2012; Gantzhom et al., 2014). In the same
context, adaptation to some alkaline biocides, e.g. via exposure to sublethal levels, i.e. less than the ‘in use’ concentrations,
may select for decreased susceptibility to antimicrobials in Salmonella. In particular, Gantzhom et al. (2014) recorded de-
creased susceptibility to disinfectants, particularly the commercial formulation Incimaxx DES,24 of strains belonging to se-
rovars Livingstone, Typhimurium and Derby, from a pig processing plant. The fact that the resistance to antimicrobials and
alkaline disinfectants share common mechanisms suggests that AMR could potentially assist in the selection of strains
persisting in the FFPE via, among others, their resistance to C&D, but this needs to be further explored in the future in order
to draw solid conclusions.
The AMR in strains of serovars Rissen and Infantis from studies in pork and poultry, respectively, was shown to be
plasmid-mediated (Medina-Santana et al., 2022; Prasertsee et al., 2019). A proportion of 97.5% of S. Infantis isolates from
the meat sector harboured the pESI-like mega plasmid that plays an important role in global AMR dissemination (Medina-
Santana et al., 2022; Vinueza-Burgos et al., 2019).
The second most frequently studied phenotype for S. enterica persistent isolates was biofilm formation. (Dantas
et al., 2020; Jakociune et al., 2014) studied together with AMR for S. Typhimurium, however no specific genes were reported
(Piras et al., 2015). Biofilm community leads to cell ageing and induces general stress response mechanisms that help bac-
teria to maintain their viability and persist under adverse conditions (Alvarez-Ordóñez et al., 2019). This characteristic has
been highlighted as a potential mechanism for persistence within the FePE (Milanov et al., 2017; Velhner et al., 2018), for
months or even years (Prunic et al., 2016; Schonewille et al., 2012; Vestby et al., 2009). Strains from the serovars Enteritidis,
Heidelberg, Ohio, Tennessee and Agona, isolated from polystyrene materials and canvas in poultry plants, and fittings in
egg processing plants, were reported to form biofilms as well as being virulent (Dantas et al., 2020; Jakociune et al., 2014).
This may be partly explained by the fact that invasion in epithelial cells requires pre-establishment of Salmonella on gut
mucosa, which practically involves the expression of mechanisms for biofilm formation (Bai et al., 2021). Of note is that S.
Tennessee strains persisting in egg processing plants had higher growth capacity in pasteurised egg product than the
presumptive non-persistent strains (Jakociune et al., 2014). Quorum sensing (e.g. autoinducer 2 and 3) may further contrib-
ute to virulence expression of cells colonising abiotic or biotic surfaces, and thus, as part of biofilm communities (Horn &
Bhunia, 2018). Nevertheless, as mentioned above, biofilm formation is a generic phenotype expressed by most Salmonella
strains, regardless of being characterised as persistent or not. As such, this phenotype should not be viewed as a definite
feature contributing to persistence. Regarding virulence, Rissen strains were genotypically considered less virulent than
other serovars (Prasertsee et al., 2019). On the contrary, the pathogenicity island genes avrA, ssaQ, mgtC, siiD and sopB and
the fimbrial gene bcfC were present in all Weltevreden and Agona strains inhabiting a poultry slaughterhouse (Dantas
et al., 2020; Ren et al., 2016).
Considering the 36 Salmonella serotypes found in the literature search as linked to persistence in the FFPE, 25 had SNP clus-
ters in the NCBI Pathogen Detection database with at least 100 isolates (Figure 6). The remaining 11 persistent serotypes did
not have clusters meeting this criterion (i.e. 13,23,i:-; 4,12,d:-; Agama; Bredeney, Cubana, Larochelle, Lexington, Livingstone,
Paratyphi var Java, Singapore and Tennesse). The two predominant serotypes reported most frequently in the database
were Typhimurium and Enteritidis (89 and 55 SNP clusters with ≥ 100 isolates, respectively). These clusters typically include
a diverse geographical distribution of isolates, and include samples from clinical, food, FFPE and environmental sources
24
The disinfectant Incimaxx DES contains a mixture of peroxyacids and hydrogen peroxide as active ingredients.
(although clinical isolates are typically the most represented among individual SNP clusters). Of the top 8 serotypes most
frequently identified as persistent among the studies retrieved in the literature search, 6 were also among the top 10 larg-
est SNP clusters in the NCBI Pathogen Detection database, including strains from both clinical and environmental/other
sources (S. Typhimurium, S. Infantis, S. Agona, S. Anatum, S. Heidelberg and S. Mbandaka), which evidences their clinical
relevance and widespread distribution.
F I G U R E 6 Number of SNP clusters in the NCBI Pathogen Detection database with at least 100 isolates, for Salmonella serotypes identified as
persistent in the literature screening.
3.2.3 | C. sakazakii subtypes and features

Studies retrieved in the literature search isolating and subtyping C. sakazakii strains from LMF (mainly powdered infant for-
mula) manufacturing plants have identified indistinguishable strains persisting in the FoPE over wide time periods which
have been characterised as belonging to some particular clonal complexes: CC1, CC3, CC4, CC40, CC58, CC64 and CC83
(Figure 7). Of these, CC1, CC64 and CC83 are the pathovars that have been reported as linked to persistence in the FoPE
on more than one occasion, considering published studies since 2010. In addition, review articles and book chapters have
highlighted the frequent isolation of persistent C. sakazakii CC4 strains from powdered infant formula facilities. For exam-
ple, Muller et al. (2013) undertook a 13-month survey of a powdered infant formula manufacturing plant in Switzerland and
recovered indistinguishable strains from CC4 which were later found to have persisted for up to 4 years (2011–2016). Jacobs
et al. (2011) analysed environmental and final product samples from a milk powder manufacturing plant in Germany over
a 4-year period (2005–2009) from the spray-drying and roller-drying areas. The recovered strains were further profiled
by Sonbol et al. (2013) as four ST of C. sakazakii including the pathovar CC4. A well-characterised persistent strain from a
powdered infant formula production facility in Ireland initially recovered from the powdered infant formula production
environment and persisting for at least 30 months was also from CC4 (Power et al., 2013). In other studies, retrieved from
the literature, strain persistence was demonstrated using PFGE as a typing method, which provides information that is not
interoperable or does not follow universally harmonised terminology for the subtypes.
F I G U R E 7 Overview of the various C. sakazakii subtypes (clonal complexes; CC) that have been found to persist in the FoPE of the low moisture
food sector. The numbers indicate the cases of persistence being identified from the experimental studies retrieved in the literature search for each
CC and sector. Note: Based on (Chase et al., 2017; Fei et al., 2015; Gan et al., 2021; Lu et al., 2019; Negrete et al., 2022; Pei et al., 2019; Yan et al., 2015).

Various studies and review articles have speculated about the features responsible for the special ability shown by C. saka-
zakii to survive for long time periods and persist in the dry conditions of LMF FoPE and the following characteristics have
been highlighted:
• The ability to form biofilms on a variety of abiotic surfaces including silicon, latex, polycarbonate, stainless steel, glass or
polyvinyl chloride (Du et al., 2012; Lehner et al., 2005).
• A high desiccation resistance and a high heat tolerance, as compared to that of other related microorganisms, such as
other members of the Enterobacteriaceae family. Survival under desiccation is related to the high osmotolerance medi-
ated by the synthesis/accumulation of a range of compatible solutes, such as trehalose. Thermotolerance can be due
to the presence of genomic islands (Orieskova et al., 2016). The locus of heat resistance is one of the genomic islands
conferring heat resistance in C. sakazakii (Wang et al., 2021). Interestingly, desiccation resistance is positively correlated
with heat resistance (Fakruddin et al., 2014) and biofilm formation, but apparently it is not linked to particular ST (Du
et al., 2018).
• The production of a capsule that aids attachment in biofilm formation, provides resistance to biocides and contributes
to survival following desiccation due to the entrapment of a shell of water within the capsule (Barron & Forsythe, 2007;
Craven et al., 2010; Fei et al., 2015; Iversen et al., 2004; Osaili & Forsythe, 2009; Yan et al., 2015).
• The production of a yellow carotenoid pigment which stabilises cell membranes and removes reactive oxygen spe-
cies providing protection against oxidative stress, direct UV-radiation and desiccation (Johler et al., 2010; Vojkovska
et al., 2015).
Some of the experimental studies retrieved in the literature search have characterised persistent strains of C. sakazakii in
relation to their biofilm formation ability and associated characteristics (mobility, congo red binding, cellulose production),
and virulence potential. However, apart from disclosing some differences among strains from different CC, such as the ones
observed in biofilm formation ability between CC1 and CC4 strains by Yan et al. (2015), these experiments do not allow to
extract conclusions in relation to the presence of specific features associated with a higher persistence ability by some par-
ticular C. sakazakii subtypes. Some of the strains associated with persistence events in powdered infant formula processing
plants have been also characterised in detail by WGS (Chase et al., 2017; Negrete et al., 2022). These studies have shown
that persistent strains from CC64 and CC83 bear multiple plasmids harbouring virulence genetic determinants (secretion
systems gene clusters), prophages or arsenic resistance determinants, among others, which could be conferring to the host
strain properties of relevance for environmental persistence.
Only 17 clusters contained genomes from both clinical and environmental/other sources, most of them with very few
strains (in some cases just one clinical and one environmental strain), and only three cross-sectoral clusters contained > 10
genomes. Eleven out of these seventeen clusters of closely related genomes (64.7%) were typed as belonging to CC4, with
three of the CC4 clusters spanning multiple years (1982–2005; 2004–2017; and 1950–1980, respectively). The other clusters
contained strains from CC8 (three clusters; 17.6%), CC1, CC13 and unassigned (one cluster each; 5.9%). With the exception of
CC13, clinical isolates have been noted among these CCs (Costa et al., 2021).
3.2.4 | Concluding remarks related to (sub)types and features
• There is a wide range (a total of 43 CC, belonging to two lineages) of L. monocytogenes subtypes reported to be involved
in persistence in the FoPE. Persistence was most commonly reported for lineage II (with the top CC 121, 8, 9 covering 46%
of persistence events for this lineage) followed by lineage I (with the top CC 5, 6, 2 covering 59% of persistence events
for this lineage). These most commonly persisting CC are found in the FoPE from many sectors, with some exceptions
(e.g. CC9 linked mostly to the meat sector). Of the top 10 CC most frequently identified as persistent, 5 were also among
the top 10 largest SNP clusters in the NCBI Pathogen Detection database including strains from both clinical and envi-
ronmental/other sources (CC6, CC8, CC9, CC121 and CC321), which evidences their clinical relevance and/or widespread
distribution.
• Several phenotypic and genomic features for L. monocytogenes have been investigated in relation to persistence in
the FoPE. The markers that have been identified as supporting improved environmental fitness, and possibly associ-
ated with persistence, are: stress survival islets SSI-1 and SSI-2, genomic islands LGI1 and LGI2, heavy metal (cadmium
and arsenic) and biocide (bcrABC, gacC, qacH, emrE and emrC) resistance determinants, often located on mobile genetic
elements (including genomic islands as is the case for emrE, transposons as is the case for qacH or plasmids), and bac-
teriophage regions (comK), globally linked to increased environmental robustness, tolerance to disinfection and/or bio-
film formation. Different combinations of these genetic markers were found in large SNP clusters in the NCBI Pathogen
Detection database from CCs frequently linked to persistence (e.g. CC121, CC9) but not from CCs not yet associated with
persistence (e.g. CC573, CC639). However, it is not clear yet what is the individual contribution of each of these markers
for persistence.
• A wide range of Salmonella serotypes (a total of 35) were reported to be involved in persistence in the FFPE, with
S. Typhimurium and S. Agona being the ones most frequently reported. S. Typhimurium was mainly associated with
persistence in the meat sector while S. Agona was linked to the LMF, meat and feed sectors. Other serotypes frequently
reported as persistent were S. Anatum, S. Senftenberg and S. Mbandaka, mainly linked to the feed sector; and S. Derby,
S. Heidelberg and S. Infantis, only linked to the meat sector.
• Of the top 8 serotypes most frequently identified as persistent, 6 were also among the top 10 largest SNP clusters in the NCBI
Pathogen Detection database including strains from both clinical and environmental/other sources (S. Typhimurium,
S. Infantis, S. Agona, S. Anatum, S. Heidelberg and S. Mbandaka), which provides further evidence for their clinical relevance
and widespread distribution.
• The set of phenotypic and genomic features that have been investigated for Salmonella in relation to persistence in the
FFPE is incomplete. Most of the studies focused on features inherent to most infectious foodborne hazards (e.g. AMR,
virulence, growth/survival in foods and biofilm formation), and reported resistance of some persistent strains to one or
more antimicrobials/disinfectants, carriage of plasmid-mediated virulence factors, biofilm formation ability or reduced
susceptibility to alkaline disinfectants. As such, it is difficult to deduce certain features, that are either indispensable for,
or may markedly contribute to, persistence, alone or in combination with other key genotypic and phenotypic elements.
• The subtypes of C. sakazakii most frequently isolated as persistent clones in the FoPE of powdered infant formula pro-
cessing plants are the pathovars CC1, CC4, CC64 and CC83.
• C. sakazakii CC4 is the subtype most widely represented in the main clusters of related genome sequences (including
strains from both clinical and environmental/other sources) of the NCBI Pathogen Detection database, evidencing its
clinical relevance and widespread distribution.
• Several features have been associated with the ability of C. sakazakii to survive for long time periods and persist in the
dry conditions of LMF FoPE, includingthe ability to form biofilms on a variety of abiotic surfaces; a high heat tolerance
and desiccation resistance; the production of a capsule that aids attachment to surfaces, provides resistance to biocides
and contributes to survival following desiccation; and the production of a yellow carotenoid pigment which stabilises
cell membranes and provides protection against stress. However, none of these features seem to be specifically linked
to those CCs most frequently isolated as persistent clones in FPE of powdered infant formula processing plants.
• Overall, no universal markers or features, responsible for persistence have been identified for any of the three bacte-
rial hazards under consideration. Although the carriage of different combinations of genetic determinants linked to
increased environmental robustness possibly predisposes some particular subtypes to have a better chance of persist-
ing in the FFPE, persistence is a multifactorial process that also depends on specific environmental conditions and risk
factors (discussed below) to take place.
3.3 | Factors at facility level that increase the likelihood of persistence of the most relevant
bacterial hazards in the FFPE (AQ4)
The contamination of FCS or NFCS is the first in a series of events that may lead to persistence, depending on the hygienic
conditions in a FFPE and the capacity of the hazards to persist, as determined by their relevant genetic and phenotypic
features (traits). As such, accidental actions, practices or hygiene failures that favour colonisation of surfaces, instead of
preventing, eliminating or controlling it, are the primary risk factors that increase the likelihood of persistence of most of
the relevant bacterial hazards in the FFPE (e.g. lack of hygiene barriers between unclean (‘dirty’) and clean(er) areas (i.e.
inadequate zoning), uncontrolled movement of personnel or product flow, frequent receival of highly contaminated raw
materials, poor hygienic design or hygienic status of processing equipment or ineffective C&D). The following paragraphs
review the literature evidence as to why and how the aforementioned risk factors are pertained to the different hazards
assessed, in certain food sectors, also detailing the niches (harbourage sites) where persistence has been reported to occur.
3.3.1 | L. monocytogenes in FoPE of four sectors
3.3.1.1 | Environmental niches or site of persistent L. monocytogenes strains

The retrieved studies frequently reported the site (or location) where the persistent strains were isolated but did not always
further investigate the harbourage site (niche) or true location of the persistent strain. Therefore, in many studies the spe-
cific niche is not identified, due to limiting sampling, persistent clones found in several sample sites, etc.
L. monocytogenes was found to persist in a wide variety of sites in the FoPE, demonstrating a comparable split across FCS
and NFCS. As can be seen in Figure 8, FCS sites were frequently identified which provide a direct opportunity for contact
and thus contamination of the associated food products produced in the FoPE. The most common FCS site where the main
subtypes of L. monocytogenes linked to persistence were isolated, across sectors, were conveyor systems/belts. Examples
of other FCS sites where persistent L. monocytogenes was isolated are: crates/buckets/trays (fish and seafood, fruit and
vegetables), carcass cutters/splitters and grinders (meat), gutting-, head/tail removing-, fileting-and skinning machines
(fish and seafood), tables and slicing-and deboning machines (fish and seafood, meat), ice cream-and milk shaking ma-
chines and smear/brine (dairy), packaging lines and mycelium scraping machines25 (fruit and vegetables). The top NFCS
sites where persistent L. monocytogenes was isolated from were drains and floors. The repeated isolation of a clone from
drains and floors may not always indicate that drains/floors are the niches (harbourage sites) of the persistent strain(s), as
these sites may act as collector sites from other sources/niches.
In the review by Belias et al. (2022), persistent Listeria were most commonly isolated from floors (20 studies), drains
(N = 14), conveyor belts (N = 14), slicers (N = 9 studies and additional 11 studies mentioned other cutting machines) and
tables (N = 8). The authors stated that, while these sites are likely to harbour Listeria, the high number of studies that men-
tioned persistent Listeria being isolated from these top five sites may be biased by the fact that these sites were commonly
sampled among all relevant studies.
25
Machine used in mushroom production. Once inoculated, early on in the cultivation of mushrooms, surface scraping is performed before the mushrooms then grow
and are harvested.
F I G U R E 8 Sites associated with Listeria monocytogenes persistence in FoPE. The size of the bubble is proportional to the number of studies
reporting persistence in a given site. (A) breakdown of persistence reported by clonal complex. (B) breakdown of persistence by sector. Site type (FCS or
NFCS) is shown in green, specific site within the FoPE in yellow. Note: FCS, Food contact surface; NFCS, non-food contact surface; CC, clonal complex.
3.3.1.2 | Risk factors for persistence of L. monocytogenes

In general L. monocytogenes is ubiquitous in nature and there are multiple routes for its introduction to the FoPE. Factors
that increase introduction to and spreading of L. monocytogenes in the FoPE may increase contamination, but it is often dif-
ficult to distinguish between repeated reintroduction and persistence. Factors that create niches where L. monocytogenes
is protected against disinfection may increase the likelihood for persistence. In the review by Belias et al. (2022), equipment
cleanability and lack of hygienic zoning were identified as the two most common risk factors for persistent Listeria by the
included studies.
The risk factors or factors that were reported to increase the likelihood of persistence in the studies in our assessment
(since 2010) were mainly related to poor hygienic design of equipment, inadequate C&D of facilities, inadequate zoning/
hygienic barriers, raw materials and humidity.
Poor hygienic design of equipment. The most reported risk factor for persistence of L. monocytogenes was poorly
designed equipment and machines, where moisture and nutrients can accumulate, thus creating a niche in the equip-
ment where L. monocytogenes can persist. A list of equipment where persistence of L. monocytogenes has been observed
is given in Section 3.3.1.1. Common for these niches is often that they are difficult to clean due to poor design and/or low
accessibility and/or to the fact that they contain worn materials with for example scratches, crevices or porous materials.
Irregularities in the equipment and difficult to reach places are suggested to hold appropriate conditions for microorgan-
isms to grow, adhere and adapt. Examples of persistent L. monocytogenes on difficult to clean equipment are a carcass
splitter (Demaitre et al., 2021), a mycelium scraping machine (Sun et al., 2021), floor cracks (Chen et al., 2022), or porous
wall, worn surfaces and cleaning tools (Guidi et al., 2022). Several outbreaks of listeriosis were linked to difficult to clean
equipment where L. monocytogenes persisted. For the caramel apple outbreak in USA in 2014–2015 the outbreak strain
was isolated from wear and damaged equipment, and a wooden bin (Angelo et al., 2017). A six-months outbreak in 2008
in Canada linked to delicatessen meats likely involved a slicer that was difficult to dismantle and clean (Currie et al., 2015).
Introduction or instalment of used and difficult to clean equipment is a risk factor for persistence of L. monocytogenes,
as exemplified for conveyor systems (Fagerlund et al., 2016), a slicer (Fagerlund et al., 2020), and a dicing machine (Lundén
et al., 2002). Also, the large L. monocytogenes outbreak with cantaloupes in the USA occurred after instalment of a line for
washing and drying of melons, previously used for other agricultural products. The outbreak report refers to inadequate
design that precluded effective C&D of the processing line as a likely source of the outbreak in addition to a lack of canta-
loupe precooling (McCollum et al., 2013).
Inadequate C&D of facilities. Several studies report isolation of persistent clones of L. monocytogenes after C&D and
conclude that inadequate C&D was the cause of persistence (Chen et al., 2022; Chen, Wang, et al., 2016; Veghova et al., 2017).
Some studies pointed to poor C&D procedures (Beccalli et al., 2019; Camargo et al., 2015; Chen et al., 2022; Chen, Wang,
et al., 2016) as a risk factor for persistence without providing more specific information, while a few studies provided details
about the discrepancies in the procedures used. For example, a study from a meat processing facility in Slovakia identified
inappropriate floor cleaning using a full-pressure steam system as a likely risk factor (Veghova et al., 2017).
In an outbreak in Switzerland in 2018, where the outbreak clone was found in the cheese processing environment,
the dairy plant had disinfection shortcomings, but no further details were given (Nüesch-Inderbinen et al., 2021). In an
outbreak in the USA with imported Ricotta salata cheese, cross-contamination between cheeses with the outbreak strain
through cutting and repackaging was observed. The need for routinely using validated disinfection protocols and to clean
and disinfect cutting equipment between blocks or wheels of cheese was stated (Heiman et al., 2016).
Self-reporting from five Italian meat processing plants showed that the most commonly neglected C&D hygienic actions
were the following: the correct use of disinfectant concentration, the correct exposure time to cleaning agents, the con-
trol of rinsing water temperature, the appropriate use of cleaning nozzles and avoidance of aerosol formation (Conficoni
et al., 2016). However, this study did not consider persistence.
Inadequate zoning/hygienic barriers. L. monocytogenes can be introduced to the processing environment through
many routes, such as personnel, equipment, animals, dust, water and raw materials. To limit the introduction and spread
of L. monocytogenes the industry is dependent on hygienic zones and barriers between outside and inside and between
low-and high-risk areas in the processing facility. An increased introduction and/or spread of L. monocytogenes may not
necessarily lead to persistence in the FoPE as many strains may be transient, but a high rate of introduction/spread
of L. monocytogenes may increase the likelihood of a given strain to reach a niche and become persistent. For facilities or
sectors with low or insufficient hygienic barriers there may be a continuous introduction of new L. monocytogenes strains
to the facility, in some cases the same clones may be introduced over time making it difficult to distinguish between per-
sistence in the FoPE of the plant or continuous reintroduction of the same strain (it may indicate persistence at a supplier
or in raw materials/outdoor environment). There may be differences between and within sectors regarding zoning and
hygienic barriers. The fruit and vegetable sector often has less strict barriers than other sectors. Departments for produc-
tion of soft cheeses and RTE meat products typically have strict barriers against introduction from departments handling
raw materials such as raw milk and meat. In a newly opened meat facility, a persistent L. monocytogenes strain was believed
to be widespread in the facility due to the lack of hygienic barriers within the facility (Bolocan et al., 2016). Similarly, in a
newly opened dairy processing plant, lack of hygienic barriers and uncontrolled personnel flow led to the spread of a per-
sistent clone (ST204) within the building (Melero, Stessl, et al., 2019). A reconstruction event aimed at an expansion of the
main building of a meat processing facility increased the probability of breaching hygienic barriers and has been linked to
increased introduction of L. monocytogenes (Stessl et al., 2020). In addition, lack or failure of systems creating overpressure
in RTE zones, may enable contaminated air to move directly from unclean areas to cleaner areas, which can contribute to
the spread of L. monocytogenes within the establishment via air (aerosols) (Nastasijevic et al., 2017). In a survey of manage-
ment practices at 32 food producers in Ireland, no management practices correlated with persistence, while separation of
hygiene control areas correlated with a reduction in L. monocytogenes occurrence (Alvarez-Ordonez et al., 2018).
Raw materials. L. monocytogenes is commonly found in many raw materials and may be introduced to the FoPE through
contaminated raw materials. It may be difficult to distinguish between persistent strains and strains repeatedly introduced
with raw materials, the latter may indicate persistence earlier in the food chain or in the outdoor environment. However,
repeated introduction will increase the contamination level and the likelihood of strains reaching niches where they can
become persistent (EFSA BIOHAZ Panel, 2018). Raw materials that are not heat treated are more likely to contain L. mono-
cytogenes and be a source for L. monocytogenes than heat-treated raw materials. In an outbreak linked to stone fruits in the
USA in 2014, it was difficult to evaluate if the outbreak strain was persistent in the facility or was a transient contamination
originating from sources outside the facility, e.g. fruit orchards (Chen, Burall, et al., 2016). Raw materials as a risk factor for
persistence were found most commonly reported for the fruit and vegetable sector, but it is also a risk factor in other
sectors. There are several reports of the same L. monocytogenes clone persisting in several processing plants in the meat
sector. This has been shown for several subtypes (e.g. CC9, CC7 and CC19) in the Norwegian meat sector, and it has been
suggested that such clones persisting in several processing plants can be regarded as pervasive (persistent strains isolated
from different processing plants). The original source of the pervasive strains was not reported, but it was suggested that
raw meat from the same suppliers may have been the original source of introduction (Fagerlund et al., 2022). Also in the
USA, the same persistent subtypes were found in several factories which received raw meats from the same slaughter
plants (Berrang et al., 2010). Lucchini et al. (2023) traced the persistent strains back to the raw meat from slaughterhouses
which are not inspected for L. monocytogenes and seem to be hot spots for its occurrence and persistence. In Germany,
two outbreaks with the same L. monocytogenes clone but associated with meat products from two different producers,
that had common suppliers, lead to the speculation that the outbreak strain could have been introduced to both plants
with raw meat (Luth et al., 2020). In the salmon sector gutted salmon from a producer may be sold as raw material to pro-
ducers of fileted or smoked salmon, and strains persisting in the salmon slaughterhouse may then be transferred via gut-
ted salmon to fileting or smoking plants. In Norwegian salmon processing plants, pervasive clones were found in several
processing plants (Fagerlund et al., 2022). In a cold smoked salmon processing facility in Ireland, which used raw (gutted)
salmon as raw material, the same MLVA type was repeatedly isolated from the raw materials (Dass et al., 2010).
Humidity. L. monocytogenes is commonly isolated from humid niches where it may grow and persist, thus the presence
of such humid niches is a risk factor. Ruckerl et al. (2014) reported intense humidity and steam and water residues on floors
after C&D due to intensified cleaning activities and speculated on this as a possible risk factor for L. monocytogenes per-
sistence on floor associated niches (floor, drains, boots). High humidity in cheese ripening rooms has also been suggested
as a risk factor (Tirloni et al., 2020).
3.3.2 | Salmonella enterica in the FFPE of four sectors
3.3.2.1 | Environmental niches or location of persistent Salmonella enterica strains

As for L. monocytogenes (Section 3.3.1.1), the niche for persistence of Salmonella in the FFPE was often not identified in the
studies retrieved, due to limiting sampling or due to persistent serotypes being found at several sampling sites. Sampling
sites where Salmonella was repeatedly isolated, and therefore may favour persistence, were reported to be NFCS (e.g.
mainly drains, whipping machines, evisceration or pre-cooling areas and personnel clothing) slightly more than FCS (e.g.
feather plucking rubber fingers, evisceration equipment, plucking machines), although, in a number of these studies, the
specific NFCS niche of persistent strain(s) is not clear. Figure 9 presents an overview of persistence of Salmonella in the FFPE
in sites, where the hazard was isolated. Contrary to the relevant evidence for L. monocytogenes, the niches where persistent
strains of Salmonella have been recovered are of lower resolution, pinpointing to broad (e.g. floors, drains, equipment,
lairage, NFCS/FCS) rather than specific sites within a FoPE.
For the meat sector, where detailed, NFCS included floors, drains and matting; FCS included scalding, splitting and
feather plucking equipment, and more generically the slaughter line.
For LMF the location of persistence was generally unclear; two studies identified the drying process as an area of per-
sistence. Very few studies reported Salmonella persistence in egg and egg products, and those who did so were unclear
about the location of persistence. Nonetheless, in egg processing plants, Salmonella is commonly isolated from various
sites, including floor drains, breaker egg diverters or breaker egg belt surfaces. For feed, exact locations for persistence
were often not reported, with data being collated as feed mill equipment or mill environment, or where sampling locations
were provided, data for persistent serotypes was not available (Musgrove et al., 2008).
F I G U R E 9 Sites associated with Salmonella enterica persistence in the FFPE, by serotype. Circle size is proportional to the number of isolates. The
colours indicate the food or feed production sector. Note: FCS: food contact surface; NFCS: non-food contact surface; LMF: low moisture foods.
3.3.2.2 | Risk factors for persistence of Salmonella enterica

Salmonella can survive in dry and dusty conditions. Therefore, animal feed and LMF processing environments support
persistence of this pathogen. However, Salmonella may also persist in the FoPE of high moisture foods, such as eggs, meat
and poultry. Hygiene failures in the reception of raw materials, the zoning of the processing lines, i.e. insufficient separation
of ‘dirty’ from ‘clean’ areas (especially in slaughter houses), with carry-over of the contamination to adjacent cleaner areas,
deficiencies in the hygienic design of the equipment, including the use of food contact materials that may serve as surface
of biofilms, lack of proper ventilation and flaws in the C&D protocols, are the primary risk factors for Salmonella persistence.
The following paragraphs detail the literature evidence associated with each of the aforementioned risk factors considered
contributory, particularly to the risk of Salmonella persistence in the FFPE. Their control may limit or prevent the persistence
of this hazard in the relevant FFPE.
Raw materials and inadequate zoning/hygiene barriers. Contamination of pork during slaughter can occur both
in the ‘dirty’ and the ‘clean’ zones (Arguello, Alvarez-Ordonez, et al., 2013; van Hoek et al., 2012) may eventually lead
to persistence. In the dirty zone, the stages where carcasses are most prone to contamination include stunning and
bleeding, dehairing and polishing (to a lesser extent). The contamination is mainly associated with the hides and the
accumulation of organic matter in scalding water, or the dehairing and polishing equipment. In the clean zone, contam-
ination may occur during evisceration, splitting, trimming and fabrication (preparing meat cuts and deboning) and is
mainly linked to leakage of intestinal content and the use of improperly cleaned equipment (Arguello, Alvarez-Ordonez,
et al., 2013). Once Salmonella enters a slaughterhouse, it may become part of the resident microbiota, inhabiting cer-
tain niches and constitute a renewable contamination reservoir, independent of incoming raw materials (Arguello,
Alvarez-Ordonez, et al., 2013; van Hoek et al., 2012). In poultry slaughterhouses, the most critical contamination sites
are the neck-cutting knife blade, the scald tank, the defeathering and the immersion chill tank (Hiett, 2010). Like pork,
the source of poultry meat contamination is the equipment and the accumulated organic matter released from car-
casses during slaughter, mainly during evisceration. Failure to replenish and disinfect water, or maintain sufficiently
high temperatures (e.g. > 60°C for pork) in water-using processes, increases the likelihood of contamination (Arguello,
Alvarez-Ordonez, et al., 2013; Hiett, 2010). Contamination of the lairage environment over time highlights the impor-
tance of hygienic handling of animals through the slaughter process, as this presents a risk of cross-contamination of
such strains into food processing lines. Within the FFPE, reports of persistent strains detected on floors and mats also
highlights the potential role of footfall of personnel around the FFPE as a means of cross-contamination; this highlights
the importance of PRPs covering workflows and hygiene, for example, the use of disinfecting foot baths, as highlighted
below in Section 3.4.2.
Storing and handling raw materials and finished goods in the same area is a detrimental risk factor that, although rarely
occurs nowadays, has been provenly associated with outbreaks, such as the US peanut butter outbreak in 2008–2009.26
Egg contact surfaces have been shown to protect Salmonella cells, especially within a biofilm community from direct
contact with disinfectants (Yang et al., 2017). In particular, egg belts have been identified as one of the means by which
Salmonella can spread from laying hen houses to egg processing plants (Murase et al., 2001).
The hygiene level of NFCS is also critical for preventing persistence, because persistent strains from NFCS have been
linked to foodborne outbreaks. A typical example is the 2018 sweetened puffed wheat cereal outbreak in Illinois, caused
by S. Mbandaka that originated from drains and the external surfaces of food processing equipment (Keaton et al., 2022).
Poor hygienic design of equipment. The use of equipment that is not well-designed and maintained poses a signif-
icant cross-contamination risk. Crevices in machinery, flooring and walls, and dead ends in piping are potential areas for
pathogen accumulation and subsequent contamination of the final product, especially of LMF. An outbreak of S. Agona
from a toasted oat cereal occurred due to an inadequate design of the manufacturing environment that was subsequently
discovered upon investigation. In this example, the processing machinery was open to the atmosphere (Breuer, 1999).
Jones (2011) suggested contamination in coolers may be elevated because condensation on interior surfaces increases
moisture, which encourages Salmonella growth. In a major outbreak of salmonellosis linked to peanut butter and peanut
paste in the US in 2008–2009, more than 700 cases of illness were reported, and the outbreak may have contributed to the
death of nine individuals. The contamination of the product implicated in this outbreak was partly attributed to the peanut
crackers used as raw materials and further to bad hygiene practices in the processing plant that either did not eliminate the
incoming contamination or contributed to the establishment and dissemination of Salmonella in the processing environ-
ment. Examples of bad practices that could have played a role as risk factors for persistence in this outbreak are the flawed
equipment maintenance and factors discussed in the following paragraphs, or already discussed above, e.g. insufficient
ventilation, lack of lethality treatments and improper cleaning of containers and utensils.
Aeration/ventilation/dust. In feed and LMF processing plants, persistence can be favoured by insufficient aeration
and ventilation or by the dispersion of dust. One study performed in a feed processing plant identified the intake pit as a
location for persistence (or continued reintroduction) (Davies & Wray, 1997), another study reported persistent Salmonella
in the pellet cooler and aspiration system (Häggblom, 2012). Even though the exact source of Salmonella contamination
of peanut butter paste and peanut butter-containing crackers in the 2008 US outbreak was not clearly determined, it was
concluded that Salmonella strains re-occurred in these products possibly due to a persistent contamination reservoir and
prevalence in raw materials. In addition, maintaining moisture of the processing environment, especially of dry foods, as
low as possible minimises the growth and spread of microbial contamination (Grasso et al., 2015).
Inadequate C&D of facilities. This risk factor applies to all sectors, and even more in those associated with foods of
animal origin, especially slaughterhouses. The prolonged survival of Salmonella in the FoPE and the resistance to disinfec-
tants may be further enhanced by the sheltering of cells in biofilms formed on FCS and NFCS that have been inadequately
cleaned and disinfected (den Besten et al., 2016; van Hoek et al., 2012; Wang, 2019). The ability of Salmonella to form biofilms
is widely recognised, with various serovars of Salmonella spp. being characterised as good biofilm formers. Flawed C&D
may enhance formation and maintenance of living biofilms that can persist for long in FoPE. In addition, the reduction of
potential commensal Salmonella competitors by decontamination interventions, e.g. in the poultry chain, in combination
with the suppression of certain serovars by vaccination, has favoured the emergence of new serovars and their establish-
ment in FoPE (Kipper et al., 2022; van Hoek et al., 2012). Inadequate C&D of cutting equipment, such as carcass splitters, may
further enhance the survival and establishment of Salmonella (van Hoek et al., 2012). In dry food processing environments,
such as in feed and LMF producing plants, dry C&D is recommended (Grasso et al., 2015). If chemical (wet) disinfection is
needed, the methods should be reviewed and validated for its efficacy in removing the contamination, without acciden-
tally increasing the condensation of water on FCS and NFCS that may support microbial growth.
3.3.3 | C. sakazakii in the FoPE of LMF sector
3.3.3.1 | Environmental niches or location of persistent C. sakazakii strains

C. sakazakii has been isolated from a wide range of powdered infant formula production environments, including roller dry-
ers, spray dryers, drying towers, tanker bays, packing machines, air filters, vacuum cleaners, tubing, ventilators, fluidised-
bed areas, powder lumps, floors, shoes, trucks or roofs, and has been shown to persist in various of these environments
due to its resistance to desiccation and ability to survive spray drying (Barron & Forsythe, 2007; Osaili & Forsythe, 2009; Yan
et al., 2013). In some studies, the persistent strain(s) were related to dryers and drying towers, vacuum cleaners, tubing or
powder lumps, whereas the niche of the persistent strain(s) was not thoroughly investigated or was not clearly identified/
reported in other studies. Overall, spray-drying, fluidised-bed drying and packing areas of production have been charac-
terised as the main risk areas for contamination (Iversen et al., 2004; Mullane et al., 2007; Mullane et al., 2008). In addition, in
the powdered infant formula production process, textile filters used for exhaust of spray-drying towers have been found to
26
https://w ww.cdc.gov/salmonella/2009/peanut-butter-2008-2009.html
be a particularly risky location for colonisation by pathogenic bacteria, including C. sakazakii, which have caused contami-
nation of the final product (Jacobs et al., 2011). Moreover, C. sakazakii has been isolated at high frequency from locations
in the external factory environment which highlights the need for vigilance in preventing cross-contamination into spray-
drying facilities from external sources.
3.3.3.2 | Risk factors for persistence of C. sakazakii

The contamination of powdered infant formula may take place via the addition of potentially contaminated heat labile
ingredients including starches, proteins, lecithin or gums. Ingredient dry-mix processes are at higher risk of contamination
of the final product. Alternatively, contamination can occur via the production environment. The risk factors identified are
mainly related to inadequate zoning/hygienic barriers; aeration/ventilation/dust; and inadequate C&D of facilities.
Inadequate zoning/hygienic barriers. Several studies have reported hygienic failures as a risk factor associated with
FoPE contamination and/or persistence by C. sakazakii, including violations of the hygienic zoning concept, apertures for
the aeration of the plant, lack of proper control for the exchange of goods through roller shutters, or personnel, air and
powder movement as significant routes of transmission.
Aeration/ventilation/dust. The opening of filters for mechanical cleaning at regular intervals has been reported as
contamination source of the environment with contaminated milk powder (Jacobs et al., 2011). Certain components of
powdered infant formula such as lactose, milk fats and proteins may have a protective effect on pathogens during drying
and the residual nutrients on the equipment and other FoPE may facilitate formation of biofilms by C. sakazakii (Henry &
Fouladkhah, 2019). Aerosolised organisms in dust particles are a route of dissemination of C. sakazakii in powdered infant
formula production facilities. The use of vacuum cleaners in the production area in dry cleaning operations can facilitate
the transmission of C. sakazakii from FoPE through dust (Pei et al., 2019).
Inadequate C&D of facilities. Dry cleaning of FoPE is the preferred option in the LMF sector and encompasses sweep-
ing, brushing, scraping, vacuuming and wiping with cloths to remove residues. However, the use of limited amounts of
water may be required in certain circumstances, for example through controlled wet cleaning followed by a rapid and
thorough drying of the cleaned surfaces immediately after cleaning. Uncontrolled wet cleaning should be avoided. The
presence of water in the dry FoPE of LMF, whether from wet cleaning, condensation or other sources, will lead to significant
and rapid growth of the microorganisms colonising FoPE, especially if temperatures are optimal. Even small quantities of
water may have a significant impact. Moist residues in spots and niches, such as cracks or crevices, which cannot be rapidly
and thoroughly dried, have a particular risk of growth and hence may lead to the build-up of a reservoir of microorganisms.
Cordier (2007) reported a rapid increase of Enterobacteriaceae counts following the presence of water in the environment.
A thorough investigation carried out following an outbreak of Cronobacter spp. Coignard (2006) showed that incriminated
lots contaminated with the same C. sakazakii subtype were manufactured over a period of about 6 months during which
wet cleaning had taken place on several occasions. Condensation and water droplets may be generated through tempera-
ture gradients within the facility or within equipment. The cooling air for LMF such as powders comes into direct contact
with the food. Significant differences between the air temperature and the temperature of the processing plant can cause
condensation in ducts or on FCS if the relative humidity is not correctly maintained.
3.3.4 | Concluding remarks related to factors at facility level that increase the likelihood of
persistence
• The most common sites where persistent L. monocytogenes was isolated are equipment (FCS), drains (NFCS), floors
(NFCS) and conveyor belts (FCS). Repeated isolation of L. monocytogenes from drains and floors may sometimes indi-
cate persistence elsewhere in the processing environment as drains/floors may act as collector sites. Specific equip-
ment where persistence has been described include carcass cutters (meat sector), deboning machines (fish and seafood,
meat), gutting, head/tail removing, skinning and fileting machines (fish and seafood), slicing machines (fish and seafood,
meat), ice cream-milk shaking machines and smear/brine (dairy) or mycelium scraping machines (fruit and vegetable).
• The main risk factor for persistence of L. monocytogenes on FoPE is poor hygienic design of equipment/machines. This
leads to niches which are difficult to clean and disinfect and where food debris and moisture can accumulate, and
L. monocytogenes can persist. Inadequate C&D of facilities and inadequate zoning/hygienic barriers and introduction
through raw materials may also lead to the introduction and spread of persistent clones in the processing environment
increasing the likelihood of strains reaching niches where they can become persistent. In addition, humidity favours
persistence of L. monocytogenes.
• Persistence of Salmonella is most frequently reported in meat and feed processing, where contamination and/or colo-
nisation of equipment is frequently reported. Contaminated equipment such as cutting/slicing, feather plucking equip-
ment, cyclones, etc. contact food/feed directly during processing, and can lead to cross-contamination of end products.
• The main risk factors for Salmonella persistence in a FFPE are: (i) inadequate zoning and hygiene barriers, that enables
contamination of clean areas with residues from contaminated areas (‘dirty’), especially in meat, poultry and egg pro-
cessing plants; (ii) flawed hygienic design, which may lead to the harbouring of strains in niches, and protect them from
exposure to disinfectants; (iii) lack of sufficient aeration/ventilation or presence of dust (especially in dry, LMF/feed pro-
cessing environments); and (iv) the inadequate C&D of the facilities, which enables Salmonella to become a member of
the resident microbiota and constitute a standing reservoir of recontamination of incoming product batches.
• The most common sites for persistent contamination with C. sakazakii are the drying and packing areas of production,
while the dryers and drying towers, vacuum cleaners, tubing and powder lumps have been identified as harbourage
sites for persistent strains in LMF (especially powdered infant formula) factories.
• The main risk factors for C. sakazakii persistence and/or cross-contamination in LMF factories are related to inadequate
zoning/hygienic barriers (e.g. violations of the hygienic zoning concept, apertures for the aeration of the plant, lack of
proper control for the exchange of goods or personnel, air and powder movement); aeration/ventilation/dust (e.g. air-
borne transmission by the opening of filters for mechanical cleaning or the use of vacuum cleaners in dry cleaning op-
erations); and inadequate C&D of facilities (e.g. presence of water in the FoPE, whether from wet cleaning, condensation
generated through temperature gradients within the facility or within equipment or other sources, which would allow
the pathogen to grow).
3.4 | Measures for monitoring, preventing and/or controlling the persistence of the most
relevant bacterial hazards in the FFPE (AQ5)
An overview of the measures for prevention and/or control of persistence of bacterial hazards in the FFPE can be found
in Figure 10. FSMS are the result of the obligation to comply with the general hygiene rules mentioned in Regulation
(EC) No 852/2004, the requirement to establish a permanent procedure based on the HACCP principles mentioned in the
same Regulation, and general aspects such as the precaution principle and traceability mentioned in Regulation (EC) No
178/2002. Before implementing HACCP principles, general hygiene needs to be on point using basic rules necessary to op-
erate hygienically (PRP). The number and type of PRPs depend on the sector, but overall, many of the proposed PRPs in the
Commission Notice (EC) No 2016/C 278/01 contribute to prevent and/or control microbial persistence in the FFPE through
avoiding the entry of the hazards to the processing plant or their spread across the facility.
F I G U R E 1 0 Measures for prevention and/or control of persistence of bacterial hazards in the FFPE (based on Commission Notice (EC) No 2016/C
278/01, EFSA BIOHAZ Panel (2017, 2020) and Tuytschaever et al. (2023)).
The verification that the FFPE is free from hazards is achieved through regular sampling and testing activities. When a
biological hazard is detected, it is recommended to promptly react. A ‘seek-and-destroy’ approach has been frequently
recommended for that (Belias et al., 2022; Malley et al., 2015; Tuytschaever et al., 2023). This is a systematic way to find sites
or niches of persistent strains in food/feed processing plants, with the goal of eradicating them or mitigating their effect
(Malley et al., 2015). It includes intensified sampling and testing, the introduction of measures to control the event and the
continuation of the intensified monitoring programme. Alternatively, systematic ‘root cause analyses’ can be applied to
identify the most probable factors/sites within the facilities contributing to the problem and define the most appropriate
interventions to eliminate the pathogen from the premises, as proposed by Belias et al. (2021). Although the protocol for
performing root cause analysis focused on elimination of Listeria in an apple packinghouse, the protocol and procedures
are expected to be broadly applicable to different food processing operations. A key step for seek-and-destroy and root
cause analyses is identifying the source of contamination, which is commonly referred as source tracking. To enable effec-
tive source tracking and to understand if persistence in the FFPE is involved, methods that are able to subtype isolates are
critical to provide the discriminatory power to determine if isolates coming from different environments or sampling times
are the same or closely related to each other (Baert et al., 2021).
The subsections below describe the main activities for sampling and testing and the PRP activities that can be applied
in processing plants to detect (Section 3.4.1) and prevent (Section 3.4.2) persistence, respectively, and summarise the evi-
dence found in the literature on the interventions triggered by persistence events (Section 3.4.3).
3.4.1 | Sampling and testing in the FFPE
The regular microbiological testing of FFPE is widely recognised as a requirement in the production of many types of food
as a critical component of any FSMS as a means of verification that food safety measures in place are effective. A well-
designed environmental sampling and testing programme can allow manufacturers to early detect hazards that have be-
come persistent and verify that existing control measures are effective to prevent or remove contamination or persistence.
Moreover, trend analysis of results can serve as an early warning, which will allow producers to rectify problems before they
become a serious risk (Bourdichon et al., 2021).
When designing an environmental sampling and testing programme, it is important to consider and fully document: (i)
the identification of sample locations and the reasons why these were selected; (ii) the target organism(s); (iii) the fre-
quency and timing of sampling and the number of samples; (iv) the sampling protocol and defined test methods; (v) and
the recording and evaluation of results, with defined limits for acceptable and unacceptable results and follow-up actions
in case of non-compliant results (Bourdichon et al., 2021; EFSA BIOHAZ Panel, 2020). Many guidance documents and litera-
ture review articles have been published providing recommendations for an effective routine testing in food processing
environments and can be consulted for more detailed information on these aspects. Most of them apply to L. monocyto-
genes. Examples include, among others: (1) the EURL Lm Technical Guidance Documents on Sampling the Food Processing
Area and Equipment for the Detection of L. monocytogenes' (EURL-L. monocytogenes and ANSES, 2023);27 (2) the document
from the Food and Drug Administration (FDA) ‘Testing Methodology for Listeria species or L. monocytogenes in Environmental
Samples’ (FDA, 2015); and (3) the ‘Guidance on environmental monitoring and control of listeria for the fresh produce in-
dustry’ (UFPA, 2018). A previous EFSA scientific opinion on the public health risk posed by L. monocytogenes in frozen fruit
and vegetables (EFSA BIOHAZ Panel, 2020) and technical report (EFSA, 2018), also provided recommendations on process-
ing environment testing for L. monocytogenes in a freezing plant or handling facility for frozen vegetables. Most of the
recommendations included in this former EFSA opinion are also fully applicable to processing plants from other food sec-
tors. Likewise, guidance is also available in the Code of Hygienic Practice for Powdered Formulae for Infants and Young
Children for the establishment of monitoring programmes for Salmonella, Cronobacter species and other Enterobacteriaceae
in high hygienic processing areas and in powdered preparation units (CAC, 2008; FAO and WHO, 2008). The recent review
by Tuytschaever et al. (2023) on L. monocytogenes in food businesses highlights the intervention strategies embedded in
the food hygiene regulation and provides guidance on hygiene, control measures and FoPE testing strategies of L. mono-
cytogenes in the food industry.
Specifically talking about persistence, the selection of sampling points should take into account areas that have tested
positive previously or that are likely to be contaminated, such as wet areas, hard to reach places and poorly cleanable equip-
ment, and FFPE more frequently linked to persistence for each specific hazard and production sector (see Section 3.3). It is
recommended to randomly rotate sampling sites across sampling times. Furthermore, in the case of special events linked
to increased risk of persistence (e.g. construction), a specific sampling plan should be developed to investigate the po-
tential presence of harbourage niches that could be accessible due to the modification, even if temporary, of the process.
Likewise, a specific sampling plan should be established following ‘non-conformities’, where intensified samplings around
the initial contamination site, considering also different categories of product proximity, should be performed to assess
how far the contamination is spread and to identify potential harbourage sites.
Although sampling and testing activities commonly include the enumeration of ‘hygiene indicator’ microorganisms (e.g.
aerobic mesophilic counts, coliform and/or Enterobacteriaceae counts, yeast and mould counts, or Listeria spp. counts), to
detect persistence, particular pathogenic microorganism(s) must be targeted, selected depending on the product being
produced and its intended use. Considering the outcomes of AQ1 (Section 3.1), the most obvious choice would be L. mono-
cytogenes in plants producing RTE foods from different sectors, Salmonella in facilities processing feed, meat, egg products
or LMF and C. sakazakii in powdered infant formula processing plants.
Investigations require the detailed characterisation of isolates of the specific hazard(s) recovered from positive samples
through subtyping methods with enough resolution, preferably through WGS-based subtyping schemes, which provide
a higher resolution than traditional serotyping or older genotyping methods, previously considered as the gold standard,
such as PFGE, MLST or MLVA typing. This detailed analysis will facilitate confirmation of the presence of a persistent strain
and identification of its niche within the facility, providing very valuable information for the design of control approaches,
and can also be used in outbreak investigations (Pightling et al., 2018). Storage of isolates over time will be advantageous
when studying persistence, as isolates from different time periods can be typed and compared. For L. monocytogenes it has
27
This update includes new features adding the presentation of important concepts, such as persistence, biofilm or viable but non-culturable (VBNC), and new guidance
for practical implementation (link to data sheets and video tutorials).
been shown that more than one clone can be found in a single environmental sample. This means that it is advisable to type
several isolates from each sample/enrichment to capture the whole diversity of the sample (Sullivan & Wiedmann, 2020).
Although pre-defined thresholds are often applied to identify case clusters and their potential sources in epidemiolog-
ical surveillance, known outbreak-specific features such as pathogen mutation rate and duration of source contamination
are rarely considered. Duval et al. (2023) developed a hypothesis-based model that estimates genetic distance thresholds
and mutation rates for point-source single-strain food or environmental outbreaks. This forward model, applicable to food-
borne or environmental source single point case clusters or outbreaks, is useful for epidemiological surveillance and may
inform control measures.
Recent technological advances are paving the way to the design of novel approaches of environmental testing based
on the untargeted analysis of the microbiome of FFPE through metagenomics, which can also allow the reconstruction and
characterisation of Metagenome Assembled Genomes (MAGs) from multiple taxa in a single analysis, with huge potential
benefits for source tracking and investigation of persistence. These approaches are still under development and have some
associated challenges and limitations, partly discussed in a previous EFSA scientific opinion (EFSA BIOHAZ Panel, 2019.
However, considering recent major improvements (Barcenilla et al., in press), it is foreseen that these innovative technolo-
gies will be integrated soon in environmental testing programmes.
Recent studies have shown that the number of samples to be taken can be calculated following mathematical or sta-
tistical approaches (Zoellner et al., 2018), and that the performance of effective testing programmes can be also validated
through mathematical modelling. For instance, (Zoellner et al., 2019) applied an agent-based modelling, developing a
model called EnABLe (‘Environmental monitoring with an Agent-Based Model of Listeria’) which allowed an in-silico ap-
proach to map Listeria persistence and dispersal and to evaluate interventions using a data-driven methodology. Likewise,
for trend analysis, statistical tools implementing binomial tests based on subtype frequencies or on previous positive
results have been applied to support identification and management of persistent L. monocytogenes contamination in
smoked fish processing plants (Malley et al., 2013).
3.4.2 | Hygienic measures
Various activities of the PRPs, which include basic conditions and actions to maintain a hygienic environment, are neces-
sary to avoid the entry of hazards into the processing plant or their establishment and/or spread within the facility, hence,
to also reduce the risk of persistence. Among all the available PRPs, the following prerequisites are of particular importance
in relation to bacterial persistence in the FFPE: infrastructure (building, equipment), C&D, technical maintenance and cali-
bration, water and air control, personnel (hygiene, health status), working methodology and food safety culture. Some PRP
(i.e. pest control, raw materials (supplier selection, specifications), food redistribution and donation, waste management,
temperature control of working and storage environment, product information and consumer awareness, and traceabil-
ity) are considered of a lesser importance in relation to microbial persistence in the FFPE, while two PRP (i.e. physical and
chemical contaminations from production environment and allergens) are not applicable.
All these PRP activities may be specific for each production sector, or factory type, with some documents being avail-
able capturing the most relevant ones in some food productions, such as in the recent EFSA scientific opinion on the
public health risk posed by L. monocytogenes in frozen fruit and vegetables (EFSA BIOHAZ Panel, 2020), which can be
consulted for more detailed information on these aspects. More information can be found in the Commission Notice (EC)
No 2016/C 278/01 and the recent review by Tuytschaever et al. (2023).
3.4.2.1 | PRP infrastructure (building, equipment)

According to Commission Notice (EC) No 2016/C 278/01, the proximity of potential contamination sources should be con-
sidered when assessing the risk from the location and surrounding areas. The factory lay-out should strictly separate con-
taminated (high risk) from clean areas (low risk). Floors and walls should be waterproof, non-absorbent, washable and
without fissures. Automatic door opening should be considered. Defined storage facilities should be available for raw
material. Attention should be paid to the possibilities whereby the use of equipment can result in (cross-) contamination
of food aiming to prevent contamination: (i) of the equipment by the environment, e.g. condensation dripping from ceil-
ings; (ii) within the food handling equipment, e.g. accumulation of food residues in slicing devices; (iii) by raw materials, e.g.
separate equipment or clean and disinfect the equipment between uses, for example when using it for raw and cooked
products.
Design and organisation of infrastructure, equipment and devices is thus important to prevent bacterial persistence in
the FFPE, in particular the prioritisation of areas of a facility according to levels or required hygienic care (hygienic areas)
and surfaces within each area designated into zones (also known as zoning). In addition, the following requirements re-
lated to the hygienic design of equipment are important: the materials used for construction, the surface roughness, the
accessibility of all parts of the equipment for inspection, the self-draining (i.e. no liquid collection) and the absence of
niches (e.g. welds should be flush and free of pits, occlusions and corrosion).
3.4.2.2 | PRP cleaning and disinfection

Cleaning and disinfection are critical operations to prevent bacterial persistence in the FFPE. According to Commission
Notice (EC) No 2016/C 278/01, it needs to be considered what, when, how and by whom to clean and disinfect. Typical steps
should be the removal of visible dirt, followed by cleaning, rinsing, disinfection and rinsing again. Cleaning should start
in high-risk areas and end in low-risk areas and different materials and equipment should be used for low and high-risk
areas. Special attention must be paid to the contamination of already disinfected surfaces due to splash when rinsing other
surfaces. Potable water and/or cleaning agent or disinfectant should be used as much as needed to gain the desired effect.
The water should be at an appropriate temperature and the chemicals should be used as per the manufacturer's instruc-
tions using available technical information (e.g. instructions for use, active component, contact time, concentration). Visual
checks on cleaning and sampling for analysis should be used.
Therefore, it is important to have a well-documented C&D programme, reporting if the equipment must be disassem-
bled, the method of C&D (e.g. foam cleaning, cleaning-in place (CIP)), types and concentration of cleaning compounds and
disinfectants, and times/temperatures/pressures to be used (PROFEL, 2019). The efficacy of the disinfection programme
should be verified to avoid harbourage sites of bacterial pathogens. C&D is commonly carried out following wet clean-
ing approaches, as disinfectants require the presence of water or traces of moisture for the active agents to be effective.
Nevertheless, rooms should be kept as dry as possible afterwards as moisture fosters growth and transfer of surviving
bacteria. This is particularly important in plants producing LMF, where dry cleaning procedures are mostly followed. Dry
cleaning can encompass a simple rinsing or flushing of the processing line with the subsequent product or with a neutral
matrix. Alternatively, the removal of product residues with a vacuum cleaner followed by scraping or brushing of the FCS
is performed in situations where simple flushing is insufficient. Vacuum cleaners can also be used to collect loose powder
and dust residues as well as residues dislodged by brushing and scraping.
Particular attention should be paid to the cleaning step, as disinfection of an improperly cleaned surface is ineffective.
The rotation of disinfectants can be considered to avoid adaptation and development of tolerance or resistance by
surviving bacteria.
Special attention should be paid to C&D of potential niches of persistent strains, for instance drains. Drains should be
cleaned and disinfected in a manner that prevents contamination of other surfaces in the room. Utensils for cleaning drains
should be easily distinguishable and be dedicated to that purpose to minimise the potential for contamination. Floor
drains should not be cleaned during production. If a drain backup occurs in finished product areas, production should stop
until the water has been removed and the areas have been cleaned and disinfected. Employees who have been cleaning
drains should not contact or clean FCS without changing clothes, washing and disinfecting hands (FAO, 2007).
3.4.2.3 | PRP technical maintenance and calibration

A preventive maintenance plan, drawn with appropriate inspection frequencies to minimise the risks, is also important to
prevent bacterial persistence in the FFPE. According to Commission Notice (EC) No 2016/C 278/01, the maintenance plan
should be considered with a technical specialist. The plan should include ‘emergency’ procedures when equipment is
defective and instructions for preventive replacement of seals, gaskets, etc. Attention should be paid to hygiene during
maintenance operations. Calibration of monitoring devices is important in controlling food safety and hygiene.
Inspection of equipment for damage is critical and should be conducted during pre-operational checks and during
preventive maintenance activities.
3.4.2.4 | PRP water and air control

Water quality is a relevant factor to be controlled to prevent transmission of hazards and their persistent colonisation of the
FFPE. Large volumes of water are commonly used in some food processing sectors for washing, cooling and transport of
food, among other uses, and for the cleaning of the FFPE. If the water quality is not well maintained, this can cause cross-
contamination with biological hazards (Gil et al., 2015).
In addition to the quite detailed requirements in Chapter VII of Annex II to Regulation (EC) No 852/2004, according to
Commission Notice (EC) No 2016/C 278/01, regular own microbiological and chemical analysis of water directly in contact
with food (unless community potable water) should be carried out. If community water is held in a tank prior to use, the
tank must be part of a regular cleaning schedule. At least clean water, or where applicable clean sea water,28 should be
used in other cases.
A water safety management plan needs to be elaborated in function of the source and quality of the water and the
water disinfection technologies (PROFEL, 2019) and should include the sampling and analytical procedures for the verifi-
cation of the quality of the water. Good manufacturing practices (GMP) and GHP related to a water management plan and
the implementation of a water management system are critical to maintain the microbiological quality of the water used
in handling and processing operations. This has been concluded for the post-harvest handling and processing operations
of fresh and frozen fruit, vegetables and herbs. Identified hygienic practices included technical maintenance of infrastruc-
ture, training of staff and cooling of post-harvest process water and intervention strategies (e.g. use of water disinfection
treatments and water replenishment) suggested to maintain the microbiological quality of process water. These water
disinfection treatments must be undertaken following an appropriate water management strategy including validation,
operational monitoring and verification (EFSA BIOHAZ Panel, 2023). Other examples are available for the safety and quality
28
Clean water: ‘water that does not compromise food safety in the circumstances of its use’. It is clean seawater (natural, artificial or purified seawater or brackish water
that does not contain microorganisms, harmful substances or toxic marine plankton in quantities capable of directly or indirectly affecting the health quality of food) and
fresh water of a similar quality according to the Regulation (EC) 852/2004.
of water use and reuse in the production and processing of fish and fishery products (FAO and WHO, 2023b) and dairy
products (FAO and WHO, 2023a).
Regarding air quality, according to Commission Notice (EC) No 2016/C 278/01, ventilation systems should be kept clean.
For high risk/care areas requiring air control, implementation of positive air pressure systems and appropriate air filtering
systems should be considered. Condensation is mostly the result of poor ventilation and should be avoided in areas where
food is being produced, handled or stored, especially if exposed or not packed.
Filtering air that enters production zones may also be an effective measure to avoid the entry of hazards to the pro-
cessing plant or their spread across the facility (Podolak et al., 2010), hence to prevent their persistence. Filters should be
cleaned and replaced on a regular basis and the system validated for removal of microorganisms (Mullane et al., 2008),
especially when the air comes directly into contact with the food product.
3.4.2.5 | PRP personnel (hygiene, health status)

Training programmes for personnel in proper handling and cleaning practices are essential to raise awareness about per-
sistence of bacterial hazards in the FFPE and assure hygienic practices are accomplished, mainly among the operators
involved in C&D or maintenance activities. According to Commission Notice (EC) No 2016/C 278/01, personnel should be
aware of hazards from gastro-intestinal infections, hepatitis and wounds and should report relevant health problems to
the manager. Hands should be washed regularly and disinfected if necessary. Disposable gloves used hygienically can be
effective in preventing cross-contamination when handling RTE foods.
Avoiding traffic of personnel between different areas (especially from dirty to clean areas) is also very important, as the
personnel can transfer bacteria across the facility and to the end product.
3.4.2.6 | PRP working methodology

According to Commission Notice (EC) No 2016/C 278/01, clear instructions should be provided on proper operation of
equipment. Work instructions or standard operation procedures (SOP)29 should be clear, accurate and simple, and easily
accessible.
Personnel should be supervised when following SOP for C&D procedures, with regular maintenance, inspection of
cleaned and disinfected equipment and audits of the whole process. The development and use of SOP are an integral
part of a successful quality system as they provide individuals with the information to perform a job properly and facilitate
consistency in the quality and integrity of a product or end results. Traffic flow patterns for personnel, food products, food
packaging materials and equipment need to be controlled.
3.4.2.7 | PRP food safety culture (FSC)

According to Commission Notice (EC) No 2016/C 278/01, the components of a FSC are (i) commitment of the management
and all employees to the safe production and distribution of food; (ii) leadership in the production of safe food and to en-
gage all employees in food safety practices; (iii) awareness of food safety hazards and of the importance of food safety by
all employees in the business; (iv) open and clear communication between all employees in the business and (v) availability
of sufficient resources to ensure the safe and hygienic handling of food.
3.4.3 | Actions triggered by persistence
Considering the recommendations made by various authors and in different guidelines to industry, the detection of bacte-
rial persistence in a food processing plant, or even the suspicion, should trigger an immediate response, following a ‘seek-
and-destroy’ or a ‘root cause analysis’, approach. Intensified sampling and testing (preferable including typing), following
the general principles described in Section 3.4.1, can allow an assessment of how far the contamination is spread across the
facility and identify of the niche. Vector sampling involves collection and testing of additional samples in various directions
(i.e. vectors) from the sample location(s) that tested positive. If the initial positive samples were collected during process-
ing, follow-up samplings after C&D can allow an assessment of the efficacy of the protocols applied and detect potential
persistent contamination (Malley et al., 2015).
Different follow-up activities can be implemented with the aim of introducing measures to control the event, and the
choice of control measures must be decided in a case-by-case manner, i.e. specific for each processing plant.
Finally, the continuation of the intensified monitoring programme is required to confirm the efficacy of the measures
taken or the need to take new ones.
Relevant studies describing actions triggered by persistence with the aim to eliminate persistent L. monocytogenes from
the processing environment are shown in Table 5. Actions included the introduction of new or specialised (deep) C&D (of
more difficult to clean equipment) (Biguzzi et al., 2012; Blatter et al., 2010; Stessl et al., 2020); the implementation of work-
flows (Dalmasso & Jordan, 2013); the installation of a new drainage system with back-flow prevention (Jordan et al., 2012);
the implementation of structural changes and renovations (Nakari et al., 2014); the control of the contamination of raw
ingredients and the improvement of the compartmentalisation (Ortiz et al., 2010); the reduction of all contamination
29
A set of written instructions that documents a routine or repetitive activity followed by an organisation.
pathways of the manufacturing area to the maximum extent practicable (Ortiz et al., 2014); the use of various chemical in-
terventions (Malley et al., 2013), the implementation of one or a combination of hygienic measures (Murugesan et al., 2015)
or the simultaneous implementation of various corrective actions (Pennone et al., 2020; Strydom et al., 2016).
More information can also be found in the review by (Belias et al., 2022) on factors that contribute to persistent Listeria
in food processing facilities and relevant interventions, and (Malley et al., 2015), on the ‘seek-and-destroy’ approach related
to L. monocytogenes process controls in the RTE meat and poultry industry.
Some of the interventions that can be implemented to control a persistence event rely on measures aimed at destroying
hazards, thus avoiding their long-term establishment in the FFPE or facilitating the prompt removal. These bactericidal
interventions can be classified considering their nature as chemical, physical or biological and are summarised below in
Sections 3.4.3.1, 3.4.3.2 and 3.4.3.3, respectively. Examples are given of studies assessing their performance in industrial
settings or on model systems very closely resembling real industrial settings.
TA B L E 5 Actions triggered by persistence with the aim to eliminate persistent Listeria monocytogenes from the processing environment in relevant studies.
Study Persistence locations Interventions description Impact of the intervention

Blatter et al. (2010) Particular slicers and conveyor belts of a Specialised deep C&D of more difficult to clean equipment (details not L. monocytogenes was no longer found on slicers, conveyor
sandwich-producing plant provided) belts or in products
Biguzzi et al. (2012) The drying-cooling tunnel in a chicken wurstel New and more accurate C&D were introduced, including C&D of A contamination average of L. monocytogenes on carcasses
(a frankfurter sausage made from chicken) evaporative cooling pads, front and rear, and fan coolers, at the dropped from about 23%–45% to 7% or below, with
production plant beginning and at the end of each working day in the slaughterhouse. a maximum count of 10 CFU/g. After the additional
Further on, the process was modified, including a pasteurisation step precautionary actions (pasteurisation step), L.
(72°C for 15 min) after the final packaging of the product monocytogenes was detected in wurstel samples, but at a
level that complied with the microbiological criteria
Dalmasso and The processing area, the washroom, a storeroom Drastic revision of C&D procedures and implementation of workflows Two out of three persistent L. monocytogenes subtypes were
Jordan (2013) and the packing room; drains, floors, wheels (more rigorous cleaning procedures with additional staff, including eliminated and there was a reduction in the number
of trolleys or other mobile equipment, tables use of peracetic acid as a disinfectant) of L. monocytogenes positive NFCS samples. The third
and boots in the changing room of a food subtype was still found, but its prevalence was reduced
processing facility
Jordan et al. (2012) Surface areas and drains in the storage rooms of (i) Thorough cleaning of the facility, workflow changes and staff (i) The problem with persistent L. monocytogenes was not
an artisan RTE food processing facility movement restrictions resolved
(ii) Installing a new drainage system with back-f low prevention and (ii) The problem was controlled but not solved
changing the cleaning detergent in use
Malley et al. (2013) Smoked fish processing plants Interventions were implemented in two plants. Various of the Elimination of persistent L. monocytogenes was found
interventions used were chemical interventions, including dry extremely challenging, but a reduction in the number
QAC granules on floors; treatment of drains with peracetic acid of samples positive for a given, presumably persistent
foam; weekly cleaning of drains with foamers, detergent and water; subtype, was often achieved without complete
improved forklift fork disinfection; intensified room C&D; and daily elimination of the subtype from the plant
drain foaming instead of weekly
Murugesan A trench drain and a floor crack during the Improvements made to C&D procedures, including among other There was a significant reduction of L. monocytogenes
et al. (2015) 1.3 years sampling period in a mushroom changes, use of granulated QAC product on the floor, filling and in the washing and slicing area and packaging
slicing and packaging facility sealing floor cracks and larger crevices, manual cleaning of floors and area. The persistent subtype was still isolated after
equipment with brushes and low pressure hosesa implementation from the trench drain and a floor crackb
Nakari et al. (2014) Two fishery production plants Intensive cleaning procedures, structural changes and renovations L. monocytogenes was no longer detected in the production
environments or products of these two plants
Ortiz et al. (2010) The environment and equipment in the Controlling the contamination of raw ingredients, improving the Two persistent L. monocytogenes PFGE types were eliminated
cutting and manufacturing room, the compartmentalisation and changing the cleaning protocols from the processing plant, although eradication of other
slaughterhouse, the loin-marinating and adapted strains was not achieved
grinding machine in an Iberian pork-
processing plant producing RTE meat
products
Ortiz et al. (2014) An Iberian pork-processing plant that produced Reducing all contamination pathways of the manufacturing area (e.g. Significant effect on subsequent contamination of
RTE meat products raw ingredient contamination, compartmentalisation of fresh manufactured products and was probably the cause of
meat environments and continuous quality control of machines the elimination of two persistent L. monocytogenes PFGE
when manufactured foods are produced) to the maximum extent types at the end of the study
practicable
(Continues)
TA B L E 5 (Continued)
Study Persistence locations Interventions description Impact of the intervention
Pennone et al. (2020) Mushroom casing production Various corrective actions were implementedc L. monocytogenes positive samples during the mushroom
casing production stage were 31% (18/39) and 22% (8/37)
before and after corrective actions, respectively. After
hygienic measures were adopted, L. monocytogenes was
not detected in the first batch of casing samples, but
3 months later, the occurrence in the casing was 20%,
suggesting that despite adopting harsher measures
recontamination occurred
Stessl et al. (2020) Drains (storage rooms, raw material processing A basic and an intensified (deep) C&D event were performed. Returning Initially persistent L. monocytogenes ST121 strains were
areas, heat treatment areas, corridors) and to normal production after a construction event (wall breaks throughs identified. There was a decrease in Listeria prevalence
overshoes were closed, return to normal work and C&D) after implementation of both cleaning and disinfection
events. During the construction work more genetic
lineage I strains were introduced into the factory (ST1,
6). After returning to normal production there was an
increase in the initially detected persistent genetic
lineage II strains (e.g. ST121)
Strydom et al. (2016) An avocado processing facility Introduction of new strategies (incl. a combination of hygienic measures, A drastic reduction of Listeria spp. in final products and the
intensive environmental testing, structural interventions and higher processing facility was achieved
quality of raw materials)d
Abbreviations: C&D, cleaning and disinfection; NFCS, non-food contact surface(s); PFGE, pulsed-f ield gel electrophoresis; QAC, quaternary ammonium compound; RTE, ready-to-eat; ST, sequence type.
a
e.g. adoption of QAC and hydrogen peroxide disinfectant, twice-a-week application of QAC product on the floor, removal of power hose for cleaning floors coupled with minimalisation of aerosol generation during floor cleaning, filling and sealing of
floor cracks and crevices with cement sealer.
b
A possible explanation for the prolonged survival below the floor surface was that L. monocytogenes persisted within the porous concrete matrix, which acted as a protection towards intermittent and short contact with surface C&D chemicals while a
continuous supply of water and nutrients was available.
c
e.g. washing procedures for the conveyor belts and transport lorries, the introduction of pools for boot disinfection at the entrance and the exit of all areas, structural interventions to contain more storage bays indoors rather than outdoors.
d
The strategies included a combination of hygienic measures (double sanitation period of 10 min for fruits entering the facility, disposable aprons), intensive environmental testing (e.g. personnel, surfaces and equipment sampled three times/week
during processing and after cleaning), structural interventions (physical separation of the hygiene box for avocado pulp handling from the rest of the facility, and enclosure of the box with only one opening coming directly from the high-risk boot
captive and openings for the conveyor belts and waste disposal window, construction of a new canteen) and higher quality of raw materials (wash step with detergent and water rinse before disinfection, strict monitoring of fruit receiving upon the
facility).
3.4.3.1 | Chemical interventions

Food industries rely on the use of cleaning agents and disinfectants to establish barriers to the entry of foodborne patho-
gens and avoid their colonisation of FCS and equipment. Most commonly a two-step approach is used with a cleaning
phase followed by disinfection. For open C&D of surfaces and equipment, foaming agents are commonly used to facilitate
sufficient contact time. Alkaline cleaners like chlorinated alkaline or acid cleaners are frequently used, while disinfect-
ants contain one or more active compounds like alcohols, aldehydes, chlorine and chlorine releasing agents (e.g. sodium
hypochlorite, chlorine dioxide), peroxygen compounds (e.g. hydrogen peroxide, peracetic acid), quaternary ammonium
compounds (e.g. benzalkonium chloride), amphoterics, bases (sodium hydroxide, potassium hydroxide, sodium carbonate)
or acids (mineral and organic acids). CIP (cleaning-in-place) is commonly used for closed systems (e.g. pipes and vessels in
dairies), where acids and bases are the most widely used agents. The utilisation of cleaning agents incorporating enzymes,
such as proteases, lipases, DNAses, amylases, cellulases or pectinases, can help degrade extracellular polymeric substances
before disinfection, facilitating the removal of bacterial biofilms. For disinfection, whole room disinfection is also an option,
where the room is filled with a fogging or gaseous disinfectant, e.g. hydrogen peroxide or ozone.
Studies under laboratory settings have assessed the survival of the main foodborne pathogenic bacteria (in planktonic
state and/or as biofilms) to a wide range of industrially used disinfectants or their active compounds at their in-house use
concentration, and some studies have shown that these are not always capable of completely inactivating target micro-
organisms when grown in single or multiple species biofilms, as reviewed by Alvarez-Ordóñez et al. (2019). A limitation of
many studies is that disinfectants are tested directly on biofilms, without a prior cleaning step, since most disinfectants
are designed to work on cleaned surfaces. Although these studies can provide the scientific basis on the most appropriate
substances and optimal concentrations to be used in industrial settings, validation activities in real industry settings should
be conducted. Validation of the formulations used is essential and may identify improper usage, leading to inadequate
contact concentrations or contact times, for instance, through erroneous formulation, inappropriate storage, inadequate
rinsing of recently cleaned and disinfected areas or biocide application to wet surfaces, with a consequent dilution of the
compound to concentrations that may be sublethal for microorganisms. However, standardised protocols for validation of
C&D in situ in processing plants are not available and the results of validation activities are not frequently published and/or
made publicly available, contributing to a general lack of information on the efficacy of alternative chemical formulations
in C&D regimes in food processing plants.
Numerous experimental studies, mainly undertaken in laboratory settings, have tested novel agents capable of remov-
ing bacterial biofilms (e.g. enzymatic agents) and identified novel antimicrobial compounds suggested to be included in
formulations for use as green disinfectants. These include electrolysed water, plasma activated water, ozone and com-
pounds of natural origin, including essential oils or extracts obtained from plants, foods and by-products and exerting
either a direct bactericidal effect or an indirect biofilm inhibition activity mainly related to the inhibition of quorum sensing
and their regulated phenotypes. Further details about research activities in these areas can be found in the reviews by
Ashraf et al. (2014), Coughlan et al. (2016) and Alvarez-Ordóñez et al. (2019).
The scarce experimental studies, retrieved from the literature search, performed in industry settings (or in model sys-
tems closely resembling industry surfaces), testing chemical interventions to prevent or control contamination or per-
sistence of hazards in the FoPE, all regarded L. monocytogenes (summarised in Appendix D).
The experimental studies in industry settings consisted of:
• testing disinfectants in different parts of a dessert-processing factory by (Campdepadros et al., 2012) and in blue crab
meat and crab processing plants by (Pagadala et al., 2012).
• adding citric acid powder to floors where water tends to accumulate in meat processing plants (Moretro et al., 2017).
• ozonation as an adjunct to the disinfection regimes across all production areas in a cheese processing facility (Eglezos &
Dykes, 2018).
The experimental studies in model systems closely resembling industry surfaces involved:
• using self-contained chlorine dioxide (ClO2)-generating and delivery pods to disinfect floor drains; the authors con-
cluded that commercially available ClO2 pods may have practical utility for killing L. monocytogenes during periodic
disinfection of floor drains (Berrang et al., 2017).
• adding antimicrobial additives in conveyor belt material, which may help to reduce L. monocytogenes on belts at low
temperatures when food residues are absent, and belts are not rapidly dried (Chaitiemwong et al., 2010).
• comparing C&D protocols of rubber blades simulating procedures used in food processing, showing that rubber blades
can be cleaned more efficiently than foam blades and leading to the recommendation of using a full procedure (deter-
gent and rinse, followed by disinfectant) including a scrubbing step (Martinez et al., 2021).
3.4.3.2 | Physical interventions

In the experimental studies retrieved, physical interventions applied to prevent persistence or eliminate persistent strains
from FoPE were mainly based on the use of non-ionising radiation and heat, with L. monocytogenes being the target micro-
organism for testing. These are summarised below and in Appendix D.
UV-C radiation has been tested to prevent or eliminate contamination of surfaces and the FoPE by L. monocytogenes.
Bernbom et al. (2011) and Morey et al. (2010) concluded that UV light can be effectively used to reduce L. monocytogenes
contamination on food processing surfaces in a fish smoke house after the routine C&D procedure and on conveyor belts,
respectively. Heat treatments, applied following different approaches, have been shown to be effective interventions to
remove L. monocytogenes from the FoPE. Eglezos and Dykes (2011) used electrical air-blowing heaters to heat and dry out
holding chillers used for post-cook commercial processed meats. The incorporation of two simple chiller heating proto-
cols into these facilities' GMPs effectively reduced Listeria prevalence in chillers. Steam cookout of the floors of mushroom
growing units can minimise the chances of contamination of subsequent batches, as shown by Pennone et al. (2020). Tobin
et al. (2020) illustrated the feasibility of hot water disinfection treatments of commercial mushroom slicers to minimise
L. monocytogenes food safety risks.
Some novel non-thermal treatments (e.g. plasma treatments) could also be applied for the decontamination of air or
surfaces (Alvarez-Ordóñez et al., 2019; Coughlan et al., 2016), but have not been tested yet in industrial settings or in rela-
tion to microbial persistence in the FFPE.
3.4.3.3 | Biological interventions

Biological interventions to prevent persistence or eliminate persistent strains from the FoPE can be based on the use of live
microorganisms or their metabolites for competitive exclusion or inactivation of the targeted hazards, or on the applica-
tion of bacteriophages active against them. Some studies have highlighted that the structure of microbial communities in
food processing facilities can impact their colonisation by pathogenic bacteria, mainly L. monocytogenes, and that influenc-
ing the microbiome in favour of certain species may limit the likelihood of product/process contamination with them (Fox
et al., 2014). Moreover, in the production of some traditional fermented products, it has been shown that the development
of ad hoc biofilms of lactic acid bacteria on vats used for production can reduce microbial variability in the product (Gaglio
et al., 2016) and that natural biofilms present on shelves used for ripening can prevent the growth of hazards like L. mono-
cytogenes (Mariani et al., 2011).
Biological interventions have been frequently tested as biocontrol strategies in laboratory settings to inhibit biofilm
formation or remove biofilms formed by different pathogenic bacteria, where the antifouling activity of various lactic acid
bacteria, mainly bacteriocin producers, bacteriocins and bacteriophage, has been monitored in detail (Alvarez-Ordóñez
et al., 2019; Coughlan et al., 2016).
However, the efficacy of biological interventions to prevent or control contamination or persistence of hazards in the
FoPE has been rarely tested in industry settings, and only for L. monocytogenes. For example, Zhao et al. (2013) illustrated
that L. monocytogenes could be efficiently reduced using competitive exclusion (using a combination of Lactococcus lactis
and Enterococcus durans) applied onto floor drains of a RTE poultry processing plant. Likewise, (Schobitz et al., 2014) showed
that L. monocytogenes can be successfully eliminated from the walls of floor gutters in a salmon processing plant using a
so-called ‘biocontroller’ composed of a thermally treated fermentate from two Carnobacterium maltaromaticum strains
and a strain of Enterococcus mundtii, plus nisin. The application of bacteriophage P100 (Listex™) was shown to significantly
reduce the incidence of Listeria spp. on NFCS in the RTE environment of refrigerated (4°C) and ambient (20°C) temperature
facilities (Reinhard et al., 2020). More details about these experimental studies are shown in Appendix D.
The main advantage of biological interventions is that biocontrol agents have generally a rather narrow spectrum of
activity. This would facilitate the design of more hazard-targeted interventions, which would better preserve harmless
taxa within the processing environment microbiota, which at the long term can result in positive ecologic effects (Alvarez-
Ordóñez et al., 2019).
3.4.4 | Concluding remarks related to measures for monitoring, preventing and/or controlling
persistence
• A well-designed environmental sampling and testing programme is the most effective strategy to identify contamina-
tion sources and detect potentially persistent hazards. It should be designed following a risk-based approach, should
be fully documented, defining limits for acceptable and unacceptable results and outlining follow-up actions in case of
non-compliant results, and should be regularly revised based on trend analysis. Key aspects are the identification of sam-
ple locations; the target organism(s); the frequency and timing of sampling and the number of samples; the sampling
protocol and defined test methods; and the recording and evaluation of results.
• The establishment of hygienic barriers and measures within the FSMS, during implementation of HACCP, is key to pre-
vent and/or control bacterial persistence in the FFPE through avoiding the entry of the hazard(s) to the processing plant
and/or their spread across the facility. The following prerequisites are of particular importance in relation to bacterial
persistence in the FFPE: infrastructure (building, equipment), C&D, technical maintenance and calibration, water and air
control, personnel (hygiene, health status), working methodology and food safety culture.
• To confirm the presence of a persistent strain and identify its niche within the facility, the testing programme must in-
volve the detailed characterisation of isolates of the specific hazard(s) recovered from positive samples using subtyping
methods with enough resolution, preferably through WGS-based subtyping schemes, which provide a higher resolution
than traditional serotyping or older genotyping methods.
• Once persistence is suspected in a food or feed processing plant, a ‘seek-and-destroy’ approach has been frequently
recommended, which includes:
i) intensified monitoring to assess how far the contamination is spread across the facility and to identify a potential
harbourage site;
ii) the introduction of measures to control the event, which can involve from drastic decisions (e.g. closure of a line,
change of equipment) to other progressive actions such as the intensification or change of the C&D scheme (in-
cluding increased dismantling) or the introduction of a novel bactericidal intervention; and
iii) the continuation of the intensified monitoring programme to confirm the efficacy of the measure(s) taken or to
identify the requirement for additional measures.
Alternatively, systematic ‘root cause analyses’ can be applied to identify the most probable factors/sites within the fa-
cilities contributing to the problem and define the most appropriate interventions to eliminate the pathogen from the
premises.
• Successful actions triggered by persistence in the FoPE were found for L. monocytogenes only. For example, these included
the introduction of new or specialised (deep) C&D, the implementation of workflows, the installation of a new drainage
system; the implementation of structural changes and renovations; the control of the contamination of raw ingredients
and the improvement of the compartmentalisation, or the simultaneous implementation of various corrective actions.
• There are several options of interventions with direct bactericidal activity that can be implemented to eliminate the per-
sistent hazard(s). These interventions can be classified considering their nature as chemical (biocides, electrolysed water,
plasma activated water, essential oils, natural extracts, directly applied or included in the formulation of conventional or
novel (green) disinfectants,) physical (based on the use of non-ionising radiation, heat or novel non-thermal technolo-
gies) or biological (use of live microorganisms for competitive exclusion or their metabolites, or the application of bac-
teriophage). While some of these interventions are already in use in the food industry and their efficacy is known from
practical experience or scientific studies, other interventions are not yet commercially available and/or their efficacy is
not yet fully validated under industrial conditions. For the majority of the listed interventions, a thorough cleaning (often
involving dismantling of equipment) is needed before the biocidal step, in order for the latter to be efficient.
3.5 | Perspectives of integrating the information gathered in risk assessment (AQ6)
It was considered that the aim of a risk assessment on the role of persistence for food safety PH risks could be to assess the
number or percentage of foodborne illness cases attributable to persistence for a certain population, hazard and/or food
product(s) and/or to estimate how different factors (e.g. the features of the hazard or the implementation of mitigation in-
terventions) can impact on them. Such assessments can be performed by two approaches: a bottom-up (or forward) and/
or a top-down (or backward) approach.
The bottom-up approach (food chain Quantitative Microbiological Risk Assessment (QMRA)) adheres to the standard
microbial risk assessment paradigm and follows the agent through the food chain to produce a prediction of risk to human
health. In the top-down approach, the level of risk associated with specific foods, hazards or their combinations is assessed
based on information gathered from epidemiological systems such as disease reporting and outbreak databases (EFSA
BIOHAZ Panel, 2012). The perspectives of these two approaches to assess the role of persistence for food safety PH risks
are described and discussed below.
3.5.1 | Bottom-up approach and data needs
The bottom-up approach is based on the Codex Alimentarius guidelines for microbiological risk assessment (FAO and
WHO, 2021a), and encompasses hazard identification, hazard characterisation, exposure assessment and risk characteri-
sation. The transfer, growth and survival of a pathogen is modelled, and the risk (e.g. number of cases) is assessed using
an exposure assessment and a DR relation. In the context of studying the role of persistence for food safety risks, this ap-
proach can be used for comparing the potential impact of persistence on the model estimates of the PH risks with that of
other factors, for example bacterial growth potential in the food, storage conditions, food hygiene or virulence potential
of the hazard.
Appendix E provides a detailed description of the perspectives for integrating persistence in a QMRA model. Available
‘persistence’ models are usually presented as transfer models or cross-contamination models, where the principle is that
bacteria in the FoPE are transferred to a food being processed. Some generic – one-and two-compartment – models are
presented in Appendix E, including a model with two linked compartments based on (Mokhtari & Van Doren, 2019). This
model was adopted as the ‘basic model for persistence’ in this scientific opinion. The model dynamics are illustrated in
Figure 11. This model considers:
• the continuous transfer of the hazard from a contaminated food contact surface shown as FoPE1 to food products (with
transfer rate b1) and to a non-food contact surface shown as FoPE2 (with transfer rate b2) and the flow back from FoPE2
to FoPE1 (with transfer rate b3)
• the change in population size in FoPE2 in the time frame between two food products passing in FoPE1, represented by
the ‘persistence parameter’ a2.
◦ A positive value implies net growth in FoPE2 with a growth rate a2;
◦ When a2 = 0, growth is equal to inactivation and there is no change in population size; and
◦ A negative value represents a net inactivation rate implying less than 100% survival.
Note that the chosen unit of time (i.e. the time between two food products being processed in FoPE1) implies that the
model does not use a common growth or inactivation rate, expressed per minute or per hour.
• the decrease in population size in FoPE1 (which, if negative, might represent growth), also in the time frame between
two food products passing in FoPE1, represented by c1.
F I G U R E 11 The two linked compartments model considered as the ‘basic model for persistence’ to evaluate the perspectives of QMRA to assess
the role of persistence for food safety public health risks. From a contaminated food contact surface FoPE1 there is continuous transfer to food
products (with transfer rate b1) and to FoPE2 (with transfer rate b2). There is also a flow back from FoPE2 to FoPE1 (with transfer rate b3). The persistence
parameter a2 indicates the growth or inactivation rate in the hidden environment FoPE2, whereas there may be limited survival in FoPE1, described by
inactivation rate c1.
To better understand the basic model for persistence and how it can be used, two illustrative examples are presented
below:
• Imagine a situation where a L. monocytogenes strain persists in a drain of the packaging room of a plant producing
roast beef. Contamination starts at the drain and is transferred to a FCS in the packaging machine, contaminating the
end product. The strain persisting in the drain belongs to CC9, has low virulence (with several truncations in virulence
factors), but has an enhanced survival in the environment and resistance to C&D, due to the carriage of several efflux
pumps, heavy metals and heat resistance genes and stress survival islet I. In this example, FoP1 is the packing machine,
FoP2 is the drain in the packaging room, and the food product is roast beef. The decrease in population size in FoPE1 (c1)
is expected to be very low, while the change in population size in FoPE2 (the ‘persistence parameter’ a2) is expected to
be zero or positive. The transfer rates are not known.
• A S. Agona strain persists in the spray drier of a plant producing powdered infant formula. Contamination starts at the
FCS of the spray drier, is then transferred to hidden compartments of the spray drier where it persists. The strain has un-
known properties in terms of virulence, biofilm formation or survival in the environment. Considering the low moisture
found both in the end product and on the surfaces of the spray drier, it is assumed the strain does not grow neither in
the hidden compartments of the spray drier nor in the powdered infant product. In this example, FoP1 is the FCS of the
spray drier, FoP2 is the hidden compartments of the spray drier, and the food product is the powdered infant product.
The decrease in population size in FoPE1 (c1) is expected to be very low, while the change in population size in FoPE2 (the
‘persistence parameter’ a2) is expected to be zero. The transfer rates are expected to be very low.
This basic model for persistence is deterministic, assuming constant rates for transfer between environments and food,
growth and inactivation in FoPE1 and/or FoPE2. This simplifies the model but may not be realistic. The alternative is to
develop a stochastic model, which may be more realistic but more complex, containing more (unknown) parameters.
Random stochastic variation can be included (assuming constant rates, but with discrete bacterial cells being sampled) or
even the rates can vary with every transfer.
The basic model assumes that inactivation and growth take place gradually (at a constant rate). This is not necessarily re-
alistic, as for example, with C&D, a larger amount of the bacterial contamination will be removed, which can be modelled by
an instantaneous inactivation step, implemented by a few logs decrease in concentration. Typically, for persistence to occur,
it can be assumed that the C&D is effective in FoPE1, but not in FoPE2. Similarly, growth may occur during a period that the
food processing is halted, e.g. overnight, giving ample time for an increase in concentration, for example in a biofilm.
The persistence model can be used in conjunction with other models in a larger food chain model for QMRA. As an
example, a QMRA model for L. monocytogenes was developed inspired by the production of cooked ham. In the L. mono-
cytogenes model, persistence was considered to occur in a slicer. It was assumed that cooking of the ham inactivates all
L. monocytogenes, that the slicer has been persistently contaminated with L. monocytogenes and that there is growth
during storage of the ham in subsequent parts of the food chain. Another example, following the same food chain model
structure, without the storage step, was explored for Salmonella in a LMF, where growth is not expected to occur in the
food product during storage.
Performance of the QMRA model for L. monocytogenes is illustrated in Figure 12, that shows an example of the dynamics
of the persistence model and the resulting impact on the risk per serving. Appendix E further illustrates the performance of
the model and shows how such a model can be used to investigate the expected impact of several factors associated with
persistence on the risk estimates, compared to other factors affecting the risk. Note that the QMRA model is a simplification
in many respects, not only because stochastic processes in the FoPE are not included, but also because variability in storage
times and temperatures is not included. It is to be expected that observed levels of contamination in the FoPE, observed
concentrations in implicated food products and observed cases in an outbreak are much impacted by this stochasticity.
As an example, Figure 13 compares the PH risk estimates for a hypothetical set of strains with that of a strain that, in
terms of persistence, is characterised by a2 = 0, and has an average virulence and growth capacity during storage (the
baseline shown in Figure 12). Here, a PH risk estimate is the expected number of human cases in an outbreak following the
contamination of the FoPE modelled. The alternative strains are more persistent (a2 > 0), more virulent and/or have a larger
growth capacity during storage than the baseline strain. In this particular example, all these characteristics increase the PH
risk estimate, where virulence seems to have most impact on the risk (the orange bar). It also shows that persistence, de-
fined by an increased growth capacity in FoPE2, leads to a longer maintenance of the risk level (the grey bar), which means
that, in the course of time, the bacterial load and the probability of illness per serving are unaltered (vertical blue line) or
are decreasing slowly.30 This maintenance (which in practice would lead to the observation of persistence) is to be ex-
pected for a strain that can grow in FoPE2. The example shows that more persistent strains are associated with larger risks,
especially in the longer term, but also that factors like virulence and growth capacity in the food contribute much to these
PH risks.
F I G U R E 12 Performance of the baseline QMRA model with persistence for L. monocytogenes in a slicer. Note: The modelled process is shown
at the left, the persistence model is included as the model for the slicer. The graphics at the right show the dynamics of (above) the levels of
contamination in FoPE1 (blue line, N1) and FoPE2 (red line, N2) and the concentration in the food after storage (black line, Nafter), and (below) the
probability of illness per serving of the food. The horizontal axis shows the consecutive food items processed. In this baseline, the persistence
parameter is a2 = 0, there is no growth or inactivation in the FoPE's. Every 20,000 food items there is C&D in FoPE1, resulting in a 5 log reduction of the
contamination level. Further details and alternative scenarios are explained in Appendix E.
Which factor contributes most will depend on the specific case study and scenarios at hand. Although the figure may
suggest that persistence is not a very important factor for the risk, this is not necessarily so. The baseline assumes that
there is 100% survival (and not growth) in FoPE2, without an effect of C&D in this environment. This is a form of persistence
as well, that is required to get an effect of growth during storage and virulence. A baseline without any persistence is not
useful for the model comparison, as it is arbitrary and would give a very low risk. However, it may be realistic in many cases.
30
In the model, an increasing risk (a grey bar beyond the blue vertical line in Figure 13) will in the course of time lead to a situation with unrealistically high concentrations
and is not sustainable (see Appendix E).
F I G U R E 13 Results of the L. monocytogenes model, comparing a ‘common’ strain (baseline) with strains that are more persistent (P and P+),
virulent (V) or fast growing in the food (F) (potentially with a higher maximum population density (FH)), or combinations thereof. Note: The orange bar
is the relative risk (i.e. the log of the expected PH risk of the indicated strain divided by the expected PH risk of the baseline strain); the grey bar is the
relative maintenance of risk of the indicated strain, which is the case ratio (the expected PH risk of the last 20% of food products divided by that of the
first 20%) of this strain divided by the case ratio of the baseline strain; the blue vertical line indicates a relative maintenance of risk corresponding to
100% (a case ratio of 1, no change in PH risk associated to the first and last 20% of food products). For details, see Appendix E.
A bottom-up risk assessment approach is not only useful to assess and compare PH risks in different scenarios, it can also
be particularly useful to clarify the potential role of persistence in food safety, and the importance of different elements
of persistence. For example, the analyses described in Appendix E using a range of plausible input values for the different
model parameters showed that:
• Persistence is a complex phenomenon, involving bacterial growth, inactivation and transfer, several compartments in
the FoPE (such as FCS and NFCS and the food products themselves) and stochastic processes. A model of the process is
necessarily a simplification, which will not allow an explanation of all observations of the process.
• It is feasible to include persistence in a QMRA model and explore its impact as compared to for example virulence or
growth in the food product.
• Growth in the FoPE is required to explain a long-term outbreak of L. monocytogenes due to persistence of a strain in the
FoPE. This was concluded using the cooked meat example presented above. Without growth in the FoPE, a high survival
rate must be associated with low transfer rates to maintain a long-term presence of the hazard in the FoPE, meaning that
relatively low initial doses will be attained. The QMRA model suggests, under the conditions used for model develop-
ment, that this will not result in illness of consumers over a longer period of time.
• In the LMF example of Salmonella, the requirement for growth in the FoPE is less stringent because more infections are
expected after exposure to much lower doses due to the non-linearity of the DR model.
• If growth in the FoPE is too extensive, the contamination of the environment will continuously increase and most likely
result in early detection of the hazard in the FoPE or in the food, followed by interventions. There is actually a very small
range of growth rates in FoPE 2 (persistence parameter a2) that can explain long term persistence without a decrease or
increase of the level of contamination in the FoPE, and a consequential decrease or increase of the associated PH risk.
This suggests that the ‘basic model for persistence’ may be too simple to explain long-term persistence. This model is
deterministic and does not include all the relevant biology, it lacks for example the complex mechanistic dynamics asso-
ciated with biofilms.
• Stochastic processes can explain seemingly chaotic dynamics and the re-appearance and detection of a persistent strain
over longer time periods. Although stochastic processes seem to play an important role in persistence in the FoPE, inclu-
sion of stochasticity increases the model complexity and the number of model parameters for which data are lacking.
• In the process of developing the persistence model and the QMRA model, it became clear that the definition of persistence
is crucial for conducting and communicating about a risk assessment on the role of persistence. Persistence, defined in
this scientific opinion as ‘the ability of a given organism to be established in niches within the FFPE for a long term, despite
the frequent application of C&D’, is a characteristic of the hazard in a given environment that, in the model, is expressed
as the persistence parameter a2 and other model parameters. The values of these parameters defining persistence are
an input of the model. However, in the literature, persistence is also referred to as for example ‘repeated isolation of the
same strain for months or even years at the same sites’ which can be translated into a model output as continuous high
levels of bacterial contamination in the FoPE (Larsen et al., 2014; Unnerstad et al., 1996). These continuous high levels can
result in a maintained level of risk that implies that cases can occur over a long period, potentially resulting in a long-term
outbreak. The model analyses show that the input (the ability to be established for a long time) and output (long-term
maintenance of high levels of contamination and an extended time period of increased risk) are strongly associated, but
they are not the same.
• On top of that, the analyses show that persistence is not a black/white, yes/no, presence/absence concept, but it is grad-
ual, as gradually increased persistence implies gradually better survival or longer presence in the FoPE, without a critical
threshold. This has for example implications for the classification of strains as being persistent or not: the ability to be
established is not present or absent, and will depend on an interaction between the strain, the specific FoPE and the
implementation of hygienic measures such as C&D.
• There is a gap between the interpretation of molecular and genotypic data, as presented and discussed in Section 3.2,
and QMRA. It would, for example, be highly relevant if the virulence as characterised by genotypic data could be trans-
lated to values of DR model parameters, such as the r-parameter for L. monocytogenes. At the moment, the virulence
of subtypes that are considered ‘hyper-or hypovirulent’ can only hypothetically be translated in virulence in a QMRA
model. The same is true for persistence: it is unclear how all features responsible for persistence (e.g. the carriage of
some of the genetic markers identified in Section 3.2 as possibly associated with persistence) are to be translated in the
a2 parameter or other parameters in a persistence model.
The information gathered in the preceding part of this Scientific Opinion indicates for which pathogens persistence is
considered to play an important role for specific food sectors. This clarifies for which hazard/food product combinations a
QMRA involving persistence could be useful. Also, it may be feasible to develop QMRA for different strains of the same haz-
ards, if the relevant traits of these strains in terms of growth and survival in the FoPE and the food, as well as the virulence,
are characterised. So far, this information is however incomplete.
The data needs for risk assessment are large, as the PH risk is a function of many parameters, and the values of these
parameters can only be obtained from bacterial concentration data that need to be frequently collected over a long pe-
riod. On top of that, the parameter values are probably very process specific (Møller et al., 2016), which is in line with the
observation that comparison of risk factors between food processing facilities is difficult (Belias et al., 2022). This challenges
the data acquisition required for estimating the parameter values and model validation.
The increasing availability of genotypic data is scarcely applied for the purpose of risk assessment. These data are par-
ticularly useful to understand the molecular mechanisms of persistence and to identify strains and subtypes that can be
associated with observed persistence and larger PH risks. However, so far, we have found no evidence of examples where
this type of data has been applied in bottom-up risk assessments that aim to compare the reduction in PH risks that can
be achieved by specific interventions or control measures. The application of this type of data in QMRA would require a
translation of the genetic information into model parameter values that are required for risk assessment.
3.5.2 | Top-down approach and data needs
In a top-down approach, the assessment starts with an analysis of epidemiological human disease data, for example re-
ported human sporadic listeriosis cases and/or outbreak cases. From these, to attribute the cases to their sources, the ac-
tual food source/vehicle carrying the pathogenic organism is to be retrieved. For the sporadic cases, this information is not
available in the EU-wide databases (TESSy). The EFSA zoonoses database includes data on ‘strong and weak’ evidence FBOs
(as defined in the Directive 2003/99/EC) occurring in MSs, including those caused by any virus, bacterium, alga, fungus,
parasite and their products, toxins and biological amines (e.g. histamine), not just zoonotic agents. It captures the causative
agent and the food vehicle. It is mandatory to report the food vehicle as a general food vehicle category (e.g. ‘Eggs and
egg products’) and since 2020 more details about the food vehicle can be reported e.g. ‘Cheeses, made from unspecified
milk or other animal milk -unspecified -made from raw or low heat-treated milk’. Optionally (and this was the only possibil-
ity before 2020), a free text data element can be used to give more detailed information on the food vehicle (for example
‘salad of raw carrots’). Unfortunately, the actual food is for many (past) outbreaks not available. More information about
the reporting on FBOs can be found in the technical report titled ‘Zoonoses, antimicrobial resistance and food-borne out-
breaks guidance for reporting 2022 data’ (EFSA, 2023).
To assess the proportion of human cases that can be attributed to persistence in the FoPE for different food categories,
not only the food vehicle needs to be identified, but also whether there is persistence involved. As persistence implies long
term survival in the environment, it is mainly of interest for its potential to cause long-term outbreaks, so the focus would
be on these. The aim of the assessment would therefore be to identify the percentage of outbreaks for a specified food that
is associated with persistence. Unfortunately, this is not (well) documented in the reporting of outbreaks. For example, in
the EFSA zoonoses database, contributory factors can be reported in the optional data element. Contributory factors may
include deficiencies in food handling or the use of contaminated material. Such contributory factors leading to FBOs are
frequently unknown, and it should be considered that ‘persistence’ does not have a common definition. For example, long
term isolation of the same strain in a food product is not necessarily a result of its persistence in the FoPE but can also be
caused by repeated reintroduction in the FoPE, which, in this Scientific Opinion, is not interpreted as persistence. It should
be noted that it is currently possible to report under ‘General’ two different terms: ‘Continuation of an outbreak reported
last/previous year’ and ‘Part of multi-country outbreak’.
A top-down risk assessment approach for persistence would require a database of outbreaks for the causative agent
under investigation that allows identification of the food vehicle, but also gives an indication of the likelihood that such
outbreak is due to persistence in the FoPE, using an agreed unambiguous definition of persistence. With an estimate of
the percentage of outbreaks associated with persistence obtained in this way, the relative risk of persistence for specific
hazards and/or food products (or categories) could be assessed.
A top-down approach can also be used to study to what extent ‘persistent strains’, i.e. genotypes associated with per-
sistence, are found in outbreaks and/or in sporadic cases. As the bottom-up risk assessment model showed, attributes
like virulence and growth capacity in foods may also be important for strains to be linked with outbreaks. If the genetic
characteristics of outbreak strains are known, the relative importance of persistence can be assessed. This requires the
availability of genetic data of outbreak strains together with the characteristics of the outbreaks for which they are found.
Unfortunately, however, despite an increased use in last years of WGS for outbreak investigation, the strains involved in
cases of infection or outbreaks are not always characterised in detail, and therefore it is not easy to know whether they
have some features likely related to persistence. Such an approach would require the use of genotypic persistence markers
(e.g. particular genes associated with increased survival in the FoPE) truly inducing phenotypic persistence. Unfortunately,
however, the ‘presence or absence of genes thought to promote persistence, was not found to be useful for predicting persistence’
in previous reports (Møller Nielsen et al., 2017) and, as highlighted in Section 3.2, no universal markers or features, respon-
sible for persistence can be identified yet with the available evidence.
3.5.3 | Concluding remarks related to perspectives of integrating the information gathered in

risk assessment
• Risk assessments may be performed for different relevant combinations of hazard and food product to assess the rela-
tive PH risks that can be associated with persistence.
• For that purpose, in bottom-up food chain QMRA, transfer or cross-contamination models involving one or two (linked
or not) FoPE compartments may be used. In this assessment, a basic two linked compartments model for persistence has
been developed and its performance explored using a range of plausible input values for the model parameters.
• In a simple, hypothetical example, it was shown that this type of model can be used to study the role of persistence in
the PH risk for a specific food production process. However, since the dynamics of persistence and its role in PH risk is
affected by complex processes, linked to the factors listed in Section 3.3, the current model may be too simple to capture
important biological processes, such as biofilm formation.
• With the currently available data, top-down risk assessment, where food vehicles are linked to sporadic human cases and
outbreaks, cannot be used to assess the relative PH risk that can be associated with persistence. It would require that the
occurrence of persistence in FoPE is collected and reported along with the other outbreak data.
• Risk assessment cannot fully exploit the data gathered to support answering the previous AQs, and the data needs for
risk assessment are not well covered. There is a need for a better translation of genotypic information of strains into
phenotypic characteristics that can be converted into parameters of risk assessment models, as well as for extensive
quantitative data to describe the dynamics of transfer, survival and growth of bacterial hazards in different FoPE niches.
• The ultimate objective of risk assessment is to provide decision support for risk managers. Risk managers would benefit
from the knowledge of the relationship between persistence in the FoPE (as defined in this scientific opinion) and PH
risk, especially if the risk assessment approach allows an assessment of the efficacy of intervention measures that reduce
persistence and the associated risks. In that respect, the risk assessment model structure described here can help to
suggest potential mitigation strategies against persistence of biological hazards in the food chain, whereas it is unclear
how the current knowledge about identified relevant (sub)types and their main features can be practically used in risk
management. So far, there seems no solid basis to manage strains differently depending on their subtype or genetic
features.
• It is important to apply clear definitions of persistence in all studies that involve persistence (observational, experimen-
tal, epidemiological, etc.) and it would be preferable to use the same unambiguous definition for all of them.
3.6 | Knowledge gaps and priorities for future research (AQ7 and 8)
The knowledge gaps and the recommendations/priorities for future research related to bacterial food safety hazards as-
sociated with persistence in the FFPE are shown in Table 6. Most of the recommendations for research would involve
activities in industrial settings. Therefore, their feasibility will depend on the availability and willingness of food industries
to share data on persistence in their facilities and participate in research actions aimed at improving the knowledge on per-
sistence in the FFPE. As an alternative, some of the recommended research activities could be performed using industrial-
like model systems of certain niches (e.g. drains, conveyors, slicers), where different strains (including knock-out variants),
environmental conditions and potential interventions can be tested, which would also allow generating quantitative data
to describe the dynamics of transfer, survival and growth of bacterial hazards and to obtain strain-or subtype-specific
parameter value estimates for QMRA.
TA B L E 6 List of knowledge gaps and priorities for future research related to bacterial food safety hazards associated with persistence in the FFPE.
Knowledge gaps Recommendations/priorities for research
In many studies the source of the contamination Longitudinal detailed, well-designed and reported studies for the most relevant hazards
is unclear, especially the relative importance associated with persistence addressing multiple sources (e.g. raw materials, primary
of persistence versus reintroduction production settings from suppliers, multiple FFPE within the processing plant, end products,
water, personnel, etc.) to distinguish persistence and reintroduction, and collecting
metadata associated with the matrices, the process, the suppliers, the end product etc.
Such studies will benefit from a common use of the term persistence and from harmonised
approaches and SOPs for bacterial typing in relation to persistence using preferably WGS
Limited knowledge on the importance of Studies targeting hazards other than L. monocytogenes regarding their possible persistence
persistence in the FFPE (and on niches and in the FFPE. Such studies can particularly focus on S. enterica and C. sakazakii, given
features of persistent strains) for bacterial their relevance in the FFPE from specific food and feed sectors and the insufficient data
hazards other than L. monocytogenes available on their features linked to persistence, but can also consider other hazards
for which studies addressing the FoPE are very scarce (e.g. Campylobacter, pathogenic
E. coli, S. aureus or B. cereus s. l.). They can include a detailed genomic and phenotypic
characterisation of strains of the main subtypes recovered from FFPE
Contribution of specific genetic markers and Systematic studies with persistent and presumed non-persistent strains harbouring specific
their link to phenotypes associated with genotypic markers with detailed characterisation of phenotypes relevant for persistence
persistence for the most relevant bacterial for QMRA. Such studies would benefit from the availability of a validated panel of
hazards and/or subtypes persistent and presumed non-persistent strains and of industrial-like model systems of
certain niches (e.g. drains, conveyors, slicers), where different strains (including knock-out
Relationship between AMR and biocide
variants) and environmental conditions can be tested
resistance of pathogens and its relevance for
persistence of hazards in the FFPE
Assessment of persistence in the FFPE from Ad hoc ecosystem studies at industry level longitudinally analysing the main niches linked to
a microbial community (microbiome) persistence of hazards in the FFPE following a holistic approach, with characterisation of
perspective the resident microbiome, to understand the dynamics of the interactions between food
and environmental microbiomes, and of a wide range of physico-chemical parameters
and with determination of the features of persisting strains
Factors promoting persistence at facility level in Various facility-specific studies specifically designed for the identification of risk factors for
different sectors persistence in different sectors
Efficacy of interventions at industrial scale to Systematic studies, ideally at industrial scale or using industry-like model systems, monitoring
control/remove persistent strains from the the impact of interventions in reducing or preventing persistence, particularly targeting
main biological hazards the identified sites/niches. Such studies would benefit from: (i) the availability of surrogate
organisms of persistent pathogenic strains that could be tested in processing plants, (ii) of
industrial-like model systems of certain niches (e.g. drains, conveyors, slicers), (iii) and the
development of harmonised protocols and/or rapid methods to in situ assess the efficacy
of C&D, to be used for validation/operational monitoring/verification activities in relation
to the control of persistence in processing plants
Translation of genotypic information into Research into the efficient use of available data to improve risk assessment models and
phenotypic characteristics that can be subsequently support risk management
further converted into parameters of risk
assessment models
Extensive quantitative data to describe the Studies generating quantitative data to describe the dynamics of transfer, survival and
dynamics of transfer, survival and growth of growth of bacterial hazards in different FFPE niches and to define strain-or subtype-
bacterial hazards in different FFPE niches specific parameters for QMRA
Abbreviations: AMR, antimicrobial resistance; C&D, cleaning and disinfection; FCS, food contact surface; FFPE, food and feed processing environment; FoPE, food
processing environment; NFCS, non-food contact surface; PH, public health; QMRA, quantitative microbiological risk assessment; WGS, whole genome sequencing.
4 | CO NCLUSIO NS
ToR1 (AQ1). To identify the most relevant microbiological food safety hazards associated with persistence in the FFPE
• The most relevant bacterial food safety hazards associated with persistence in the FFPE of each of the considered sectors
in the EU/EEA are:
◦ S. enterica in the feed for food animal production sector;

◦ L. monocytogenes and S. enterica in the meat processing sector;
◦ L. monocytogenes in the fish and seafood processing sector;
◦ L. monocytogenes in the dairy sector;
◦ S. enterica in the eggs and egg processing sector;
◦ L. monocytogenes in the fruit and vegetables processing sector; and
◦ S. enterica and C. sakazakii in the LMF sector.
• Other bacterial hazards were either not of highest PH relevance in the specified/specific sector or were of highest PH
relevance but not considered as most relevant bacterial food safety hazards associated with persistence in the FFPE in
the specified/specific sector based on the available information.
ToR2 (AQ2-3). To identify the main (sub)types of the most relevant hazards involved in persistence and the main fea-
tures responsible for their persistence in the FFPE
• For the three most relevant hazards, there is a wide range of subtypes reported to be involved in persistence in the FFPE.
Some specific subtypes are more commonly reported as persistent:
◦ for L. monocytogenes, especially CC 121, CC8, CC9 from lineage II and CC 5, CC6, CC2 from lineage I;
◦ for S. enterica, S. Typhimurium and S. Agona; and also S. Derby, S. Anatum, S. Infantis, S. Heidelberg, S. Mbandaka and
S. Senftenberg; and
◦ for C. sakazakii, CC64, CC1, CC83 and CC4.
• Some of these subtypes (CC6, CC8, CC9, CC121 and CC321 for L. monocytogenes; S. Typhimurium, S. Infantis, S. Agona,
S. Anatum, S. Heidelberg and S. Mbandaka for S. enterica; and CC4 for C. sakazakii) have clinical relevance and are widely
distributed according to the analysis of clusters available in the NCBI Pathogen Detection database.
• For L. monocytogenes, some markers have been identified as possibly associated with persistence: stress survival islets
SSI-1 and SSI-2, genomic islands LGI-1 and LGI-2, heavy metal (cadmium and arsenic) and biocide (bcrABC, qacC, qacH,
emrE and emrC) resistance determinants, often located on mobile genetic elements (mainly plasmids), and bacteriophage
regions (comK), globally linked to increased environmental robustness, tolerance to disinfection and/or biofilm formation.
• The set of phenotypic and genomic features that have been investigated for Salmonella and C. sakazakii in relation to
persistence in the FFPE is incomplete. As such, it is difficult to deduce certain features, that are either indispensable for,
or may markedly contribute to, persistence, alone or in combination with other key genotypic and phenotypic elements.
• For Salmonella, most studies focused on features inherent to most infectious foodborne hazards (e.g. AMR, virulence,
growth/survival in foods and biofilm formation), and reported resistance of some strains to one or more antimicrobials,
carriage of plasmid-mediated virulence factors, biofilm formation ability or reduced susceptibility to alkaline disinfectants.
• Several features have been associated with the ability of C. sakazakii to survive for long time periods and persist in the dry
conditions of the LMF FoPE, including the ability to form biofilms on a variety of abiotic surfaces; a high heat tolerance
and desiccation resistance; the production of a capsule that aids attachment to surfaces, provides resistance to biocides
and contributes to survival following desiccation; and the production of a yellow carotenoid pigment which stabilises
cell membranes and provides protection against stress. However, none of these features seem to be specifically linked
to particular subtypes frequently associated with persistence.
• No universal markers or features, responsible for persistence have been identified. Although the carriage of different
combinations of genetic determinants linked to increased environmental robustness possibly confers the ability to per-
sist on particular subtypes, persistence is a multifactorial process that also depends on specific environmental condi-
tions and risk factors (discussed below).
ToR3 (AQ4). To identify the risk factors at facility level responsible for the persistence of the most relevant hazards in
the FFPE
• The main risk factor at facility level responsible for the persistence of the three bacterial hazards in the FFPE is poor
hygienic design of equipment and machines. This leads to niches (or harbourage sites) which are difficult to clean and
disinfect and where food debris and moisture can accumulate, and the hazards can survive and persist. Examples of such
niches on FCS are slicers and cutters for L. monocytogenes, feather plucking-and evisceration equipment for Salmonella
and dryers and drying towers for C. sakazakii.
• Other important factors are: (i) inadequate zoning and hygiene barriers, that enables the spread of contamination from
contaminated to clean areas; (ii) inadequate C&D of the facilities; (iii) introduction of the hazards through raw materials,
which may lead to the colonisation and spread of persistent clones in the processing environment; and (iv) humidity,
which favour persistence.
• Specifically for hazards of relevance in dry (LMF/feed) processing environments (S. enterica and C. sakazakii), additional
risk factors are airborne transmission through dust, the limited use of disinfectants due to dry cleaning operations, or
the presence of water in the FoPE, whether from wet cleaning, condensation generated through temperature gradients
within the facility or within equipment, or other sources.
ToR4 (AQ5). To assess available and enhanced measures or interventions for monitoring, preventing and/or controlling
the persistence of the most relevant microbiological food safety hazards in the FFPE
• A well-designed environmental sampling and testing programme, following a risk-based approach, is the most effective
strategy to identify potential contamination sources and detect potentially persistent hazards.
• The establishment of hygienic barriers and measures within the FSMS, during implementation of HACCP, is key to pre-
vent and/or control bacterial persistence in the FFPE through avoiding the entry of the hazard(s) to the processing plant
and/or their spread across the facility. The following prerequisites are of particular importance: infrastructure (building,
equipment), C&D, technical maintenance and calibration, water and air control, personnel (hygiene, health status), work-
ing methodology and food safety culture.
• The confirmation of the presence of a persistent strain, and the identification of its niche within the facility, require the
detailed characterisation of isolates of the specific hazard(s) recovered from positive samples using subtyping methods
with enough resolution, preferably WGS-based subtyping schemes.
• Once persistence is suspected in a food and feed processing plant, a ‘seek-and-destroy’ approach has been frequently
recommended, which includes: (i) intensified monitoring; (ii) the introduction of measures to control the event; (iii) and
the continuation of the intensified monitoring programme to confirm the efficacy of the measures taken or to identify
the requirement for additional measures. Alternatively, systematic ‘root cause analyses’ can be applied to identify the
most probable factors/sites within the facilities contributing to the problem and define the most appropriate interven-
tions to eliminate the pathogen from the premises.
• Successful actions triggered by persistence of L. monocytogenes in the FoPE, for example, included the introduction of
new or specialised (deep) C&D, the implementation of workflows, the installation of a new drainage system; the imple-
mentation of structural changes and renovations; the control of the contamination of raw ingredients and the improve-
ment of the compartmentalisation, or the simultaneous implementation of various corrective actions.
• Some options of interventions to eliminate the persistent hazard(s) with direct bactericidal activity and of different na-
ture (i.e. as chemical (e.g. use of biocides), physical (e.g. heat or novel non-thermal technologies) or biological (e.g. com-
petitive exclusion, phage)) are described but in some cases these are not yet commercially available and/or their efficacy
is not yet fully validated under industrial conditions.
ToR5 (AQ6-8). To identify knowledge gaps and priorities for future research and develop the perspectives of integrat-
ing the information gathered in the previous ToR in risk assessment
• Perspectives are provided for the use of risk assessment for relevant combinations of hazard and food product to assess
the relative PH risks that can be associated with persistence, based on bottom-up and top-down approaches.
◦ The proposed basic model for persistence to be used in bottom-up food chain QMRA can be used to study the role
of persistence in the PH risk for a specific food production process. The dynamics of persistence and its role in PH risk
will however be very food process specific, and the current model may be too simple to capture important biological
processes, such as biofilm formation.
◦ With the currently available data, top-down risk assessment cannot be used to assess the relative PH risk that can be
attributed to persistence.
• Risk assessment cannot fully exploit the data gathered to support answering the AQs of this scientific opinion, and the data
needs for risk assessment are not well covered. Application of these data would require better translation of genotypic
information of strains into phenotypic characteristics that can be converted into parameters of risk assessment models, as
well as extensive quantitative data to describe the dynamics of transfer, survival and growth of bacteria in the FoPE.
• Nine specific knowledge gaps have been identified and translated into recommendations for research for filling those
knowledge gaps.
• Most of these recommendations would involve activities at industry settings, but some of the research activities could
be performed using industrial-like model systems of certain niches, where different strains, environmental conditions
and potential interventions can be tested.
• These research activities would enable to establish the contribution of specific genetic markers and their link to phe-
notypes associated with persistence, and to monitor the impact of particular interventions in reducing or preventing
persistence. They would also allow the generation of quantitative data to describe the dynamics of transfer, survival and
growth of bacterial hazards and to define strain-or subtype-specific parameters for QMRA.
5 | R ECOM M E N DATIO NS
• To apply clear definitions of persistence in all involved research areas (observational, experimental, epidemiological,
etc.), aiming at the same unambiguous definition for all of them.
• The environmental sampling and testing programme should be robust and carefully planned by the food business op-
erators and ensure an adequate surveillance of higher risk niches for target bacterial hazards, preferably supported by
molecular subtyping, to control contamination by persistent strains.
• During outbreak investigation, to optimise the sampling strategy (e.g. frequency, critical points) and to improve data
reporting of official and industrial sampling, in order to strengthen the link between FFPE and the outbreak.
• To promote the use of interoperable standards to collect and report metadata associated with WGS data to ensure au-
ditability, to streamline data sharing and to reduce uncertainty.
• To promote the open access both to WGS data, and to complete and unambiguous associated metadata related to the
strain isolation, including strains from both industrial and official sampling, respecting data confidentiality and the inter-
ests of different partners in the food chain, for investigating persistence in the FFPE.
GLOSSARY
Biocide a chemical substance or microorganism intended to destroy, deter, render
harmless or exert a controlling effect on any harmful organism by chemical
or biological means.31
Cleaning the removal of soil, food residue, dirt, grease or other objectionable
matter.32
Critical control point(s) (CCP) a step at which control can be applied and is essential to prevent or elimi-
nate a food safety hazard or reduce it to an acceptable level. Most typical
CCP to control microbiological hazards are temperature requirements e.g.
the time/temperature conditions to reduce or eliminate a hazard (e.g. pas-
teurisation). Other CCP may be checking for micro-lesions in canned food,
checking for physical hazards by sieving or metal detection or checking
time/temperature of frying oil to avoid chemical process contaminants (EU
Commission Notice, 2022/C355/01.33
Disinfecting/disinfection to destroy or irreversibly inactivate specified fungi, bacteria and/or viruses,
but not necessarily bacterial spores
Disinfectant chemical agent or combination of chemical agents that is used on inanimate
objects or surfaces. Some chemicals may function as both sanitisers and dis-
infectants. Disinfectants can be sporostatic but are not necessarily sporicidal.
Within the remit of this opinion, disinfectant agents are defined as those de-
contamination agents applied to eliminate microorganisms on surfaces
Food Safety Management system (FSMS) Prerequisite programmes, supplemented with control measures at CCP,
as appropriate, that when taken as a whole, ensure that food is safe and
suitable for its intended use. The FSMS is also the combination of control
measures and assurance activities. The latter aims at providing evidence
that control measures are working properly such as validation and verifi-
cation, documentation and record keeping (EU Commission Notice, 2022/
C355/0133)
Good Hygiene Practices (GHP) Fundamental measures and conditions applied at any step within the food
chain to provide safe and suitable food. GHP include also good manufac-
turing practice(s) (GMP, stressing correct work methodologies e.g. correct
dosage of ingredients, appropriate processing temperature, checking that
packages are clean and non-damaged), good agriculture practice(s) (GAP,
e.g. use of water of appropriate quality for irrigation, all in/all out system
in animal rearing), good veterinarian practice(s) (GVP), good production
practice(s) (GPP), good distribution practice(s) (GDP) and good trading
practice(s) (GTP) (EU Commission Notice, 2022/C355/0133)
Monitor The act of conducting a planned sequence of observations or measure-
ments of control parameters to assess whether a control measure is under
control (EU Commission Notice 2022/C355/0133)
Niche the harbourage site of persistent strains
Operational Prerequisite Programme(s) (oPRP) control measure or combination of control measures applied to prevent or
reduce a significant food safety hazard to an acceptable level and where
action criterion and measurement or observation enable effective control
of the process and/or product. They are typically linked to the production
process and are identified by the hazard analysis as essential, in order to
control the likelihood of the introduction, survival and/or proliferation of
food safety hazards in the product(s) or in the processing environment (EU
Commission Notice, 2022/C355/0133)
Prerequisite programme(s) (PRP) Preventive practices and conditions including all GHP, as well as other
practices and procedures such as training and traceability, that establish
the basic environmental and operating conditions that set the foundation
for implementation of HACCP-based procedures (EU Commission Notice,
2022/C355/0133)
31
CAC (Codex Alimentarius Commission), 2003. Code of Hygienic Practice for Fresh Fruits and Vegetables CXC 53–2003 p. 1–39.
32
CAC (Codex Alimentarius Commission), 2022. General Principles of Food Hygiene CXC 1–1969 p. 1–38.
33
European Commission, 2022. Commission Notice on the implementation of food safety management systems covering Good Hygiene Practices and procedures based
on the HACCP principles, including the facilitation/flexibility of the implementation in certain food businesses (2022/C 355/01). 16.9.2022, p. 1–58. https://eur-lex.europa.
eu/legal-content/EN/TXT/PDF/?uri=CELEX:52022XC0916(01)
Presumed non-persistent strain a strain that has been identified as sporadically (not repeatedly) contami-
nating the FFPE of a processing plant, as a more intensified or a longer
sampling campaign could result in their repeated isolation from the FFPE
Persistent strain a strain found to be established in niches within the FFPE for a long term,
despite the frequent application of C&D. It requires prolonged existence
usually with multiplication of the microorganism in the specific FFPE. It is a
phenomenon which may lead to recurrent food contamination events and
is normally detected through the repeated isolation from the same prem-
ises or equipment on different dates (spanning months or years) of strains
that are subsequently identified as highly related subtypes (as determined
by phenotypic or genotypic methods). Persistence does not include con-
tinuous reintroduction in the facility of the same organism, although in
practice it is often not possible to distinguish between both phenomena
Pervasive strain a persistent strain isolated from different processing plants
Sanitation Used to reduce, but not necessarily eliminate, microorganisms from the
inanimate environment to levels considered safe as determined by pub-
lic health codes or regulations. Process of reducing microbiological con-
tamination on an effectively cleaned surface using a bactericidal treatment
such as heat or chemicals, to a level that is acceptable to local health regu-
lations. For effectiveness, this must be preceded by cleaning (a mix of de-
tergent and disinfectant or a disinfectant).34
Site the location of persistent strains (sampling sites positive for persistent
strains)
Surveillance the systematic ongoing collection, collation and analysis of information
related to food safety and the timely dissemination of information for as-
sessment and response as necessary (FAO, 2022).35
Validation Obtaining evidence that a control measure or combination of control meas-
ures, if properly implemented in the HACCP-based procedures and by the
oPRP, can control the hazard to a specified outcome. Revalidation may be
required in case of changes affecting the control measure. Detailed ex-
amples can be found in CAC/GL 69-200836 (EU Commission Notice, 2022/
C355/0133)
Verification The application of methods, procedures, tests and other evaluations, in
addition to monitoring to determine compliance with the HACCP-based
procedures, i.e. to determine whether a control measure is or has been op-
erating as intended. Verification is conducted periodically to demonstrate
that the HACCP system and the management of the oPRP are working as
planned (EU Commission Notice 2022/C355/0133)
A B B R E V I AT I O N S
ADI arginine deiminase
AFLP amplified fragment length polymorphism
AhpCF alkyl hydroperoxidase
AMR antimicrobial resistance
AQ assessment question(s)
ATR acid tolerance response
BapL biofilm associated protein
BC benzalkonium chloride
BIOHAZ Panel EFSA Panel on Biological Hazards
CAC Codex Alimentarius Commission
Cat catalase
CC clonal complex(es)
CDC Centers for Disease Control and Prevention (United States of America)
cgMLST core genome multi-locus sequence type
CEA Controlled Environment Agriculture
CFU colony forming unit(s)
CIP cleaning-in place
34
Environmental Protection Agency (EPA). (n.d.). What are antimicrobial pesticides? https://www.epa.gov/pesticide-registration/whatare-antimicrobial-pesticides
35
FAO (Food and Agriculture Organization), 2022. Listeria monocytogenes in ready-to-eat (RTE) foods: attribution, characterisation and monitoring. 202 pp. Available
online: https://www.fao.org/3/cc2400en/cc2400en.pdf
36
CAC (Codex Alimentarius), 2008. Guidelines for the validation of food safety control measures. CAC/GL 69-2008.p.1–16.
CP control point(s)
Csp cold shock protein
CT clonal type(s)
C&D cleaning and disinfection
DALY disability adjusted life year
DR dose–response
ECDC European Centre for Disease Prevention and Control
EMP environmental monitoring programme(s)
EnABLe Environmental monitoring with an Agent-Based Model of Listeria
ESBL Extended Spectrum Beta-Lactamase
FAO Food and Agriculture Organization of the United Nations
FBO foodborne outbreak(s)
FBOp food business operator(s)
FCS food contact surface(s)
FDA Food and Drug Administration (United States of America)
FePE feed processing environment(s)
FFPE food and feed processing environment(s)
FoPE food processing environment(s)
fri ferritin-like protein
FSANZ Food Standards Australia New Zealand
FSIS United States Department of Agriculture (USDA), Food Safety and Inspection Services
FSMS food safety management system(s)
GAD glutamate decarboxylase
GMP good manufacturing practice(s)
HACCP hazard analysis and critical control points
HUS haemolytic uremic syndrome
JNS joint notification summary
LGI Listeria Genomic Island
LIPI Listeria Pathogenicity Island
LMF low moisture food
MAG Metagenome Assembled Genome
MLST multi-locus sequence typing
MLVA multi-locus variable number tandem repeat analysis
MLVST multi-virulence-locus sequence typing
MPD maximum population density
NCBI National Center for Biotechnology Information
NFCS non-food contact surface(s)
oPRP operational prerequisite programme(s)
PFGE pulsed-field gel electrophoresis
PH Public health
PMSC premature stop codon
PRP prerequisite programme(s)
QAC quaternary ammonium compound
QMRA quantitative microbiological risk assessment
RAD restriction site-associated DNA
RASFF Rapid Alert System for Food and Feed
ROA rapid outbreak assessment(s)
ROS reactive oxygen species
RR relative risk
RTE ready-to-eat
SNP single nucleotide polymorphism
Sod superoxide dismutase
SQ sub-question(s)
SSI stress survival islet
ST sequence type(s)
STEC Shiga toxin-producing Escherichia coli
Ti/Ab title/abstract
ToR Terms of Reference
US FDA United States Food and Drug Administration
VBNC viable but non-culturable
wgMLST whole genome multi-locus sequence type
WGS whole genome sequencing

WHO World Health Organization
AC K N OW L ED G EM E N T S

The Panel wishes to thank the EFSA staff Mirko Rossi and Irene Muñoz Guajardo for their contributions to this scientific output.
CONFLIC T OF INTEREST
If you wish to access the declaration of interests of any expert contributing to an EFSA scientific assessment, please contact
interestmanagement@efsa.europa.eu.
REQUESTOR
EFSA
QUESTION NUMBER
EFSA-Q -2022-00217
PA N E L M E M B E R S
Ana Allende, Avelino Alvarez-Ordóñez, Declan Bolton, Sara Bover-Cid, Marianne Chemaly, Alessandra De Cesare, Lieve
Herman, Friederike Hilbert, Konstantinos Koutsoumanis, Roland Lindqvist, Maarten Nauta, Romolo Nonno, Luisa Peixe,
Giuseppe Ru, Marion Simmons, Panagiotis Skandamis, and Elisabetta Suffredini.
C O P Y R I G H T F O R N O N - E F S A C O N T E N T
EFSA may include images or other content for which it does not hold copyright. In such cases, EFSA indicates the copyright
holder and users should seek permission to reproduce the content from the original source.
MAP DISCLAIMER
The designations employed and the presentation of material on any maps included in this scientific output do not imply
the expression of any opinion whatsoever on the part of the European Food Safety Authority concerning the legal status of
any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries.
REFERENCES
Abdallah, M., Benoliel, C., Drider, D., Dhulster, P., & Chihib, N.-E. (2014). Biofilm formation and persistence on abiotic surfaces in the context of food and
medical environments. Archives of Microbiology, 196, 453–472. https://doi.org/10.1007/s00203- 014- 0983-1
Acciari, V. A., Iannetti, L., Gattuso, A., Sonnessa, M., Scavia, G., Montagna, C., Addante, N., Torresi, M., Zocchi, L., Scattolini, S., Centorame, P., Marfoglia, C.,
Prencipe, V. A., & Gianfranceschi, M. V. (2016). Tracing sources of Listeria contamination in traditional Italian cheese associated with a US outbreak:
Investigations in Italy. Epidemiology and Infection, 144, 2719–2727. https://doi.org/10.1017/s095026881500254x
Alvarez-Molina, A., Cobo-Diaz, J. F., Lopez, M., Prieto, M., de Toro, M., & Alvarez-Ordonez, A. (2021). Unraveling the emergence and population diversity
of Listeria monocytogenes in a newly built meat facility through whole genome sequencing. International Journal of Food Microbiology, 340, 13.
https://doi.org/10.1016/j.ijfoodmicro.2021.109043
Alvarez-Ordóñez, A., Coughlan, L. M., Briandet, R., & Cotter, P. D. (2019). Biofilms in food processing environments: Challenges and opportunities. Annual
Review of Food Science and Technology, 10, 173–195. https://doi.org/10.1146/annurev-food- 032818-121805
Alvarez-Ordonez, A., Leong, D., Hunt, K., Scollard, J., Butler, F., & Jordan, K. (2018). Production of safer food by understanding risk factors for L-
monocytogenes occurrence and persistence in food processing environments. Journal of Food Safety, 38, 7. https://doi.org/10.1111/jfs.12516
Anast, J. M., Etter, A. J., & Schmitz-Esser, S. (2022). Comparative analysis of Listeria monocytogenes plasmid transcriptomes reveals common and plasmid-
specific gene expression patterns and high expression of noncoding RNAs. Microbiology, 11, 17. https://doi.org/10.1002/mbo3.1315
Angelidis, A. S., & Smith, G. M. (2003). Role of the glycine betaine and carnitine transporters in adaptation of Listeria monocytogenes to chill stress in
defined medium. Applied and Environmental Microbiology, 69, 7492–7498. https://doi.org/10.1128/aem.69.12.7492-7498.2003
Angelo, K. M., Conrad, A. R., Saupe, A., Dragoo, H., West, N., Sorenson, A., Barnes, A., Doyle, M., Beal, J., Jackson, K. A., Stroika, S., Tarr, C., Kucerova, Z.,
Lance, S., Gould, L. H., Wise, M., & Jackson, B. R. (2017). Multistate outbreak of Listeria monocytogenes infections linked to whole apples used in
commercially produced, prepackaged caramel apples: United States, 2014–2015. Epidemiology and Infection, 145, 848–856. https://doi.org/10.1017/
s0950268816003083
APHA (Animal and Plant Health Agency), 2022. Salmonella in animals and feed in Great Britain. 2022. https://assets.publishing.service.gov.uk/gover
nment/uploads/system/uploads/attachment_data/f ile/1120012/salmonella-animals-feed-gb-2021-v.2__003_.pdf
Arguello, H., Alvarez-Ordonez, A., Carvajal, A., Rubio, P., & Prieto, M. (2013). Role of slaughtering in Salmonella spreading and control in pork production.
Journal of Food Protection, 76, 899–911. https://doi.org/10.4315/0362-028x.Jfp-12-404
Arguello, H., Carvajal, A., Naharro, G., Arcos, M., Rodicio, M. R., Martin, M. C., & Rubio, P. (2013). Sero-and genotyping of Salmonella in slaughter pigs,
from farm to cutting plant, with a focus on the slaughter process. International Journal of Food Microbiology, 161, 44–52. https://doi.org/10.
1016/j.ijfoodmicro.2012.11.013
Ashraf, M. A., Ullah, S., Ahmad, I., Qureshi, A. K., Balkhair, K. S., & Rehman, M. A. (2014). Green biocides, a promising technology: Current and future appli-
cations to industry and industrial processes. Journal of the Science of Food and Agriculture, 94, 388–403. https://doi.org/10.1002/jsfa.6371
Baert, L., McClure, P., Winkler, A., Karn, J., Bouwknegt, M., & Klijn, A. (2021). Guidance document on the use of whole genome sequencing (WGS) for
source tracking from a food industry perspective. Food Control, 130, 108148. https://doi.org/10.1016/j.foodcont.2021.108148
Bai, X. J., Nakatsu, C. H., & Bhunia, A. K. (2021). Bacterial biofilms and their implications in pathogenesis and food safety. Food, 10, 26. https://doi.org/
10.3390/foods10092117
Bangieva, D., & Rusev, V. (2017). Prevalence of Listeria monocytogenes in raw cow milk -a review. Bulgarian Journal of Veterinary Medicine, 20, 430–436.
Baquero, F., Lanza, V. F., Duval, M., & Coque, T. M. (2020). Ecogenetics of antibiotic resistance in Listeria monocytogenes. Molecular Microbiology, 113,
570–579. https://doi.org/10.1111/mmi.14454
Barata, A. R., Nunes, B., Oliveira, R., Guedes, H., Almeida, C., Saavedra, M. J., da Silva, G. J., & Almeida, G. (2022). Occurrence and seasonality of campylo-
bacter spp. in Portuguese dairy farms. International Journal of Food Microbiology, 383, 109961. https://doi.org/10.1016/j.ijfoodmicro.2022.109961
Barcenilla, C., Cobo-Diaz, J. F., De Filippis, F., Valentino, V., Cabrera Rubio, R., O'Neil, D., Mahler de Sanchez, L., Armanini, F., Carlino, N., Blanco-Míguez, A.,
Pinto, F., Calvete-Torre, I., Sabater, C., Delgado, S., Ruas-Madiedo, P., Quijada, N. M., Dzieciol, M., Skírnisdóttir, S., … Alvarez-Ordóñez, A. (in press).
Improved sampling and DNA extraction procedures for microbiome analysis in food processing environments. Nature Protocols.
Barría, C., Singer, R. S., Bueno, I., Estrada, E., Rivera, D., Ulloa, S., Fernández, J., Mardones, F. O., & Moreno-Switt, A. I. (2020). Tracing Listeria monocyto-
genes contamination in artisanal cheese to the processing environments in cheese producers in southern Chile. Food Microbiology, 90, 103499.
https://doi.org/10.1016/j.fm.2020.103499
Barron, J. C., & Forsythe, S. J. (2007). Dry stress and survival time of Enterobacter sakazakii and other Enterobacteriaceae in dehydrated powdered infant
formula. Journal of Food Protection, 70, 2111–2117. https://doi.org/10.4315/0362-028x-70.9.2111
Beales, N. (2004). Adaptation of microorganisms to cold temperatures, weak acid preservatives, Low pH, and osmotic stress: A review. Comprehensive
Reviews in Food Science and Food Safety, 3, 1–20. https://doi.org/10.1111/j.1541- 4337.2004.tb00057.x
Beccalli, M. P., Picozzi, C., Mangieri, N., Vigentini, I., & Foschino, R. (2019). Assessment of Microbial populations in the manufacture of vacuum-packaged
ready-to-eat roast beef and in a related production plant. Journal of Food Protection, 82, 58–64. https://doi.org/10.4315/0362-028x.Jfp-18-147
Becker, L. A., Evans, S. N., Hutkins, R. W., & Benson, A. K. (2000). Role of sigma(B) in adaptation of Listeria monocytogenes to growth at low temperature.
Journal of Bacteriology, 182, 7083–7087. https://doi.org/10.1128/jb.182.24.7083-7087.2000
Begley, M., & Hill, C. (2015). Stress adaptation in foodborne pathogens. In M. P. Doyle & T. R. Klaenhammer (Eds.), Annual review of food science and tech-
nology (Vol. 6, pp. 191–210). Palo Alto.
Belias, A., Sullivan, G., & Wiedmann, M. (2021). Root cause analysis can be used to identify and reduce a highly diverse Listeria population in an apple
packinghouse: A case study. Food Protection Trends, 41, 555–567.
Belias, A., Sullivan, G., Wiedmann, M., & Ivanek, R. (2022). Factors that contribute to persistent Listeria in food processing facilities and relevant interven-
tions: A rapid review. Food Control, 133, 12. https://doi.org/10.1016/j.foodcont.2021.108579
Bell, R. L., Kase, J. A., Harrison, L. M., Balan, K. V., Babu, U., Chen, Y., Macarisin, D., Kwon, H. J., Zheng, J., Stevens, E. L., Meng, J., & Brown, E. W. (2021).
The persistence of bacterial pathogens in surface water and its impact on global food safety. Pathogens, 10. https://doi.org/10.3390/pathogens1
0111391
Bennett, S. D., Littrell, K. W., Hill, T. A., Mahovic, M., & Behravesh, C. B. (2015). Multistate foodborne disease outbreaks associated with raw tomatoes,
United States, 1990–2010: A recurring public health problem. Epidemiology and Infection, 143, 1352–1359. https://doi.org/10.1017/s095026881
4002167
Bergholz, T. M., den Bakker, H. C., Katz, L. S., Silk, B. J., Jackson, K. A., Kucerova, Z., Joseph, L. A., Turnsek, M., Gladney, L. M., Halpin, J. L., Xavier, K.,
Gossack, J., Ward, T. J., Frace, M., & Tarrc, C. L. (2016). Determination of evolutionary relationships of outbreak-associated Listeria monocytogenes
strains of serotypes 1/2a and 1/2b by whole-genome sequencing. Applied and Environmental Microbiology, 82, 928–938. https://doi.org/10.1128/
aem.02440-15
Bernbom, N., Vogel, B. F., & Gram, L. (2011). Listeria monocytogenes survival of UV-C radiation is enhanced by presence of sodium chloride, organic food
material and by bacterial biofilm formation. International Journal of Food Microbiology, 147, 69–73. https://doi.org/10.1016/j.ijfoodmicro.2011.03.
009
Berrang, M. E., Harrison, M. A., Meinersmann, R. J., & Gamble, G. R. (2017). Self-contained chlorine dioxide generation and delivery pods for decontami-
nation of floor drains. Journal of Applied Poultry Research, 26, 410–415. https://doi.org/10.3382/japr/pfx009
Berrang, M. E., Meinersmann, R. J., Frank, J. F., & Ladely, S. R. (2010). Colonization of a newly constructed commercial chicken further processing plant
with Listeria monocytogenes. Journal of Food Protection, 73, 286–291.
Bersot, L. D., Carbonera, N. R., Valcanaia, C. D. R., Viana, C., & Nero, L. A. (2021). Multidrug-resistant and extended-Spectrum β-lactamase-producing
Salmonella enterica serotype Heidelberg is widespread in a poultry processing Facility in Southern Brazil. Journal of Food Protection, 84, 2053–2058.
https://doi.org/10.4315/jfp-21-140
Bertrand, S., Dierick, K., Heylen, K., De Baere, T., Pochet, B., Robesyn, E., Lokietek, S., Van Meervenne, E., Imberechts, H., De Zutter, L., & Collard, J. M.
(2010). Lessons learned from the Management of a National Outbreak of Salmonella Ohio linked to pork meat processing and distribution. Journal
of Food Protection, 73, 529–534. https://doi.org/10.4315/0362- 028x-73.3.529
Beuchat, L. R. (2002). Ecological factors influencing survival and growth of human pathogens on raw fruits and vegetables. Microbes and Infection, 4,
413–423. https://doi.org/10.1016/S1286-4579(02)01555-1
Biguzzi, V., Pezzotti, G., Gianfranceschi, M. V., & De Medici, D. (2012). Tracing the source of persistent contamination of ready-to-eat chicken products
with Listeria monocytogenes in Italy. Case Studies in Food Safety and Authenticity, 102–109.
Binter, C., Straver, J. M., Haggblom, P., Bruggeman, G., Lindqvist, P. A., Zentek, J., & Andersson, M. G. (2011). Transmission and control of Salmonella in
the pig feed chain: A conceptual model. International Journal of Food Microbiology, 145, S7–S17. https://doi.org/10.1016/j.ijfoodmicro.2010.09.001
Blatter, S., Giezendanner, N., Stephan, R., & Zweifel, C. (2010). Phenotypic and molecular typing of Listeria monocytogenes isolated from the processing
environment and products of a sandwich-producing plant. Food Control, 21, 1519–1523. https://doi.org/10.1016/j.foodcont.2010.04.025
Bolocan, A. S., Nicolau, A. I., Alvarez-Ordonez, A., Borda, D., Oniciuc, E. A., Stessl, B., Gurgu, L., Wagner, M., & Jordan, K. (2016). Dynamics of Listeria
monocytogenes colonisation in a newly-opened meat processing facility. Meat Science, 113, 26–34. https://doi.org/10.1016/j.meatsci.2015.10.016
Bonifait L, Thepault A, Baugé L, Markwitz B and Bris V, 2022. Salmonella Kedougou, from feed to poultry farms using wgs. International symposium
Salmonella and salmonellosis. 40 pp.
Boubendir, S., Arsenault, J., Quessy, S., Thibodeau, A., Fravalo, P., Thériault, W. P., Fournaise, S., & Gaucher, M. L. (2021). Salmonella contamination of
broiler chicken carcasses at critical steps of the slaughter process and in the environment of two slaughter plants: Prevalence, genetic profiles, and
association with the final carcass status. Journal of Food Protection, 84, 321–332. https://doi.org/10.4315/jfp-20-250
Bourdichon, F., Betts, R., Dufour, C., Fanning, S., Farber, J., McClure, P., Stavropoulou, D. A., Wemmenhove, E., Zwietering, M. H., & Winkler, A. (2021).
Processing environment monitoring in low moisture food production facilities: Are we looking for the right microorganisms? International Journal
of Food Microbiology, 356, 19. https://doi.org/10.1016/j.ijfoodmicro.2021.109351
Brackett, R. E. (2007). Incidence and behavior of Listeria monocytogenes in products of plant origin. Listeria, Listeriosis, and Food Safety, Third Edition,
CRC Press., 655–680.
Breuer, T. (1999). CDC investigations: The May 1998 outbreak of Salmonella Agona linked to cereal. Cereal Foods World, 44, 185–186.
Bridier, A., Le Grandois, P., Moreau, M. H., Prénom, C., Le Roux, A., Feurer, C., & Soumet, C. (2019). Impact of cleaning and disinfection procedures on micro-
bial ecology and Salmonella antimicrobial resistance in a pig slaughterhouse. Scientific Reports, 9, 13. https://doi.org/10.1038/s41598- 019- 49464-8
Brown, P., Chen, Y., Siletzky, R., Parsons, C., Jaykus, L. A., Eifert, J., Ryser, E., Logue, C. M., Stam, C., Brown, E., & Kathariou, S. (2021). Harnessing whole
genome sequence data for facility-specific signatures for Listeria monocytogenes: A case study with Turkey processing plants in the United States.
Frontiers in Sustainable Food Systems, 5. https://doi.org/10.3389/fsufs.2021.742353
Buchrieser, C., Rusniok, C., Kunst, F., Cossart, P., Glaser, P., & Listeria, C. (2003). Comparison of the genome sequences of Listeria monocytogenes and
Listeria innocua: Clues for evolution and pathogenicity. FEMS Immunology and Medical Microbiology, 35, 207–213. https://doi.org/10.1016/s0928-
8244(02)00448-0
Bucur, F. I., Grigore-Gurgu, L., Crauwels, P., Riedel, C. U., & Nicolau, A. I. (2018). Resistance of Listeria monocytogenes to stress conditions encountered in
food and food processing environments. Frontiers in Microbiology, 9, 18. https://doi.org/10.3389/fmicb.2018.02700
Burnett, E., Kucerova, Z., Freeman, M., Kathariou, S., Chen, J., & Smikle, M. (2022). Whole-genome sequencing reveals multiple subpopulations of domi-
nant and persistent lineage I isolates of Listeria monocytogenes in two meat processing facilities during 2011–2015. Microorganisms, 10. https://doi.
org/10.3390/microorganisms10051070
Byun, K. H., & Kim, H. J. (2023). Survival strategies of Listeria monocytogenes to environmental hostile stress: Biofilm formation and stress responses.
Food Science and Biotechnology, 32, 1631–1651. https://doi.org/10.1007/s10068-023-01427-6
Cabrita, P., Trigo, M. J., Ferreira, R. B., & Brito, L. (2015). Differences in the expression of cold stress-related genes and in the swarming motility among
persistent and sporadic strains of Listeria monocytogenes. Foodborne Pathogens and Disease, 12, 576–584. https://doi.org/10.1089/fpd.2014.1918
CAC (Codex Alimentarius Commission). (2008). Code of Hygienic Practice for Powdered Formulae for Infants and Young Children. CAC/RCP 66-2008.
29 pp. https://w ww.fao.org/fao-whocodexalimentarius/-codexalimentarius/codex-tests/cpdes-of-practice/en/
Camargo, A. C., Dias, M. R., Cossi, M. V. C., Lanna, F., Cavicchioli, V. Q., Vallim, D. C., Pinto, P. S. D., Hofer, E., & Nero, L. A. (2015). Serotypes and Pulsotypes
diversity of Listeria monocytogenes in a beef-processing environment. Foodborne Pathogens and Disease, 12, 323–326. https://doi.org/10.1089/
fpd.2014.1875
Campdepadros, M., Stchigel, A. M., Romeu, M., Quilez, J., & Sola, R. (2012). Effectiveness of two sanitation procedures for decreasing the microbial
contamination levels (including Listeria monocytogenes) on food contact and non-food contact surfaces in a dessert-processing factory. Food
Control, 23, 26–31. https://doi.org/10.1016/j.foodcont.2011.05.017
Campos, B., Pickering, A. C., Rocha, L. S., Aguilar, A. P., Fabres-Klein, M. H., de Oliveira Mendes, T. A., Fitzgerald, J. R., & de Oliveira Barros Ribon, A.
(2022). Diversity and pathogenesis of Staphylococcus aureus from bovine mastitis: Current understanding and future perspectives. BMC Veterinary
Research, 18:115. https://doi.org/10.1186/s12917- 022- 03197-5
Carrascosa, C., Raheem, D., Ramos, F., Saraiva, A., & Raposo, A. (2021). Microbial biofilms in the food industry-a comprehensive review. International
Journal of Environmental Research and Public Health, 18, 31. https://doi.org/10.3390/ijerph18042014
Centorotola, G., Guidi, F., D'Aurizio, G., Salini, R., Di Domenico, M., Ottaviani, D., Petruzzelli, A., Fisichella, S., Duranti, A., Tonucci, F., Acciari, V. A., Torresi,
M., Pomilio, F., & Blasi, G. (2021). Intensive environmental surveillance plan for Listeria monocytogenes in food producing plants and retail Stores
of Central Italy: Prevalence and genetic diversity. Food, 10, 16. https://doi.org/10.3390/foods10081944
Chaitiemwong, N., Hazeleger, W. C., & Beumer, R. R. (2010). Survival of Listeria monocytogenes on a conveyor belt material with or without antimicrobial
additives. International Journal of Food Microbiology, 142, 260–263. https://doi.org/10.1016/j.ijfoodmicro.2010.06.021
Chaitiemwong, N., Hazeleger, W. C., Beumer, R. R., & Zwietering, M. H. (2014). Quantification of transfer of Listeria monocytogenes between cooked ham
and slicing machine surfaces. Food Control, 44, 177–184. https://doi.org/10.1016/j.foodcont.2014.03.056
Charlebois, S., & Horan, H. (2010). Institutional and relational determinants in high-and medium-extent food product crises: The inner perspective of a
public health crisis. International Journal of Environmental Health Research, 20, 299–312. https://doi.org/10.1080/09603123.2010.482151
Chase, H. R., Gopinath, G. R., Eshwar, A. K., Stoller, A., Fricker-Feer, C., Gangiredla, J., Patel, I. R., Cinar, H. N., Jeong, H., Lee, C., Negrete, F., Finkelstein, S.,
Stephan, R., Tall, B. D., & Lehner, A. (2017). Comparative genomic characterization of the highly persistent and potentially virulent Cronobacter
sakazakii ST83, CC65 Strain H322 and other ST83 strains. Frontiers in Microbiology, 8, 9. https://doi.org/10.3389/fmicb.2017.01136
Chen, B. Y., Wang, C. Y., Wang, C. L., Fan, Y. C., Weng, I. T., & Chou, C. H. (2016). Prevalence and persistence of Listeria monocytogenes in ready-to-eat
tilapia sashimi processing plants. Journal of Food Protection, 79, 1898–1903. https://doi.org/10.4315/0362-028x.Jfp-16-149
Chen, M. T., Wu, Q. P., Zhang, J. M., Guo, W. P., Wu, S., & Yang, X. B. (2014). Prevalence and contamination patterns of Listeria monocytogenes in Flammulina
velutipes plants. Foodborne Pathogens and Disease, 11, 620–627. https://doi.org/10.1089/fpd.2013.1727
Chen, Y., Burall, L. S., Luo, Y., Timme, R., Melka, D., Muruvanda, T., Payne, J., Wang, C., Kastanis, G., Maounounen-Laasri, A., De Jesus, A. J., Curry, P. E.,
Stones, R., K'Aluoch, O., Liu, E., Salter, M., Hammack, T. S., Evans, P. S., Parish, M., … Brown, E. W. (2016). Listeria monocytogenes in Stone fruits linked
to a multistate outbreak: Enumeration of cells and whole-genome sequencing. Applied and Environmental Microbiology, 82, 7030–7040. https://doi.
org/10.1128/aem.01486-16
Chen, Y., Luo, Y., Curry, P., Timme, R., Melka, D., Doyle, M., Parish, M., Hammack, T. S., Allard, M. W., Brown, E. W., & Strain, E. A. (2017). Assessing the genome
level diversity of Listeria monocytogenes from contaminated ice cream and environmental samples linked to a listeriosis outbreak in the United
States. PLoS One, 12, 19. https://doi.org/10.1371/journal.pone.0171389
Chen, Y., Simonetti, T., Peter, K., Jin, Q., Brown, E., LaBorde, L. F., & Macarisin, D. (2022). Genetic diversity of Listeria monocytogenes isolated from three
commercial tree fruit packinghouses and evidence of persistent and transient contamination. Frontiers in Microbiology, 12, 15. https://doi.org/
10.3389/fmicb.2021.756688
Cherifi, T., Arsenault, J., Pagotto, F., Quessy, S., Côté, J. C., Neira, K., Fournaise, S., Bekal, S., & Fravalo, P. (2020). Distribution, diversity and persistence of
Listeria monocytogenes in swine slaughterhouses and their association with food and human listeriosis strains. PLoS One, 15, 19. https://doi.org/
10.1371/journal.pone.0236807
Cherifi, T., Carrillo, C., Lambert, D., Miniaï, I., Quessy, S., Larivière-Gauthier, G., Blais, B., & Fravalo, P. (2018). Genomic characterization of Listeria monocy-
togenes isolates reveals that their persistence in a pig slaughterhouse is linked to the presence of benzalkonium chloride resistance genes. BMC
Microbiology, 18, 220. https://doi.org/10.1186/s12866- 018-1363-9
Chiara, M., D'Erchia, A. M., Manzari, C., Minotto, A., Montagna, C., Addante, N., Santagada, G., Latorre, L., Pesole, G., Horner, D. S., & Parisi, A. (2014).
Draft genome sequences of Six Listeria monocytogenes strains isolated from dairy products from a processing Plant in Southern Italy. Genome
Announcements, 2. https://doi.org/10.1128/genomeA.00282-14
Chlebicz, A., & Slizewska, K. (2018). Campylobacteriosis, salmonellosis, Yersiniosis, and Listeriosis as zoonotic foodborne diseases: A review. International
Journal of Environmental Research and Public Health, 15, 28. https://doi.org/10.3390/ijerph15050863
Christidis, T., Pintar, K. D., Butler, A. J., Nesbitt, A., Thomas, M. K., Marshall, B., & Pollari, F. (2016). Campylobacter spp. prevalence and levels in raw Milk: A
systematic review and meta-analysis. Journal of Food Protection, 79, 1775–1783. https://doi.org/10.4315/0362-028x.Jfp-15-480
Coignard B, 2006. Infections à Enterobacter sakazakii associées à la consommation d'une préparation en poudre pour nourrissons. France, octobre à décembre
2004. Rapport d'investigation. https://w ww.santepubliquefrance.fr/docs/infections-a-enterobacter-sakazakii-associees-a-la-consommation-d-
une-preparation-en-poudre-pour-nourrissons.-france-octobre-a-decembre-2004.-rapp.place
Colagiorgi, A., Bruini, I., Di Ciccio, P. A., Zanardi, E., Ghidini, S., & Ianieri, A. (2017). Listeria monocytogenes biofilms in the wonderland of food industry.
Pathogens, 6, 9. https://doi.org/10.3390/pathogens6030041
Collins, Á. B., Floyd, S., Gordon, S. V., & More, S. J. (2022). Prevalence of Mycobacterium bovis in milk on dairy cattle farms: An international systematic
literature review and meta-analysis. Tuberculosis, 132, 102166. https://doi.org/10.1016/j.tube.2022.102166
Conficoni, D., Losasso, C., Cortini, E., Di Cesare, A., Cibin, V., Giaccone, V., Corno, G., & Ricci, A. (2016). Resistance to biocides in Listeria monocytogenes
collected in meat-processing environments. Frontiers in Microbiology, 7, 9. https://doi.org/10.3389/fmicb.2016.01627
Corcoran, M., Morris, D., De Lappe, N., O'Connor, J., Lalor, P., Dockery, P., & Cormican, M. (2013). Salmonella enterica biofilm formation and density in the
Centers for Disease Control and Prevention's biofilm reactor model is related to Serovar and substratum. Journal of Food Protection, 76, 662–667.
https://doi.org/10.4315/0362-028x.Jfp-12-303
Cordier, J.-L. (2007). Production of powdered infant formulae and microbiological control measures. Enterobacter Sakazakii, 145–185.
Cortimiglia, C., Luini, M., Bianchini, V., Marzagalli, L., Vezzoli, F., Avisani, D., Bertoletti, M., Ianzano, A., Franco, A., & Battisti, A. (2016). Prevalence of
Staphylococcus aureus and of methicillin-resistant S. Aureus clonal complexes in bulk tank milk from dairy cattle herds in Lombardy region (north-
ern Italy). Epidemiology and Infection, 144, 3046–3051. https://doi.org/10.1017/s0950268816001576
Costa, P. V., Vasconcellos, L., Forsythe, S. J., & Brandão, M. L. L. (2021). Diversity of Cronobacter genus isolated between 1970 and 2019 on the American
continent and genotyped using multi-locus sequence typing. FEMS Microbiology Letters, 368. https://doi.org/10.1093/femsle/fnab027
Coughlan, L. M., Cotter, P. D., Hill, C., & Alvarez-Ordonez, A. (2016). New weapons to fight old enemies: Novel strategies for the (bio)control of bacterial
biofilms in the food industry. Frontiers in Microbiology, 7, 21. https://doi.org/10.3389/fmicb.2016.01641
Craven, H. M., McAuley, C. M., Duffy, L. L., & Fegan, N. (2010). Distribution, prevalence and persistence of Cronobacter (Enterobacter sakazakii) in the
nonprocessing and processing environments of five milk powder factories. Journal of Applied Microbiology, 109, 1044–1052. https://doi.org/
10.1111/j.1365-2672.2010.04733.x
Currie, A., Farber, J. M., Nadon, C., Sharma, D., Whitfield, Y., Gaulin, C., Galanis, E., Bekal, S., Flint, J., Tschetter, L., Pagotto, F., Lee, B., Jamieson, F., Badiani,
T., MacDonald, D., Ellis, A., May-Hadford, J., McCormick, R., Savelli, C., … Sierpinska, U. (2015). Multi-Province Listeriosis outbreak linked to contam-
inated deli meat consumed primarily in institutional settings, Canada, 2008. Foodborne Pathogens and Disease, 12, 645–652. https://doi.org/
10.1089/fpd.2015.1939
Cutler, S. J. (2014). Bacteria: Coxiella burnetii. In Y. Motarjemi (Ed.), Encyclopedia of food safety (pp. 410–416). Academic Press.
Daeschel, D., Pettengill, J. B., Wang, Y., Chen, Y., Allard, M., & Snyder, A. B. (2022). Genomic analysis of Listeria monocytogenes from US food processing en-
vironments reveals a high prevalence of QAC efflux genes but limited evidence of their contribution to environmental persistence. BMC Genomics,
23, 14. https://doi.org/10.1186/s12864- 022- 08695-2
Dallman, T. J., Greig, D. R., Gharbia, S. E., & Jenkins, C. (2021). Phylogenetic structure of Shiga toxin-producing Escherichia coli O157:H7 from sub-lineage
to SNPs. Microbial Genomics, 7. https://doi.org/10.1099/mgen.0.000544
Dalmasso, M., & Jordan, K. (2013). Process environment sampling can help to reduce the occurrence of Listeria monocytogenes in food processing
facilities. Irish Journal of Agricultural and Food Research, 52, 93–100.
Dantas, S. T. A., Camargo, C. H., Tiba-Casas, M. R., Vivian, R. C., Pinto, J., Pantoja, J. C. F., Hernandes, R. T., Fernandes, A., & Rall, V. L. M. (2020). Environmental
persistence and virulence of Salmonella spp. isolated from a poultry slaughterhouse. Food Research International, 129, 8. https://doi.org/10.1016/j.
foodres.2019.108835
Dass, S. C., Abu-Ghannam, N., Antony-Babu, S., & Cummins, E. J. (2010). Ecology and molecular typing of L. monocytogenes in a processing plant for
cold-smoked salmon in the Republic of Ireland. Food Research International, 43, 1529–1536. https://doi.org/10.1016/goodres.2010.04.030
Davies, R. H., Nicholas, R. A. J., McLaren, I. M., Corkish, J. D., Lanning, D. G., & Wray, C. (1997). Bacteriological and serological investigation of persistent
Salmonella enteritidis infection in an integrated poultry organisation. Veterinary Microbiology, 58, 277–293. https://doi.org/10.1016/s0378-1135(97)
00157-0
Davies, R. H., & Wray, C. (1997). Distribution of salmonella contamination in ten animal feedmills. Veterinary Microbiology, 57, 159–169. https://doi.org/
10.1016/j.tube.2022.102166
De Cesare, A., Parisi, A., Mioni, R., Comin, D., Lucchi, A., & Manfreda, G. (2017). Listeria monocytogenes circulating in rabbit meat products and slaugh-
terhouses in Italy: Prevalence data and comparison among typing results. Foodborne Pathogens and Disease, 14, 167–176. https://doi.org/10.1089/
fpd.2016.2211
De Sousa, V. M., Podeur, G., Michel, V., Guillier, L., Radomski, N., Lailler, R., Le Hello, S., Weill, F.-X., Mistou, M.-Y., & Mallet, L. (2022). A retrospective and
regional approach assessing the genomic diversity of Salmonella Dublin. NAR Genomics and Bioinformatics, 4. https://doi.org/10.1093/nargab/
lqac047
Demaitre, N., De Reu, K., Haegeman, A., Schaumont, D., De Zutter, L., Geeraerd, A., & Rasschaert, G. (2021). Study of the transfer of Listeria monocyto-
genes during the slaughter of cattle using molecular typing. Meat Science, 175, 9. https://doi.org/10.1016/j.meatsci.2021.108450
Demaître, N., Rasschaert, G., De Zutter, L., Geeraerd, A., & De Reu, K. (2021). Genetic Listeria monocytogenes types in the pork processing plant environ-
ment: From occasional introduction to plausible persistence in harborage sites. Pathogens, 10. https://doi.org/10.3390/pathogens10060717
den Bakker, H. C., Didelot, X., Fortes, E. D., Nightingale, K. K., & Wiedmann, M. (2008). Lineage specific recombination rates and microevolution in Listeria
monocytogenes. BMC Evolutionary Biology, 8, 13. https://doi.org/10.1186/1471-2148-8-277
den Besten, H. M. W., Ding, Y. C., Abee, T., & Yang, L. (2016). Emergent strategies for detection and control of biofilms in food processing environments. John
Wiley & Sons Ltd.
Desai, A. N., Anyoha, A., Madoff, L. C., & Lassmann, B. (2019). Changing epidemiology of Listeria monocytogenes outbreaks, sporadic cases, and recalls
globally: A review of ProMED reports from 1996 to 2018. International Journal of Infectious Diseases, 84, 48–53. https://doi.org/10.1016/j.ijid.2019.
04.021
Dos Santos Rosario, A. I. L., da Silva, M. Y., Castro, V. S., da Silva, M. C. A., Conte-Junior, C. A., & da Costa, M. P. (2021). Everybody loves cheese: Crosslink
between persistence and virulence of Shiga-toxin Escherichia coli. Critical Reviews in Food Science and Nutrition, 61, 1877–1899. https://doi.org/
10.1080/10408398.2020.1767033
Douarre, P. E., Sévellec, Y., Le Grandois, P., Soumet, C., Bridier, A., & Roussel, S. (2022). FepR as a central genetic target in the adaptation to quaternary
ammonium compounds and cross-resistance to ciprofloxacin in Listeria monocytogenes. Frontiers in Microbiology, 13, 864576. https://doi.org/
10.3389/fmicb.2022.864576
Du, X.-j., Wang, F., Lu, X., Rasco, B. A., & Wang, S. (2012). Biochemical and genetic characteristics of Cronobacter sakazakii biofilm formation. Research in
Microbiology, 163, 448–456. https://doi.org/10.1016/j.resmic.2012.06.002
Du, X. J., Wang, X. Y., Dong, X., Li, P., & Wang, S. (2018). Characterization of the desiccation tolerance of Cronobacter sakazakii strains. Frontiers in
Microbiology, 9, 12. https://doi.org/10.3389/fmicb.2018.02867
Dubois-Brissonnet, F., Annous, B. A., & Gurtler, J. B. (2012). Adaptation of Salmonella to antimicrobials in food-processing environments.
Duffy, L. L., O'Callaghan, D., McAuley, C. M., Fegan, N., & Craven, H. M. (2009). Virulence properties of Escherichia coli isolated from Australian dairy pow-
der factory environments. International Dairy Journal, 19, 178–179. https://doi.org/10.1016/j.idair yj.2008.09.002
Duggan, S. J., Mannion, C., Prendergast, D. M., Leonard, N., Fanning, S., Gonzales-Barron, U., Egan, J., Butler, F., & Duffy, G. (2010). Tracking the Salmonella
status of pigs and pork from lairage through the slaughter process in the Republic of Ireland. Journal of Food Protection, 73, 2148–2160. https://doi.
org/10.4315/0362-028x-73.12.2148
Duranti, A., Sabbatucci, M., Blasi, G., Acciari, V. A., Ancora, M., Bella, A., Busani, L., Centorame, P., Camma, C., Conti, F., De Medici, D., Di Domenico, M., Di
Marzio, V., Filippini, G., Fiore, A., Fisichella, S., Gattuso, A., Gianfranceschi, M., Graziani, C., … Pomilio, F. (2018). A severe outbreak of listeriosis in
Central Italy with a rare pulsotype associated with processed pork products. Journal of Medical Microbiology, 67, 1351–1360. https://doi.org/
10.1099/jmm.0.000785
Duval, A., Opatowski, L., & Brisse, S. (2023). Defining genomic epidemiology thresholds for common-source bacterial outbreaks: A modelling study. The
Lancet Microbe, 4, e349–e357. https://doi.org/10.1016/S2666-5247(22)00380-9
ECDC and EFSA (European Centre for Disease Prevention and Control and European Food Safety Authority). (2014a). Cluster of monophasic Salmonella
Typhimurium with previously unseen MLVA pattern in the EU/EEA. EFSA supporting publication, EN-657. https://doi.org/10.2903/sp.efsa.2014.
EN-657
ECDC and EFSA (European Centre for Disease Prevention and Control and European Food Safety Authority), 2014b. Multi-country outbreak of Salmonella
Enteritidis infections associated with consumption of eggs from Germany. EFSA supporting publication, EN-646. https://doi.org/10.2903/sp.efsa.
2014.EN-646
ECDC and EFSA (European Centre for Disease Prevention and Control and European Food Safety Authority). (2014c). Multi-country outbreak of
Salmonella Stanley infections – Third update. EFSA supporting publication, EN-592. https://doi.org/10.2903/sp.efsa.2014.EN-592
ECDC and EFSA (European Centre for Disease Prevention and Control and European Food Safety Authority). (2019a). Multi-country outbreak of Listeria
monocytogenes clonal complex 8 infections linked to consumption of cold-smoked fish products– 4 June 2019. EFSA supporting publication, EN-
1665. https://doi.org/10.2903/sp.efsa.2019.EN-1665
ECDC and EFSA (European Centre for Disease Prevention and Control and European Food Safety Authority). (2019b). Multi-country outbreak of Listeria
monocytogenes sequence type 6 infections linked to ready-to-eat meat products – 25 November 2019. EFSA supporting publication, EN-1745.
https://doi.org/10.2903/sp.efsa.2019.EN-1745
Salmonella Poona infections linked to consumption of infant formula. EFSA supporting publication, EN-1594. https://doi.org/10.2903/sp.efsa.2019.
EN-1594
ECDC and EFSA (European Centre for Disease Prevention and Control and European Food Safety Authority). (2020a). Multi-country outbreak of
Salmonella Enteritidis infections linked to eggs, third update – 6 February 2020. EFSA supporting publication, EN-1799. https://doi.org/10.2903/sp.
efsa.2019.EN-1799
ECDC and EFSA (European Centre for Disease Prevention and Control and European Food Safety Authority). (2020b). Multi-country outbreak of
Salmonella Typhimurium and S. Anatum infections linked to Brazil nuts – 21 October 2020. EFSA supporting publication, EN-1944. https://doi.org/
10.2903/sp.efsa.2020.EN-1944
ECDC and EFSA (European Centre for Disease Prevention and Control and European Food Safety Authority). (2021a). Multi-country outbreak of
Salmonella Braenderup ST22, presumed to be linked to imported melons −20 July 2021. EFSA supporting publication, EN- 6807. https://doi.org/
10.2903/sp.efsa.2021.EN- 6807
ECDC and EFSA (European Centre for Disease Prevention and Control and European Food Safety Authority). (2021b). Multi-country outbreak of multiple
Salmonella enterica serotypes linked to imported sesame-based products. EFSA supporting publication, EN-6922. https://doi.org/10.2903/sp.efsa.
2021.EN-6922
Salmonella Enteritidis sequence type (ST)11 infections linked to poultry products in the EU/EEA and the United Kingdom. EFSA supporting publica-
tion, EN-6486. https://doi.org/10.2903/sp.efsa.2021.EN- 6486
ECDC and EFSA (European Centre for Disease Prevention and Control and European Food Safety Authority). (2022a). Multi-country outbreak of mono-
phasic Salmonella Typhimurium sequence type 34 linked to chocolate products – first update. EFSA supporting publication, EN-7318. https://doi.
org/10.2903/sp.efsa.2022.EN-7318
Salmonella Enteritidis sequence type (ST)11 infections linked to eggs and egg products – 8 February 2022. EFSA supporting publication, EN-7180.
Salmonella Mbandaka ST413 possibly linked to consumption of chicken meat in the EU/EEA Israel and the UK – 30 November 2022. EFSA supporting
publication, EN-7707. https://doi.org/10.2903/sp.efsa.2022
ECDC and EFSA (European Centre for Disease Prevention and Control and European Food Safety Authority). (2023a). Multi-country outbreak of of
Salmonella Senftenberg ST14 infections,possibly linked to cherry-like tomatoes. EFSA supporting publication, EN-8211. https://doi.org/10.2903/sp.
efsa.2023
Salmonella Virchow ST16 infections linked to the consumption of meat products containing chicken meat. EFSA supporting publication, EN-7983.
EFSA (European Food Safety Authority). (2018). Urgent scientific and technical assistance to provide recommendations for sampling and testing in the
processing plants of frozen vegetables aiming at detecting Listeria monocytogenes. EFSA supporting publication, EN-1445. https://doi.org/10.2903/
sp.efsa.2018.EN-1445
EFSA (European Food Safety Authority). (2020). Draft framework for protocol development for EFSA's scientific assessments. EFSA supporting publication,
EN-1843. https://doi.org/10.2903/sp.efsa.2020.EN-1843
EFSA (European Food Safety Authority). (2023). Zoonoses, foodborne outbreaks and antimicrobial resistance guidance for reporting 2022 data. EFSA
supporting publication, EN-7827. https://doi.org/10.2903/sp.efsa.2023.EN-7827
EFSA and ECDC (European Food Safety Authority and European Centre for Disease Prevention and Control). (2012). Multi-country outbreak of Salmonella
Stanley infections update. EFSA Journal, 10(9), 2893. https://doi.org/10.2903/j.efsa.2012.2893
EFSA and ECDC (European Food Safety Authority and European Centre for Disease Prevention and Control). (2015). Technical report on a fatal human
case of Bacillus anthracis infection and bovine meat contamination in Bulgaria. First update, 7 August 2015. EFSA supporting publication, EN-863.
EFSA and ECDC (European Food Safety Authority and European Centre for Disease Prevention and Control). (2016a). Multi-country outbreak of
Salmonella Enteritidis phage type 8, MLVA type 2–9–7-3-2 and 2–9–6-3-2 infections. EFSA supporting publication, EN-1110. https://doi.org/10.2903/
sp.efsa.2016.EN-1110
EFSA and ECDC (European Food Safety Authority and European Centre for Disease Prevention and Control). (2016b). Multi-country outbreak of Shiga
toxin-producing Escherichia coli infection associated with haemolytic uraemic syndrome. EFSA supporting publication, EN-1017. https://doi.org/
10.2903/sp.efsa.2016.EN-1017
EFSA and ECDC (European Food Safety Authority and European Centre for Disease Prevention and Control). (2016c). Type E botulism associated with fish
product consumption – Germany and Spain. EFSA supporting publication, EN-1157. https://doi.org/10.2903/sp.efsa.2016.EN-1157
EFSA and ECDC (European Food Safety Authority and European Centre for Disease Prevention and Control). (2017a). Multi-country outbreak of new
Salmonella enterica 11:z41:e,n,z15 infections associated with sesame seeds. EFSA supporting publication, EN-1256. https://doi.org/10.2903/sp.efsa.
2017.EN-1256
EFSA and ECDC (European Food Safety Authority and European Centre for Disease Prevention and Control). (2017b). Multi-country outbreak of
Salmonella Enteritidis infections linked to Polish eggs. EFSA supporting publication, EN-1353. https://doi.org/10.2903/sp.efsa.2017.EN-1353
EFSA and ECDC (European Food Safety Authority and European Centre for Disease Prevention and Control). (2017c). Multi-country outbreak of
Salmonella Enteritidis phage type 8, MLVA profile 2–9–7-3-2 and 2–9–6-3-2 infections. EFSA supporting publication, EN-1188. https://doi.org/
10.2903/sp.efsa.2017.EN-1188
EFSA and ECDC (European Food Safety Authority and European Centre for Disease Prevention and Control). (2018a). Multi-country outbreak of Listeria
monocytogenes sequence type 8 infections linked to consumption of salmon products. EFSA supporting publication, EN-1496. https://doi.org/
10.2903/sp.efsa.2018.EN-1496
EFSA and ECDC (European Food Safety Authority and European Centre for Disease Prevention and Control). (2018b). Multi-country outbreak of Listeria
monocytogenes serogroup IVb, multi-locus sequence type 6, infections linked to frozen corn and possibly to other frozen vegetables – first up-
date. EFSA supporting publication, EN-1448. https://doi.org/10.2903/sp.efsa.2018.EN-1448
EFSA and ECDC (European Food Safety Authority and European Centre for Disease Prevention and Control). (2018c). Multi-country outbreak of
Salmonella Agona infections linked to infant formula. EFSA Supporting Publication, EN-1365. https://doi.org/10.2903/sp.efsa.2018.EN-1365
EFSA and ECDC (European Food Safety Authority and European Centre for Disease Prevention and Control). (2018d). Multi-country outbreak of
Salmonella Agona infections possibly linked to ready-to-eat food. EFSA Supporting publication, EN-1465. https://doi.org/10.2903/sp.efsa.2018.
EN-1465
EFSA and ECDC (European Food Safety Authority and European Centre for Disease Prevention and Control). (2022). The European Union one health 2021
Zoonoses report. EFSA Journal, 20(12), 7666. https://doi.org/10.2903/j.efsa.2022.7666
EFSA BIOHAZ Panel (EFSA Panel on Biological Hazards). (2008). Scientific opinion of the panel on biological hazards on a request from the health and
consumer protection, directorate general, European Commission on microbiological risk assessment in feedingstuffs for food-producing animals.
EFSA Journal, 720. https://doi.org/10.2903/j.efsa.2008.720
EFSA BIOHAZ Panel (EFSA Panel on Biological Hazards). (2012). Scientific opinion on the development of a risk ranking framework on biological hazards.
EFSA Journal, 10(6), 2724. https://doi.org/10.2903/j.efsa.2012.2724
EFSA BIOHAZ Panel (EFSA Panel on Biological Hazards). (2014a). Scientific opinion on the public health risks of table eggs due to deterioration and de-
velopment of pathogens. EFSA Journal, 12(7), 3782. https://doi.org/10.2903/j.efsa.2014.3782
EFSA BIOHAZ Panel (EFSA Panel on Biological Hazards). (2014b). Scientific opinion on the risk posed by pathogens in food of non-animal origin. Part 2
(Salmonella and norovirus in leafy greens eaten raw as salads). EFSA Journal, 12(3), 3600. https://doi.org/10.2903/j.efsa.2014.3600
EFSA BIOHAZ Panel (EFSA Panel on Biological Hazards). (2015). Scientific opinion on the public health risks related to the consumption of raw drinking
milk. EFSA Journal, 13(1), 3940. https://doi.org/10.2903/j.efsa.2015.3940
EFSA BIOHAZ Panel (EFSA Panel on Biological Hazards). (2017). Hazard analysis approaches for certain small retail establishments in view of the applica-
tion of their food safety management systems. EFSA Journal, 15(3), 4697. https://doi.org/10.2903/j.efsa.2017.4697
EFSA BIOHAZ Panel (EFSA Panel on Biological Hazards). (2018). Scientific opinion on the Listeria monocytogenes contamination of ready-to-eat foods and
the risk for human health in the EU. EFSA Journal, 16(1), 5134. https://doi.org/10.2903/j.efsa.2018.5134
EFSA BIOHAZ Panel (EFSA Panel on Biological Hazards). (2019). Scientific opinion on the whole genome sequencing and metagenomics for outbreak
investigation, source attribution and risk assessment of food-borne microorganisms. EFSA Journal, 17(12), 5898. https://doi.org/10.2903/j.efsa.
2019.5898
EFSA BIOHAZ Panel (EFSA Panel on Biological Hazards). (2020). Scientific opinion on the public health risk posed by Listeria monocytogenes in frozen fruit
and vegetables, including herbs, blanched during processing. EFSA Journal, 18(4), 6092. https://doi.org/10.2903/j.efsa.2020.6092
EFSA BIOHAZ Panel (EFSA Panel on Biological Hazards). (2023). Scientific opinion on the microbiological hazards associated with the use of water in the
post-harvest handling and processing operations of fresh and frozen fruits, vegetables and herbs (ffFVHs). Part 1 (outbreak data analysis, literature
review and stakeholder questionnaire). EFSA Journal, 21(11), 8332. https://doi.org/10.2903/j.efsa.2023.8332
EFSA Scientific Committee. (2018a). Guidance on uncertainty analysis in scientific assessments. EFSA Journal, 16(1), 5123. https://doi.org/10.2903/j.efsa.
2018.5123
EFSA Scientific Committee. (2018b). Scientific opinion on the principles and methods behind EFSA's guidance on uncertainty analysis in scientific as-
sessment. EFSA Journal, 16(1), 5122. https://doi.org/10.2903/j.efsa.2018.5122
Eglezos, S., & Dykes, G. A. (2011). Application of heat in Postcook meat chillers reduces Listeria. Journal of Food Protection, 74, 999–1002. https://doi.org/
10.4315/0362-028x.Jfp-10-532
Eglezos, S., & Dykes, G. A. (2018). Reduction of environmental Listeria using gaseous ozone in a cheese processing facility. Journal of Food Protection, 81,
795–798. https://doi.org/10.4315/0362-028x.Jfp-17-361
Elson, R., Awofisayo-Okuyelu, A., Greener, T., Swift, C., Painset, A., Amar, C. F. L., Newton, A., Aird, H., Swindlehurst, M., Elviss, N., Foster, K., Dallman, T. J.,
Ruggles, R., & Grant, K. (2019). Utility of whole genome sequencing to describe the persistence and evolution of Listeria monocytogenes strains
within crabmeat processing environments linked to two outbreaks of Listeriosis. Journal of Food Protection, 82, 30–38. https://doi.org/10.4315/
0362-028x.Jfp-18-206
Eriksson, J., Löfström, C., Aspán, A., Gunnarsson, A., Karlsson, I., Borch, E., de Jong, B., & Rådström, P. (2005). Comparison of genotyping methods by
application to Salmonella Livingstone strains associated with an outbreak of human salmonellosis. International Journal of Food Microbiology, 104,
93–103. https://doi.org/10.1016/j.ijfoodmicro.2005.01.011
Eshwar, A. K., Guldimann, C., Oevermann, A., & Tasara, T. (2017). Cold-shock domain family proteins (Csps) are involved in regulation of virulence, cellular
aggregation, and flagella-based motility in Listeria monocytogenes. Frontiers in Cellular and Infection Microbiology, 7, 453. https://doi.org/10.3389/
fcimb.2017.00453
Estrada, E. M., Hamilton, A. M., Sullivan, G. B., Wiedmann, M., Critzer, F. J., & Strawn, L. K. (2020). Prevalence, persistence, and diversity of Listeria mono-
cytogenes and Listeria species in produce packinghouses in three US states. Journal of Food Protection, 83, 277–286. https://doi.org/10.4315/0362-
028x.Jfp-19-411
EURL-L. monocytogenes and ANSES (European Union Reference Laboratory for Listeria monocytogenes and French Agency for Food, Environmental
and Occupational Health Safety ). (2023). Guidelines on sampling the food processing area and equipment for the detection of Listeria monocyto-
genes. Version 4–3 October 2023. 13 pp. https://zenodo.org/records/8406616
European Commission (Directorate-General for Health and Food Safety). (2019). RASFF – The Rapid Alert System for Food and Feed – Annual Report
2018. https://doi.org/10.2875/914558
2019. https://doi.org/10.2875/993888
2020. https://food.ec.europa.eu/system/f iles/2021- 08/rasff_pub_annual-report_ 2020.pdf
Fagerlund, A., Langsrud, S., & Moretro, T. (2020). In-depth longitudinal study of Listeria monocytogenes ST9 isolates from the meat processing industry:
Resolving diversity and transmission patterns using whole-genome sequencing. Applied and Environmental Microbiology, 86, 18. https://doi.org/
10.1128/aem.00579-20
Fagerlund, A., Langsrud, S., & Moretro, T. (2021). Microbial diversity and ecology of biofilms in food industry environments associated with Listeria
monocytogenes persistence. Current Opinion in Food Science, 37, 171–178. https://doi.org/10.1016/j.cofs.2020.10.015
Fagerlund, A., Langsrud, S., Schirmer, B. C. T., Moretro, T., & Heir, E. (2016). Genome analysis of Listeria monocytogenes sequence type 8 strains persist-
ing in Salmon and Poultry processing environments and comparison with related strains. PLoS One, 11, 22. https://doi.org/10.1371/journal.pone.
0151117
Fagerlund, A., Wagner, E., Moretro, T., Heir, E., Moen, B., Rychli, K., & Langsrud, S. (2022). Pervasive Listeria monocytogenes is common in the Norwegian
food system and is associated with increased prevalence of stress survival and resistance determinants. Applied and Environmental Microbiology,
88, 21. https://doi.org/10.1128/aem.00861-22
Fakruddin, M., Rahaman, M., Ahmed, M. M., & Hoque, M. M. (2014). Stress tolerant virulent strains of Cronobacter sakazakii from food. Biological Research,
47, 63. https://doi.org/10.1186/0717-6287-47-63
FAO (Food and Agriculture Organization). (2007). Guidelines on the Application of General Principles of Food Hygiene to the Control of Listeria
Monocytogenes in Foods, 28 pp. https://w ww.fda.gov/regulatory-information/search-fda-guidance-documents/draft-guidance-industry-contr
ol-listeria-monocy togenes-ready-eat-foods
FAO and WHO (Food and Agriculture Organization of the United Nations and World Health Organization). (2004). Enterobacter sakazakii and other mi-
croorganisms in powdered infant formula -Meeting report. Microbiological Risk Assessment Series (MRA) 6. Rome, Italy. 80 pp. https://w ww.fao.
org/documents/card/en/c/3ffdb8bc-8c2c-5d73-87a4- 6e9698b164ae
FAO and WHO (Food and Agriculture Organization of the United Nations and World Health Organization). (2008). Enterobacter sakazakii (Cronobacter
spp.) in powdered follow-up formulae -Meeting report. Microbiological Risk Assessment Series (MRA) 15. Rome, Italy. 105 pp. https://w ww.fao.
org/f ileadmin/templates/agns/pdf/jemra/mra15_sakazaki.pdf
FAO and WHO (Food and Agriculture Organization of the United Nations and World Health Organization). (2021a). Microbiological Risk Assessment -
Guidance for Food. Microbiological Risk Assessment Series No. 36. Rome. https://w ww.fao.org/documents/card/en/c/cb5006en
FAO and WHO (Food and Agriculture Organization of the United Nations and World Health Organization). (2021b). Safety and quality of water used
with fresh fruits and vegetables. In Microbiological risk assessment series (MRA) 37. Rome, Italy. 153 pp. https://w ww.fao.org/documents/card/en/c/
cb7678en
FAO and WHO (Food and Agriculture Organization of the United Nations and World Health Organization). (2022). Ranking of low-moisture foods in
support of microbiological risk management. Meeting report and systematic review. Microbiological Risk Assessment Series No. 26. Rome. https://
doi.org/10.4060/cc0763en
FAO and WHO (Food and Agriculture Organization of the United Nations and World Health Organization). (2023a). Safety and quality of water use and
reuse in the production and processing of dairy products: meeting report. Microbiological Risk Assessment Series No. 40. Rome. 204 pp. https://
FAO and WHO (Food and Agriculture Organization of the United Nations and World Health Organization). (2023b). Safety and quality of water used in
the production and processing of fish and fishery products: meeting report. Microbiological Risk Assessment Series No. 41. Rome. 140 pp. https://
Fatica, M. K., & Schneider, K. R. (2011). Salmonella and produce survival in the plant environment and implications in food safety. Virulence, 2, 573–579.
https://doi.org/10.4161/viru.2.6.17880
FDA (U.S. Food and Drug Administration). (2015). Testing Methodology for Listeria species or L. monocytogenes in Environmental Samples. 11 pp. https://
www.fda.gov/media/94358/download
FDA (U.S. Food and Drug Administration). (2022). Investigation Report: Factors Potentially Contributing to the Contamination of Packaged Leafy Greens
Implicated in the Outbreak of Salmonella Typhimurium During the Summer of 2021. 8 pp. https://w ww.fda.gov/food/outbreaks-foodborne-illne
ss/factors-potentially-contributing-contamination-packaged-leafy-greens-implicated-outbreak-salmonella
Fei, P., Man, C. X., Lou, B. B., Forsythe, S. J., Chai, Y. L., Li, R., Niu, J. T., & Jiang, Y. J. (2015). Genotyping and source tracking of Cronobacter sakazakii and C.
Malonaticus isolates from powdered infant formula and an infant formula production factory in China. Applied and Environmental Microbiology, 81,
5430–5439. https://doi.org/10.1128/aem.01390-15
Fenner, K., Elsner, M., Lueders, T., McLachlan, M. S., Wackett, L. P., Zimmermann, M., & Drewes, J. E. (2021). Methodological advances to study contami-
nant biotransformation: New prospects for understanding and reducing environmental persistence? ACS ES&T Water, 1, 1541–1554. https://doi.org/
10.1021/acsest water.1c00025
Ferreira, V., Wiedmann, M., Teixeira, P., & Stasiewicz, M. J. (2014). Listeria monocytogenes persistence in food-associated environments: Epidemiology,
Strain characteristics, and implications for public health. Journal of Food Protection, 77, 150–170. https://doi.org/10.4315/0362-028x.Jfp-13-150
Figueras, M. J., Levican, A., Pujol, I., Ballester, F., Rabada Quilez, M. J., & Gomez-Bertomeu, F. (2014). A severe case of persistent diarrhoea associated
with Arcobacter cryaerophilus but attributed to campylobacter sp. and a review of the clinical incidence of Arcobacter spp. New Microbes New
Infections, 2, 31–37. https://doi.org/10.1002/2052-2975.35
Filipello, V., Bonometti, E., Campagnani, M., Bertoletti, I., Romano, A., Zuccon, F., Campanella, C., Losio, M. N., & Finazzi, G. (2020). Investigation and fol-
low-up of a staphylococcal food poisoning outbreak linked to the consumption of traditional hand-crafted Alm cheese. Pathogens, 9. https://doi.
org/10.3390/pathogens9121064
Filipello, V., Gallina, S., Amato, E., Losio, M. N., Pontello, M., Decastelli, L., & Lomonaco, S. (2017). Diversity and persistence of Listeria monocytogenes
within the Gorgonzola PDO production chain and comparison with clinical isolates from the same area. International Journal of Food Microbiology,
245, 73–78. https://doi.org/10.1016/j.ijfoodmicro.2017.01.012
Finn, L., Onyeaka, H., & Oneill, S. (2023). Listeria monocytogenes biofilms in food-associated environments: A persistent enigma. Food, 12, 22. https://doi.
org/10.3390/foods12183339
Finn, S., Condell, O., McClure, P., Amézquita, A., & Fanning, S. (2013). Mechanisms of survival, responses and sources of Salmonella in low-moisture envi-
ronments. Frontiers in Microbiology, 4, 331. https://doi.org/10.3389/fmicb.2013.00331
Forsythe, S. J. (2013). Microbiologia da Segurança dos Alimentos United Kingdon. Wiley-Blackwell, 2, 496.
Fox, E., Hunt, K., O'Brien, M., & Jordan, K. (2011). Listeria monocytogenes in Irish farmhouse cheese processing environments. International Journal of
Food Microbiology, 145, S39–S45. https://doi.org/10.1016/j.ijfoodmicro.2010.10.012
Fox, E. M., Solomon, K., Moore, J. E., Wall, P. G., & Fanning, S. (2014). Phylogenetic profiles of in-house microflora in drains at a food production facil-
ity: Comparison and biocontrol implications of Listeria-positive and -negative bacterial populations. Applied and Environmental Microbiology, 80,
3369–3374. https://doi.org/10.1128/aem.00468-14
Franciosa, G., Maugliani, A., Scalfaro, C., Floridi, F., & Aureli, P. (2009). Expression of Internalin a and biofilm formation among Listeria Monocytogenes
clinical isolates. International Journal of Immunopathology and Pharmacology, 22, 183–193. https://doi.org/10.1177/039463200902200121
Fretz, R., Sagel, U., Ruppitsch, W., Pietzka, A. T., Stöger, A., Huhulescu, S., Heuberger, S., Pichler, J., Much, P., Pfaff, G., Stark, K., Prager, R., Flieger, A., &
Feenstra, O. (2010). Listeriosis outbreak caused by acid curd cheese ‘Quargel’, Austria and Germany 2009. Euro Surveillance, 15(5), pii=19477.
Friesema, I., de Jong, A., Hofhuis, A., Heck, M., van den Kerkhof, H., de Jonge, R., Hameryck, D., Nagel, K., van Vilsteren, G., van Beek, P., Notermans,
D., & van Pelt, W. (2014). Large outbreak of Salmonella Thompson related to smoked salmon in The Netherlands, august to December 2012.
Eurosurveillance, 19, 13–20. https://doi.org/10.2807/1560-7917.Es2014.19.39.20918
Gaglio, R., Cruciata, M., Di Gerlando, R., Scatassa, M. L., Cardamone, C., Mancuso, I., Sardina, M. T., Moschetti, G., Portolano, B., & Settanni, L. (2016).
Microbial activation of wooden vats used for traditional cheese production and evolution of neoformed biofilms. Applied and Environmental
Microbiology, 82, 585–595. https://doi.org/10.1128/AEM.02868-15
Gale, P., Kelly, L., Mearns, R., Duggan, J., & Snary, E. L. (2015). Q fever through consumption of unpasteurised milk and milk products -a risk profile and
exposure assessment. Journal of Applied Microbiology, 118, 1083–1095. https://doi.org/10.1111/jam.12778
Gan, X., Li, M. H., Yan, S. F., Wang, X. F., Wang, W., & Li, F. Q. (2021). Genomic landscape and phenotypic assessment of Cronobacter sakazakii isolated from
raw material, environment, and production facilities in powdered infant formula factories in China. Frontiers in Microbiology, 12, 10. https://doi.org/
10.3389/fmicb.2021.686189
Gantzhom, M. R., Pedersen, K., Olsen, J. E., & Thomsen, L. E. (2014). Biocide and antibiotic susceptibility of Salmonella isolates obtained before and after
cleaning at six Danish pig slaughterhouses. International Journal of Food Microbiology, 181, 53–59. https://doi.org/10.1016/j.ijfoodmicro.2014.04.021
Garcia-Sanchez, L., Melero, B., Jaime, I., Hanninen, M. L., Rossi, M., & Rovira, J. (2017). Campylobacter jejuni survival in a poultry processing plant environ-
ment. Food Microbiology, 65, 185–192. https://doi.org/10.1016/j.fm.2017.02.009
Gardan, R., Duché, O., Leroy-Sétrin, S., & Labadie, J. (2003). Role of ctc from Listeria monocytogenes in osmotolerance. Applied and Environmental
Microbiology, 69, 154–161. https://doi.org/10.1128/aem.69.1.154-161.2003
Garner, D., & Kathariou, S. (2016). Fresh produce-associated Listeriosis outbreaks, sources of concern, teachable moments, and insights. Journal of Food
Protection, 79, 337–344. https://doi.org/10.4315/0362-028x.Jfp-15-387
Gelbícová, T., Florianová, M., Tomástíková, Z., Pospísilová, L., Kolácková, I., & Karpísková, R. (2019). Prediction of persistence of Listeria monocytogenes
ST451 in a rabbit meat processing Plant in The Czech Republic. Journal of Food Protection, 82, 1350–1356. https://doi.org/10.4315/0362-028x.
Jfp-19-030
Gelbicova, T., Zobanikova, M., Tomastikova, Z., Van Walle, I., Ruppitsch, W., & Karpiskova, R. (2018). An outbreak of listeriosis linked to Turkey meat prod-
ucts in The Czech Republic, 2012-2016. Epidemiology and Infection, 146, 1407–1412. https://doi.org/10.1017/s0950268818001565
Germinario, C., Germinario, C., Caprioli, A., Giordano, M., Chironna, M., Gallone, M., Tafuri, S., Minelli, F., Maugliani, A., Michelacci, V., Santangelo, L.,
Mongelli, O., Montagna, C., Scavia, G., & on behalf of all participants of the outbreak investigation t. (2016). Community-wide outbreak of haemo-
lytic uraemic syndrome associated with Shiga toxin 2-producing E. Coli O26-H11. Euro Surveillance. https://doi.org/10.2807/1560-7917.ES.2016.21.
38.303432
Gieraltowski, L., Higa, J., Peralta, V., Green, A., Schwensohn, C., Rosen, H., Libby, T., Kissler, B., Marsden-Haug, N., Booth, H., Kimura, A., Grass, J., Bicknese,
A., Tolar, B., Defibaugh-Chavez, S., Williams, I., Wise, M., & Salmonella Heidelberg Investigation, T. (2016). National Outbreak of multidrug resistant
Salmonella Heidelberg infections linked to a single poultry company. PLoS One, 11, e0162369. https://doi.org/10.1371/journal.pone.0162369
Gil, M. I., Selma, M. V., Suslow, T., Jacxsens, L., Uyttendaele, M., & Allende, A. (2015). Pre-and postharvest preventive measures and intervention strate-
gies to control Microbial food safety hazards of fresh leafy vegetables. Critical Reviews in Food Science and Nutrition, 55, 453–468. https://doi.org/
10.1080/10408398.2012.657808
Gorski, L., Cooley, M. B., Oryang, D., Carychao, D., Nguyen, K., Luo, Y., Weinstein, L., Brown, E., Allard, M., Mandrell, R. E., & Chen, Y. (2022). Prevalence and
clonal diversity of over 1,200 Listeria monocytogenes isolates collected from public access waters near produce production areas on the Central
California coast during 2011 to 2016. Applied and Environmental Microbiology, 88, 24. https://doi.org/10.1128/aem.00357-22
Gosling, R., Oastler, C., Nichols, C., Jackson, G., Wales, A. D., & Davies, R. H. (2022). Investigations into Salmonella contamination in feed Mills producing
rations for the broiler industry in Great Britain. Veterinary Sciences, 9, 12. https://doi.org/10.3390/vetsci9070307
Gould, L. H., Mungai, E., & Behravesh, C. B. (2014). Outbreaks attributed to cheese: Differences between outbreaks caused by unpasteurized and pasteur-
ized dairy products, United States, 1998–2011. Foodborne Pathogens and Disease, 11, 545–551. https://doi.org/10.1089/fpd.2013.1650
Grasso, E. M., Grove, S. F., Halik, L. A., Arritt, F., & Keller, S. E. (2015). Cleaning and sanitation of Salmonella-contaminated peanut butter processing equip-
ment. Food Microbiology, 46, 100–106. https://doi.org/10.1016/j.fm.2014.03.003
Grimont, P. A. D., & Weill, F. X. (2007). Antigenic formulae of the Salmonella serovars. WHO Collaborating Centre for Reference and Research on Salmonella.
https://w ww.pasteur.fr/sites/default/f iles/veng_0.pdf
Gu, D., Wang, Z. Y., Tian, Y. Q., Kang, X. L., Meng, C., Chen, X., Pan, Z. M., & Jiao, X. N. (2020). Prevalence of Salmonella isolates and their distribution based
on whole-genome sequence in a chicken slaughterhouse in Jiangsu. China. Frontiers in Veterinary Science, 7. https://doi.org/10.3389/f vets.2020.
00029
Guidi, F., Lorenzetti, C., Centorotola, G., Torresi, M., Camma, C., Chiaverini, A., Pomilio, F., & Blasi, G. (2022). Atypical Serogroup IVb-v1 of Listeria mono-
cytogenes assigned to new ST2801, widely spread and persistent in the environment of a pork-meat producing Plant of Central Italy. Frontiers in
Guidi, F., Orsini, M., Chiaverini, A., Torresi, M., Centorame, P., Acciari, V. A., Salini, R., Palombo, B., Brandi, G., Amagliani, G., Schiavano, G. F., Massacci, F. R.,
Fisichella, S., Di Domenico, M., Ancora, M., Di Pasquale, A., Duranti, A., Cammà, C., Pomilio, F., & Blasi, G. (2021). Hypo-and hyper-virulent Listeria
monocytogenes clones persisting in two different food processing plants of Central Italy. Microorganisms, 9, 21. https://doi.org/10.3390/micro
organisms9020376
Hachler, H., Marti, G., Giannini, P., Lehner, A., Jost, M., Beck, J., Weiss, F., Bally, B., Jermini, M., Stephan, R., & Baumgartner, A. (2013). Outbreak of listerosis
due to imported cooked ham, Switzerland 2011. Eurosurveillance, 18, 7–13.
Häggblom, P. (2012). How does Sweden control Salmonella before it enters the food chain? In J. Hoorfar (Ed.), Case studies in food safety and authenticity:
Lessons from real-life situations (pp. 198–205). Woodhead Publ Ltd.
Hald, T., Wingstrand, A., Swanenburg, M., von Altrock, A., & Thorberg, B. M. (2003). The occurrence and epidemiology of Salmonella in European pig
slaughterhouses. Epidemiology and Infection, 131, 1187–1203. https://doi.org/10.1017/s0950268803001171
Haque, E., Hussain, J., Shankar, S., Haldar, S., & Chatterjee, S. (2023). Chapter 15 -implication of vibrio biofilms in human and seafood sector. In S. Das &
N. A. Kungwani (Eds.), Understanding Microbial Biofilms (pp. 247–260). Academic Press.
Harrand, A. S., Jagadeesan, B., Baert, L., Wiedmann, M., & Orsi, R. H. (2020). Evolution of Listeria monocytogenes in a food processing plant involves limited
single-nucleotide substitutions but considerable diversification by gain and loss of prophages. Applied and Environmental Microbiology, 86, 22.
https://doi.org/10.1128/aem.02493-19
Harter, E., Wagner, E. M., Zaiser, A., Halecker, S., Wagner, M., & Rychli, K. (2017). Stress survival islet 2, predominantly present in Listeria monocytogenes
strains of sequence type 121, is involved in the alkaline and oxidative stress responses. Applied and Environmental Microbiology, 83, e00827–e00817.
https://doi.org/10.1128/AEM.00827-17
Heiman, K. E., Garalde, V. B., Gronostaj, M., Jackson, K. A., Beam, S., Joseph, L., Saupe, A., Ricotta, E., Waechter, H., Wellman, A., Adams-Cameron, M., Ray,
G., Fields, A., Chen, Y., Datta, A., Burall, L., Sabol, A., Kucerova, Z., Trees, E., … Silk, B. J. (2016). Multistate outbreak of listeriosis caused by imported
cheese and evidence of cross-contamination of other cheeses, USA, 2012. Epidemiology and Infection, 144, 2698–2708. https://doi.org/10.1017/
s095026881500117x
Henry, M., & Fouladkhah, A. (2019). Outbreak history, biofilm formation, and preventive measures for control of Cronobacter sakazakii in infant formula
and infant care settings. Microorganisms, 7, 22. https://doi.org/10.3390/microorganisms7030077
Hiett, K. L. (2010). Tracing pathogens in poultry and egg production and at the abattoir. In S. Brul, P. M. Fratamico, & T. A. McMeekin (Eds.), Tracing patho-
gens in the food chain (pp. 465–502). Woodhead Publ Ltd.
Hiko, A., Irsigler, H., Ameni, G., Zessin, K. H., & Fries, R. (2016). Salmonella serovars along two beef chains in Ethiopia. Journal of Infection in Developing
Countries, 10, 1168–1176. https://doi.org/10.3855/jidc.6354
Hingston, P., Brenner, T., Truelstrup Hansen, L., & Wang, S. (2019). Comparative analysis of Listeria monocytogenes plasmids and expression levels of
plasmid-encoded genes during growth under salt and acid stress conditions. Toxins (Basel), 11. https://doi.org/10.3390/toxins11070426
Hobbs, J. L., Warshawsky, B., Maki, A., Zittermann, S., Murphy, A., Majury, A., & Middleton, D. (2017). Nuggets of wisdom: Salmonella Enteritidis outbreaks
and the case for new rules on uncooked frozen processed chicken. Journal of Food Protection, 80, 703–709. https://doi.org/10.4315/0362-028X.
JFP-16-431
Hoffmann, M., Miller, J., Melka, D., Allard, M. W., Brown, E. W., & Pettengill, J. B. (2020). Temporal dynamics of Salmonella enterica subsp. enterica Serovar
Agona isolates from a recurrent multistate outbreak. Frontiers in Microbiology, 11, 10. https://doi.org/10.3389/fmicb.2020.00478
Holch, A., Webb, K., Lukjancenko, O., Ussery, D., Rosenthal, B. M., & Gram, L. (2013). Genome sequencing identifies two nearly unchanged strains of per-
sistent Listeria monocytogenes isolated at two different fish processing plants sampled 6 years apart. Applied and Environmental Microbiology, 79,
2944–2951. https://doi.org/10.1128/aem.03715-12
Holschbach, C. L., & Peek, S. F. (2018). Salmonella in dairy cattle. The Veterinary Clinics of North America. Food Animal Practice, 34, 133–154. https://doi.org/
10.1016/j.c vfa.2017.10.005
Horn, N., & Bhunia, A. K. (2018). Food-associated stress primes foodborne pathogens for the gastrointestinal phase of infection. Frontiers in Microbiology,
9, 16. https://doi.org/10.3389/fmicb.2018.01962
Hu, J., Shuai, W., Sumner, J. T., Moghadam, A. A., & Hartmann, E. M. (2022). Clinically relevant pathogens on surfaces display differences in survival and
transcriptomic response in relation to probiotic and traditional cleaning strategies. Npj Biofilms and Microbiomes, 8(1), 72.
Hurley, D., Luque-Sastre, L., Parker, C. T., Huynh, S., Eshwar, A. K., Nguyen, S. V., Andrews, N., Moura, A., Fox, E. M., Jordan, K., Lehner, A., Stephan, R., &
Fanning, S. (2019). Whole-genome sequencing-based characterization of 100 Listeria monocytogenes isolates collected from food processing en-
vironments over a four-year period. Msphere, 4. https://doi.org/10.1128/mSphere.00252-19
Hussein, H. S., & Sakuma, T. (2005). Invited review: Prevalence of Shiga toxin-producing Escherichia coli in dairy cattle and their products. Journal of Dairy
Science, 88, 450–465. https://doi.org/10.3168/jds.S0022- 0302(05)72706-5
ILSI (international Life Sciences Institute). (2011). Persistence and survival of pathogens in dry foods and dry food processing environments. ILSI Europe
Report Series, 1–48. https://ilsi.eu/publication/persistence-and-survival-of-pathogens-in-dry-foods-and-dry-food-processing-environments/
Iversen, C., Lane, M., & Forsythe, S. J. (2004). The growth profile, thermotolerance and biofilm formation of Enterobacter sakazakii grown in infant for-
mula milk. Letters in Applied Microbiology, 38, 378–382. https://doi.org/10.1111/j.1472-765X.2004.01507.x
Jaakkonen, A., Kivistö, R., Aarnio, M., Kalekivi, J., & Hakkinen, M. (2020). Persistent contamination of raw milk by campylobacter jejuni ST-883. PLoS One,
15, e0231810. https://doi.org/10.1371/journal.pone.0231810
Jackson, K. A., Biggerstaff, M., Tobin-D'Angelo, M., Sweat, D., Klos, R., Nosari, J., Garrison, O., Boothe, E., Saathoff-Huber, L., Hainstock, L., & Fagan, R.
P. (2011). Multistate outbreak of Listeria monocytogenes associated with Mexican-style cheese made from pasteurized milk among pregnant,
Hispanic women. Journal of Food Protection, 74, 949–953. https://doi.org/10.4315/0362-028x.Jfp-10-536
Jacobs, C., Braun, P., & Hammer, P. (2011). Reservoir and routes of transmission of Enterobacter sakazakii (Cronobacter spp.) in a milk powder-producing
plant. Journal of Dairy Science, 94, 3801–3810. https://doi.org/10.3168/jds.2011- 4318
Jakociune, D., Bisgaard, M., Pedersen, K., & Olsen, J. E. (2014). Demonstration of persistent contamination of a cooked egg product production facility
with Salmonella enterica serovar Tennessee and characterization of the persistent strain. Journal of Applied Microbiology, 117, 547–553. https://doi.
org/10.1111/jam.12536
Jensen, A. K., Nielsen, E. M., Björkman, J. T., Jensen, T., Müller, L., Persson, S., Bjerager, G., Perge, A., Krause, T. G., Kiil, K., Sorensen, G., Andersen, J. K.,
Molbak, K., & Ethelberg, S. (2016). Whole-genome sequencing used to investigate a Nationwide outbreak of Listeriosis caused by ready-to-eat
delicatessen meat, Denmark, 2014. Clinical Infectious Diseases, 63, 64–78. https://doi.org/10.1093/cid/ciw192
Jiang, Y., Ma, Y., Liu, Q., Li, T., Li, Y., Guo, K., & Zhang, Y. (2022). Tracing Clostridium perfringens strains from beef processing of slaughter house by pulsed-
field gel electrophoresis, and the distribution and toxinotype of isolates in Shaanxi province. China. Food Microbiology, 101, 103887. https://doi.
org/10.1016/j.fm.2021.103887
Joensen, K. G., Schjorring, S., Gantzhorn, M. R., Vester, C. T., Nielsen, H. L., Engberg, J. H., Holt, H. M., Ethelberg, S., Muller, L., Sando, G., & Nielsen, E. M.
(2021). Whole genome sequencing data used for surveillance of campylobacter infections: Detection of a large continuous outbreak, Denmark,
2019. Eurosurveillance, 26, 8. https://doi.org/10.2807/1560-7917.Es.2021.26.22.2001396
Johler, S., Stephan, R., Hartmann, I., Kuehner, K. A., & Lehner, A. (2010). Genes involved in yellow pigmentation of Cronobacter sakazakii ES5 and influ-
ence of pigmentation on persistence and growth under environmental stress. Applied and Environmental Microbiology, 76, 1053–1061. https://doi.
org/10.1128/aem.01420- 09
Johler, S., Weder, D., Bridy, C., Huguenin, M. C., Robert, L., Hummerjohann, J., & Stephan, R. (2015). Outbreak of staphylococcal food poisoning among
children and staff at a Swiss boarding school due to soft cheese made from raw milk. Journal of Dairy Science, 98, 2944–2948. https://doi.org/10.
3168/jds.2014-9123
Jones, F. T. (2011). A review of practical Salmonella control measures in animal feed. Journal of Applied Poultry Research, 20, 102–113. https://doi.org/10.
3382/japr.2010-00281
Jones, G., Dela Gandara, M. P., Herrera-Leon, L., Herrera-Leon, S., Martinez, C. V., Hureaux-Roy, R., Abdallah, Y., Nisavanh, A., Fabre, L., Renaudat, C.,
Mossong, J., Mattheus, W., Huard, C., Le Borgne, C., De Valk, H., Weill, F. X., & Jourdan-Da Silva, N. (2019). Outbreak of Salmonella enterica serotype
Poona in infants linked to persistent Salmonella contamination in an infant formula manufacturing facility, France, august 2018 to February 2019.
Eurosurveillance, 24, 2–8. https://doi.org/10.2807/1560-7917.Es.2019.24.13.1900161
Jordan, K. N., Hunt, K., & Fox, E. M. (2012). Troubleshooting the environmental source of contamination with Listeria monocytogenes in a typical small
food manufacturing plant in Ireland. Case Studies in Food Safety and Authenticity., 95–101.
Jung, Y. J., Jang, H., & Matthews, K. R. (2014). Effect of the food production chain from farm practices to vegetable processing on outbreak incidence.
Microbial Biotechnology, 7, 517–527. https://doi.org/10.1111/1751-7915.12178
Kaszoni-Rückerl, I., Mustedanagic, A., Muri-Klinger, S., Brugger, K., Wagner, K. H., Wagner, M., & Stessl, B. (2020). Predominance of distinct Listeria Innocua
and Listeria Monocytogenes in recurrent contamination events at dairy processing facilities. Microorganisms, 8. https://doi.org/10.3390/micro
organisms8020234
Kawakami, V., Bottichio, L., Lloyd, J., Carleton, H., Leeper, M., Olson, G., Li, Z., Kissler, B., Angelo, K. M., Whitlock, L., Sinatra, J., Defibaugh-Chavez, S., Bicknese,
A., Kay, M., Wise, M. E., Basler, C., & Duchin, J. (2019). Multidrug-resistant Salmonella I 4, 5, 12:i:-and Salmonella Infantis infections linked to whole roasted
pigs from a single slaughter and processing facility. Journal of Food Protection, 82, 1615–1624. https://doi.org/10.4315/0362-028x.Jfp-19-048
Kayode, A. J., & Okoh, A. I. (2022). Assessment of the molecular epidemiology and genetic multiplicity of Listeria monocytogenes recovered from ready-
to-eat foods following the south African listeriosis outbreak. Scientific Reports, 12, 11. https://doi.org/10.1038/s41598- 022-20175-x
Keaton, A. A., Schwensohn, C. A., Brandenburg, J. M., Pereira, E., Adcock, B., Tecle, S., Hinnenkamp, R., Havens, J., Bailey, K., Applegate, B., Whitney, P.,
Gibson, D., Manion, K., Griffin, M., Ritter, J., Biskupiak, C., Ajileye, K., Golwalkar, M., Gosciminski, M., … Gieraltowski, L. B. (2022). Multistate outbreak
of Salmonella Mbandaka infections linked to sweetened puffed wheat cereal -United States, 2018. Epidemiology and Infection, 150, 6. https://doi.
org/10.1017/s095026882200108x
Kim, Y. J., Chon, J. W., Lim, J. S., Song, B. R., Seo, K. H., Heo, E. J., Park, H. J., Wee, S. H., Oh, D. H., & Moon, J. S. (2015). Traceback Investigation for Salmonella
contamination at egg processing plants in South Korea: Prevalence, antibiotic resistance, and epidemiological tracing by rep-PCR fingerprinting.
Journal of Food Science, 80, M759–M764. https://doi.org/10.1111/1750-3841.12731
King, L. A., Loukiadis, E., Mariani-Kurkdjian, P., Haeghebaert, S., Weill, F. X., Baliere, C., Ganet, S., Gouali, M., Vaillant, V., Pihier, N., Callon, H., Novo, R., Gaillot,
O., Thevenot-Sergentet, D., Bingen, E., Chaud, P., & de Valk, H. (2014). Foodborne transmission of sorbitol-fermenting Escherichia coli O157:[H7] via
ground beef: An outbreak in northern France, 2011. Clinical Microbiology and Infection, 20, O1136–O1144. https://doi.org/10.1111/1469-0691.12736
Kingsbury, J. M., Thom, K., Erskine, H., Olsen, L., & Soboleva, T. (2019). Prevalence and genetic analysis of Salmonella enterica from a cross-sectional
survey of the New Zealand egg production environment. Journal of Food Protection, 82, 2201–2214. https://doi.org/10.4315/0362-028x.Jfp-19-159
Kinross, P., van Alphen, L., Urtaza, J. M., Struelens, M., Takkinen, J., Coulombier, D., Mäkelä, P., Bertrand, S., Mattheus, W., Schmid, D., Kanitz, E., Rücker,
V., Krisztalovics, K., Pászti, J., Szögyényi, Z., Lancz, Z., Rabsch, W., Pfefferkorn, B., Hiller, P., … Gossner, C. (2014). Multidisciplinary investigation
of a multicountry outbreak of Salmonella Stanley infections associated with Turkey meat in the European Union, august 2011 to January 2013.
Eurosurveillance, 19, 6–15.
Kipper, D., Mascitti, A. K., De Carli, S., Carneiro, A. M., Streck, A. F., Fonseca, A. S. K., Ikuta, N., & Lunge, V. R. (2022). Emergence, dissemination and anti-
microbial resistance of the Main poultry-associated Salmonella Serovars in Brazil. Veterinary Sciences, 9, 24. https://doi.org/10.3390/vetsci9080405
Knudsen, G. M., Nielsen, J. B., Marvig, R. L., Ng, Y., Worning, P., Westh, H., & Gram, L. (2017). Genome-wide-analyses of Listeria monocytogenes from food-
processing plants reveal clonal diversity and date the emergence of persisting sequence types. Environmental Microbiology Reports, 9, 428–440.
https://doi.org/10.1111/1758-2229.12552
Kotwal, G., & Cannon, J. L. (2014). Environmental persistence and transfer of enteric viruses. Current Opinion in Virology, 4, 37–43. https://doi.org/10.1016/j.
coviro.2013.12.003
Kovacevic, J., Ziegler, J., Walecka-Zacharska, E., Reimer, A., Kitts, D. D., & Gilmour, M. W. (2016). Tolerance of Listeria monocytogenes to quaternary am-
monium sanitizers is mediated by a novel efflux pump encoded by emrE. Applied and Environmental Microbiology, 82, 939–953. https://doi.org/10.
1128/aem.03741-15
Kuenne, C., Billion, A., Mraheil, M. A., Strittmatter, A., Daniel, R., Goesmann, A., Barbuddhe, S., Hain, T., & Chakraborty, T. (2013). Reassessment of the
Listeria monocytogenespan-genome reveals dynamic integration hotspots and mobile genetic elements as major components of the accessory
genome. BMC Genomics, 14, 47. https://doi.org/10.1186/1471-2164-14- 47
Kuenne, C., Voget, S., Pischimarov, J., Oehm, S., Goesmann, A., Daniel, R., Hain, T., & Chakraborty, T. (2010). Comparative analysis of plasmids in the genus
Listeria. PLoS One, 5, 7. https://doi.org/10.1371/journal.pone.0012511
Kuhn, K. G., Sorensen, G., Torpdahl, M., Kjeldsen, M. K., Jensen, T., Gubbels, S., Bjerager, G. O., Wingstrand, A., Porsbo, L. J., & Ethelberg, S. (2013). A
long-lasting outbreak of Salmonella Typhimurium U323 associated with several pork products, Denmark, 2010. Epidemiology and Infection, 141,
260–268. https://doi.org/10.1017/s0950268812000702
Labska, K., Spackova, M., Daniel, O., Vcelak, J., Vlasakova, V., Cerny, T., Gelbicova, T., Florianova, M., Karpiskova, R., Gavacova, D., Stefkovicova, M., &
Orlikova, H. (2021). A cross-border outbreak of Salmonella Bareilly cases confirmed by whole genome sequencing, Czech Republic and Slovakia,
2017 to 2018. Eurosurveillance, 26, 10–20. https://doi.org/10.2807/1560-7917.Es.2021.26.14.2000131
Lachmann, R., Halbedel, S., Adler, M., Becker, N., Allerberger, F., Holzer, A., Boone, I., Falkenhorst, G., Kleta, S., Al Dahouk, S., Stark, K., Luber, P., Flieger, A.,
& Wilking, H. (2021). Nationwide outbreak of invasive listeriosis associated with consumption of meat products in health care facilities, Germany,
2014–2019. Clinical Microbiology and Infection, 27, 1035.e5. https://doi.org/10.1016/j.cmi.2020.09.020
Lachmann, R., Halbedel, S., Luth, S., Holzer, A., Adler, M., Pietzka, A., Al Dahouk, S., Stark, K., Flieger, A., Kleta, S., & Wilking, H. (2022). Invasive listeriosis outbreaks
and salmon products: A genomic, epidemiological study. Emerging Microbes & Infections, 11, 1308–1315. https://doi.org/10.1080/22221751.2022.2063075
Lake, F., van Overbeek, L. S., Baars, J. J. P., Koomen, J., Abee, T., & den Besten, H. M. W. (2021). Genomic characteristics of Listeria monocytogenes isolated
during mushroom (Agaricus bisporus) production and processing. International Journal of Food Microbiology, 360, 10. https://doi.org/10.1016/j.ijfoo
dmicro.2021.109438
Lakicevic, B. Z., den Besten, H. M. W., & De Biase, D. (2022). Landscape of stress response and virulence genes among Listeria monocytogenes strains.
Frontiers in Microbiology, 12, 16. https://doi.org/10.3389/fmicb.2021.738470
Larsen, M. H., Dalmasso, M., Ingmer, H., Langsrud, S., Malakauskas, M., Mader, A., Moretro, T., Mozina, S. S., Rychli, K., Wagner, M., Wallace, R. J., Zentek,
J., & Jordan, K. (2014). Persistence of foodborne pathogens and their control in primary and secondary food production chains. Food Control, 44,
92–109. https://doi.org/10.1016/j.foodcont.2014.03.039
Lassen, S. G., Ethelberg, S., Bjorkman, J. T., Jensen, T., Sorensen, G., Jensen, A. K., Muller, L., Nielsen, E. M., & Molbak, K. (2016). Two listeria outbreaks
caused by smoked fish consumption-using whole-genome sequencing for outbreak investigations. Clinical Microbiology and Infection, 22, 620–
624. https://doi.org/10.1016/j.cmi.2016.04.017
Lee, B. H., Cole, S., Badel-Berchoux, S., Guillier, L., Felix, B., Krezdorn, N., Hebraud, M., Bernardi, T., Sultan, I., & Piveteau, P. (2019). Biofilm formation of
Listeria monocytogenes strains under food processing environments and Pan-genome-wide association study. Frontiers in Microbiology, 10. https://
doi.org/10.3389/fmicb.2019.02698
Lehner, A., Riedel, K., Eberl, L., Breeuwer, P., Diep, B., & Stephan, R. (2005). Biofilm formation, extracellular polysaccharide production, and cell-to-cell sig-
naling in various Enterobacter sakazakii strains: Aspects promoting environmental persistence. Journal of Food Protection, 68, 2287–2294. https://
doi.org/10.4315/0362-028X-68.11.2287
Lianou, A., Panagou, E. Z., & Nychas, G.-J. E. (2023). Chapter 16 -meat safety—I foodborne pathogens and other biological issues. In F. Toldrá (Ed.),
Lawrie's meat science (ninth edition) (pp. 549–590). Woodhead Publishing.
Lienemann, T., Niskanen, T., Guedes, S., Siitonen, A., Kuusi, M., & Rimhanen-Finne, R. (2011). Iceberg lettuce as suggested source of a Nationwide out-
break caused by two Salmonella serotypes, Newport and Reading, in Finland in 2008. Journal of Food Protection, 74, 1035–1040. https://doi.org/10.
4315/0362-028x.Jfp-10-455
Liu, X., Chen, W. J., Fang, Z. X., Yu, Y., Bi, J., Wang, J., Dong, Q. L., & Zhang, H. Z. (2022). Persistence of Listeria monocytogenes ST5 in ready-to-eat food
processing environment. Food, 11. https://doi.org/10.3390/foods11172561
Liu, Y., Ge, W. P., Zhang, J., Li, X. P., Wu, X. Y., Li, T. L., Zhang, X. J., & Wang, X. N. (2018). Detection of Bacillus cereus sensu lato from environments associated
with goat milk powdered infant formula production facilities. International Dairy Journal, 83, 10–16. https://doi.org/10.1016/j.idair yj.2018.02.005
Löfström, C., Eriksson, J., Aspán, A., Häggblom, P., Gunnarsson, A., Borch, E., & Rådström, P. (2006). Improvement and validation of RAPD in combination
with PFGE analysis of Salmonella enterica ssp enterica serovar Senftenberg strains isolated from feed mills. Veterinary Microbiology, 114, 345–351.
https://doi.org/10.1016/j.vetmic.2005.12.015
Lomonaco, S., Decastelli, L., Nucera, D., Gallina, S., Manila Bianchi, D., & Civera, T. (2009). Listeria monocytogenes in Gorgonzola: Subtypes, diversity and
persistence over time. International Journal of Food Microbiology, 128, 516–520. https://doi.org/10.1016/j.ijfoodmicro.2008.10.009
Louha, S., Meinersmann, R. J., Abdo, Z., Berrang, M. E., & Glenn, T. C. (2020). An open-source program (Haplo-ST) for whole-genome sequence typ-
ing shows extensive diversity among Listeria monocytogenes isolates in outdoor environments and poultry processing plants. Applied and
Environmental Microbiology, 87. https://doi.org/10.1128/aem.02248-20
Lu, Y., Liu, P., Li, C. G., Sha, M., Fang, J. Q., Gao, J. W., Xu, X. X., & Matthews, K. R. (2019). Prevalence and genetic diversity of Cronobacter species isolated
from four infant formula production factories in China. Frontiers in Microbiology, 10, 8. https://doi.org/10.3389/fmicb.2019.01938
Lucchini, R., Carraro, L., Pauletto, M., Gallo, M., Andreani, N. A., Weiss, G., Tessaro, C., Babbucci, M., & Cardazzo, B. (2023). Molecular typing and genome
sequencing allow the identification of persistent Listeria monocytogenes strains and the tracking of the contamination source in food environ-
ments. International Journal of Food Microbiology, 386, 13. https://doi.org/10.1016/j.ijfoodmicro.2022.110025
Lundén, J., Tolvanen, R., & Korkeala, H. (2008). Acid and heat tolerance of persistent and nonpersistent Listeria monocytogenes food plant strains. Letters
in Applied Microbiology, 46, 276–280. https://doi.org/10.1111/j.1472-765X.2007.02305.x
Lundén, J. M., Autio, T. J., & Korkeala, H. J. (2002). Transfer of persistent Listeria monocytogenes contamination between food-processing plants associ-
ated with a dicing machine. Journal of Food Protection, 65, 1129–1133. https://doi.org/10.4315/0362-028x-65.7.1129
Luth, S., Halbedel, S., Rosner, B., Wilking, H., Holzer, A., Roedel, A., Dieckmann, R., Vincze, S., Prager, R., Flieger, A., Al Dahouk, S., & Kleta, S. (2020).
Backtracking and forward checking of human listeriosis clusters identified a multiclonal outbreak linked to Listeria monocytogenesin meat prod-
ucts of a single producer. Emerging Microbes & Infections, 9, 1600–1608. https://doi.org/10.1080/22221751.2020.1784044
Madden, R. H., Hutchison, M., Jordan, K., Pennone, V., Gundogdu, O., & Corcionivoschi, N. (2018). Prevalence and persistence of Listeria monocytogenes
in premises and products of small food business operators in Northern Ireland. Food Control, 87, 70–78. https://doi.org/10.1016/j.foodcont.2017.
12.020
Magalhaes, R., Almeida, G., Ferreira, V., Santos, I., Silva, J., Mendes, M. M., Pita, J., Mariano, G., Mancio, I., Sousa, M. M., Farber, J., Pagotto, F., & Teixeira, P.
(2015). Cheese-related listeriosis outbreak, Portugal, march 2009 to February 2012. Eurosurveillance, 20, 14–19. https://doi.org/10.2807/1560-7917.
Es2015.20.17.21104
Magalhaes, R., Ferreira, V., Brandao, T. R. S., Palencia, R. C., Almeida, G., & Teixeira, P. (2016). Persistent and non-persistent strains of Listeria monocyto-
genes: A focus on growth kinetics under different temperature, salt, and pH conditions and their sensitivity to sanitizers. Food Microbiology, 57,
103–108. https://doi.org/10.1016/j.fm.2016.02.005
Maggio, F., Rossi, C., Chiaverini, A., Ruolo, A., Orsini, M., Centorame, P., Acciari, V. A., López, C. C., Salini, R., Torresi, M., Serio, A., Pomilio, F., & Paparella, A.
(2021). Genetic relationships and biofilm formation of Listeria monocytogenes isolated from the smoked salmon industry. International Journal of
Food Microbiology, 356, 109353. https://doi.org/10.1016/j.ijfoodmicro.2021.109353
Malley, T. J. V., Butts, J., & Wiedmann, M. (2015). Seek and destroy process: Listeria monocytogenes process controls in the ready-to-eat meat and poultry
industry. Journal of Food Protection, 78, 436–445. https://doi.org/10.4315/0362-028x.Jfp-13-507
Malley, T. J. V., Stasiewicz, M. J., Grohn, Y. T., Roof, S., Warchocki, S., Nightingale, K., & Wiedmann, M. (2013). Implementation of statistical tools to support
identification and Management of Persistent Listeria monocytogenes contamination in smoked fish processing plants. Journal of Food Protection,
76, 796–811. https://doi.org/10.4315/0362-028x.Jfp-12-236
Mannion, C., Fanning, J., McLernon, J., Lendrum, L., Gutierrez, M., Duggan, S., & Egan, J. (2012). The role of transport, lairage and slaughter processes in the
dissemination of Salmonella spp. in pigs in Ireland. Food Research International, 45, 871–879. https://doi.org/10.1016/j.foodres.2011.02.001
Manso, B., Melero, B., Stessl, B., Jaime, I., Wagner, M., Rovira, J., & Rodríguez-Lázaro, D. (2020). The response to oxidative stress in Listeria monocytogenes
is temperature dependent. Microorganisms, 8, 10. https://doi.org/10.3390/microorganisms8040521
Mariani, C., Oulahal, N., Chamba, J. F., Dubois-Brissonnet, F., Notz, E., & Briandet, R. (2011). Inhibition of Listeria monocytogenes by resident biofilms
present on wooden shelves used for cheese ripening. Food Control, 22, 1357–1362. https://doi.org/10.1016/j.foodcont.2011.02.012
Martinez, B. A., Bianchini, A., Stratton, J., Raabe, O., & Swanson, S. (2021). Condensation removal practices and their potential for contributing to environ-
mental pathogen contamination in food processing facilities. Journal of Food Protection, 84, 1047–1054. https://doi.org/10.4315/jfp-20-337
Martínez-Suárez, J. V., Ortiz, S., & López-Alonso, V. (2016). Potential impact of the resistance to quaternary ammonium disinfectants on the persistence
of Listeria monocytogenes in food processing environments. Frontiers in Microbiology, 7, 8. https://doi.org/10.3389/fmicb.2016.00638
Martinez-Urtaza, J., & Liebana, E. (2005). Use of pulsed-f ield gel electrophoresis to characterize the genetic diversity and clonal persistence of Salmonella
Senftenberg in mussel processing facilities. International Journal of Food Microbiology, 105, 153–163. https://doi.org/10.1016/j.ijfoodmicro.2005.04.
006
Matereke, L. T., & Okoh, A. I. (2020). Listeria monocytogenes virulence, antimicrobial resistance and environmental persistence: A review. Pathogens, 9,
12. https://doi.org/10.3390/pathogens9070528
Maurella, C., Gallina, S., Ru, G., Adriano, D., Bellio, A., Bianchi, D. M., Chiavacci, L., Crescio, M. I., Croce, M., D'Errico, V., Dupont, M. F., Marra, A., Natangelo,
U., Pomilio, F., Romano, A., Stanzione, S., Zaccaria, T., Zuccon, F., Caramelli, M., & Decastelli, L. (2018). Outbreak of febrile gastroenteritis caused by
Listeria monocytogenes 1/2a in sliced cold beef ham, Italy, May 2016. Eurosurveillance, 23, 13–21. https://doi.org/10.2807/1560-7917.Es.2018.23.10.
17-00155
Maury, M. M., Bracq-Dieye, H., Huang, L., Vales, G., Lavina, M., Thouvenot, P., Disson, O., Leclercq, A., Brisse, S., & Lecuit, M. (2019). Hypervirulent Listeria
monocytogenes clones' adaption to mammalian gut accounts for their association with dairy products. Nature Communications, 10, 13. https://
doi.org/10.1038/s41467-019-10380-0
Maury, M. M., Tsai, Y. H., Charlier, C., Touchon, M., Chenal-Francisque, V., Leclercq, A., Criscuolo, A., Gaultier, C., Roussel, S., Brisabois, A., Disson, O., Rocha,
E. P. C., Brisse, S., & Lecuit, M. (2016). Uncovering Listeria monocytogenes hypervirulence by harnessing its biodiversity. Nature Genetics, 48, 308–313.
https://doi.org/10.1038/ng.3501
McCollum, J. T., Cronquist, A. B., Silk, B. J., Jackson, K. A., O'Connor, K. A., Cosgrove, S., Gossack, J. P., Parachini, S. S., Jain, N. S., Ettestad, P., Ibraheem, M.,
Cantu, V., Joshi, M., DuVernoy, T., Fogg, N. W., Gorny, J. R., Mogen, K. M., Spires, C., Teitell, P., … Mahon, B. E. (2013). Multistate outbreak of Listeriosis
associated with cantaloupe. New England Journal of Medicine, 369, 944–953. https://doi.org/10.1056/NEJMoa1215837
McLauchlin, J., Aird, H., Amar, C., Barker, C., Dallman, T., Elviss, N., Jorgensen, F., & Willis, C. (2020). Listeria monocytogenes in cooked chicken: Detection
of an outbreak in the United Kingdom (2016 to 2017) and analysis of L. monocytogenes from unrelated monitoring of foods (2013 to 2017). Journal
of Food Protection, 83, 2041–2052. https://doi.org/10.4315/jfp-20-188
McLauchlin, J., Aird, H., Amar, C., Barker, C., Dallman, T., Lai, S., Painset, A., & Willis, C. (2021). An outbreak of human listeriosis associated with frozen sweet
corn consumption: Investigations in the UK. International Journal of Food Microbiology, 338, 9. https://doi.org/10.1016/j.ijfoodmicro.2020.108994
Medina-Santana, J. L., Ortega-Paredes, D., de Janon, S., Burnett, E., Ishida, M., Sauders, B., Stevens, M., & Vinueza-Burgos, C. (2022). Investigating the
dynamics of Salmonella contamination in integrated poultry companies using a whole genome sequencing approach. Poultry Science, 101, 12.
https://doi.org/10.1016/j.psj.2021.101611
Melero, B., Juntunen, P., Hanninen, M. L., Jaime, I., & Rovira, J. (2012). Tracing campylobacter jejuni strains along the poultry meat production chain from
farm to retail by pulsed-f ield gel electrophoresis, and the antimicrobial resistance of isolates. Food Microbiology, 32, 124–128. https://doi.org/10.
1016/j.fm.2012.04.020
Melero, B., Manso, B., Stessl, B., Hernández, M., Wagner, M., Rovira, J., & Rodríguez-Lázaro, D. (2019). Distribution and persistence of Listeria monocyto-
genes in a heavily contaminated poultry processing facility. Journal of Food Protection, 82, 1524–1531. https://doi.org/10.4315/0362-028x.Jfp-19-087
Melero, B., Stessl, B., Manso, B., Wagner, M., Esteban-Carbonero, Ó. J., Hernández, M., Rovira, J., & Rodriguez-Lázaro, D. (2019). Listeria monocytogenes
colonization in a newly established dairy processing facility. International Journal of Food Microbiology, 289, 64–71. https://doi.org/10.1016/j.ijfoo
dmicro.2018.09.003
Mellou, K., Kyritsi, M., Chrysostomou, A., Sideroglou, T., Georgakopoulou, T., & Hadjichristodoulou, C. (2019). Clostridium perfringens foodborne out-
break during an athletic event in northern Greece, June 2019. International Journal of Environmental Research and Public Health, 16. https://doi.org/
10.3390/ijerph16203967
Miao, J., Liang, Y., Chen, L., Wang, W., Wang, J., Li, B., Li, L., Chen, D., & Xu, Z. (2017). Formation and development of staphylococcus biofilm: With focus on
food safety. Journal of Food Safety, 37, e12358. https://doi.org/10.1111/jfs.12358
Milanov, D., Ljubojevic, D., Cabarkapa, I., Karabasil, N., & Velhner, M. (2017). Biofilm as risk factor for Salmonella contamination in various stages of poultry
production. European Poultry Science, 81, 14. https://doi.org/10.1399/eps.2017.190
Mirena, I., Martin Laage, K., Judit, S., Elif Seyda, T., Natacha Friis, H., Alexander, G., Corinne, A., Claudia, G., TuAnh, N. H., Renáta, K., Carmen Rota, G.,
Diego, G., Eurydice, A., Andrea, E., Patrizia, C., Marina, T., Marina Elisabetta, A., Francesco, P., Anders Hauge, O., … Pimlapas, L. (2023). Large-scale
phenotypic and genomic characterization of Listeria monocytogenes susceptibility to quaternary ammonium compounds. bioRxiv. https://doi.org/
10.1101/2023.09.07.556668
Moffatt, C. R. M., Musto, J., Pingault, N., Combs, B., Miller, M., Stafford, R., Gregory, J., Polkinghorne, B. G., & Kirk, M. D. (2017). Recovery of Salmonella
enterica from Australian layer and processing environments following outbreaks linked to eggs. Foodborne Pathogens and Disease, 14, 478–482.
https://doi.org/10.1089/fpd.2016.2268
Mohan, V., Cruz, C. D., van Vliet, A. H. M., Pitman, A. R., Visnovsky, S. B., Rivas, L., Gilpin, B., & Fletcher, G. C. (2021). Genomic diversity of Listeria monocyto-
genes isolates from seafood, horticulture and factory environments in New Zealand. International Journal of Food Microbiology, 347, 109166. https://
doi.org/10.1016/j.ijfoodmicro.2021.109166
Mokhtari, A., & Van Doren, J. M. (2019). An agent-based model for pathogen persistence and cross-contamination dynamics in a food facility. Risk
Analysis, 39, 992–1021. https://doi.org/10.1111/risa.13215
Møller, C. O., Nauta, M., & Hansen, T. B. (2020). Modelling cross-contamination in food processing. In F. Perez-Rodriguez (Ed.), Risk assessment methods for
biological and chemical hazards in food (p. 355, 386). CRC Press. https://doi.org/10.1201/9780429083525-13
Møller, C. O., Nauta, M. J., Christensen, B. B., Dalgaard, P., & Hansen, T. B. (2012). Modelling transfer of Salmonella Typhimurium DT104 during simulation
of grinding of pork. Journal of Applied Microbiology, 112, 90–98. https://doi.org/10.1111/j.1365-2672.2011.05177.x
Møller, C. O., Sant'Ana, A. S., Hansen, S. K., Nauta, M. J., Silva, L. P., Alvarenga, V. O., Maffei, D., Silva, F. F., Lopes, J. T., Franco, B. D., Aabo, S., & Hansen, T. B.
(2016). Evaluation of a cross contamination model describing transfer of Salmonella spp. and Listeria monocytogenes during grinding of pork and
beef. International Journal of Food Microbiology, 226, 42–52. https://doi.org/10.1016/j.ijfoodmicro.2016.03.016
Møller Nielsen, E., Björkman, J. T., Kiil, K., Grant, K., Dallman, T., Painset, A., Amar, C., Roussel, S., Guillier, L., Félix, B., Rotariu, O., Perez-Reche, F., Forbes, K.,
& Strachan, N. (2017). Closing gaps for performing a risk assessment on Listeria monocytogenes in ready-to-eat (RTE) foods: Activity 3, the com-
parison of isolates from different compartments along the food chain, and from humans using whole genome sequencing (WGS) analysis. EFSA
supporting publications, 1151E. https://doi.org/10.2903/sp.efsa.2017.EN-1151
Monte, D. F. M., Andrigheto, C., Ribeiro, V. B., Landgraf, M., & Destro, M. T. (2020). Highly clonal relationship among Salmonella Enteritidis isolates in a
commercial chicken production chain, Brazil. Brazilian Journal of Microbiology, 51, 2049–2056. https://doi.org/10.1007/s42770-020-00372-4
Monte, D. F. M., Nethery, M. A., Berman, H., Keelara, S., Lincopan, N., Fedorka-Cray, P. J., Barrangou, R., & Landgraf, M. (2022). Clustered regularly inter-
spaced short palindromic repeats genotyping of multidrug-resistant Salmonella Heidelberg strains isolated from the poultry production chain
across Brazil. Frontiers in Microbiology, 13, 14. https://doi.org/10.3389/fmicb.2022.867278
Moretro, T., Midtgaard, E. S., Nesse, L. L., & Langsrud, S. (2003). Susceptibility of Salmonella isolated from fish feed factories to disinfectants and air-
drying at surfaces. Veterinary Microbiology, 94, 207–217. https://doi.org/10.1016/s0378-1135(03)00105-6
Moretro, T., Schirmer, B. C. T., Heir, E., Fagerlund, A., Hjemli, P., & Langsrud, S. (2017). Tolerance to quaternary ammonium compound disinfectants may
enhance growth of Listeria monocytogenes in the food industry. International Journal of Food Microbiology, 241, 215–224. https://doi.org/10.1016/j.
ijfoodmicro.2016.10.025
Morey, A., McKee, S. R., Dickson, J. S., & Singh, M. (2010). Efficacy of ultraviolet light exposure against survival of Listeria monocytogenes on conveyor
belts. Foodborne Pathogens and Disease, 7, 737–740. https://doi.org/10.1089/fpd.2009.0464
Morganti, M., Bolzoni, L., Scaltriti, E., Casadei, G., Carra, E., Rossi, L., Gherardi, P., Faccini, F., Arrigoni, N., Sacchi, A. R., Delledonne, M., & Pongolini, S. (2018).
Rise and fall of outbreak-specific clone inside endemic pulsotype of Salmonella 4, 5, 12:i:-; insights from high-resolution molecular surveillance in
Emilia-Romagna, Italy, 2012 to 2015. Eurosurveillance, 23, 19–29. https://doi.org/10.2807/1560-7917.Es.2018.23.13.17- 00375
Morita, T., Ohyagi, N., Matsuura, J., Kawaguchi, T., & Ishizaki, N. (2022). Salmonella contamination and hazard analysis in a storage facility for feed mate-
rials in Japan. Journal of Applied Microbiology, 133, 2966–2978. https://doi.org/10.1111/jam.15744
Mottola, A., Bonerba, E., Bozzo, G., Marchetti, P., Celano, G. V., Colao, V., Terio, V., Tantillo, G., Figueras, M. J., & Di Pinto, A. (2016). Occurrence of emerging
food-borne pathogenic Arcobacter spp. isolated from pre-cut (ready-to-eat) vegetables. International Journal of Food Microbiology, 236, 33–37.
https://doi.org/10.1016/j.ijfoodmicro.2016.07.012
Mottola, A., Ciccarese, G., Sinisi, C., Savarino, A. E., Marchetti, P., Terio, V., Tantillo, G., Barrasso, R., & Di Pinto, A. (2021). Occurrence and characterization of
Arcobacter spp. from ready-to-eat vegetables produced in southern Italy. Italian Journal of Food Safety, 10, 8585. https://doi.org/10.4081/ijfs.2021.
8585
Muchaamba, F., Eshwar, A. K., Stevens, M. J. A., Stephan, R., & Tasara, T. (2022). Different shades of Listeria monocytogenes: Strain, serotype, and lineage-
based variability in virulence and stress tolerance profiles. Frontiers in Microbiology, 12, 23. https://doi.org/10.3389/fmicb.2021.792162
Muhterem-Uyar, M., Ciolacu, L., Wagner, K.-H., Wagner, M., Schmitz-Esser, S., & Stessl, B. (2018). New aspects on Listeria monocytogenes ST5-ECVI pre-
dominance in a heavily contaminated cheese processing environment. Frontiers in Microbiology, 9. https://doi.org/10.3389/fmicb.2018.00064
Mullane, N., Healy, B., Meade, J., Whyte, P., Wall, P. G., & Fanning, S. (2008). Dissemination of Cronobacter spp. (Enterobacter sakazakii) in a powdered milk
protein manufacturing facility. Applied and Environmental Microbiology, 74, 5913–5917. https://doi.org/10.1128/aem.00745- 08
Mullane, N. R., Whyte, P., Wall, P. G., Quinn, T., & Fanning, S. (2007). Application of pulsed-f ield gel electrophoresis to characterise and trace the prev-
alence of Enterobacter sakazakii in an infant formula processing facility. International Journal of Food Microbiology, 116, 73–81. https://doi.org/
10.1016/j.ijfoodmicro.2006.12.036
Muller, A., Stephan, R., Fricker-Feer, C., & Lehner, A. (2013). Genetic diversity of Cronobacter sakazakii (I)solates collected from a Swiss infant formula
production facility. Journal of Food Protection, 76, 883–887. https://doi.org/10.4315/0362-028x.Jfp-12-521
Murase, T., Senjyu, K., Maeda, T., Tanaka, M., Sakae, H., Matsumoto, Y., Kaneda, Y., Ito, T., & Otsuki, K. (2001). Monitoring of chicken houses and an attached
egg-processing facility in a laying farm for Salmonella contamination between 1994 and 1998. Journal of Food Protection, 64, 1912–1916. https://
doi.org/10.4315/0362-028x-64.12.1912
Murugesan, L., Kucerova, Z., Knabel, S. J., & LaBorde, L. F. (2015). Predominance and distribution of a persistent Listeria monocytogenes clone in a com-
mercial fresh mushroom processing environment. Journal of Food Protection, 78, 1988–1998. https://doi.org/10.4315/0362-028x.Jfp-15-195
Musgrove, M. T., Berrang, M. E., & Liljebjelke, K. A. (2008). Attached bacterial cell contamination of Shell egg processing facilities. Proceedings of the UJNR
Food & Agricultural Panel Proceedings, 66–67.
Myintzaw, P., Pennone, V., McAuliffe, O., Begley, M., & Callanan, M. (2023). Association of Virulence, biofilm, and antimicrobial resistance genes with
specific clonal complex types of Listeria monocytogenes. Microorganisms, 11. https://doi.org/10.3390/microorganisms11061603
Nakari, U. M., Rantala, L., Pihlajasaari, A., Toikkanen, S., Johansson, T., Hellsten, C., Raulo, S. M., Kuusi, M., Siitonen, A., & Rimhanen-Finne, R. (2014).
Investigation of increased listeriosis revealed two fishery production plants with persistent Listeria contamination in Finland in 2010. Epidemiology
and Infection, 142, 2261–2269. https://doi.org/10.1017/s095026881300349x
Napoleoni, M., Villa, L., Barco, L., Busani, L., Cibin, V., Lucarelli, C., Tiengo, A., Dionisi, A. M., Conti, F., Da Silva Nunes, F. R., Tantucci, L., Staffolani, M.,
Silenzi, V., Fraticelli, R., Morandi, B., Blasi, G., Rocchegiani, E., Fisichella, S., & E-N and Enter-Vet Peripheral Laboratories Referents For Marche R.
(2021). A strong evidence outbreak of Salmonella Enteritidis in Central Italy linked to the consumption of contaminated raw sheep Milk cheese.
Microorganisms, 9. https://doi.org/10.3390/microorganisms9122464
Nastasijevic, I., Milanov, D., Velebit, B., Djordjevic, V., Swift, C., Painset, A., & Lakicevic, B. (2017). Tracking of Listeria monocytogenes in meat establish-
ment using whole genome sequencing as a food safety management tool: A proof of concept. International Journal of Food Microbiology, 257,
157–164. https://doi.org/10.1016/j.ijfoodmicro.2017.06.015
Nauta, M. J. (2005). Microbiological risk assessment models for partitioning and mixing during food handling. International Journal of Food Microbiology,
Nauta, M. J. (2007). The modular process risk model (MPRM): A structured approach to food chain exposure assessment. Microbial Risk Analysis of Foods.,
99–136.
Negrete, F. J., Ko, K., Jang, H., Hoffmann, M., Lehner, A., Stephan, R., Fanning, S., Tall, B., & Gopinath, G. R. (2022). Complete genome sequences and
genomic characterization of five plasmids harbored by environmentally persistent Cronobacter sakazakii strains ST83 H322 and ST64 GK1025B
obtained from powdered infant formula manufacturing facilities. Gut Pathogens, 14, 12. https://doi.org/10.1186/s13099- 022- 00500-5
Nelson, K. E., Fouts, D. E., Mongodin, E. F., Ravel, J., DeBoy, R. T., Kolonay, J. F., Rasko, D. A., Angiuoli, S. V., Gill, S. R., Paulsen, I. T., Peterson, J., White, O.,
Nelson, W. C., Nierman, W., Beanan, M. J., Brinkac, L. M., Daugherty, S. C., Dodson, R. J., Durkin, A. S., … Fraser, C. M. (2004). Whole genome compar-
isons of serotype 4b and 1/2a strains of the food-borne pathogen Listeria monocytogenes reveal new insights into the core genome components
of this species. Nucleic Acids Research, 32, 2386–2395. https://doi.org/10.1093/nar/gkh562
Nesbakken, T., Kapperud, G., & Caugant, D. A. (1996). Pathways of Listeria monocytogenes contamination in the meat processing industry. International
Journal of Food Microbiology, 31, 161–171. https://doi.org/10.1016/0168-1605(96)00978-6
Nesse, L. L., Nordby, K., Heir, E., Bergsjoe, B., Vardund, T., Nygaard, H., & Holstad, G. (2003). Molecular analyses of Salmonella enterica isolates from fish
feed factories and fish feed ingredients. Applied and Environmental Microbiology, 69, 1075–1081. https://doi.org/10.1128/aem.69.2.1075-1081.2003
Nesse, L. L., Refsum, T., Heir, E., Nordby, K., Vardund, T., & Holstad, G. (2005). Molecular epidemiology of Salmonella spp. isolates from gulls, fish-meal
factories, feed factories, animals and humans in Norway based on pulsed-f ield gel electrophoresis. Epidemiology and Infection, 133, 53–58. https://
doi.org/10.1017/s0950268804003279
Nichols, M., Conrad, A., Whitlock, L., Stroika, S., Strain, E., Weltman, A., Johnson, L., Dement, J., Reporter, R., & Williams, I. (2019). Short communication:
Multistate outbreak of Listeria monocytogenes infections retrospectively linked to unpasteurized milk using whole-genome sequencing. Journal
of Dairy Science, 103, 176–178. https://doi.org/10.3168/jds.2019-16703
Nilsson, R. E., Latham, R., Mellefont, L., Ross, T., & Bowman, J. P. (2012). MudPIT analysis of alkaline tolerance by Listeria monocytogenes strains recovered
as persistent food factory contaminants. Food Microbiology, 30, 187–196. https://doi.org/10.1016/j.fm.2011.10.004
Nilsson, R. E., Ross, T., & Bowman, J. P. (2011). Variability in biofilm production by Listeria monocytogenes correlated to strain origin and growth condi-
tions. International Journal of Food Microbiology, 150, 14–24. https://doi.org/10.1016/j.ijfoodmicro.2011.07.012
Nowak, J., Cruz, C. D., Tempelaars, M., Abee, T., van Vliet, A. H. M., Fletcher, G. C., Hedderley, D., Palmer, J., & Flint, S. (2017). Persistent Listeria monocyto-
genes strains isolated from mussel production facilities form more biofilm but are not linked to specific genetic markers. International Journal of
Food Microbiology, 256, 45–53. https://doi.org/10.1016/j.ijfoodmicro.2017.05.024
Nowak, J., Visnovsky, S. B., Pitman, A. R., Cruz, C. D., Palmer, J., Fletcher, G. C., & Flint, S. (2021). Biofilm formation by Listeria monocytogenes 15G01, a per-
sistent isolate from a seafood-processing plant, is influenced by inactivation of multiple genes belonging to different functional groups. Applied
and Environmental Microbiology, 87, 19. https://doi.org/10.1128/aem.02349-20
Nüesch-Inderbinen, M., Bloemberg, G. V., Müller, A., Stevens, M. J. A., Cernela, N., Kollöffel, B., & Stephan, R. (2021). Listeriosis caused by persistence of
Listeria monocytogenes serotype 4b sequence type 6 in cheese production environment. Emerging Infectious Diseases, 27, 284–288. https://doi.
org/10.3201/eid2701.203266
Ojeniyi, B., Christensen, J., & Bisgaard, M. (2000). Comparative investigations of Listeria monocytogenes isolated from a Turkey processing plant, Turkey
products, and from human cases of listeriosis in Denmark. Epidemiology and Infection, 125, 303–308. https://doi.org/10.1017/s0950268899004483
Orieskova, M., Kajsik, M., Szemes, T., Holy, O., Forsythe, S., Turna, J., & Drahovska, H. (2016). Contribution of the thermotolerance genomic Island to
increased thermal tolerance in Cronobacter strains. Antonie Van Leeuwenhoek International Journal of General and Molecular Microbiology, 109,
405–414. https://doi.org/10.1007/s10482-016-0645-1
Orsi, R. H., Borowsky, M. L., Lauer, P., Young, S. K., Nusbaum, C., Galagan, J. E., Birren, B. W., Ivy, R. A., Sun, Q., Graves, L. M., Swaminathan, B., & Wiedmann,
M. (2008). Short-term genome evolution of Listeria monocytogenes in a non-controlled environment. BMC Genomics, 9, 17. https://doi.org/10.1186/
1471-2164-9-539
Ortiz, S., Lopez, V., & Martinez-Suarez, J. V. (2014). Control of Listeria monocytogenes contamination in an Iberian pork processing plant and selection of
benzalkonium chloride-resistant strains. Food Microbiology, 39, 81–88. https://doi.org/10.1016/j.fm.2013.11.007
Ortiz, S., Lopez, V., Villatoro, D., Lopez, P., Davila, J. C., & Martinez-Suarez, J. V. (2010). A 3-year surveillance of the genetic diversity and persistence of
Listeria monocytogenes in an Iberian pig slaughterhouse and processing plant. Foodborne Pathogens and Disease, 7, 1177–1184.
Ortiz, S., Lopez-Alonso, V., Rodriguez, P., & Martinez-Suarez, J. V. (2016). The connection between persistent, disinfectant-resistant Listeria monocy-
togenes strains from two geographically separate Iberian pork processing plants: Evidence from comparative genome analysis. Applied and
Environmental Microbiology, 82, 308–317. https://doi.org/10.1128/aem.02824-15
Osaili, T., & Forsythe, S. (2009). Desiccation resistance and persistence of Cronobacter species in infant formula. International Journal of Food Microbiology,
Osek, J., Lachtara, B., & Wieczorek, K. (2022). Listeria monocytogenes -how this pathogen survives in food-production environments? Frontiers in
Oswaldi, V., Lüth, S., Dzierzon, J., Meemken, D., Schwarz, S., Fessler, A. T., Félix, B., & Langforth, S. (2022). Distribution and characteristics of Listeria spp.
Pigs and Pork Production Chains in Germany. Microorganisms, 10. https://doi.org/10.3390/microorganisms10030512
Oxaran, V., Lee, S. H. I., Chaul, L. T., Corassin, C. H., Barancelli, G. V., Alves, V. F., de Oliveira, C. A. F., Gram, L., & De Martinis, E. C. P. (2017). Listeria monocy-
togenes incidence changes and diversity in some Brazilian dairy industries and retail products. Food Microbiology, 68, 16–23. https://doi.org/10.
1016/j.fm.2017.06.012
Pagadala, S., Parveen, S., Rippen, T., Luchansky, J. B., Call, J. E., Tamplin, M. L., & Porto-Fett, A. C. S. (2012). Prevalence, characterization and sources of
Listeria monocytogenes in blue crab (Callinectus sapidus) meat and blue crab processing plants. Food Microbiology, 31, 263–270. https://doi.org/
10.1016/j.fm.2012.03.015
Painset, A., Bjorkman, J. T., Kiil, K., Guillier, L., Mariet, J. F., Felix, B., Amar, C., Rotariu, O., Roussel, S., Perez-Reche, F., Brisse, S., Moura, A., Lecuit, M., Forbes,
K., Strachan, N., Grant, K., Moller-Nielsen, E., & Dallman, T. J. (2019). LiSEQ -whole-genome sequencing of a cross-sectional survey of Listeria mono-
cytogenes in ready-to-eat foods and human clinical cases in Europe. Microbial Genomics, 5. https://doi.org/10.1099/mgen.0.000257
Palacios, A., Otto, M., Flaherty, E., Boyle, M. M., Malec, M., Holloman, K., Low, M., Wellman, A., Newhart, C., Gollarza, L., Weeks, T., Muyombwe, A., Kristen
Lozinak, K., Erin Kafka, E., Daniel O'Halloran, D., Teresa Rozza, T., David Nicholas, D., Stranjae Ivory, S., Katherine Kreil, K., … Amanda Conrad, A.
(2022). Multistate outbreak of Listeria monocytogenes infections linked to fresh, soft Hispanic-style cheese—United States, 2021. https://w ww.
cdc.gov/mmwr/volumes/71/wr/mm7121a3.htm?s_cid=mm7121a3_w
Palaiodimou, L., Fanning, S., & Fox, E. M. (2021). Genomic insights into persistence of Listeria species in the food processing environment. Journal of
Applied Microbiology, 131, 2082–2094. https://doi.org/10.1111/jam.15089
Palma, F., Brauge, T., Radomski, N., Mallet, L., Felten, A., Mistou, M. Y., Brisabois, A., Guillier, L., & Midelet-Bourdin, G. (2020). Dynamics of mobile genetic
elements of Listeria monocytogenes persisting in ready-to-eat seafood processing plants in France. BMC Genomics, 21, 20. https://doi.org/10.1186/
s12864-020-6544-x
Palma, F., Pasquali, F., Lucchi, A., De Cesare, A., & Manfreda, G. (2017). Whole genome sequencing for typing and characterisation of Listeria monocyto-
genes isolated in a rabbit meat processing plant. Italian Journal of Food Safety, 6, 125–130. https://doi.org/10.4081/ijfs.2017.6879
Parker, E. M., Edwards, L. J., Mollenkopf, D. F., Ballash, G. A., Wittum, T. E., & Parker, A. J. (2019). Salmonella monitoring programs in Australian feed mills:
A retrospective analysis. Australian Veterinary Journal, 97, 336–342. https://doi.org/10.1111/avj.12851
Parker, E. M., Valcanis, M., Edwards, L. J., Andersson, P., Mollenkopf, D. F., & Wittum, T. E. (2022). Antimicrobial-resistant Salmonella is detected more
frequently in feed milling equipment than in raw feed components or processed animal feed. Australian Veterinary Journal, 100, 213–219. https://
doi.org/10.1111/avj.13146
Pasquali, F., Palma, F., Guillier, L., Lucchi, A., De Cesare, A., & Manfreda, G. (2018). Listeria monocytogenes sequence types 121 and 14 repeatedly iso-
lated within one year of sampling in a rabbit meat processing plant: Persistence and ecophysiology. Frontiers in Microbiology, 9, 12. https://doi.
org/10.3389/fmicb.2018.0 0596
Pei, X. Y., Li, Y., Zhang, H. N., Zhan, L., Yu, X. J., Lan, G., Jia, H. Y., Li, N., Yang, D. J., & Mei, L. L. (2019). Surveillance and characterisation of Cronobacter in
powdered infant formula processing factories. Food Control, 96, 318–323. https://doi.org/10.1016/j.foodcont.2018.09.009
Pellegrini, D. D. P., Paim, D. S., de Lima, G., Pissetti, C., Kich, J. D., & Cardoso, M. R. D. (2015). Distribution of Salmonella clonal groups in four Brazilian feed
mills. Food Control, 47, 672–678. https://doi.org/10.1016/j.foodcont.2014.08.013
Pennone, V., Dygico, K. L., Coffey, A., Gahan, C. G. M., Grogan, H., McAuliffe, O., Burgess, C. M., & Jordan, K. (2020). Effectiveness of current hygiene prac-
tices on minimization of Listeria monocytogenes in different mushroom production-related environments. Food Science & Nutrition, 8, 3456–3468.
https://doi.org/10.1002/fsn3.1629
Pennone, V., Lehardy, A., Coffey, A., McAuliffe, O., & Jordan, K. (2018). Diversity of Listeria monocytogenes strains isolated from Agaricus bisporus mush-
room production. Journal of Applied Microbiology, 125, 586–595. https://doi.org/10.1111/jam.13773
Pérez-Baltar, A., Pérez-Boto, D., Medina, M., & Montiel, R. (2021). Genomic diversity and characterization of Listeria monocytogenes from dry-cured ham
processing plants. Food Microbiology, 99, 103779. https://doi.org/10.1016/j.fm.2021.103779
Pexara, A., Solomakos, N., & Govaris, A. (2018). Q fever and prevalence of Coxiella burnetii in milk. Trends in Food Science & Technology, 71, 65–72. https://
doi.org/10.4315/0362-028X-73.10.1919
Pightling, A. W., Pettengill, J. B., Luo, Y., Baugher, J. D., Rand, H., & Strain, E. (2018). Interpreting whole-genome sequence analyses of foodborne bacteria
for regulatory applications and outbreak investigations. Frontiers in Microbiology, 9, 13. https://doi.org/10.3389/fmicb.2018.01482
Piras, F., Brown, D. J., Meloni, D., Mureddu, A., & Mazzette, R. (2011). Investigation of Salmonella enterica in Sardinian slaughter pigs: Prevalence, serotype
and genotype characterization. International Journal of Food Microbiology, 151, 201–209. https://doi.org/10.1016/j.ijfoodmicro.2011.08.025
Piras, F., Fois, F., Consolati, S. G., Mazza, R., & Mazzette, R. (2015). Influence of temperature, source, and serotype on biofilm formation of Salmonella
enterica isolates from pig slaughterhouses. Journal of Food Protection, 78, 1875–1878. https://doi.org/10.4315/0362-028x.Jfp-15-085
Podolak, R., Enache, E., Stone, W., Black, D. G., & Elliott, P. H. (2010). Sources and risk factors for contamination, survival, persistence, and heat resistance
of Salmonella in Low-moisture foods. Journal of Food Protection, 73, 1919–1936. https://doi.org/10.4315/0362-028X-73.10.1919
Pouillot, R., Hoelzer, K., Chen, Y., & Dennis, S. B. (2015). Listeria monocytogenes dose response revisited—Incorporating adjustments for variability in
Strain virulence and host susceptibility. Risk Analysis, 35, 90–108. https://doi.org/10.1111/risa.12235
Pouillot, R., & Lubran, M. B. (2011). Predictive microbiology models vs. modeling microbial growth within Listeria monocytogenes risk assessment: What
parameters matter and why. Food Microbiology, 28, 720–726. https://doi.org/10.1016/j.fm.2010.06.002
Power, K. A., Yan, Q. Q., Fox, E. M., Cooney, S., & Fanning, S. (2013). Genome sequence of Cronobacter sakazakii SP291, a persistent Thermotolerant isolate
derived from a factory producing powdered infant formula. Genome Announcements, 1, 1. https://doi.org/10.1128/genomeA.00082-13
Prasertsee, T., Chuammitri, P., Deeudom, M., Chokesajjawatee, N., Santiyanont, P., Tadee, P., Nuangmek, A., Tadee, P., Sheppard, S. K., Pascoe, B., &
Patchanee, P. (2019). Core genome sequence analysis to characterize Salmonella enterica serovar Rissen ST469 from a swine production chain.
International Journal of Food Microbiology, 304, 68–74. https://doi.org/10.1016/j.ijfoodmicro.2019.05.022
Prendergast, D. M., O'Grady, D., Fanning, S., Cormican, M., Delappe, N., Egan, J., Mannion, C., Fanning, J., & Gutierrez, M. (2011). Application of multiple
locus variable number of tandem repeat analysis (MLVA), phage typing and antimicrobial susceptibility testing to subtype Salmonella enterica
serovar Typhimurium isolated from pig farms, pork slaughterhouses and meat producing plants in Ireland. Food Microbiology, 28, 1087–1094.
https://doi.org/10.1016/j.fm.2011.02.013
PROFEL (European Association of Fruit and Vegetable Processors). (2019). Draft Hygiene guidelines for the production of quick-frozen fruits, vegetables
and herbs in control of Listeria monocytogenes. 45 pp.
Prunic, B., Milanov, D., Velhner, M., Pajic, M., Pavlovic, L., & Misic, D. (2016). Clonal persistence of Salmonella enterica serovars Montevideo, Tennessee,
and Infantis in feed factories. Journal of Infection in Developing Countries, 10, 662–666. https://doi.org/10.3855/jidc.7313
Quintana-Hayashi, M. P., & Thakur, S. (2012). Longitudinal study of the persistence of antimicrobial-resistant campylobacter strains in distinct swine
production systems on farms, at slaughter, and in the environment. Applied and Environmental Microbiology, 78, 2698–2705. https://doi.org/
10.1128/aem.07723-11
Ramees, T. P., Dhama, K., Karthik, K., Rathore, R. S., Kumar, A., Saminathan, M., Tiwari, R., Malik, Y. S., & Singh, R. K. (2017). Arcobacter: An emerging food-
borne zoonotic pathogen, its public health concerns and advances in diagnosis and control -a comprehensive review. The Veterinary Quarterly, 37,
136–161. https://doi.org/10.1080/01652176.2017.1323355
Reinhard, R. G., Kalinowski, R. M., Bodnaruk, P. W., Eifert, J. D., Boyer, R. R., Duncan, S. E., & Bailey, R. H. (2020). Practical application of bacteriophage in
food manufacturing facilities for the control of Listeria sp. Journal of Food Safety, 43. https://doi.org/10.1111/jfs.12871
Ren, X., Li, M., Xu, C., Cui, K., Feng, Z., Fu, Y., Zhang, J., & Liao, M. (2016). Prevalence and molecular characterization of Salmonella enterica isolates
throughout an integrated broiler supply chain in China. Epidemiology and Infection, 144, 2989–2999. https://doi.org/10.1017/s0950268816001515
Rinsky, J. L., Berl, E., Greene, V., Morrow, J., Didomenico, A., MacFaerquar, J., Gómez, G. A., Lúquez, C., & Williams, C. (2016). Notes from the field: Clostridium
perfringens gastroenteritis outbreak associated with a catered lunch — North Carolina, November 2015. MMWR. Morbidity and Mortality Weekly
Report, 2016(65), 1300–1301. https://doi.org/10.15585/mmwr.mm6546a5
Rios, E. A., Ramos-Pereira, J., Santos, J. A., López-Díaz, T. M., Otero, A., & Rodríguez-Calleja, J. M. (2020). Behaviour of non-O157 STEC and atypical EPEC
during the manufacturing and ripening of raw Milk cheese. Food, 9, 1215.
Robinson, E., Travanut, M., Fabre, L., Larréché, S., Ramelli, L., Pascal, L., Guinard, A., Vincent, N., Calba, C., Meurice, L., Le Thien, M. A., Fourgere, E., Jones,
G., Fournet, N., Smith-Palmer, A., Brown, D., Le Hello, S., Pardos de la Gandara, M., Weill, F. X., & Jourdan-Da Silva, N. (2020). Outbreak of Salmonella
Newport associated with internationally distributed raw goats' milk cheese, France, 2018. Epidemiology and Infection, 148, e180. https://doi.org/
10.1017/s0950268820000904
Rodríguez-Campos, D., Rodríguez-Melcón, C., Alonso-Calleja, C., & Capita, R. (2019). Persistent Listeria monocytogenes isolates from a poultry-processing
facility form More biofilm but do not have a greater resistance to disinfectants than sporadic strains. Pathogens, 8, 13. https://doi.org/10.3390/
pathogens8040250
Rodríguez-López, P., Rodríguez-Herrera, J. J., Vázquez-Sánchez, D., & Cabo, M. L. (2018). Current knowledge on Listeria monocytogenes biofilms in food-
related environments: Incidence, resistance to biocides, Ecology and Biocontrol. Foods, 7, 19. https://doi.org/10.3390/foods7060085
Rossetti, C. A., Maurizio, E., & Rossi, U. A. (2022). Comparative review of brucellosis in small domestic ruminants. Frontiers in Veterinary Science, 9, 887671.
https://doi.org/10.3389/f vets.2022.887671
Ruckerl, I., Muhterem-Uyar, M., Muri-Klinger, S., Wagner, K. H., Wagner, M., & Stessl, B. (2014). L. Monocytogenes in a cheese processing facility: Learning
from contamination scenarios over three years of sampling. International Journal of Food Microbiology, 189, 98–105. https://doi.org/10.1016/j.ijfoo
dmicro.2014.08.001
Ruegg, P. L. (2017). A 100-year review: Mastitis detection, management, and prevention. Journal of Dairy Science, 100, 10381–10397. https://doi.org/10.
3168/jds.2017-13023
Ruiz-Llacsahuanga, B., Hamilton, A., Zaches, R., Hanrahan, I., & Critzer, F. (2021). Prevalence of Listeria species on food contact surfaces in Washington
state apple packinghouses. Applied and Environmental Microbiology, 87, 13. https://doi.org/10.1128/aem.02932-20
Russo, E. T., Biggerstaff, G., Hoekstra, R. M., Meyer, S., Patel, N., Miller, B., Quick, R., & Salmonella Agona Outbreak Invest, T. (2013). A recurrent, multistate
outbreak of Salmonella serotype Agona infections associated with dry, unsweetened cereal consumption, United States, 2008. Journal of Food
Ryan, S., Begley, M., Hill, C., & Gahan, C. G. M. (2010). A five-gene stress survival islet (SSI-1) that contributes to the growth of Listeria monocytogenes in
suboptimal conditions. Journal of Applied Microbiology, 109, 984–995. https://doi.org/10.1111/j.1365-2672.2010.04726.x
Sahu, R., Rawool, D. B., Dhaka, P., Yadav, J. P., Mishra, S. P., Kumar, M., Vergis, J., Malik, S. S., & Barbuddhe, S. B. (2021). Current perspectives on the oc-
currence of Q fever: Highlighting the need for systematic surveillance for a neglected zoonotic disease in Indian subcontinent. Environmental
Microbiology Reports, 13, 138–158. https://doi.org/10.1111/1758-2229.12918
Salimena, A. P., Lange, C. C., Camussone, C., Signorini, M., Calvinho, L. F., Brito, M. A., Borges, C. A., Guimarães, A. S., Ribeiro, J. B., Mendonça, L. C., & Piccoli,
R. H. (2016). Genotypic and phenotypic detection of capsular polysaccharide and biofilm formation in Staphylococcus aureus isolated from bovine
milk collected from Brazilian dairy farms. Veterinary Research Communications, 40, 97–106. https://doi.org/10.1007/s11259- 016-9658-5
Sarno, E., Pezzutto, D., Rossi, M., Liebana, E., & Rizzi, V. (2021). A review of significant European foodborne outbreaks in the last decade. Journal of Food
Protection, 84, 2059–2070. https://doi.org/10.4315/jfp-21- 096
Schiavano, G. F., Ateba, C. N., Petruzzelli, A., Mele, V., Amagliani, G., Guidi, F., De Santi, M., Pomilio, F., Blasi, G., Gattuso, A., Di Lullo, S., Rocchegiani, E.,
& Brandi, G. (2022). Whole-genome sequencing characterization of virulence profiles of Listeria monocytogenes food and human isolates and
in vitro adhesion/invasion assessment. Microorganisms, 10, 13. https://doi.org/10.3390/microorganisms10010062
Schierstaedt, J., Jechalke, S., Nesme, J., Neuhaus, K., Sorensen, S. J., Grosch, R., Smalla, K., & Schikora, A. (2020). Salmonella persistence in soil de-
pends on reciprocal interactions with indigenous microorganisms. Environmental Microbiology, 22, 2639–2652. https://doi.org/10.1111/1462-
2920.14972
Schmitz-Esser, S., Anast, J. M., & Cortes, B. W. (2021). A large-scale sequencing-based survey of plasmids in Listeria monocytogenes reveals global dissem-
ination of plasmids. Frontiers in Microbiology, 12, 17. https://doi.org/10.3389/fmicb.2021.653155
Schmitz-Esser, S., Muller, A., Stessl, B., & Wagner, M. (2015). Genomes of sequence type 121 Listeria monocytogenes strains harbor highly conserved
plasmids and prophages. Frontiers in Microbiology, 6, 10. https://doi.org/10.3389/fmicb.2015.00380
Schnitt, A., Lienen, T., Wichmann-Schauer, H., Cuny, C., & Tenhagen, B. A. (2020). The occurrence and distribution of livestock-associated methicillin-
resistant Staphylococcus aureus ST398 on German dairy farms. Journal of Dairy Science, 103, 11806–11819. https://doi.org/10.3168/jds.2020-18958
Schobitz, R., Gonzalez, C., Villarreal, K., Horzella, M., Nahuelquin, Y., & Fuentes, R. (2014). A biocontroller to eliminate Listeria monocytogenes from the
food processing environment. Food Control, 36, 217–223. https://doi.org/10.1016/j.foodcont.2013.08.017
Schonewille, E., Nesse, L. L., Hauck, R., Windhorst, D., Hafez, H. M., & Vestby, L. K. (2012). Biofilm building capacity of Salmonella enterica strains from the
poultry farm environment. FEMS Immunology & Medical Microbiology, 65, 360–365. https://doi.org/10.1111/j.1574-695X.2012.00966.x
Schroeder, S., Harries, M., Prager, R., Hofig, A., Ahrens, B., Hoffmann, L., Rabsch, W., Mertens, E., & Rimek, D. (2016). A prolonged outbreak of Salmonella Infantis
associated with pork products in Central Germany, April-October 2013. Epidemiology and Infection, 144, 1429–1439. https://doi.org/10.1017/s095026881
5002629
Sebastianski, M., Bridger, N. A., Featherstone, R. M., & Robinson, J. L. (2022). Disease outbreaks linked to pasteurized and unpasteurized dairy
products in Canada and the United States: A systematic review. Canadian Journal of Public Health, 113, 569–578. https://doi.org/10.17269/
s41997-022-0 0614-y
Self, J. L., Conrad, A., Stroika, S., Jackson, A., Whitlock, L., Jackson, K. A., Beal, J., Wellman, A., Fatica, M. K., Bidol, S., Huth, P. P., Hamel, M., Franklin, K.,
Tschetter, L., Kopko, C., Kirsch, P., Wise, M. E., & Basler, C. (2019). Multistate outbreak of Listeriosis associated with packaged leafy Green salads,
United States and Canada, 2015-2016. Emerging Infectious Diseases, 25, 1461–1468. https://doi.org/10.3201/eid2508.180761
Shang, K., Wei, B., Jang, H. K., & Kang, M. (2019). Phenotypic characteristics and genotypic correlation of antimicrobial resistant (AMR) Salmonella
isolates from a poultry slaughterhouse and its downstream retail markets. Food Control, 100, 35–45. https://doi.org/10.1016/j.foodcont.2018.
12.0 46
Shedleur-Bourguignon, F., Thériault, W. P., Longpré, J., Thibodeau, A., & Fravalo, P. (2021). Use of an ecosystem-based approach to shed light on the
heterogeneity of the contamination pattern of Listeria monocytogenes on Conveyor Belt surfaces in a swine slaughterhouse in the province of
Quebec, Canada. Pathogens, 10. https://doi.org/10.3390/pathogens10111368
Simonetti, T., Peter, K., Chen, Y., Jin, Q., Zhang, G. D., LaBorde, L. F., & Macarisin, D. (2021). Prevalence and distribution of Listeria monocytogenes in three
commercial tree fruit packinghouses. Frontiers in Microbiology, 12, 12. https://doi.org/10.3389/fmicb.2021.652708
Sinde, E., & Carballo, J. (2000). Attachment of Salmonella spp. and Listeria monocytogenes to stainless steel, rubber and polytetrafluorethylene: The
influence of free energy and the effect of commercial sanitizers. Food Microbiology, 17, 439–447. https://doi.org/10.1006/fmic.2000.0339
Smith, A. M., Tau, N. P., Smouse, S. L., Allam, M., Ismail, A., Ramalwa, N. R., Disenyeng, B., Ngomane, M., & Thomas, J. (2019). Outbreak of Listeria monocyto-
genes in South Africa, 2017–2018: Laboratory activities and experiences associated with whole-genome sequencing analysis of isolates. Foodborne
Pathogens and Disease, 16, 524–530. https://doi.org/10.1089/fpd.2018.2586
Sofos, J. N. (2008). Challenges to meat safety in the 21st century. Meat Science, 78, 3–13. https://doi.org/10.1016/j.meatsci.2007.07.027
Sonbol, H., Joseph, S., McAuley, C. M., Craven, H. M., & Forsythe, S. J. (2013). Multilocus sequence typing of Cronobacter spp. from powdered infant for-
mula and milk powder production factories. International Dairy Journal, 30, 1–7. https://doi.org/10.1016/j.idair yj.2012.11.004
Soni, K. A., Nannapaneni, R., & Tasara, T. (2011). The contribution of transcriptomic and proteomic analysis in elucidating stress adaptation responses of
Listeria monocytogenes. Foodborne Pathogens and Disease, 8, 843–852. https://doi.org/10.1089/fpd.2010.0746
Stessl, B., Fricker, M., Fox, E., Karpiskova, R., Demnerova, K., Jordan, K., Ehling-Schulz, M., & Wagner, M. (2014). Collaborative survey on the colonization of
different types of cheese-processing facilities with Listeria monocytogenes. Foodborne Pathogens and Disease, 11, 8–14. https://doi.org/10.1089/fpd.
2013.1578
Stessl, B., Szakmary-Brandle, K., Vorberg, U., Schoder, D., & Wagner, M. (2020). Temporal analysis of the Listeria monocytogenes population structure in
floor drains during reconstruction and expansion of a meat processing plant. International Journal of Food Microbiology, 314, 7. https://doi.org/10.
1016/j.ijfoodmicro.2019.108360
Stoller, A., Stevens, M. J. A., Stephan, R., & Guldimann, C. (2019). Characteristics of Listeria Monocytogenes strains persisting in a meat processing facility
over a 4-year period. Pathogens, 8, 16. https://doi.org/10.3390/pathogens8010032
Strydom, A., Vorster, R., Gouws, P. A., & Witthuhn, R. C. (2016). Successful management of Listeria spp. in an avocado processing facility. Food Control, 62,
208–215. https://doi.org/10.1016/j.foodcont.2015.10.043
Sullivan, G., & Wiedmann, M. (2020). Detection and prevalence of Listeria in US produce packinghouses and fresh-cut facilities. Journal of Food Protection,
83, 1656–1666. https://doi.org/10.4315/jfp-20- 094
Sullivan, G., Zoellner, C., Wiedmann, M., & Ivanek, R. (2021). In Silico models for design and optimization of science-based Listeria environmental moni-
toring programs in fresh-cut produce facilities. Applied and Environmental Microbiology, 87, e0079921. https://doi.org/10.1128/aem.00799-21
Sun, Q. F., Cai, S. Z., Cheng, J. H., Zhang, Y., Lin, R. Q., Ye, Q. H., Xue, L., Zeng, H. Y., Lei, T., Zhang, S. H., Luo, X. T., Wu, K. G., Wu, Q. P., Chen, M. T., & Zhang,
J. M. (2021). Distribution, contamination routes, and seasonal influence of persistent Listeria monocytogenes in a commercial fresh Hypsizigus
marmoreus production facility. Food Control, 127, 8. https://doi.org/10.1016/j.foodcont.2021.108118
Tadee, P., Boonkhot, P., Pornruangwong, S., & Patchanee, P. (2015). Comparative phenotypic and genotypic characterization of Salmonella spp. in pig
farms and slaughterhouses in two provinces in northern Thailand. PLoS One, 10, e0116581. https://doi.org/10.1371/journal.pone.0116581
Teunis, P. F. M. (2022). Dose response for Salmonella Typhimurium and Enteritidis and other nontyphoid enteric salmonellae. Epidemics, 41, 100653.
https://doi.org/10.1016/j.epidem.2022.100653
Thomas, C., Daughtry, B., Padula, D., Jordan, D., Arzey, G., Davey, K., Holds, G., Slade, J., & Pointon, A. (2006). An egg: Salmonella quantitative risk assessment
model for the Australian egg industry. Australian Egg Corporation Limited.
Thomas, J., Govender, N., McCarthy, K., Erasmus, L., Doyle, T., Allam, M., Ismail, A., Ramalwa, N., Sekwadi, P., Ntshoe, G., Shonhiwa Nee Rakgantso, A.,
Essel, V., Tau, N., Smouse, S., Ngomane, H., Disenyeng, B., Page, N., Govender, N., Duse, A., & Blumberg, L. (2020). Outbreak of Listeriosis in South
Africa associated with processed meat. New England Journal of Medicine., 382, 632–643. https://doi.org/10.1056/NEJMoa1907462
Thomassen, G. M. B., Krych, L., Knochel, S., & Mehli, L. (2021). ON-rep-seq as a rapid and cost-effective alternative to whole-genome sequencing for
species-level identification and strain-level discrimination of Listeria monocytogenes contamination in a salmon processing plant. Microbiology, 10.
https://doi.org/10.1002/mbo3.1246
Tirloni, E., Bernardi, C., Pomilio, F., Torresi, M., De Santis, E. P. L., Scarano, C., & Stella, S. (2020). Occurrence of Listeria spp. and Listeria monocytogenes
isolated from PDO Taleggio production plants. Food, 9, 14. https://doi.org/10.3390/foods9111636
Tobin, H. M., Lele, S. R., Cutter, C. N., Anantheswaran, R. C., & LaBorde, L. F. (2020). Hot water sanitization of a commercial mushroom disk slicer to inacti-
vate Listeria monocytogenes. Food Control, 109, 10. https://doi.org/10.1016/j.foodcont.2019.106900
Torralbo, A., Borge, C., García-Bocanegra, I., Méric, G., Perea, A., & Carbonero, A. (2015). Higher resistance of campylobacter coli compared to campylo-
bacter jejuni at chicken slaughterhouse. Comparative Immunology, Microbiology and Infectious Diseases, 39, 47–52. https://doi.org/10.1016/j.cimid.
2015.02.003
Townsend, A., Strawn, L. K., Chapman, B. J., & Dunn, L. L. (2021). A systematic review of Listeria species and Listeria monocytogenes prevalence, per-
sistence, and diversity throughout the fresh produce supply chain. Food, 10, 28. https://doi.org/10.3390/foods10061427
Trinetta, V., Magossi, G., Allard, M. W., Tallent, S. M., Brown, E. W., & Lomonaco, S. (2020). Characterization of Salmonella enterica isolates from selected US
swine feed Mills by whole-genome sequencing. Foodborne Pathogens and Disease, 17, 126–136. https://doi.org/10.1089/fpd.2019.2701
Truchado, P., Gil, M. I., Querido-Ferreira, A. P., Capón, C. L., Álvarez-Ordoñez, A., & Allende, A. (2022). Frozen vegetable processing plants can harbour
diverse Listeria monocytogenes populations: Identification of critical operations by WGS. Food, 11. https://doi.org/10.3390/foods11111546
Tuytschaever, T., Raes, K., & Sampers, I. (2023). Listeria monocytogenes in food businesses: From persistence strategies to intervention/prevention strat-
egies-a review. Comprehensive Reviews in Food Science and Food Safety:41., 22, 3910–3950. https://doi.org/10.1111/1541-4337.13219
UFPA (United Fresh Produce Association). (2018). Guidance on Environmental Monitoring and Control of Listeria for the Fresh Produce Industry (2nd
ed.). 36 pp. pp. https://bfff.co.uk/wp-content/uploads/2018/09/BFFF- Guide-to-the-Management-of-Risk-of-Listeriosis-from-Frozen-Vegetables-
FINAL.pdf
Ung, A., Baidjoe, A. Y., Van Cauteren, D., Fawal, N., Fabre, L., Guerrisi, C., Danis, K., Morand, A., Donguy, M. P., Lucas, E., Rossignol, L., Lefevre, S.,
Vignaud, M. L., Cadel-Six, S., Lailler, R., Jourdan-da Silva, N., & Le Hello, S. (2019). Disentangling a complex nationwide Salmonella Dublin out-
break associated with raw-milk cheese consumption, France, 2015 to 2016. Eurosurveillance, 24, 32–41. https://doi.org/10.2807/1560-7917.Es.
2019.24.3.1700703
Unnerstad, H., Bannerman, E., Bille, J., Danielsson-Tham, M.-L., Tham, W., & Waak, E. (1996). Prolonged contamination of a dairy with Listeria monocyto-
genes. Proceedings of the food associated pathogens, Uppsala (Sweden), 6-8 May 1996.
Valkovska, L., Mališevs, A., Kovaļenko, K., Bērziņš, A., & Grantiņa-Ieviņa, L. (2021). Coxiella Burnetii DNA in Milk, Milk products, and fermented dairy prod-
ucts. Journal of Veterinary Research, 65, 441–447. https://doi.org/10.2478/jvetres-2021- 0055
van Hoek, A., de Jonge, R., van Overbeek, W. M., Bouw, E., Pielaat, A., Smid, J. H., Malorny, B., Junker, E., Lofstrom, C., Pedersen, K., Aarts, H. J. M., & Heres,
L. (2012). A quantitative approach towards a better understanding of the dynamics of Salmonella spp. in a pork slaughter-line. International Journal
of Food Microbiology, 153, 45–52. https://doi.org/10.1016/j.ijfoodmicro.2011.10.013
Vazquez-Boland, J. A., Wagner, M., & Scortti, M. (2020). Why are some Listeria monocytogenes genotypes More likely to cause invasive (brain, placental)
infection? MBio, 11, 7. https://doi.org/10.1128/mBio.03126-20
Veghova, A., Minarovicova, J., Korenova, J., Drahovska, H., & Kaclikova, E. (2017). Prevalence and tracing of persistent Listeria monocytogenes strains in
meat processing facility production chain. Journal of Food Safety, 37, 10. https://doi.org/10.1111/jfs.12315
Velhner, M., Milanov, D., & Kozoderovic, G. (2018). Salmonella spp. in poultry: A constant challenge and new insights. Journal of the Hellenic Veterinary
Medical Society, 69, 899–910. https://doi.org/10.12681/jhvms.18012
Verghese, B., Lok, M., Wen, J., Alessandria, V., Chen, Y., Kathariou, S., & Knabel, S. (2011). comK prophage junction fragments as markers for Listeria mono-
cytogenes genotypes unique to individual meat and poultry processing plants and a model for rapid niche-specific adaptation, biofilm formation,
and persistence. Applied and Environmental Microbiology, 77, 3279–3292. https://doi.org/10.1128/aem.00546-11
Verraes, C., Vlaemynck, G., Van Weyenberg, S., De Zutter, L., Daube, G., Sindic, M., Uyttendaele, M., & Herman, L. (2015). A review of the microbiological
hazards of dairy products made from raw milk. International Dairy Journal, 50, 32–44. https://doi.org/10.1016/j.idair yj.2015.05.011
Vestby, L. K., Moretro, T., Langsrud, S., Heir, E., & Nesse, L. L. (2009). Biofilm forming abilities of Salmonella are correlated with persistence in fish meal-
and feed factories. BMC Veterinary Research, 5, 6. https://doi.org/10.1186/1746- 6148-5-20
Viazis, S., Beal, J. K., Monahan, C., Lanier, W. A., Kreil, K. R., Melka, D. C., Boden, W. D., Dion, J. L., Miller, Z. A., Nguyen, T. A., Gieraltowski, L. B., & Zink, D. L.
(2015). Laboratory, environmental, and epidemiologic Investigation and regulatory enforcement actions in response to an outbreak of Salmonella
Bredeney infections linked to Peanut butter. Open Forum Infectious Diseases, 2, 8. https://doi.org/10.1093/ofid/ofv114
Vinueza-Burgos, C., Baquero, M., Medina, J., & De Zutter, L. (2019). Occurrence, genotypes and antimicrobial susceptibility of Salmonella collected from
the broiler production chain within an integrated poultry company. International Journal of Food Microbiology, 299, 1–7. https://doi.org/10.
1016/j.ijfoodmicro.2019.03.014
Viswanath, P., Murugesan, L., Knabel, S. J., Verghese, B., Chikthimmah, N., & LaBorde, L. F. (2013). Incidence of Listeria monocytogenes and Listeria spp. in
a small-scale mushroom production facility. Journal of Food Protection, 76, 608–615. https://doi.org/10.4315/0362-028x.Jfp-12-292
Vojkovska, H., Karpiskova, R., Oriešková, M., & Drahovska, H. (2015). Characterization of Cronobacter spp. isolated from food of plant origin and envi-
ronmental samples collected from farms and from supermarkets in The Czech Republic. International Journal of Food Microbiology, 217, 130–136.
https://doi.org/10.1016/j.ijfoodmicro.2015.10.017
Wahl, E., Rømma, S., & Granum, P. E. (2013). A Clostridium perfringens outbreak traced to temperature-abused beef stew, Norway, 2012. Euro Surveillance,
18(9), pii=20408.
Wang, C. L., Chen, B. Y., Hsu, C. T., Wu, R., & Chou, C. H. (2019). Salmonella contamination in ready-to-eat tilapia sashimi processing plants. Journal of Food
Wang, H. H., Shu, R. H., Zhao, Y., Zhang, Q. Q., Xu, X. L., & Zhou, G. H. (2014). Analysis of ERIC-PCR genomic polymorphism of Salmonella isolates from
chicken slaughter line. European Food Research and Technology, 239, 543–548. https://doi.org/10.1007/s00217- 014-2277-x
Wang, L., Forsythe, S. J., Yang, X., Fu, S., Man, C., & Jiang, Y. (2021). Invited review: Stress resistance of Cronobacter spp. affecting control of its growth
during food production. Journal of Dairy Science, 104, 11348–11367. https://doi.org/10.3168/jds.2021-20591
Wang, R. (2019). Biofilms and meat safety: A mini-review. Journal of Food Protection, 82, 120–127. https://doi.org/10.4315/0362-028x.Jfp-18-311
Weatherill, S. (2009). Report of the independent investigator into the 2008 listeriosis outbreak. Agriculture and Agri-Food Canada July, 181.
Wierup, M., & Kristoffersen, T. (2014). Prevention of Salmonella contamination of finished soybean meal used for animal feed by a Norwegian produc-
tion plant despite frequent Salmonella contamination of raw soy beans, 1994-2012. Acta Veterinaria Scandinavica, 56, 9. https://doi.org/10.1186/
s13028-014-0041-7
Wiktorczyk-Kapischke, N., Skowron, K., Grudlewska-Buda, K., Walecka-Zacharska, E., Korkus, J., & Gospodarek-Komkowska, E. (2021). Adaptive response
of Listeria monocytogenes to the stress factors in the food processing environment. Frontiers in Microbiology, 12, 18. https://doi.org/10.3389/fmicb.
2021.710085
Williamson, K., Pao, S., Dormedy, E., Phillips, T., Nikolich, G., & Li, L. (2018). Microbial evaluation of automated sorting systems in stone fruit packing-
houses during peach packing. International Journal of Food Microbiology, 285, 98–102. https://doi.org/10.1016/j.ijfoodmicro.2018.07.024
Wingstrand, A., Jensen, T., Sørensen, G., Gubbels, S., Sweden, L. J., Ethelberg, S., Bjerager, G. O., Kjeldsen, K., Torpdahl, M., & Kuhn, K. G. (2012). A proactive
attempt to control an outbreak of foodborne disease in humans after detection of Salmonella during surveillance and control of pork in Denmark.
In J. Hoorfar (Ed.), Case studies in food safety and authenticity (pp. 141–150). Woodhead Publishing.
Wulff, G., Gram, L., Ahrens, P., & Vogel, B. F. (2006). One group of genetically similar Listeria monocytogenes strains frequently dominates and persists in
several fish slaughter and smokehouses. Applied and Environmental Microbiology, 72, 4313–4322. https://doi.org/10.1128/AEM.02288- 05
Xi, M., Feng, Y., Li, Q., Yang, Q., Zhang, B., Li, G., Shi, C., & Xia, X. (2015). Prevalence, distribution, and diversity of Escherichia coli in plants manufacturing
goat milk powder in Shaanxi, China. Journal of Dairy Science, 98, 2260–2267. https://doi.org/10.3168/jds.2014-9057
Yan, Q., Power, K., Cooney, S., Fox, E., Gopinath, G., Grim, C., Tall, B., McCusker, M., & Fanning, S. (2013). Complete genome sequence and phenotype
microarray analysis of Cronobacter sakazakii SP291: A persistent isolate cultured from a powdered infant formula production facility. Frontiers in
Microbiology, 4. https://doi.org/10.3389/fmicb.2013.00256
Yan, Q. Q., Wang, J., Gangiredla, J., Cao, Y., Martins, M., Gopinath, G. R., Stephan, R., Lampel, K., Tall, B. D., & Fanning, S. (2015). Comparative genotypic
and phenotypic analysis of Cronobacter species cultured from four powdered infant formula production facilities: Indication of Pathoadaptation
along the food chain. Applied and Environmental Microbiology, 81, 4388–4402. https://doi.org/10.1128/aem.00359-15
Yang, C., Li, Y., Jiang, M., Wang, L., Jiang, Y., Hu, L., Shi, X., He, L., Cai, R., Wu, S., Qiu, Y., Lu, L., Zuo, L., Chen, Q., Wu, Y., Martinez-Urtaza, J., Wan, C., Yang,
R., Cui, Y., & Hu, Q. (2022). Outbreak dynamics of foodborne pathogen Vibrio parahaemolyticus over a seventeen year period implies hidden res-
ervoirs. Nature Microbiology, 7, 1221–1229. https://doi.org/10.1038/s41564- 022- 01182- 0
Yang, H., Hoffmann, M., Allard, M. W., Brown, E. W., & Chen, Y. (2020). Microevolution and gain or loss of Mobile genetic elements of outbreak-related
Listeria monocytogenes in food processing environments identified by whole genome sequencing analysis. Frontiers in Microbiology, 11. https://
doi.org/10.3389/fmicb.2020.00866
Yang, Y. S., Hoe, Y. W., Zheng, Q. W., Chung, H. J., & Yuk, H. G. (2017). Biofilm formation by Salmonella Enteritidis in a simulated liquid egg processing
environment and its sensitivity to chlorine and hot water treatment. Food Control, 73, 595–600. https://doi.org/10.1016/j.foodcont.2016.09.008
Zastempowska, E., Grajewski, J., & Twarużek, M. (2016). Food-borne pathogens and contaminants in raw Milk – A review. Annals of Animal Science, 16,
623–639. https://doi.org/10.1515/aoas-2015-0089
Zeng, H., De Reu, K., Gabriel, S., Mattheus, W., De Zutter, L., & Rasschaert, G. (2021). Salmonella prevalence and persistence in industrialized poultry
slaughterhouses. Poultry Science, 100, 10. https://doi.org/10.1016/j.psj.2021.01.014
Zhang, H. Z., Wang, J., Chang, Z. Y., Liu, X., Chen, W. J., Yu, Y., Wang, X. G., Dong, Q. L., Ye, Y. L., & Zhang, X. (2021). Listeria monocytogenes contamination
characteristics in two ready-to-eat meat plants from 2019 to 2020 in Shanghai. Frontiers in Microbiology, 12. https://doi.org/10.3389/fmicb.2021.
729114
Zhao, T., Podtburg, T. C., Zhao, P., Chen, D., Baker, D. A., Cords, B., & Doyle, M. P. (2013). Reduction by competitive bacteria of Listeria monocytogenes in
biofilms and Listeria bacteria in floor drains in a ready-to-eat poultry processing plant. Journal of Food Protection, 76, 601–607. https://doi.org/10.
4315/0362-028x.Jfp-12-323
Zhuang, K. J., Li, H. F., Zhang, Z. W., Wu, S., Zhang, Y. S., Fox, E. M., Man, C. X., & Jiang, Y. J. (2019). Typing and evaluating heat resistance of Bacillus cereus
sensu stricto isolated from the processing environment of powdered infant formula. Journal of Dairy Science, 102, 7781–7793. https://doi.org/10.
3168/jds.2019-16392
Zoellner, C., Ceres, K., Ghezzi-Kopel, K., Wiedmann, M., & Ivanek, R. (2018). Design elements of Listeria environmental monitoring programs in food
processing facilities: A scoping review of research and guidance materials. Comprehensive Reviews in Food Science and Food Safety, 17, 1156–1171.
https://doi.org/10.1111/1541-4337.12366
Zoellner, C., Jennings, R., Wiedmann, M., & Ivanek, R. (2019). EnABLe: An agent-based model to understand Listeria dynamics in food processing facili-
ties. Scientific Reports, 9, 14. https://doi.org/10.1038/s41598- 018-36654-z
Zuber, I., Lakicevic, B., Pietzka, A., Milanov, D., Djordjevic, V., Karabasil, N., Teodorovic, V., Ruppitsch, W., & Dimitrijevic, M. (2019). Molecular characteriza-
tion of Listeria monocytogenes isolates from a small-scale meat processor in Montenegro, 2011–2014. Food Microbiology, 79, 116–122. https://doi.
org/10.1016/j.fm.2018.12.005
Zwirzitz, B., Wetzels, S. U., Dixon, E. D., Fleischmann, S., Selberherr, E., Thalguter, S., Quijada, N. M., Dzieciol, M., Wagner, M., & Stessl, B. (2021). Co-
occurrence of Listeria spp. and spoilage associated microbiota during meat processing due to cross-contamination events. Frontiers in Microbiology,
12, 14. https://doi.org/10.3389/fmicb.2021.632935
S U P P O R T I N G I N F O R M AT I O N
Additional supporting information can be found online in the Supporting Information section at the end of this article.
How to cite this article: EFSA BIOHAZ Panel (EFSA Panel on Biological Hazards), Koutsoumanis, K., Allende, A.,
Bolton, D., Bover-Cid, S., Chemaly, M., De Cesare, A., Herman, L., Hilbert, F., Lindqvist, R., Nauta, M., Nonno, R., Peixe,
L., Ru, G., Simmons, M., Skandamis, P., Suffredini, E., Fox, E., Gosling, R. B. … Alvarez-Ordóñez, A. (2024). Persistence
of microbiological hazards in food and feed production and processing environments. EFSA Journal, 22(1), e8521.
https://doi.org/10.2903/j.efsa.2024.8521
APPE N D IX A
Literature searches, screening and data extraction
A.1 | Literature searches for the assessment of the persistence of microbiological hazards in food production
and processing environments
The search was performed on 17 October 2022 in the Web of Science™ Core Collection (1975–present) to retrieve papers,
review papers, book chapters and books related to the persistence of microbiological hazards in food production and
processing environments. The search string can be found in Annex A (protocol; Table A3). It was further restricted based on
the record characteristics: only in English, post 2010 for the publication year and document types Article or Review Article
or Book Chapters.
The EndNote file was transferred into DistillerSR® Web-Based Systematic Review Software (Evidence Partners, Ottawa,
Canada) for the selection procedure.
The screening was undertaken in three steps:
(1) Screening of titles (for records classified as primary research studies only)
Screening of title was done by an EFSA staff member to identify obviously irrelevant studies, to be excluded from the
assessment, and potentially relevant or unclear studies, to be moved to title/abstract (Ti/Ab) screening.
(2) Screening of titles and abstracts
Screening of titles and abstracts was done in duplicate to identify records containing:
• irrelevant studies, to be excluded from the assessment; and

• potentially relevant or unclear studies, to be moved to full-text screening.
This was done by EFSA staff and Working Group members in duplicate (i.e. each Ti/Ab was screened for relevance by two
reviewers). If there were doubts or divergences between the two reviewers, it was solved by discussion.
A pilot sample of 25 records of primary research studies were screened by the two EFSA staff members and six Working
Group members involved in the screening. Disagreements were discussed, which helped to clarify the questions.
The question posed was:
Does the study/review include one of the below:
• sampling in a FoPE (one or more of this list: meat processing (including slaughterhouses and processing plants), fish and sea-
food processing, dairy processing, egg and egg products processing, fruit and vegetable processing, including Controlled
Environment Agriculture (CEA)/production through indoor hydroponic operations and low moisture food (LMF) processing)
with analysis of a hazard (one or more of this list: Salmonella spp., Listeria monocytogenes, pathogenic Escherichia coli,
Campylobacter jejuni, Campylobacter coli, Clostridium botulinum, Clostridium perfringens, Bacillus cereus, Staphylococcus
aureus, Vibrio parahaemolyticus, Cronobacter sakazakii) or subtypes
• features of strains associated with persistent events
• risk factors at facility level in a food processing environment related to persistence
• interventions at facility level in a food processing environment related to persistence
• Interventions at laboratory level related to persistence, for example to remove biofilms
• a study related to an outbreak in a food processing environment related to persistence.
If yes, or unclear, it was asked to classify the study.
The study/review includes:
▫ Sampling in a food processing environment with analysis of a hazard or subtypes

▫ Features of strains associated with persistent events
▫ Risk factors at facility level in a food processing environment related to persistence
▫ Interventions at facility level in a food processing environment related to persistence
▫ A study related to an outbreak in a food processing environment related to persistence
▫ Interventions at laboratory level related to persistence, for example to remove biofilms
Does the study concern one of the below sectors/processing environments?
▫ meat processing (including slaughterhouses and processing plants)

▫ fish and seafood processing
▫ dairy processing
▫ egg and egg products processing
▫ fruit and vegetable processing, including Controlled Environment Agriculture (CEA)/production through indoor hydroponic
operation
▫ low moisture food (LMF) processing
▫ none of the above
▫ unclear
Does the study concern one of the below hazards?
▫ Bacillus cereus
▫ Campylobacter jejuni or Campylobacter coli
▫ Cronobacter sakazakii
▫ Clostridium botulinum or Clostridium perfringens
▫ Listeria monocytogenes
▫ pathogenic Escherichia coli
▫ Salmonella spp.
▫ Staphylococcus aureus
▫ Vibrio parahaemolyticus
▫ none of the above
▫ unclear
(3) Screening of full-text documents
Screening of full-text documents was done in single by one of the Working Group members or EFSA staff in two parts:
Part I: to further identify records to be excluded based on criteria related to report characteristics (e.g. full text not avail-
able, not in English; other than primary research study or review) and to classify them as primary research study or review.
Part II: to confirm the relevance of the study based on the full text. The question posed was: Does the study/review
include one of the below [for one or more of this list: Salmonella spp., Listeria monocytogenes, pathogenic Escherichia coli,
Campylobacter jejuni, Campylobacter coli, Clostridium botulinum, Clostridium perfringens, Bacillus cereus, Staphylococcus au-
reus, Vibrio parahaemolyticus, Cronobacter sakazakii] Pls select.
▫ Repeated sampling with the same subtype found at different sampling occasions in a food processing environment with anal-
ysis of one of the hazards
▫ Risk factors at facility level in a food processing environment related to persistence
▫ A study related to an outbreak in a food processing environment related to persistence
▫ Interventions at laboratory level related to persistence, for example to remove biofilms (if this selected: study to be flagged for
potential later use)
▫ None of the above (if this selected: study to be excluded)
Then, two questions were posed to further classify the studies.
Which of below sectors/processing environments?
▫ meat processing (including slaughterhouses and processing plants)

▫ fish and seafood processing
▫ dairy processing
▫ egg and egg products processing
▫ fruit and vegetable processing, including Controlled Environment Agriculture (CEA)/production through indoor hydroponic
operation
▫ low moisture food (LMF) processing
▫ none of the above (in case of a laboratory study)
Which of the below hazards does the study include?
▫ Bacillus cereus
▫ Campylobacter jejuni or Campylobacter coli
▫ Cronobacter sakazakii
▫ Clostridium botulinum or Clostridium perfringens
▫ Listeria monocytogenes
▫ pathogenic Escherichia coli
▫ Salmonella spp.
▫ Staphylococcus aureus
▫ Vibrio parahaemolyticus
Data extraction
Data were extracted from the eligible full-text documents when the study concerned the most relevant bacterial food
safety hazards associated with persistence in the FFPE of the specific food production and processing sectors; i.e. S. enterica
in the food animal production, meat processing, eggs and egg processing and LMF sectors; and L. monocytogenes in the
meat processing, fish and seafood processing, dairy and fruit and vegetables processing sectors; and C. sakazakii in the
LMF sector.
A pre-defined data extraction form in DistillerSR was used and data extraction was performed by a Working Group
member or EFSA staff. Data extraction consisted of gathering information about the country where sampling took place,
the sector and plant within the sector, the reason for sampling, the hazard level 1 (i.e. Salmonella, Listeria or Cronobacter),
the hazard level 2 (i.e. serovar of Salmonella, L. monocytogenes or C. sakazakii), the hazard level 3 (subtype), the location of
the persistence (i.e. non-food and/or food contact surface) with further details, if available, the typing method(s) used, the
features of persistence investigated, the risk factors for the persistence, the interventions, the number of sampling events,
the sampling period and period of persistence.
The reviews that were eligible at full-text level and the extracted data from the primary research studies have been made
available through the Knowledge Junction under https://doi.org/10.5281/zenodo.10299549.
A.2 | Literature searches for the assessment of the persistence of microbiological hazards in feed production
and processing environments
The search was performed on 17 January 2023 in the Web of Science™ Core Collection (1975–present) to retrieve papers,
review papers, book chapters and books related to the persistence of microbiological hazards in feed production and pro-
cessing environments. The search string can be found in Annex A (protocol; Table A3). It was slightly revised to ensure more
coverage by adding an additional string using a search based on the use of Salmonella and feed in the title. The search was
restricted based on the record characteristics: only in English, post 2010 (for the broader title search only) for the publica-
tion year and document types Article or Review Article or Book Chapters.
The EndNote file was transferred into DistillerSR® Web-Based Systematic Review Software (Evidence Partners, Ottawa,
Canada) for the selection procedure.
The screening was undertaken in three steps:
(1) Screening of titles
Screening of titles was done by an EFSA staff member to identify obviously irrelevant studies, to be excluded from the
assessment, and potentially relevant or unclear studies, to be moved to Ti/Ab screening.
(2) Screening of titles and abstracts
Screening of titles and abstracts was done in duplicate to identify records containing:
• irrelevant studies, to be excluded from the assessment;

• potentially relevant or unclear studies, to be moved to full-text screening.
This was done by EFSA staff and a Working Group member in duplicate. Doubts or divergences between the two review-
ers were solved by discussion.
The question posed was: Does the study/review address the persistence of Salmonella spp. in feed processing environments,
considering one of the below?
▫ repeated sampling with the same subtype found at different sampling occasions in a feed processing environment with analy-
sis of Salmonella spp.
▫ features of strains associated with persistent events

▫ risk factors at facility level in a feed processing environment related to persistence
▫ interventions at facility level in a feed processing environment related to persistence
(3) Screening of full-text documents
Screening of full-text documents was done in single by a Working Group member to confirm the relevance of the study
based on the full text. The question posed was: Does the study/review address the persistence of Salmonella spp. in feed pro-
cessing environments, including one of the below? Pls select.
▫ Repeated sampling with the same subtype found at different sampling occasions in a feed processing environment with anal-
ysis of Salmonella spp.
▫ Risk factors at facility level in a feed processing environment related to persistence
▫ Interventions at facility level in a feed processing environment related to persistence
▫ None of the above (if this selected: study to be excluded)
Data extraction
Data were extracted from the eligible full-text documents when the study concerned the most relevant bacterial food
safety hazard associated with persistence in the FFPE of the feed production and processing sector, i.e. S. enterica. A pre-
defined data extraction form in DistillerSR was used and data extraction was performed by a Working Group member.
The reviews that were eligible at full-text level and the extracted data from the primary research studies have been made
available through the Knowledge Junction under https://doi.org/10.5281/zenodo.10299549.
Table 7 Figure 21 Figure 25,

APPE N D IX B
Uncertainty analysis
T A B L E B .1 Potential sources of uncertainty linked to specific assessment questions (AQs) in the assessment of the persistence of microbiological
hazards in food and feed production and processing environments of the various food and feed production and processing sectors in the EU/EEA.
AQ Source or location of the uncertainty Nature or cause of the uncertainty

AQ1 Food sectors: Incomplete information in the data Especially for the ‘strong evidence’ FBO from the EFSA zoonoses database and
derived from the various sources to a lesser extent for the ROA and JNS, the link between cases/outbreaks and
the incriminated food is not always established or reported and sometimes
the general description of the food does not allow to accurately understand
the type of food incriminated. The RASFF database is not an epidemiological
surveillance system as this system is primarily a communication facility
enabling many food safety risks to be averted to prevent further spread
of the risk over Europe. As it provides some understanding of the types of
hazards typically detected in particular foods, it was used to complement the
FBO data
AQ1 Feed for food animal production sector: In the feed sector, only the RASFF data was used as FBO are only very rarely
Incomplete information in the notifications linked to the feed sector. Also, the former scientific opinion of the BIOHAZ
extracted from the RASFF database panel on Microbiological Risk Assessment in feedingstuffs for food-
producing animals for both public health and animal health (EFSA BIOHAZ
Panel, 2008) was consulted
AQ1 Insufficient tracing of the contamination source in In most outbreak investigations there is a lack of clear information on whether
outbreak investigations the contamination source is the FoPE and whether the involved strains are
persisting in the FoPE
AQ1 Lack of studies addressing the FoPE for some Persistent clones may be not detected due to lack of field data or to inefficient
particular hazards and insufficient or sampling (using a low frequency of sampling, an inefficient sampling
inefficient sampling of FoPE in environmental method, not sampling relevant areas or sample types, absence of
monitoring studies longitudinal sampling, etc). This may be particularly the case for those
hazards for which studies addressing the FoPE are very scarce (e.g.
Campylobacter, S. aureus, B. cereus s. l.)
AQ1 Origin of isolates It is difficult to evaluate if re-isolation of the same clone over time is due to
persistence or reintroduction (or both)
AQ2 Use of typing methods which provide information Only persistent clones identified through some subtyping methods (such
that is not interoperable or does not follow as serotypes, epidemic clones, lineages, clades, spa types, panC types,
universally harmonised terminology for the clonal complexes, MLST types, wgMLST types and cgMLST types) can be
subtypes unequivocally assigned to known subtypes. As a result, some relevant
subtypes may be missed in the assessment
AQ2 Use of typing methods with low resolution. In some studies, the use of typing methods with low resolution makes it
Typing method does not provide enough uncertain whether isolates belong to the same clone or several clones. This is
detail to confirm persistence of the same particularly relevant for Salmonella, where serotyping is the typing method
strain over time frequently used. Serotyping methods used for typing of Salmonella lead to
the assumption that repeated isolation of the same serotype over time is
equivalent to persistence
AQ2 Insufficient characterisation of persistent strains Many of the studies that report isolation of the same subtype along time
in the FFPE did not characterise the persistent strain phenotypically or
genotypically through WGS to identify features associated with persistence.
As a result, there is no sufficient knowledge on the features of some of the
subtypes identified in persistence events
AQ2 Use of presumed non-persistent strains as control Some of the studies characterising features of persistent strains use presumed
strains in experiments trying to identify strain non-persistent strains as control. Sometimes they belong to other CC or
features associated with persistence in the serotypes and there is uncertainty on whether those strains could also persist
FFPE under different circumstances or following different sampling approaches
AQ2 Lack of harmonisation among studies and of Different studies assess diverse genotypic and phenotypic features, without an
relevant model systems to test if particular exhaustive set of features shared between studies to enable the identification
features are involved in persistence of features that are indispensable for persistence. It is difficult to assess
whether features that seem to be statistically correlated to persistent
subtypes play a role in persistence in the FFPE, as pathogenic strains cannot
be added to the FFPE to test hypotheses and it is uncertain whether studies
performed in laboratory settings reflect the situation in the FFPE satisfactorily
AQ2 Biased representation of subtypes in WGS Databases such as the NCBI Pathogen Detection database are biased in the sense
databases that they contain an uneven number of genomes from different sources,
countries, subtypes, etc, reflecting particular sequencing efforts carried
out by different instances to investigate specific contamination or outbreak
events. Thus, the number of genomes available for the different subtypes will
not necessarily reflect the relevance of that subtype
(Continues)
T A B L E B .1 (Continued)
AQ Source or location of the uncertainty Nature or cause of the uncertainty
AQ3 Harbourage site (or niche) of the persistent In many of the studies reporting repeated isolation along time of the same
strain(s) subtype the actual niche (harbourage site) of the persistent strain was
unclear, either because it was not investigated in detail or because the
persistent strain was isolated from multiple sources within the processing
plant. Hence, niches were not always identified. In addition, repeated
isolation from a FFPE such as conveyers, drains, floors may not always indicate
persistence on that specific surface as the original niche for persistence may
be another niche shedding or leaking bacteria to that surface
AQ4 Lack of studies specifically designed for the Many of the identified risk factors are reported in studies reflecting personal
identification of persistence risk factors opinion or expertise of the authors or discussion around available literature,
but there is scarcity of studies specifically designed to identify risk factors for
persistence
AQ4 Risk factors for persistence or cross-contamination It is sometimes difficult to understand/interpret from the available literature
whether some of the risk factors for a particular hazard are true risk factors for
persistence of such hazard or just risk factors for the entry of the pathogen in
the facility of for the cross-contamination of the end product
AQ5 Scarcity of studies testing interventions in While there is detailed information in the literature testing a wide range of
industry settings interventions as alternatives for the control of biofilms of hazards in lab
settings, only a few studies report results on testing interventions to address
persistence of hazards in industry settings
AQ5 Hazards targeted in the interventions Interventions tested at industry level sometimes follow indicator microorganisms
(e.g. Listeria spp.) instead of a particular hazard. In other occasions, the effect
of the intervention is assessed by monitoring changes in the incidence or
prevalence of the hazard, but with no further characterisation of the hazard.
Therefore, it is unclear whether the intervention is capable of preventing
or controlling persistence. Interventions reducing the numbers of indicator
microorganisms or the incidence/prevalence of a hazard in the industry are
expected to also prevent/control the persistence of hazards
AQ6 Deterministic versus stochastic modelling In deterministic models we omit stochastic processes and leave out
heterogeneity in processing. Neither do we include uncertainty in model
parameter estimates. Models including all these aspects would become
highly complex. Studying their perspectives was beyond the scope of this
Scientific Opinion
AQ6 Options for risk assessment approaches Many risk assessment approaches exist. We only explored bottom-up food chain
QMRA and did not apply methods like Bayesian modelling, machine learning
or agent-based modelling. These may have perspectives that have not been
covered
Abbreviations: AQ, assessment question; CC, clonal complex; cgMLST, core genome multi-locus sequence type; C&D, cleaning and disinfection; FBO, foodborne outbreak;
FFPE, food and feed processing environment; FoPE, food processing environment; JNS, joint notification summary; MLST, multi-locus sequence type, NCBI, National
Center for Biotechnology Information; NFCS, non-food contact surface; QMRA, quantitative microbiological risk assessment; RASFF, Rapid Alert System for Food and Feed;
ROA, rapid outbreak assessments; wgMLST, whole genome multi-locus sequence type; WGS, whole genome sequencing.
APPE N D IX C
Bacterial pathogens most relevant for public health in the food and feed production and processing sectors
C.1 | Feed for food animal production sector
Considering the notifications for feed from the RASSF database since 2010, more than 99.5% of the 1511 notifications
were associated with Salmonella enterica subsp. enterica (1504 notifications). Most notifications were for the serovars
S. Senftenberg (N = 138), S. Mbandaka (N = 124), S. Agona (N = 116), S. Tennessee (N = 87) and S. Infantis (N = 58). There were
three notifications for Clostridium perfringens in fish feed, two for Brucella in frozen hare by-products, one for Bacillus cereus
s. l. in a bacterial feed additive for pigs produced by Corynebacterium glutamicum, and one for the possible presence of
Bacillus anthracis in beef bones for feed.
In the scientific opinion of the BIOHAZ panel on Microbiological Risk Assessment in feedingstuffs for food-producing
animals for both public health and animal health, the panel identified Salmonella spp. as the major hazard for bacterial
contamination of animal feed. L. monocytogenes, E. coli O157:H7 and Clostridium spp. are other hazards for which feed was
regarded a far less important source (EFSA BIOHAZ Panel, 2008).
C.2 | Meat sector, excluding LMF sector
Strong evidence FBO data at EU level since 2010 showed that the most relevant pathogen involved in outbreaks related
with meat and meat products was S. enterica subsp. enterica (567 of 1050 outbreaks reported, involving 10,865 human
cases). S. Enteritidis was the most common serovar identified, causing 230 of the outbreaks, with 3298 human cases and 5
deaths. This was followed by S. Typhimurium, which caused 149 outbreaks, involving 3467 cases and 6 deaths. Pork meat,
broiler meat and meat and meat products were the most relevant vehicles of infection. Clostridium spp. was the second
pathogen in importance (196 of 1050 outbreaks with 6160 cases involved), with bovine meat, pork meat and broiler meat
most represented in outbreaks. Of these, C. perfringens predominated, causing 168 of the outbreaks with 6083 cases. In
addition, all 9 deaths were due to C. perfringens. C. botulinum caused all other outbreaks (28), with 77 cases involved. The
third pathogen was Campylobacter spp. (177 of 1050 outbreaks, involving 7936 cases and 2 deaths), with broiler meat
products as the main vehicle of infection. The species of Campylobacter was not reported for most of the outbreaks (111
of the 177 outbreaks); when available, C. jejuni was the most common (62 outbreaks, 6201 cases and 1 death). Other
hazards associated with outbreaks were B. cereus s. l. (43 outbreaks), pathogenic E. coli (26 outbreaks), L. monocytogenes
(18 outbreaks), S. aureus (14 outbreaks, with an additional 2 Staphylococcus spp. outbreaks not attributed to a species),
Yersinia enterocolitica (5 outbreaks), Shigella spp. (2 outbreaks) and Francisella (1 outbreak). S. enterica subsp. enterica was
responsible for the highest number of hospitalisations (2177 out of 2834 hospitalisations including all bacterial hazards)
while L. monocytogenes caused the highest number of deaths related to meat and meat products consumption (37 out of
67 deaths including all bacterial hazards).
Five out of the seven multi-country outbreaks published as ROAs since 2012 and linked to meat were caused by S. en-
terica subsp. enterica. One outbreak of S. Stanley was linked to turkey meat consumption (EFSA and ECDC, 2012; ECDC
and EFSA, 2014c), another one of S. Enteriditis was linked to poultry products (ECDC and EFSA, 2021c), one outbreak of
monophasic S. Typhimurium was linked to meat products (ECDC and EFSA, 2014a), one of S. Mbandaka was likely linked
to RTE chicken products (ECDC and EFSA, 2022c) and one of S. Virchow was likely linked to kebab meat products contain-
ing chicken meat (ECDC and EFSA, 2023b). L. monocytogenes was involved in one multi-country outbreak linked to RTE
meat products (ECDC and EFSA, 2019b), while a B. anthracis outbreak was linked to bovine meat contamination (EFSA and
ECDC, 2015).
All four multi-country outbreaks reported as a JNS were caused by S. enterica subsp. enterica. An outbreak of S. Bredeney
was caused by a chilled cooked pork preparation, while an outbreak of S. Typhimurium was caused by marinated pork and
minced beef (European Commission, 2020). Another outbreak was caused by S. Agona in kebab meat and S. Enteriditis was
the serotype involved in the last multi-country outbreak possibly linked to poultry products (European Commission, 2021).
S. enterica subsp. enterica has been the main bacterial pathogen involved in the RASFF notifications for meat and meat
products with a total of 4644 notifications (79.1%). The top three serovars implicated were S. Enteritidis (795 notifications,
comprising 22% of S. enterica subsp. enterica notifications, 17% of total RASFF notifications), followed by S. Infantis (364
notifications, 10% of Salmonella notifications and 8% of total RASFF notifications) and S. Typhimurium (333 notifications,
9% of Salmonella notifications and 7% of total RASFF notifications). Pathogenic E. coli was notified 460 times (9.9%).
L. monocytogenes was the third pathogen with a total of 388 notifications (8.4%). Campylobacter spp. was also relevant
with 103 notifications (2.2%). Of these, C. jejuni was most common (52 notifications), followed by pathogenic C. coli
(18 notifications). This was predominantly associated with poultry meat and poultry meat products. Other pathogens
reported with less than 1% of notifications were Clostridium spp., Staphylococcus spp., possible presence of B. anthracis,
Y. enterocolitica, Mycobacterium tuberculosis and Brucella spp.
Considering outbreak data from outside the EU, 24 of 39 confirmed or suspected reported outbreaks from USA reported
by CDC from 2010 to 2022 from meat and meat products were caused by S. enterica subsp. enterica. Pathogenic E. coli
was involved in 11 outbreaks and four outbreaks were caused by L. monocytogenes; three of them related with RTE meat
products consumption. Six outbreak reports were retrieved from Health Canada in the period from 2015 to 2022, with
S. enterica subsp. enterica involved in five of those linked to the consumption of chicken meat and L. monocytogenes in a
single outbreak connected with cooked diced chicken.
The particular relevance of the pathogens more frequently involved in outbreaks linked to meat and meat products is
further supported by previous studies and literature analyses (Lianou et al., 2023; Sofos, 2008).
C.3 | Fish and seafood sector, excluding LMF sector
Strong evidence FBO data at EU level since 2010 showed that outbreaks associated with fish and seafood caused by bac-
teria were associated with various serovars of S. enterica subsp. enterica (47 of 115 outbreaks of which 11 with S. Enteritidis;
involving 1657 human cases), B. cereus s. l. (20 outbreaks with 159 cases), L. monocytogenes (19 outbreaks with 145 cases),
C. botulinum or C. perfringens (14 outbreaks with 313 cases), Vibrio parahaemolyticus, all in crustaceans, shellfish, mol-
luscs and products thereof (7 outbreaks with 127 cases), pathogenic E. coli (3 outbreaks with 57 cases), Campylobacter
(3 outbreaks with 43 cases), Enterococcus (1 outbreak with 3 cases) and S. aureus (1 outbreak with 2 cases). S. enterica
subsp. enterica and L. monocytogenes caused the highest number of hospitalisations (102 and 90, respectively, of 265
hospitalisations including all bacterial hazards) and L. monocytogenes the majority of the deaths (14 of 18 deaths includ-
ing all bacterial hazards).
Five of the six multi-country outbreaks reported either as ROA (2012–2022) or JNS (2019–2021) were caused by L. monocy-
togenes and linked to cold smoked trout and cold smoked/marinated salmon (EFSA and ECDC, 2018a; ECDC and EFSA, 2019a)
(European Commission, 2019, 2021) and one outbreak was caused by C. botulinum (botulism) and linked to dried and salted
roach (EFSA and ECDC, 2016c).
The majority (64.5%) of the 864 RASFF notifications on bacterial pathogens in fish and seafood products were associ-
ated with L. monocytogenes (557 notifications, mostly fish and fish products; in particular, cold smoked salmon products),
S. enterica subsp. enterica (217, in a wide range of fish and seafood products) or Vibrio (73, mostly crustaceans and products
thereof; shrimp products).
Regarding outbreak data from outside the EU, of the eight confirmed or suspected reported outbreaks from bacteria
from USA reported by CDC from 2012 to 2022 from fish and seafood products, five were caused by S. enterica subsp. en-
terica (raw tuna, cooked shrimp, raw and cooked seafood) and two by V. parahaemolyticus (raw shellfish and fresh crab
meat). Health Canada reported two outbreaks of V. parahaemolyticus caused by (raw) shellfish in the period 2014–2022,
while FSANZ published a notification on V. parahaemolyticus and raw Pacific oysters.
C.4 | Dairy sector, excluding LMF sector
Since 2010, 279 strong evidence FBOs associated with milk and milk products with 3053 human cases have been docu-
mented. The bacterial hazards most often associated with these outbreaks in the dairy sector in the EU since 2010 were:
S. enterica subsp. enterica (N = 128), Campylobacter spp. (N = 89), pathogenic E. coli (N = 25), S. aureus (N = 19), B. cereus
s. l. (N = 7) and L. monocytogenes (N = 5). Consumption of milk (N = 237) and cheese (N = 287) contaminated with zo-
onotic pathogens resulted in 610 hospitalisations. Among these reported outbreaks, patients after the consumption of
cheese suffered of infections with S. enterica subsp. enterica (N = 144), S. aureus (N = 57) and L. monocytogenes (N = 51).
Individuals infected with L. monocytogenes (N = 14), pathogenic E. coli (N = 3) or S. enterica subsp. enterica (N = 3) through
consumption of milk and milk products, died as a result of the disease. Serovars Typhimurium, Enteritidis and Newport,
of S. enterica subsp. enterica, were linked to strong evidence FBOs transmitted by cheese (N = 88), followed by S. aureus
(N = 14), pathogenic E. coli (N = 10) and L. monocytogenes (N = 5). FBOs linked to milk consumption were mainly caused by
Campylobacter spp. (N = 81), S. enterica subsp. enterica (12), pathogenic E. coli (N = 10) and B. cereus s. l. (N = 4). In the
category dairy products other than cheese, S. enterica subsp. enterica (N = 28), Campylobacter spp. (N = 5), pathogenic
E. coli (N = 4) and S. aureus (N = 3) were involved in the strong-evidenced FBOs. C. perfringens was linked to two FBOs
and Y. pseudotuberculosis and Shigella flexneri to one FBO each transmitted by raw milk, dairy products and cheeses.
In RASFF notifications since 2010, 95% (N = 487/512) of those reported in the dairy products sector were associated with
cheese. In the cheese category, most notifications for pathogenic bacteria were associated with L. monocytogenes (66.3%),
followed by pathogenic E. coli (19.3%) and S. enterica subsp. enterica (12.1%). Notifications linked to bacterial contamination
in milk were linked to L. monocytogenes (41.6%) and B. cereus s. l. (33.3%). Recent notifications in the RASFF portal (May and
August 2022) are linked to FBOs suspected to be caused by S. Dublin in chilled raw milk cheese from France and L. monocy-
togenes ST155 in Asiago Pressato cheese, respectively.
A STEC O26 multi-country outbreak transmitted by cheese was reported in a ROA report (EFSA and ECDC, 2016b). It af-
fected 25 humans with 19 haemolytic uraemic syndrome (HUS) cases with a lethal outcome. A JNS was related to S. Newport
cases, potentially associated with French goats' cheese (European Commission, 2019). FBOs published in Eurosurveillance
included a L. monocytogenes outbreak in cheese in 2015 (Magalhaes et al., 2015), and infections by STEC O26:H11 in children
(Germinario et al., 2016) and by S. Dublin both after bovine raw milk products consumption (Ung et al., 2019).
CDC reported 13 outbreaks linked to cheeses (2006–2022). The majority were caused by L. monocytogenes (n = 10), and
others were linked to STEC (N = 1) and S. enterica subsp. enterica (N = 2) transmitted by fresh and soft cheese. Raw milk
was the outbreak vehicle in a single outbreak linked to L. monocytogenes. FSANZ reported on a single outbreak linked to
L. monocytogenes in soft cheeses.
In the U.S., multistate outbreaks with L. monocytogenes in pregnant Hispanic women have been linked to Mexican fresh
cheese made from pasteurised milk for decades (Jackson et al., 2011; Palacios et al., 2022). Recently, multistate outbreaks linked
to raw milk contaminated by L. monocytogenes have been reported by CDC (Nichols et al., 2019; Sebastianski et al., 2022).
Other documented multi-country FBOs, as the Quargel outbreak in Austria, were more often transmitted by raw milk or
pasteurised cheeses, either by cross-contamination or recontamination events (Desai et al., 2019; Fretz et al., 2010; Gould
et al., 2014). In Europe, most dairy-related outbreaks are caused by surface-ripened cheeses and raw milk cheeses made by
small-producers (Filipello et al., 2020; Johler et al., 2015; Napoleoni et al., 2021; Robinson et al., 2020; Ung et al., 2019)
The EFSA Opinion related to the consumption of raw drinking milk (EFSA BIOHAZ Panel, 2015) already shows the range of
pathogens that can accumulate in raw milk and pose a risk to consumers. In addition to the previously mentioned patho-
gens, M. bovis and B. melitiensis are named as relevant pathogens. The source of pathogen entry takes place in several ways,
with few pathogens being excreted through the infected udder of farm animals (e.g. S. aureus, B. melitensis, M. bovis) (Collins
et al., 2022; Rossetti et al., 2022; Ruegg, 2017). Raw milk is far more often contaminated by the farm environment (L. mono-
cytogenes) or faeces (Salmonella spp., pathogenic E. coli, Campylobacter spp.) (Bangieva & Rusev, 2017; Christidis et al., 2016;
Hussein & Sakuma, 2005). The lack of removal of biofilms in the equipment used for milk production and further process-
ing also plays a major role (Bai et al., 2021; Carrascosa et al., 2021; Chlebicz & Slizewska, 2018). The initial concentration in
raw milk is at the detection limit for most pathogens, so highly contagious microorganisms (pathogenic E. coli, Brucella,
Coxiella, Mycobacterium) play a major role in farm related outbreaks (Cutler, 2014; Gale et al., 2015; Pexara et al., 2018; Sahu
et al., 2021; Valkovska et al., 2021; Verraes et al., 2015; Zastempowska et al., 2016).
C.5 | Egg sector, excluding LMF sector
According to the assessment undertaken by the EFSA BIOHAZ Panel on the public health risks of table eggs due to dete-
rioration and development of pathogens, S. Enteritidis is considered the only pathogen posing a major risk of egg-borne
diseases in the EU, and while other different microorganisms can be found on or in eggs, eggs are not a significant vehicle
for foodborne disease other than for Salmonella (EFSA BIOHAZ Panel, 2014a). This Salmonella serovar is recognised to be
the major pathogen related to egg-borne disease because of its ability to contaminate the interior of intact eggs during
their formation within the body of infected hens.
The revision of other data sources used in the current assessment showed that the conclusions of the former EFSA as-
sessment remain valid. Indeed, most strong evidence FBOs at EU level since 2010 linked to eggs and egg products were
caused by S. enterica subsp. enterica (1137 out of 1153 outbreaks, involving 12,502 cases). 73.4% of those salmonellosis
outbreaks were known to be linked to S. Enteritidis. Other serovars of S. enterica subsp. enterica involved in outbreaks
were S. Typhimurium (N = 41), S. Infantis (N = 3), S. Newport (N = 2), S. Virchow (N = 2), S. Kottbus (N = 1), S. Mbandaka (N = 2)
and S. Muenchen (N = 1), while 20% of the outbreaks were linked to untyped Salmonella. Other biological hazards, apart
from S. enterica subsp. enterica, involved as causative agents of occasional outbreaks linked to eggs and egg products
were B. cereus s. l. (11 outbreaks), and Campylobacter spp., C. perfringens, pathogenic E. coli, S. flexneri and S. aureus, with
one outbreak each. Unfortunately, additional information is only available for one of the 11 outbreaks linked to B. cereus
s. l., which was linked to pancake.
All six multi-country FBOs published as rapid outbreak assessments and involving eggs and egg products were caused
by S. Enteritidis linked to German eggs (ECDC and EFSA, 2014b), Polish eggs (EFSA and ECDC, 2016a, 2017b, 2017c; ECDC and
EFSA, 2020a) and eggs and egg products (ECDC and EFSA, 2022b). A JNS was related to a FBO suspected to be caused by
S. Enteritidis in eggs from Poland (European Commission, 2020).
All but one of the 99 RASFF notifications on bacterial pathogens in eggs and egg products concerned S. enterica subsp.
enterica (99%). These were linked to eggs (N = 65), liquid egg (N = 31) and other egg products (N = 2). There was one notifica-
tion of L. monocytogenes in other egg products (i.e. frozen omelette strips).
Outbreak data from outside of the EU included five outbreaks of S. Enteritidis and other serovars of S. enterica subsp.
enterica in shell eggs (CDC and FSANZ), and a single outbreak of L. monocytogenes in hard boiled eggs (CDC).
C.6 | Fruit and vegetable sector, excluding LMF sector
Under natural conditions (e.g. pre-or post-harvest storage of non-injured fruits and vegetables), the outer layer of the plant
tissue consists of a hydrophobic surface providing a natural barrier for microorganisms (Beuchat, 2002; Brackett, 2007). As
such, the microbiological safety risks in the consumption of fruit and vegetables are linked to their contamination with
enteric pathogens, mainly via soil or water used in agriculture and/or post-harvest handling and processing operations of
fresh produce, although other risk factors have been identified (FAO and WHO, 2021b). Among the key bacterial hazards,
pathogenic E. coli, S. enterica subsp. enterica and L. monocytogenes have been the most common cause of disease out-
breaks (EFSA BIOHAZ Panel, 2014b, 2020; Schierstaedt et al., 2020; Bell et al., 2021). Other bacterial pathogens which are less
frequently associated with fresh produce outbreaks, including emerging ones, are Arcobacter, Bacillus and Campylobacter
(Bell et al., 2021). S. enterica subsp. enterica and pathogenic E. coli can be found in the faeces of livestock and therefore are
very likely to contaminate soil, irrigation water and finally, leafy vegetables (Bell et al., 2021). L. monocytogenes is ubiquitous
in the production and processing environment and may occur in whole fresh, fresh-cut or frozen fruits and vegetables, as
a result of environmental (post-process) contamination (EFSA BIOHAZ Panel, 2020).
According to the FBO data at EU level (period 2010–2020), the main body of strong evidence for outbreaks by bacterial
hazards linked to fruits and vegetables is associated with the following sub-categories: (i) fruit, berries and juices and other
products thereof (16 outbreaks and 346 cases in total), (ii) herbs and spices (15 outbreaks and 572 cases) and (iii) vegetables
(146 outbreaks and 8320 cases). S. enterica subsp. enterica, B. cereus s. l., Clostridium (botulinum and perfringens) and patho-
genic E. coli were the main causative agents, being responsible for 69 outbreaks (2197 cases), 36 outbreaks (849 cases), 24
outbreaks (252 cases) and 21 outbreaks (5139 cases), respectively. Fruits, berries and juices outbreaks were mainly linked
to S. enterica subsp. enterica (13 out of a total of 15 outbreaks), while (fresh-cut) vegetables, juices and products thereof
outbreaks were mainly linked to S. enterica subsp. enterica, B. cereus s. l., Clostridium and pathogenic E. coli (52, 29, 24 and
18, respectively, out of a total of 146 outbreaks). Other hazards associated with vegetables and fruits outbreaks since 2010
include L. monocytogenes (N = 8), Shigella sonnei and flexneri (N = 6), Y. enterocolitica (N = 5), Aeromonas hydrophila (N = 3) and
Campylobacter unspecified (N = 2).
Different serovars of S. enterica subsp. enterica have been involved in multi-country outbreaks. In particular, a S. Agona
contamination possibly linked to RTE food products containing cucumbers caused a five-countries outbreak with 25 cases
between 2014 and 2016 and another 122 cases from January 2017 till July 2018 (EFSA and ECDC, 2018d). S. Braenderup ST
22 was identified as the causative agent of an outbreak presumably linked to Gallia melons, imported from a Honduran
producer, with 348 cases in 12 EU/EEA countries and the UK (ECDC and EFSA, 2021a). Between August 2022 and 12 July
2023, S. Senftenberg ST14 caused a multi-country outbreak possibly linked to cherry-like tomatoes in 11 EU/EEA countries
and the UK, with 92 cases reported of which one fatality (ECDC and EFSA, 2023a).
A listeriosis multi-country outbreak with 47 cases as of June 2018, caused by serogroup Ivb, ST6, in 4 MSs and UK, was
traced to frozen corn and possibly other frozen vegetables, such as green beans and spinach, produced in a freezing plant
in Hungary during the 2016–2018 production seasons (EFSA and ECDC, 2018b).
The total number of RASFF notifications on bacterial pathogens in the fruit and vegetable sector (excluding LMF) are 539
for more than 65 different serovars of S. enterica subsp. enterica, 61 for L. monocytogenes, 35 for pathogenic E. coli, 19 for
Bacillus (18 for B. cereus s. l. and 1 for B. subtilis), 12 for Campylobacter spp., 6 for C. botulinum or C. perfringens, 3 for Shigella
sonnei and 1 for Y. enterocolitica. RASFF notifications for S. enterica subsp. enterica and L. monocytogenes mainly relate to
various fresh and frozen fruits, vegetables and herbs.
Regarding the outbreak data from outside EU, CDC has reported 33 salmonellosis outbreaks since 2006, with papayas,
cantaloupes, fresh-cut melons, tomatoes, mushrooms and various fresh-cut leafy salads being identified as vehicles of con-
tamination with one or more of 28 different serovars of S. enterica subsp. enterica. In the same period CDC has identified
17 outbreaks due to O157-and non-O157 STEC strains in fresh-cut vegetable salads and 7 listeriosis outbreaks (since 2011)
linked to packaged salads, mushrooms, bean sprouts and cantaloupes. In 2021, a multistate outbreak of S. Typhimurium
linked to packaged leafy greens produced at a Controlled Environment Agriculture (CEA) indoor hydroponic operation
took place in USA (FDA, 2022). In 2022, FDA listed numerous outbreaks associated with leafy greens and fruits (e.g. romaine
lettuce, other packaged salads, peaches, fresh-cut cantaloupe and strawberries), contaminated with L. monocytogenes,
S. Javiana, S. Typhimurium, S. Enteritidis and E. coli O157:H7. FSANZ has reported a single outbreak linked to rock melons
with S. enterica subsp. enterica. Finally, Health Canada reported 12 outbreaks since 2015, due to consumption of fresh
and frozen fruits and vegetables. More specifically, seven outbreaks were linked to fresh-cut salads contaminated with
pathogenic E. coli, and a single listeriosis outbreak was linked to packaged salads, while four salmonellosis outbreaks were
caused by contaminated frozen kernel corn, imported peaches and red onions and English cucumbers.
Although not yet officially linked to outbreaks caused by consumption of fresh fruits and vegetables, there is an increas-
ing body of evidence that Arcobacter species, particularly A. butzleri and A. cryoaerophilus, are emerging zoonotic patho-
gens with remarkable occurrence in RTE fresh-cut salads (Ramees et al., 2017). Arcobacter is a Campylobacter-like organism,
in terms of phenotypic responses at detection, a fact that introduces a systematic bias in the detection with culture-based
methods and is known to cause acute or prolonged (chronic) watery diarrhoea combined with abdominal pain (Figueras
et al., 2014; Mottola et al., 2021). Reported prevalence of the above two Arcobacter species in leafy greens can be as high as
14%–27% (Mottola et al., 2016; Mottola et al., 2021).
C.7 | LMF sector
Low moisture foods (LMF) are foods that are naturally low in moisture or are produced from foods with high moisture
through drying or dehydration processes. They all show a low water activity (aw), which contributes to a long shelf life.
Indeed, LMF are commonly defined as any food item that has a aw level of less than 0.85. However, despite they normally
do not support the growth of pathogenic microorganisms given their low aw, they are frequently involved in outbreaks of
foodborne illnesses due to the low infectious dose of some of the microorganisms that may survive in the food product or
to possible subsequent temperature abuse after rehydration that will allow the contaminant organisms to grow.
A wide range of foods and food products can be considered as LMF. A recent assessment by the Food and Agriculture
Organization of the United Nations (FAO) and the World Health Organization (WHO) for ranking LMF using multi criteria decision
analysis in support of microbiological risk management established the following categorisation of LMF (FAO and WHO, 2022):
1. cereals and grains;

2. confections and snacks;
3. dried fruits and vegetables;
4. dried protein products (dairy powders, egg powders, dried fish and fish meal/flour, gelatin and meat powders);
5. honey and preserves;
6. nuts and nut products;
7. seeds for consumption;
8. spices and dried aromatic herbs (including teas); and
9. specialised nutritional products (food supplements).
Powdered formulae for infants and young children were not included among the categories as the hazards and risks
associated with these products had been previously reviewed by FAO and WHO (FAO and WHO, 2004; CAC, 2008). In addi-
tion, oils intended for use in food were not considered in the exercise. Remarkably, although ‘cereals and grains’ scored first
in the risk ranking exercise, mainly due to the international trade and food consumption criteria, ‘dried protein products’,
which were ranked second, stood out in terms of burden of disease. This was influenced by a couple of very large outbreaks
associated with dried dairy products, which led to a high disability adjusted life year (DALY) calculation for this category.
Likewise, ‘nuts and nut products’ also showed a higher burden of disease in terms of DALYs. The burden of disease was
estimated considering data extracted from a scoping review on microbiological hazards in LMF, which showed that S. en-
terica subsp. enterica was implicated in most of the outbreaks and accounted for 44.9% of disease outbreaks across LMF
categories, followed by B. cereus s. l. (25.7%), C. botulinum (15.0%), S. aureus (7.5%), C. perfringens (4.7%) and pathogenic E. coli
(2.3%). In fact, S. enterica subsp. enterica was responsible for 93% of the outbreaks linked to ‘confections and snacks’ (mainly
related to chocolate), 100% for ‘dried fruits and vegetables’, 46.1% for ‘dried protein products’, 80% for ‘nuts and nut prod-
ucts’, 100% for ‘seeds for consumption’, 46.4% for ‘spices and dried aromatic herbs’ (including teas) and 13.8% for ‘cereals
and grains’. The scoping review also evidenced that some of the microbiological hazards are mainly associated with some
particular LMF sub-categories, such as B. cereus s. l., involved in 61.1% of the outbreaks due to ‘cereals and grains’ (mainly
related to cooked rice and pasta dishes) and 35.7% of the ones associated with ‘spices and dried aromatic herbs (including
teas)’, or C. botulinum, causative agent of all but one outbreaks due to ‘honey and preserves’, linked to infant botulism cases
through honey consumption.
Similar conclusions can be drawn from the revision of other data sources used in this assessment.
Strong evidence FBO data at EU level since 2010 showed that most outbreaks associated with cereal products including
rice and seeds/pulses (nuts, almonds) were caused by B. cereus s. l. (51 out of 72 outbreaks, with 676 associated cases) and
S. enterica subsp. enterica (12 out of 72 outbreaks, with 259 associated cases), with C. perfringens (5 outbreaks), C. botulinum
(1 outbreak), S. aureus (1 outbreak) and L. monocytogenes (1 outbreak) being also occasional causative agents of outbreaks.
In relation to sweets and chocolate, S. enterica subsp. enterica caused most of the outbreaks reported (93 out of 105, with
990 associated cases), with B. cereus s. l. (2 outbreaks), S. aureus (1 outbreak) and Vibrio spp. (1 outbreak) being involved in
occasional outbreaks.
All multi-country outbreaks reported as ROA reports and involving LMF were caused by S. enterica subsp. enterica
and included outbreaks of S. Agona and S. Poona infections linked to infant formula (EFSA and ECDC, 2018c; ECDC and
EFSA, 2019c), S. Typhimurium and S. Anatum linked to Brazil nuts (ECDC and EFSA, 2020b), S. Amsterdam, S. Havana, S.
Kintambo, S. Mbandaka, S. Orion and S. Senftenberg linked to imported sesame-based products (ECDC and EFSA, 2021b),
and monophasic S. Typhimurium linked to chocolate products (ECDC and EFSA, 2022a). Between March 2016 and May 2017,
a new serovar of S. enterica subspecies enterica with antigenic formula 11:z41:e,n,z15 infected 47 individuals in five EU coun-
tries through consumption of sesame seeds (EFSA and ECDC, 2017a). Five JNS notifications were associated with a multi-
country outbreak by S. München, with sesame seeds from Sudan being, yet weakly, the suspect vehicle of contamination,
as the same strain of the above serovar was involved in all five notifications (European Commission, 2020).
Outbreak data from outside of the EU included outbreaks of different serovars of S. enterica subsp. enterica linked to
peanut butter, tahini, snacks cereal, dried coconut, pistachios, nuts and nut butter, and chia seed powder (20 outbreaks
reported by CDC and one by Health Canada), pathogenic E. coli in flour and flour products, soynut butter and hazelnuts
(4 outbreaks reported by CDC and one by Health Canada) and Cronobacter sakazakii in powdered infant formula (FDA).
FSANZ also reported an outbreak linked to whey protein concentrate possibly contaminated with C. botulinum.
The majority of the 2216 RASFF notifications for bacterial pathogens for the product categories considered LMF con-
cerned S. enterica subsp. enterica (93.0%). Bacillus, mainly B. cereus s. l., was found in 4.5% of the notifications, and con-
cerned herbs and spices in about half of those notifications. Between 0.6 and 0.8% of LMF notifications were for C. sakazakii
(mainly dietetic foods, food supplements and fortified foods), L. monocytogenes (mainly in cereals and bakery products and
nuts, nut products and seeds) and pathogenic E. coli in various groups of LMF. About 0.3% were for S. aureus (mainly in cere-
als and bakery products) while 0.2% were for C. perfringens (of which three for dried herbs and spices).
APPE N D IX D
Interventions tested to eliminate Listeria from the processing environment
Studies assessing the performance of chemical, physical or biological interventions to destroy Listeria from the processing
environment in industrial settings or on model systems very closely resembling real industrial settings are shown in
Tables D.1–D.3.
T A B L E D .1 Chemical interventions tested to eliminate Listeria from the processing environment in relevant studies.
Study Possible persistence locations Intervention description Intervention conclusions
In industrial settings
Campdepadros FCS and NFCS of different parts of Two disinfection protocolsa L. monocytogenes was not detected
et al. (2012) a dessert-processing factory on FCS while its identification was
restricted to NFCS (mostly the floor).
Both disinfection protocols reduced
the L. monocytogenes load but did not
eradicate the microorganism
Moretro et al. (2017) Floors where water tends to Addition of citric acid powder There was a reduction of L. monocytogenes
accumulate in two meat to five floors and a floor positive floors from 59% to 13% after
processing plants gutter once a day during 4 weeks. In the citric acid test period,
4 weeks after disinfection L. monocytogenes was eradicated from
but before start of all floor areas, except for a floor that was
production positive in three out of four samplings
Pagadala et al. (2012) Seven blue crab meat and crab Feedback provided to There was a significant reduction in the
processing plants (the most processors for improving number of positive L. monocytogenes
common sites positive for L. C&D practices; for samples (samples were taken from crab
monocytogenes were raw crab example, switching to meat and environment)
coolers and receiving docks) more aggressive alkaline
detergents for removal of
biofilms
Eglezos and All production areas in a cheese Ozonationb as an adjunct to the L. monocytogenes isolations were
Dykes (2018) processing facility C&D regimes significantly reduced in all areas from
15% (27/180) in the samples taken
pre-ozonation (i.e. taken after the end
of the deep clean monthly disinfection)
to 1.67% (3/180) in the post-ozonation
samples
Using model systems
Berrang et al. (2017) Uninoculated PVC floor drains Disinfection using self- There was a significantly reduction of
and L. monocytogenes- contained chlorine dioxide L. monocytogenes counts in standing
inoculated laboratory scale (ClO2)-generating and waterc and the inner surface of treated
model PVC floor drains delivery pods. Drains were drains.d A 24-h treatment was more
exposed to ClO2 for 4 h or effective than a 4-h treatment. The most
24 h efficient treatment eliminated viable
L. monocytogenes in the drain liquid
and lead to > 6 log decrease in attached
L. monocytogenes
Chaitiemwong Conveyor belt The conveyer belt was The L. monocytogenes reduction on belt
et al. (2010) composed of polyester material with antimicrobial additives
fabric impregnated with was greater than on belt material
either thermoplastic without additives when the surfaces
polyurethane (control) were wet. The presence of food
or with thermoplastic debris neutralised the effect of the
polyurethane and the antimicrobials. This suggests that the
antimicrobial compound antimicrobial additives in conveyor
HyGUARD®e belt material could help to reduce
L. monocytogenes on belts at low
temperatures when food residues are
absent, and belts are not rapidly dried
Martinez et al. (2021) L. innocua inoculated rubber or Different C&D protocols Rubber blades were cleaned more
foam blade squeegees simulating cleaning efficiently than foam blades, possibly
procedures used in food due to reduced bacterial attachment
processingf to the structure. A full procedure
(detergent and rinse, followed by
disinfectants) including a scrubbing step
was recommended
T A B L E D .1 (Continued)
Abbreviations: C&D, cleaning and disinfection; FCS, food contact surface; NFCS, non-food contact surface; PVC, polyvinyl chloride.
a
Protocol A: Applied daily. Visible dust was removed by manual scanning. The surface was rinsed with warm water (45–50°C) at low pressure. Easyfoam VF32 (diluted to
3%–4%) was applied and left for 15–20 min followed by manual scrubbing and abundant rinsing with tap water. Suredis VT 01 L (diluted to 1%–2%) was applied and left
for 15–20 min. Finally, it was rinsed abundantly with tap water. Weekly air disinfection was performed after work (Saturday afternoon to Sunday morning). Protocol B:
Once every 3 weeks. Visible dust was removed by manual scanning. The surface was rinsed with warm water (45–50°C) at low pressure. Easyfoam VF32 (diluted to 3%–4%)
was applied and left for 15–20 min followed by manual scrubbing, and abundant rinsing with tap water. Aciplusfoam VF59 (diluted to 3%–4%) was applied and left for
15–20 min. It was then scrubbed manually and rinsed abundantly with tap water. Suredis VT 01 L (diluted to 3%–4%) was applied and left for 15–20 min. Finally, it was
rinsed abundantly with tap water.
b
5 L/min with a concentration of 20 g/min for 15 min on Mon - Fri and 120 min on Sat – Sun.
c
Planktonic cells remaining viable in the liquid.
d
Viable attached cells on the inner surface.
e
With substances silver zeolite, aluminium oxide, calcium oxide, magnesium oxide, zinc pyrithione, oxybisphenoxarsine or a combination of those.
f
The detergents used were quaternary ammonium (550 ppm), a chlorine-based disinfectant (600 ppm), a peracetic acid solution (2025 ppm) or a chlorinated alkaline
detergent (20,000 ppm).
TA B L E D. 2 Physical interventions tested to eliminate Listeria from the processing environment in relevant studies.
Possible persistence
Study locations Intervention description Intervention conclusions
UV-C radiation
Bernbom et al. (2011) Food processing surfaces Ceiling-mounted UV-C light L. monocytogenes positive samples were reduced
in a fish smoke house (wavelength 254 nm) from 44% (30/68) before the 48-h exposure
after the routine C&D to 12% (8/68). After 7 h of exposure, positive
procedure samples reduced from 34% (23/68) to 26%
(18/68). Laboratory experiments showed that
UV-C light is a useful extra bactericidal step
and that it, as all disinfecting procedures, is
hampered by the presence of organic material
Using model systems
Morey et al. (2010) Four types of conveyor UV light applied at 5.53 and L. monocytogenes was significantly reduced on all
belts made of different 5.95 mW/cm2 for 1 and 3 s belt types irrespective of UV light intensities
materials and exposure times. More survival was found at
the lowest light intensity on all conveyor belts.
Populations were reduced to below detection
limits on three types of belts after exposure to
the highest UV light intensity for 3 s
Heat treatments
Eglezos and Holding chillers used for Heating interventions using The Listeria prevalence in chiller A was significant
Dykes (2011) post-cook commercial electrical air-blowing reduced using air temperatures of 37°C for 36 h
processed meats in heaters at each of the two from 10.6% (19/180) to 1.7% (3/180), while it was
facilities facilities, with 2 weeks reduced in chiller E using 50°C for 2 h from 7.8%
of post-intervention (7/90) to 0% (0/90)
sampling after each
treatment
Pennone et al. (2020) The floors of mushroom Steam cookout processes There was a significant reduction in the number
growing units of positive L. monocytogenes samples. Before
cookout, the incidence averaged 63% (75% of
the floor swabs and 45% of the spent substrate
samples positive), while after a first cookout,
the average was 40% (67% of the floor swabs
positive and no positives in the spent substrates).
19% of floor swabs taken after a second cookout
were found positive
Using model systems
Tobin et al. (2020) Detachable mushroom Hot water disinfection using Complete elimination of L. innocua cells from each
slicer heads from water at 55°C, 65°C or 75°C slicer head treatment
industrial food slicing for 93, 16.4 or 6.5 min
equipment
Abbreviations: C&D, cleaning and disinfection; PVC, polyvinyl chloride; UV, ultraviolet.
TA B L E D. 3 Biological interventions tested to eliminate Listeria from the processing environment in relevant studies
Study Possible persistence locations Intervention description Intervention conclusions
Reinhard et al. (2020) The RTE environment of Bacteriophage P100 (Listex™) There were significant reductions (ranging
refrigerated (4°C) and using two different application from 32% to 44.4%) in the fraction of NFCS
ambient (20°C) temperature strategiesa samples positive for Listeria spp. using
facilities producing RTE meat both application strategies
and poultry products, RTE
bakery items and assembled
RTE sandwiches
Schobitz et al. (2014) The walls of floor gutters in a A so-called ‘biocontroller’ L. monocytogenes was successfully eliminated
salmon processing plant consisting of a thermally from the walls of the floor gutters in five
treated fermentate from out of the seven trials. The two failures
two Carnobacterium were linked to loss of contact of the
maltaromaticum strains biocontroller with the side wall of the floor
and a strain of Enterococcus gutter
mundtii, plus nisin at 1000 IU/
mL, entrapped in an alginate
matrix supported by a mesh-
type fabric
Zhao et al. (2013) Six floor drains of a RTE poultry Competitive exclusion using a L. monocytogenes could be efficiently reduced;
processing plant that were combination of Lactococcus it was not found in five of the floor drains
consistently found Listeria lactis and Enterococcus durans after the first week of treatment and all
positive applied for 4 weeksb floor drains were negative after 8 weeks
of treatment, whereas control drains were
still Listeria positive
Abbreviations: IU, international Units; NFCS, non-food contact surface; RTE, ready-to-eat.
a
A moderate application applied as a single treatment every 24 h over three consecutive days (2 × 107 PFU/mL) and an intensified application applied once every 6 h over a
24 h period (1 × 108 PFU/mL).
b
4 times a week during the first week and twice-a-week for the following 3 weeks.
APPE N D IX E
QMRA module for persistence
E.1 | Introduction
To better understand the importance of persistence for public health (PH) risks related to microbial food safety, as com-
pared to other factors such as bacterial growth and virulence, we developed a model that describes the process, and in-
cluded it in a hypothetical food chain model for risk assessment. The aim of the modelling is to:
• provide a generic mathematical description of the persistence process, that can be integrated in a quantitative microbi-
ological risk assessment (QMRA) and is as simple as possible, whilst capturing the essential characteristics of persistence
in a food processing environment (FoPE);
• explore the model, to increase our understanding of the essentials of the persistence process;
• explore the importance of persistence compared to other factors such microbial growth, survival and virulence; and
• compare strains with different characteristics in terms of persistence and other factors, by including the possibility to
make the model ‘strain specific’, as specific persistent strains are well known.
As AQ6 requires to develop perspectives, and not to perform a risk assessment per se, an exploration of the model was
pursued while a concrete case study was not performed.
Previously, for the development of QMRA models, Nauta (2007) defined a set of basic processes that typically describe
events leading to changes in concentrations of bacteria in the food chain: growth and inactivation; and mixing, partition-
ing, cross-contamination and removal. Of these, persistence seems to involve cross-contamination (bacterial transfer), de-
creased inactivation (increased survival) and probably growth in the FoPE.
In general, persistence can be observed if the following steps hold:
1. the FoPE is contaminated, either from (raw) food material entering the environment or through (re-)contamination
from the FoPE itself (e.g. through workers or equipment). This is a single event. Repeated contamination events
may induce different instances of persistence;
2. there is prolonged survival (no inactivation or removal, as intended in the processing) and/or growth in the FoPE; and
3. there is bacterial transfer (cross-contamination) from the FoPE to the food product, as otherwise there is no PH effect for
consumers.
The second step is the critical step, as here, in the food production process, it is foreseen that the contamination is re-
duced or eliminated, so the contamination is only temporal. For persistence to occur, the bacteria must prevail in the FoPE
longer than anticipated in the hygienic design of the food production process. Persistence can be due to a processing step
which is explicitly included to inactivate or remove potential contamination and is not effective (due to process parameters
or the typical characteristics of the bacterial strain), or due to an incident in an otherwise safe processing step (something
unexpected in the FoPE or ‘colonisation’ due to the typical characteristics of the bacterial strain). Typical for persistence is
also that it occurs over a long period. In general, this ‘long period’ is not well defined.
Obviously, when there is bacterial growth in the FoPE, for example in a biofilm, there can be a continuous influx of bac-
teria. A priori, it is unclear to what extent bacterial growth in the second step is required to observe a persistent strain over
a long period. This can be evaluated by the model.
Below, in Section E.2, we first provide an overview of potential persistence models for QMRA, based on literature known
to the Working Group, and select a model that was considered suitable. Next, in Section E.3, simple QMRA food chain
models are constructed for Listeria monocytogenes in a product like cooked ham and Salmonella in a low moisture food.
Then, in Section E.4, the performance of these models is explored. The results of Section E.4 are discussed in Section E.5.
E.2 | Potential persistence models for QMRA
Although few models in the literature are explicitly made to describe persistence in the FoPE, some available models can
be applied for that purpose. Usually, these models are presented as transfer models or cross-contamination models, where
the principle is that bacteria in the FoPE are transferred to a food being processed. As these models are usually developed
to describe a specific process for a specific (type of) food product, they contain specific factors or transfer routes. Here,
we first provide some generic models, derived from published models (Sections E2.1-E2.3). Next, based on these, a basic
deterministic persistence model is presented (Section E2.4) with variants that allow inclusion of stochasticity (Section E2.5)
and instantaneous increase and decrease in population size, due to, for example, growth overnight or C&D (Section E2.6).
E.2.1 | One compartment model
The first model describes a FoPE that contains N bacteria (due to a single accidental contamination) that can be trans-
ferred to a food product that is processed in the FoPE (See Figure E.1). It holds the basic assumption of a fixed transfer rate.
(Chaitiemwong et al., 2014), for example, use this as a model for transfer from a slicer; it is a simplified version of, for exam-
ple, the broiler processing plant model described by (Nauta, 2005). In principle, there is a continuous flow of consecutive
food products. At step I in the process, with a transfer rate b1 and Ni bacteria(in the FoPE, the number of bacteria on the food
product is b1 Ni and the bacterial number in the FoPE reduces to Ni+1 = Ni 1 − b1 . The transfer rate can be interpreted as
)
the probability for a bacterial cell in the FoPE to be transferred to the food product. The model assumes that the expected
number of bacteria is transferred, so no stochasticity is included.
F I G U R E E .1 One compartment model. From a contaminated FoPE there is continuous transfer to food products (it is assumed that there is a
constant flow of consecutive food products, at step i the ith food product passes through the FoPE).
An example of the dynamics of the process is given in Figure E.2. It shows a loglinear decline of the bacterial numbers in
both the FoPE and on the consecutive food products.
This model shows(no typical
)j−1 persistence behaviour. The jth food product passing through the environment is con-
taminated with N0 b1 1 − b1 bacteria.
F I G U R E E . 2 One compartment dynamics, N0 = 108 CFU, b1 = 0.012. Blue line: log(Ni) in the FoPE. Orange line: Number of bacteria on the
consecutive food products (in log CFU).
E.2.2 | Model with two separate compartments
After performing experiments with meat grinders, it was found that this one compartment model cannot explain observa-
tions, which show tailing (Møller et al., 2012; Møller et al., 2016). An alternative model was proposed, with two compart-
ments representing two different FoPE (see Figure E.3). With this model the observations could be explained well. Here we
present a simplified version where there is only transfer from the environments to the food products and not vice versa.
FoPE1 and 2 are contaminated with N1,i and N2,i bacteria at step i, transfer rates are b1 and b2, so Nf ,i = N1,i−1 b1 + N2,i−1 b2.
F I G U R E E . 3 Two separate compartments model. From two (simultaneously) contaminated FoPE there is continuous transfer to food products,
with different rates.
An example of the dynamics of the process is given in Figure E.4. With this model, the tailing as observed in several ex-
periments is reproduced (Møller et al., 2016).
F I G U R E E . 4 Two separate compartment dynamics, N1,0 = 107 CFU, N2,0 = 107.95 CFU, b1 = 0.012, b2 = 0.0108. Blue line: log(Ni) in the FoPE1
(solid) and 2 (dashed). Orange line: Number of bacteria on the consecutive food products (in log CFU), showing tailing.
This model can explain the tailing that is often observed and can therefore explain persistence over a longer time
than the
( one)j−1compartment ( model.)j−1 The jth food product passing through the environment is contaminated with
N1,0 b1 1 − b1 + N2,0 b2 1 − b2 bacteria.
The identification of the two separate environments in the FoPE may be challenging (Møller et al., 2020). Furthermore, it
is questionable whether two of those separate environments with direct transfer to the food products exist in every FoPE
where persistence occurs.
E.2.3 | Model with two linked compartments
A third option for a model is inspired by the agent-based model for persistence presented by (Mokhtari & Van Doren, 2019).
They describe the persistence as a process with two linked environments in the FoPE, one with bacterial cells on more
accessible areas (FoPE1) and one with bacterial cells harboured in holes and cracks that are difficult to clean and disinfect
(FoPE2) (Figure E.5). FoPE1 can be identified as a food contact environment and FoPE2 as a hidden environment. We here
omit the C&D and present the principle of their model as a model with two linked compartments.
F I G U R E E . 5 Two linked compartments model. From a contaminated FoPE1 there is continuous transfer to food products and to FoPE2. There is
also a flow back from FoPE2 to FoPE1.
In this model, transfer to the food products is only from FoPE1, whereas there is exchange between FoPE1 and FoPE2,
with two transfer rates, b2 and b3
The model system reads as:
Nf ,i = N1,i−1 b1
( )
N1,i = N1,i−1 1 − b1 − b2 + N2,i−1 b3
N2,i = N1,i−1 b2 − N2,i−1 b3
An example of the dynamics of the process is given in Figure E.6. Again, tailing is observed.
F I G U R E E . 6 Two linked compartment dynamics, N1,0 = 108 CFU, N2,0 = 0 CFU, b1 = 0.003,b2 = 0.009, b3 = 0.002. Blue line: log(Ni) in the FoPE1
(solid) and the FoPE2 (dashed); orange line: number of bacteria on the consecutive food products (in log CFU), showing tailing.
This model can explain the tailing that is often observed, and thus explain persistence over a longer time than the one
compartment model. It requires one more parameter than the two separate compartments model but may be more real-
istic in the assumptions that only one FoPE is originally contaminated, and that there is a ‘hidden’ environment (such as a
drain, or an inaccessible component of the equipment, difficult to clean and disinfect) that is connected to the other one,
but not directly to the food products.
Note that it is assumed that the initial contamination occurs in FoPE1, which is in direct contact with the food that could
be the source of the contamination.
E.2.4 | Basic model for persistence
The Working Group members agreed that the two linked compartments model is the most realistic, both in assumptions
and performance.
The version of this model described in Section E.2.3 only contains transfer between environments and food but does not
include growth potential or inactivation (less than 100% survival), which are both considered important for persistence.
These are included in a more complete version of the model, which is presented in Figure E.7. This version of the model is
considered as the ‘basic model for persistence’ in this Scientific Opinion as it includes the relevant characteristics leading
to persistence; other versions of the model, with small modifications, can be used to describe more realistic case studies.
The ‘persistence parameter’ a2 represents the change in population size in FoPE2 in the time frame between two food
products passing in FoPE1, which is presumably very short. It could be estimated from observed growth in a certain time
period, if the processing speed is known. A positive value implies growth in FoPE2, so a2 represents a growth rate which will
result in increased persistence. When a2 = 0, there is 100% survival, and only transfer to and from FoPE2 influences changes
in the population size in FoPE2. A negative value of a2 represents an inactivation rate implying less than 100% survival.
The inactivation rate parameter c1 represents the decrease in population size in FoPE1 (which, if negative, might represent
growth), also in the time frame between two food products passing in FoPE1.
The model system reads:
Nf ,i = N1,i−1 b1
( )
N1,i = N1,i−1 1 − b1 − b2 − c1 + N2,i−1 b3
( )
N2,i = N1,i−1 b2 + N2,i−1 a2 − b3
From these model equations, it can derived that the system is in equilibrium (i.e. constant values for N1 and N2) if
( )
b3 b1 + c1
a2 =
b1 + b2 + c1
The total population size (N1 + N2) is increasing if a2 is larger and decreasing if a2 is smaller. It is likely that a2 < b3, as oth-
erwise growth dominates the process too much.
F I G U R E E . 7 Extended version of the two linked compartments model depicted in Figure E.6. Persistence parameter a2 indicates the growth or
inactivation rate in the hidden environment FoPE2, whereas there may be limited survival in FoPE1, described by inactivation rate c1.
E.2.5 | Stochastic versions of the model
The basic model for persistence above is deterministic, assuming constant rates for transfer between environments and
food, growth and inactivation. This simplifies the model but may not be realistic. The alternative is to develop a stochastic
model, which may be more realistic but will contain more (unknown) parameters.
Random stochastic variation can be included. Instead of constant transfer and inactivation rates, for every passing food
item, the number of bacteria transferred or inactivated is sampled from a binomial distribution Bin(N, p), where N is N1 and
p is rate b1, b2 or c1; or N is N2 and p is rate b3. This assumes constant rates, but discrete bacterial cells are sampled, so the
model does not operate with expected values only.
Stochasticity can be taken one step further, as even the rates can vary with every transfer. This increases the variation
in numbers of transferred and inactivated bacterial cells. A Betabinomial distribution can be used, which is similar to a
Binomial distribution, but with rates that have a Beta(kp, k(1 − p)) distribution, where k is a parameter (0, ∞ ). This gives
increasing variability in transfer or inactivation with increasing k; the model is similar to the one described by (Nauta, 2005),
for clustering of cells during a partitioning process.
The persistence parameter a2 could be sampled from, for example, a Normal distribution with mean a2,mean and standard
deviation a2,sd.
Figure E.8 shows an example of the change in dynamics when a stochastic model is applied. With high concentrations,
the use of the binomial model does not add much variation. With the use of the betabinomial model the system gets more
unpredictable, as transfer goes in ‘bursts’. This may be realistic, but complicates the model analysis considerably, both in
terms of input requirements, potential outputs and scenario analysis.
F I G U R E E . 8 Examples of results of the stochastic version of the two linked compartments model, comparable to Figure E.6. Figure A applies
the binomial model; figure B applies the BetaBinomial model, with k = 1. Blue line: log(Ni) in the FoPE1 (solid) and the FoPE2 (dashed); orange line:
number of bacteria on the consecutive food products (in log CFU). Note that in these examples, the initial contamination is in FoPE1, not in FoPE2.
Although a stochastic model would be more realistic, it was not further evaluated because it gives a model with even
more unknown parameters and its analysis requires more resources, so it complicates the model analysis considerably.
E.2.6 | Instantaneous inactivation and growth
The basic model assumes that inactivation and growth take place gradually, at a constant rate. This is not necessarily real-
istic or desirable. With C&D, a larger amount of the bacterial contamination will be removed, which can be modelled by an
instantaneous inactivation step, implemented by a few logs decrease in concentration. Typically, for persistence to occur, it
can be assumed that the C&D is effective in FoPE1, but not in FoPE2. This represents a situation where the food business op-
erator's hygienic measures are targeted at the food contact areas that are in direct contact with the food, whereas ‘hidden’
areas may be missed (Mokhtari & Van Doren, 2019). Similarly, growth may occur during a period that the food processing is
halted, e.g. overnight, giving ample time for increase in concentration, for example in a biofilm.
These phenomena can be modelled by introduction of a log increase or decrease after processing of a certain number
of food products. Figure E.9 shows an example of a sudden 5 log decrease in FoPE1, and the limited effect it has on the
contamination of the processed food products.
F I G U R E E . 9 Example of the effect of C&D after processing 500 food products, in the two linked compartments model as shown in Figure E.6.
The concentration in the FoPE abruptly decreases, but increases again due to contamination from FoPE2, in which the C&D was not effective. Blue
line: log(Ni) in the FoPE1 (solid) and the FoPE2 (dashed); orange line: number of bacteria on the consecutive food products (in log CFU). Note that in
this example, the initial contamination is in FoPE1, not in FoPE2.
E.3 | A QMRA food chain model with persistence
The persistence model is to be used in conjunction with other models in a larger food chain model for QMRA. As an ex-
ample, we applied a simple food chain, inspired by the production of cooked ham that is potentially contaminated with L.
monocytogenes. Persistence was considered to occur in a slicer, as illustrated in Figure E.10.
F I G U R E E .1 0 QMRA food chain model involving persistence in the slicer.
We assume that cooking (the first step) inactivates all L. monocytogenes, and either a component of the slicer in contact
with the ham (FoPE1) or a FoPE in contact with this part of the slicer (FoPE2) has been accidently contaminated with L.
monocytogenes that became persistent. After that we can get growth during storage, exposure and possibly illness, as
defined by a dose–response (DR) model. Additionally, some assumptions on the population considered (number of con-
sumers and servings, serving sizes etc.) are required.
As an alternative, we also defined a model for Salmonella in a LMF, where growth is not expected to occur in the food
product during storage. This follows the same food chain model structure, without the storage step.
The model described in Section E.2.4 was applied as a persistence model. The food products passing through FoPE1
were assumed to have a weight of 500 g. Below, the growth model for L. monocytogenes and DR models for L. monocy-
togenes and Salmonella are described.
E.3.1 | Growth model for L. monocytogenes growth during storage
The model for growth of L. monocytogenes during storage of the food product after slicing is taken from (Pouillot &
Lubran, 2011), the triphasic linear model without lag, given by
( )
𝜇
logNend = min logN0 + × t, MPD ,
ln10
(T −Tmin ) 2
[ ]
with 𝜇 = 𝜇ref ,
(Tref −Tmin )
and Tmin = − 2.86 ◦ C ; Tref = 25 ◦ C ; 𝜇ref = 6.24 per day and MPD = 7.27 log CFU ∕g.
As default, we take T = 6 ◦ C and t = 5 days. Portion size is assumed to be 50 g.
E.3.2 | Dose–response models
E.3.2.1 | Dose–response model for L. monocytogenes
The lognormal-Poisson DR model for L. monocytogenes that was developed by (Pouillot et al., 2015) and later applied by
(EFSA BIOHAZ Panel, 2018) provides models for specific population groups, depending on susceptibility, and includes spe-
cific terms describing variability between hosts in these population groups and between strains. The probability of illness
after consumption of one cell of L. monocytogenes is expressed through the r-value. The standard deviation of the r-value
is assumed to be the same for all the population groups. It is summarising the variability between L. monocytogenes strains
and host individual susceptibilities. With a given mean value for the log r-value (typical for the exponential DR model) for
each population group, the standard deviation of this log r-value between hosts is 0.55 and between strains is 1.52 (Pouillot
et al., 2015).
For the healthy population (mean log r = −14.11) this allows the derivation of the log r-value distribution of a typically
virulent strain. According to (Pouillot et al., 2015), the variability in mean log r-values between strains can be described by
a Normal (−14.11, 1.52) distribution. A virulent strain can then be characterised as having a mean 𝜇 = − 10.574 (the 99%-ile
of the between strain distribution) and sd = 0.55.
Implementation of the lognormal-Poisson DR model requires stochastic modelling, which is rather cumbersome and
not needed. For our purposes, it was sufficient to use this model to establish a DR model providing mean probabilities of
illness as a function of dose, obtained by Monte Carlo simulation with 100,000 iterations for a series of doses ranging from
100 to 108.
Pill,healthy pop,virulent strain = D × 10−10.229
This is an approximation of the original model, but, as shown in Figure E.11, it seems to be appropriate as the correlation
is perfect (r2 = 1).
The DR model for the whole population can be derived in a similar way. Here, we use the definition of population sub-
groups given by Pouillot et al. (2015) and assume that all population groups consume the same amount of cooked ham.
Mean values of log r are sampled on the basis of the relative sizes of the different population groups, to obtain a new (empiri-
cal) distribution of r-values for the same virulent strain. This distribution is applied in a Monte Carlo simulation with 1000,000
iterations for a series of doses ranging from 100 to 1010, for 200 doses with step size 0.05 for the log dose, resulting in:
logPill, whole pop,virulent strain = − 9.038 + 0.9922 × logD − 0.00523 × logD 2 + 0.00352 × logD 3 − 0.00043 × logD 4
which is also a well-fitting approximation (see Figure E.11).

F I G U R E E .11 Data points and fitted line of the dose–response model to mean probabilities of illness obtained for 160 and 200 doses, using the
(Pouillot et al., 2015) DR model for a virulent strain and healthy people in the general population (A) and the whole population, including susceptible
population groups (B).
With the same methodology, models can be derived for the ‘median’ strain (50% of the virulence scale, 𝜇 = − 14.11, for
the healthy population) as
Pill,healthy pop,average strain = D × 10−13.766
and
logPill,whole pop,average strain = − 12.585 + 1.00795 × logD − 0.004417 × logD 2 + 0.000831 × logD 3 − 0.00005 × logD 4
E.3.2.2 | Dose–response model for Salmonella
Previously, the (EFSA BIOHAZ Panel, 2014a) applied a Beta Poisson DR model for Salmonella in their opinion on Salmonella
in eggs, based on (Thomas et al., 2006).
)−0.149344
D
(
Pill = 1 − 1 +
50.60408
In this model, variation between strains is not explicitly included, so strain virulence cannot be described.
Typically, Salmonella requires much smaller dose than L. monocytogenes to cause human infection and illness. The shape
of the DR relation is not linear.
F I G U R E E .12 Dose–response relation for probability of illness after exposure to Salmonella, based on (Thomas et al., 2006).
Other and more recent DR relations for Salmonella are published (for example Teunis, 2022), but it is outside our scope
to investigate those.
E.4 | Model performance
The L. monocytogenes and the Salmonella QMRA food chain models have been implemented in R and are available as R
shiny apps through the Knowledge Junction under https://doi.org/10.5281/zenodo.10299477. (see Section E.4.2.1).
The basic persistence model (Section E.2.4) has five unknown parameters (a2, b1, b2, b3 and c1), and unknown values of the
initial conditions. For all of them we have no or very little data available. Exploration of the model gives many possibilities
for (combinations of) inputs. In general, for risk assessment, it would be of interest to know which (combinations of) param-
eter values characterise observations of persistence and long-term outbreaks, i.e. high concentrations in the food and/or
cases of human illness over a long time period. Such (combinations of) parameter values of the persistence model would
allow a comparison of risk estimates found with high persistence versus high virulence or large growth capacity and could
identify parameters to be targeted for interventions. It is expected that the ‘persistence parameter’ a2, which expresses the
survival and/or growth in FoPE2, is of major influence, but a priori, the role of the transfer parameters (b1, b2, b3) is less clear.
Below, the results of the L. monocytogenes model are shown and then results of the Salmonella model are summarised,
for comparison, in Section E.5.3. Before the influence of the persistence model parameters is explored in Section E.4.2, it
was first considered to what extent growth is required to explain persistence. This question is not only important for a gen-
eral understanding of the persistence phenomenon, but also to evaluate whether the persistence model can be simplified
by omitting growth, whilst still being able to describe what is observed when persistence is found.
E.4.1 | The role of growth in persistence
If we consider a simple version of the two linked compartments persistence model, i.e. the one shown in Figure E.5, without
growth or inactivation in FoPE2 (a2 = 0) and with full survival in FoPE1 (c1 = 0), it is easy to see that all the bacteria that enter
the system at t = 0 will eventually be transferred to the food. This can be a long-lasting process, when the transfer rates are
low, especially when b1 and/or b3 are low. However, in that case, the number of cells transferred to the food are also low, as
is the probability of illness after consumption of such a food product. Given the (approximate) linearity of the DR relation of
L. monocytogenes (Figure E.11), under the same conditions in other parts of the food chain included in the model, the total
expected number of cases from a single contamination event of FoPE1, initiating the process, is the same, independent
of the transfer rates. Hence, without growth, and without removal of contamination in the food production process, the
model predicts that transfer rates and the consequential duration of the presence of contamination in the FoPE have no
effect on the expected number of human illness cases, as all bacteria end up in the food. In that case the contamination is a
food safety problem, but the potentially observed persistence due to survival and transfer between compartments of the
FoPE does not increase the PH risk.
If the DR relation is not linear (such as for Salmonella), persistence due to low transfer rates will (slightly) increase the
expected number of cases, as, due to the shape of the DR models that follow the single-hit concept (and on biological
grounds reject the minimum infectious dose concept), a few exposures to high doses will give less cases than many expo-
sures to low doses, for the same total number of bacteria in all doses together. Note, however, that low levels of contamina-
tion on the food and a series of seemingly sporadic cases that occur over a longer period of time, may not be observed in
surveillance of food or surveillance of human cases, and therefore the persistence may not be observed either. Similarly,
high levels of contamination and a few cases occurring in a short time period, may be more easily found. This, counterin-
tuitively, might imply that, without bacterial growth in the FoPE and with a non-linear DR relationship, persistence may
reduce the observed contamination of food and the notification of outbreaks, whilst simultaneously actually increasing the
number of cases, although that may go unnoticed.
If there is limited survival in FoPE1 (and/or FoPE2, i.e. a2 < 0 and/or c1 > 0), due to hygienic measures, such as regular C&D
or other good hygienic practices, without growth, the overall PH risks are lower as compared to a situation with immediate
transfer of contamination to a food product. So, the opposite of what we would call persistence (long term observation
of the same strain in the FoPE) would theoretically give a higher PH risk. If, on the other hand, we define ‘persistence’ as a
strain specific characteristic, so that persistent strains are those that are not removed with common C&D practices, ‘persis-
tence’ increases the risk.
These findings imply that, in general, even though persistence without growth, as modelled by this simple version of
the model, can result in the finding of positive FoPE samples over a longer time, it will not result in a higher risk than the
alternative where the bacteria do not persist in the environment, but contaminate the food anyway. So, from a theoretical
perspective, growth (increase of numbers of bacteria, potentially in biofilms) in the environment seems to be important for
persistence to have a significant PH impact.
E.4.2 | Examples of model performance
E.4.2.1 | R shiny apps
The R shiny apps give many opportunities to explore the model. The user defines the transfer parameters b1, b2, b3 and the
persistence parameter a2, which as a default is put at a2 = 0 (no growth or inactivation in FoPE2), as well as the exposed
population (only the healthy population or the whole population), the strain type considered (virulent or average) and the
place of first contamination (FoPE1 or FoPE2).37 Other inputs can be defined after clicking the ‘Show more input’ button.
The output shows graphs with the dynamics of N1 (population size in FoPE1), N2 (population size in FoPE2) and Nf ,after, the
37
virulence and population are not included in the Salmonella model.
concentration in the food after storage; and of Pill , the probability of illness per serving. Below it, inputs and outputs are
summarised.
There are several options to include growth and inactivation:
• continuous growth or inactivation in FoPE2, parameter a2. As growth in a2 typically characterises persistence in the FoPE,
this is the main persistence parameter;
• instantaneous increase, e.g. after a period where the food production is stopped, every ‘interval for growth boost’ food
products, with an indicated log increase;
• instantaneous decrease in FoPE1 only, e.g. regular C&D that does not affect FoPE2, every ‘C&D (FoPE1) food products’,
with an indicated log decrease;
• instantaneous decrease in FoPE1 and FoPE2, e.g. special thorough C&D that affects both FoPE, every ‘thorough C&D
(both FoPE) food products’, with an indicated log decrease;
A ‘standard run’ of the L. monocytogenes model, as in Section E.4.1, would give the output as shown in Figure E.13. The
model parameter values are a2 = 0; b1 = b2 = b3 = 10−4 ; c1 = 0; logN0 = 9.
E.4.2.2 | Scenarios in the L. monocytogenes model
The effect of single parameters:
• Continuous growth in FoPE2: the situation with a2 = 0 has been discussed in Section E.4.1. If a2 is lower than the equilib-
rium value, the total number of bacteria in the system is decreasing. Persistence can take a very long time, definitely if it is
close to the equilibrium, but the bacterial population will die out at some point. If a2 is larger than the equilibrium value,
the number of bacteria in the system will increase, and so will eventually the probability of illness per food product. This
is not sustainable as at some point the bacterial numbers and numbers of cases will get too high. If this would occur in
reality, it would be observed, and measures would be taken.
• A typical model output with a2 = 0.00008 (which is above the equilibrium value) is shown in Figure E.14. Concentrations
increase, and so does the probability of illness.38
• C&D, resulting in instantaneous decreases in FoPE1 only, does not affect the dynamics very much. FoPE1 is quickly filled
up with bacteria from FoPE2 (as in Figure E.9). Here, one should realise that the axis is on a log scale. It shows that inef-
fective C&D, where a hidden area is not reached, is indeed inefficient (See Figure E.15).
• Thorough C&D resulting in instantaneous decreases in both FoPE1 and FoPE2 is effective. Despite a lower log reduction,
Figure E.16 shows that the risk can be reduced to zero in this case.
F I G U R E E .13 Input and output of the model using the app, given default parameter values b1 = b2 = b3 = 10−4 ; c1 = 0; logN0 = 9) as described
in Section E.4.2.1, without growth and with full survival (a2 = 0). Initial contamination takes place in FoPE2.
38
With growth in FoPE2, a complication in the model can occur: The growth model for storage has an MPD (maximum population density) which implies that, for the
same growth conditions, increase in concentration in the food product before storage will not always lead to the same increase in the concentration after storage. The
increase gets lower with higher initial concentration. However, if the food product is more heavily contaminated than the MPD, the concentration after storage shows an
increase again as well, even without growth. This looks odd but is a consequence of the model formulation. Another complication may be that the dose is larger than 10
log CFU. In that case the DR model is no longer valid. However, obviously, in that case the concentration in the food is unrealistically high anyway.
107 of 114
F I G U R E E .15 The model of Figure E.13 with a log decrease of 5 log in FoPE1 every 20,000 food products shows that, in this case, C&D is not very
|
PERSISTENCE OF MICROBIOLOGICAL HAZARDS IN FOOD AND FEED ENVIRONMENTS
The model of Figure E.13 with a2 = 0.00008.
effective, despite the 5 log reduction.

F I G U R E E .14
F I G U R E E .1 6 The model of Figure E.13 with a log decrease of 3 log in both FoPE1 and FoPE2 every 20,000 food products shows that C&D is very
effective in this case, despite the lower effect of C&D (3 log reduction) per event.
E.5 | Risk assessment: the impact of persistence
E.5.1 | The L. monocytogenes QMRA
QMRA models can be used to compare the impact of different processes in the food chain on the PH risk. Here, we explore
how the impact of persistence could be compared with the impact of growth during storage and the impact of the DR
relationship. The aim is to show the perspectives of using this approach, not to conclude on the actual importance of per-
sistence for PH risk.
For a comparison like this, it is helpful to define a baseline model and then study the relative risk (RR) estimates of differ-
ent alternative scenarios. The baseline for the persistence model is an arbitrary choice. To allow calculations of the RRs, the
baseline should predict some cases, and allow scenarios that result in larger (and smaller) risk estimates. Here we decided
to choose the model version shown in Figure E.15 as a baseline. It assumes no growth or inactivation in FoPE1 and FoPE2
(a2 = 0; c1 = 0) and some C&D effect in FoPE1 (5 log every 20,000 food products). The baseline for the growth model during
storage is the one presented in Section E.3.1, with some minor adaptations: a storage time of 14 days, to get a risk estimate
> 0 in the baseline; the baseline for the DR model is an average strain in the whole population; as an initial population size
in the hidden environment FoPE2, due to a contamination event, we set N0 = 109 CFU.
The most relevant risk estimate is the expected number of cases, which is the mean of the probability of illness (Pill ) for all
food products modelled, multiplied by the number of food products. This expected number of cases for 100,000 products
is 0.1237 in the baseline. By dividing the risk estimate of each alternative scenario by the risk estimate of the baseline, the
RR of the alternative scenario is derived. Usually, the log of this RR is used in a comparative graphic (Figure E.17) to compare
risk increases and risk reductions in a fair way.
It is found that the largest increase in risk, as compared to the baseline, is found in the scenarios where a virulent strain is
modelled and where there is much growth in FoPE2 or during storage. Storage time and temperature may also have a high
impact on the risk estimate. The effects of changes in the transfer parameters of the persistence model (the b-parameters),
and the C&D regime are small, unless there is no transfer at all and the food products do not get contaminated. If the inac-
tivation rates in the FoPE (c1 and negative a2 values) are similar to the transfer rates, this reduces the risks a little.
These results suggest that, in this analysis, transfer and survival in the FoPE have less impact on the risk than possible
growth during storage and the virulence of the strain. However, growth in the FoPE, combined with (indirect) transfer to
the food products can have a high impact on the risk.
F I G U R E E .17 Relative risks of alternative scenarios, with changes in one or more model parameter value. Note: Upper red: change in dose–
response; orange: change in growth during storage; grey: change in persistence (growth and survival in FoPE); blue: change in transfer parameters;
green: change in C&D regime; dark blue: change in initial contamination; purple: change in serving size; lower red: initial contamination in FoPE1
instead of FoPE2. With a2 = 10−3.2, the population size increases beyond realistic limits and the risk as well (> 4 log relative risk). In ‘thorough C&D’,
a 5 log10 reduction is obtained in both FoPE1 and FoPE2.
A specific characteristic of persistence is that the risk is maintained over a long time period. By only analysing the mean
risk over the whole period, it is unclear whether the risk is equally spread out over all the food products, or it is high at the
beginning (or the end) of the process and low at the end (or the beginning). For that reason, the ‘case ratio’ was evaluated,
defined as the ratio of the expected number of cases in the last 20% of the food products and the first 20%. If this ratio
equals 1, the risk is the same over the whole period of food processing, which characterises ‘full’ persistence. A ratio below
1 indicates a decline in the risk, which implies decreasing persistence or absence of persistence with very low case ratios.
A ratio above 1 indicates an increasing risk with time which, in the course of time, cannot be sustainable as the risk will get
unrealistically high on the long run.
This case ratio for the different scenarios was again compared with the baseline, where the value is 0.0131 (which means
that the risk of the last 20,000 products is 1.3% of that of the first 20,000, a decline in the risk). Scenarios with case ratios with
higher values than the baseline have higher persistence than the baseline. The results are shown in Figure E.18.
F I G U R E E .1 8 Case ratios (expressed on a log scale) found in alternative scenarios, with changes in one or more model parameter value. Note:
Colours as in Figure E.17. The vertical blue line indicates the case ratio of the baseline. Scenarios with a case ratio = 1 (log case ratio = 0) do not show
any bar. The three scenarios with high case ratios (show as log case ratio = 1) have even higher case ratios, which imply a more than tenfold increase in
the risk.
Figure E.18 shows that case ratios close to one can occur due to persistence (loga2 positive and between −4.3 and −4). If
the growth in FoPE2 (a2) is larger than the equilibrium value,39 the risk increases during the process. A case ratio close to
one is also found in scenarios where the growth during storage reaches the maximum population density (MPD) (T = 10°C
or a higher maximum specific growth rate). Analyses where the food production is followed for a longer time (500,000 it-
erations instead of 100,000, data not shown) indicate that the persistence in high temperature scenarios is not maintained
on the long run, when the MPD is no longer reached when the contamination of processed food products declines. The
case ratio of 1 in the ‘no transfer’ scenario is an artefact, as there is no risk at all.
Figure E.19 combines Figures E.17 and E.18 and shows the association of the risk estimates of different scenarios and the
case ratios. It illustrates that both the risk estimate (log cases) and the case ratio are sensitive for the persistence parameter
a2 (white dots), and less for the transfer parameters. The risk estimate is also sensitive for the growth during storage and
the DR (virulence).
Interestingly, the transfer parameters seem to have little impact on the risk and higher impact on case ratios. This relates
to what has been discussed in Section E.4.1. The transfer rates have an impact on the observation of persistence, but not
so much on the PH risk.
39
The equilibrium value is derived in Section E.2.4. This is not exactly the equilibrium in the baseline model, as this includes cleaning, which adds an inactivation that is
not included in the equilibrium.
F I G U R E E .1 9 Scatter plot of the risk (the log of the number of cases found with the model, the basis of Figure E.17) and the case ratio (as in
Figure E.18) Red dot: baseline; white dots: results for different values of the persistence parameter a2; orange dots: scenarios with altered growth
during storage; blue dots: scenarios with altered transfer parameter values; black dots: scenarios with change in dose–response.
E.5.2 | Comparison of strains
Some of the model parameters can be interpreted as strain characteristics. This can be useful if the PH risk of persistent
strains is to be compared with that of strains with other characteristics.
Figure E.20 shows how knowledge on these characteristics translated into model parameter values could be used to com-
pare the relative risks of strains. The baseline is the same as in Section E.5.1. Further, we defined a persistent (a2 = 10−4.3)
and strongly persistent (a2 = 10−4) strain, a fast-growing strain (𝜇ref = 11.92, the 97.5th percentile of the distribution pro-
vided by Pouillot and Lubran (2011)), a fast-growing strain with high MPD (MPD = 9logCFU∕ g), a virulent strain (see Section
E.3.2.1) and combinations thereof.
In this particular example, the impact of persistence (characterised by parameter a2) on the risk is relatively small. Other
factors lead to larger risks, and larger values of a2 would induce too high concentrations of L. monocytogenes in the food
products (as shown by the grey bar).
F I G U R E E . 2 0 Comparison of risks and case ratios associated with different potential strain types. Note: P: a2 = 10−4.3; P+: a2 = 10−4; F:
𝜇ref = 11.92; V: virulent strain; H: MPD = 9logCFU ∕ g. The orange bars show the relative risks (as in Figure E.17); the grey bar is the relative maintenance
of risk of the indicated strain, which is the case ratio (the expected PH risk of the last 20% of food products divided by that of the first 20%) of this
strain divided by the case ratio of the baseline strain; the blue vertical line indicates a relative maintenance of risk corresponding to 100% (a case ratio
of 1, no change in PH risk associated to the first and last 20% of food products).
Although the figure may suggest that persistence is not a very important factor for the risk, this is not necessarily so. The base-
line assumes that there is 100% survival (and not growth) in FoPE2, without an effect of C&D in this environment. This is a form of
persistence as well, that is required to get an effect of growth during storage and virulence. A baseline without any persistence is
not useful for the model comparison, as it is arbitrary and would give a very low risk. However, it may be realistic in many cases.
E.5.3 | The Salmonella QMRA
For Salmonella in a LMF a similar approach was used as for the Listeria QMRA. As for Salmonella we have no DR model that
includes variation in virulence, and growth during storage does not occur in the LMF, the model is simpler. As a baseline,
we define a model with a2 = 0; b1 = b2 = b3 = 10−5 ; c1 = 0; logN0 = 7 and C&D of FoPE1 resulting in 5 log reduction every
20.000th food product. This results in an expected number of cases of 173.3 and a case ratio of 0.487. Hence, despite the
lower level of initial contamination of FoPE2, and the absence of growth during storage, the risk (number of cases) is higher
and the observed persistence (case ratio) is larger than for L. monocytogenes. This can be explained by the DR relation.
Results for different scenarios are shown in Figure E.21, as in Figures E.17 and E.18. Next, the association between the
number of cases and the case ratios found for different scenarios is shown in Figure E.22, as in Figure E.19.
F I G U R E E . 21 Relative risks (A) and case ratios (B) found for different scenarios in the Salmonella model.
F I G U R E E . 2 2 Scatter plot of the risk (the log of the number of cases found with the model) and the case ratio. Note: Red dot: baseline; white
dots: results for different values of the persistence parameter a2; blue dots: scenarios with altered transfer parameter values.
As for Listeria, results show that both the risk estimate (log cases) and the case ratio are sensitive for the persistence
parameter a2. Interestingly, for Salmonella the C&D of FoPE 1 has large effect on the case ratio, as without C&D it reaches a
value far above 1. If the initial contamination occurs in FoPE1, the overall risk is larger, but the case ratio is very low, which
means that all cases will occur for the first food products, so it will not appear as persistence. This effect is more pro-
nounced for Salmonella than for Listeria.
E.6 | Discussion
The model analyses above show how the persistence model can be integrated in a QMRA model, and how it can be used
to assess the impact of persistence. Although the analyses are still incomplete, they illustrate the perspectives of this type
of modelling to increase our understanding of the role of persistence in public health.
In the model analysis, a duality in the definition of persistence appeared. On the one hand, persistence defined as sur-
vival and, potentially, growth in the FoPE, is included in the model input by model parameter a2. At the other hand, persis-
tence is recognised as several model outputs, such as the long-term presence on the FoPEs, long-term contamination of
food products and ultimately by means of the case ratio, which shows whether the risk is pertained over a longer time. The
modelling results show that, as to be expected, the observed persistence is strongly related to the value of a2, the param-
eter defining the persistence (survival and growth potential in the FoPE) in the input. The transfer parameters (b) have a
less profound effect, but it was found that ‘moderate’ transfer rates are required to assure that the bacteria are not quickly
transferred out of the FPPE, or are not transferred to the food products at all.
It is notable how sensitive the model results are for the precise value of the parameter a2: for example, in the L. monocy-
togenes model, a case ratio close to 1 (say between 0.9 and 1.1) is only found for values of a2 between 104.214 and 104.194, i.e.
between 0.0000611 and 0.0000644, a very small range of values. The model predicts that, outside that range, persistence
either fades out on the (not so) long run or gives ever increasing concentrations (to unrealistically high levels) on the food
product. This small range seems to lack a biological basis, which means the model has difficulty in explaining long-lasting
persistence in the FoPE, as frequently observed. Possible explanations for this unrealistic feature of the model can be the
stochasticity of the process (see also Section E.2.5) and the fact that bacteria in a FoPE (often in biofilms) do not grow with
a constant rate as defined in the model. Too much growth and too high concentrations of the bacteria in a FoPE will be
slowed down by a lack of substrate, and parts of a biofilm may loosen from the FoPE in a stochastic manner. These types of
processes could be added to a more mechanistic version of the model.
In the analysis above, observed persistence is assessed by means of the case ratio, i.e. as the ‘persistence’ of the probabil-
ity of illness. Although persistence can be identified on the basis of long-lasting outbreaks from cases caused by the same
bacterial strain and traced back to the same source, over a longer time, it is usually defined on the basis of microbiological
samples from the FoPE. The model does not include this type of sampling from the FoPE. Next to the ambition to only do
a limited number of analyses, a reason for leaving out this sampling is its stochastic nature. Sampling cannot be included
meaningfully in a deterministic model like this, and when realistic stochastic random variation is added for the sampling,
but not for the ‘persistence dynamics’ in the FoPE (see Section E.2.5), the value of the analyses can be debated. Still, in fu-
ture work, it will be relevant to simulate sampling to identify persistence and compare it with observations.
The comparison of risks of different hypothetical L. monocytogenes strains (Section E.5.2) shows that persistent strains
(with higher values of a2) do lead to higher PH risks, but not more than virulent strains or strains that have a higher growth
potential at storage. Strains combining these traits (‘super bugs’) obviously lead to the highest risks. An interesting next
step would be to consider specific L. monocytogenes strains for which the characteristics in terms of persistence, virulence
and growth potential are known, and compare the PH risks that, according to the model, would be associated with them.
It may give insight into the actual occurrence of ‘super bugs’.
AN N E X A
Protocol for the assessment of the persistence of microbiological hazards in food and feed production and
processing environments
Annex A is available under the Supporting Information section on the online version of the scientific output.
The EFSA Journal is a publication of the European Food Safety

Authority, a European agency funded by the European Union

Persistencia de Peligros Biológicos

Uploaded by

Copyright:

Available Formats

Persistencia de Peligros Biológicos

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Persistencia de Peligros Biológicos

Uploaded by

Copyright:

Available Formats

Adopted: 6 December 2023

Persistence of microbiological hazards in food and feed

EFSA Panel on Biological Hazards (BIOHAZ) | Konstantinos Koutsoumanis | Ana Allende |

EFSA Journal. 2024;22:e8521. ﻿ efsa.onlinelibrary.wiley.com/journal/1831-4732 | 1 of 114

1 | I NTRO DUC TIO N

1.1 | Background and Terms of Reference as provided by the requestor

1.2 | Interpretation of the Terms of Reference

FIGURE 1 Conceptual map of the AQs to be addressed in the current assessment.

The AQs are as follows:

1.3 | Additional information

2.1.1.1 | Food production and processing sectors

2.1.1.2 | Feed production and processing sectors

2.1.2.1 | Food production and processing sectors

2.1.2.2 | Feed production and processing sectors

2.5 | Perspectives of integrating the information gathered in risk assessment (AQ6)

2.7 | Uncertainty analysis

Pathogens of highest PH relevance in the sector

Bacterial pathogen F M FS D E FV LMF

Pathogens of highest PH relevance in the sector

Bacterial pathogen F M FS D E FV LMF

Pathogens of highest PH relevance in sector and/or persisting in the FFPE of sector

Bacterial pathogen F M FS D E FV LMF

3.2.1 | L. monocytogenes subtypes and features

3.2.1.1 | Subtypes linked to persistence

3.2.1.2 | Features associated with persistence and link to subtypes

TA B L E 3 Overview of genetic features of L. monocytogenes lineage I possibly linked to persistence.

CC, STa 1 2 3 4 5 6 59 87 183 (ST382) 217 218 224 288 554

TA B L E 4 Overview of genetic features of L. monocytogenes lineage II possibly linked to persistence.

Hypo/ / / Hypovirulent / Hypervirulent Hypervirulent Hypovirulent / / Hypovirulent / / / / /

LIPI-­hypervirulent LIPI-­2-­inlI LIPI-­2, LIPI-­3 LIPI-­2-­inlI / / / LIPI-­2-­inlI LIPI-­2 LIPI-­2-­inlI LIPI-­2-­inlI / / / / /

3.2.2 | Salmonella enterica subtypes and features

3.2.2.1 | Subtypes linked to persistence

3.2.2.2 | Features associated with persistence and link to subtypes

3.2.3 | C. sakazakii subtypes and features

3.2.3.1 | Subtypes linked to persistence

3.2.3.2 | Features associated with persistence and link to subtypes

3.2.4 | Concluding remarks related to (sub)types and features

3.3.1 | L. monocytogenes in FoPE of four sectors

3.3.1.1 | Environmental niches or site of persistent L. monocytogenes strains

3.3.1.2 | Risk factors for persistence of L. monocytogenes

3.3.2 | Salmonella enterica in the FFPE of four sectors

3.3.2.1 | Environmental niches or location of persistent Salmonella enterica strains

3.3.2.2 | Risk factors for persistence of Salmonella enterica

3.3.3 | C. sakazakii in the FoPE of LMF sector

3.3.3.1 | Environmental niches or location of persistent C. sakazakii strains

3.3.3.2 | Risk factors for persistence of C. sakazakii

3.4.1 | Sampling and testing in the FFPE

3.4.2 | Hygienic measures

3.4.2.1 | PRP infrastructure (building, equipment)

3.4.2.2 | PRP cleaning and disinfection

3.4.2.3 | PRP technical maintenance and calibration

3.4.2.4 | PRP water and air control

3.4.2.5 | PRP personnel (hygiene, health status)

3.4.2.6 | PRP working methodology

3.4.2.7 | PRP food safety culture (FSC)

3.4.3 | Actions triggered by persistence

Study Persistence locations Interventions description Impact of the intervention

Study Persistence locations Interventions description Impact of the intervention

3.4.3.1 | Chemical interventions

3.4.3.2 | Physical interventions

EFSA Journal. 2024;22:e8521. efsa.onlinelibrary.wiley.com/journal/1831-4732 | 1 of 114

LIPI-hypervirulent LIPI-2-inlI LIPI-2, LIPI-3 LIPI-2-inlI / / / LIPI-2-inlI LIPI-2 LIPI-2-inlI LIPI-2-inlI / / / / /

3.5.1 | Bottom-up approach and data needs

3.5.2 | Top-down approach and data needs

AC K N OW L ED G EM E N T S

(3) Screening of full-text documents

(3) Screening of full-text documents

which is also a well-fitting approximation (see Figure E.11).