•Structuring of relevant terms in microbiological risk assessment modelling domain.•Definition of metadata and controlled vocabularies for annotation of QMRA models.•Harmonized model annotation will ...support knowledge exchange between software tools.•Establishment of curated food safety knowledge repository is proposed.•Need of rules for knowledge annotation essential for knowledge reusability.
In the last decades the microbial food safety community has developed a variety of valuable knowledge (e.g., mathematical models and data) and resources (e.g., databases and software tools) in the areas of quantitative microbial risk assessment (QMRA) and predictive microbiology. However, the reusability of this knowledge and the exchange of information between resources are currently difficult and time consuming. This problem has increased over time due to the lack of harmonized data format and rules for knowledge annotation. It includes the lack of a common understanding of basic terms and concepts and of a harmonized information exchange format to describe and annotate knowledge. The existence of ambiguities and inconsistencies in the use of terms and concepts in the QMRA and predictive microbial (PM) modelling necessitates a consensus on their refinement, which will allow a harmonized exchange of information within these areas. Therefore, this work aims to harmonize terms and concepts used in QMRA and PM modelling spanning from high level concepts as defined by Codex Alimentarius, Food and Agriculture Organization (FAO) and World Health Organization (WHO), up to terms generally used in statistics or data and software science. As a result, a harmonized schema for metadata that allows consistent annotation of data and models from these two domains is proposed. This metadata schema is also a key component of the Food Safety Knowledge Markup Language (FSK-ML), a harmonized format for information exchange between resources in the QMRA and PM modelling domain. This work is carried out within a research project that aims to establish a new community resource called Risk Assessment Modelling and Knowledge Integration Platform (RAKIP). This platform will facilitate the sharing and execution of curated QMRA and PM models using the foundation of the proposed harmonized metadata schema and information exchange format. Furthermore, it will also provide access to related open source software libraries, converter tools and software-specific import and export functions that promote the adoption of FSK-ML by the microbial food safety community. In the future, these resources will hopefully promote both the knowledge reusability and the high-quality information exchange between stakeholders within the areas of QMRA and PM modelling worldwide.
Superchilling entails lowering the fish temperature to between the initial freezing point of the fish and about 1–2°C lower. The temperature of superchilled fresh fishery products (SFFP) in boxes ...without ice was compared to that of products subject to the currently authorised practice in boxes with ice (CFFP) under the same conditions of on‐land storage and/or transport. A heat transfer model was developed and made available as a tool to identify under which initial configurations of SFFP the fish temperature, at any time of storage/transport, is lower or equal to CFFP. A minimum degree of superchilling, corresponding to an ice fraction in the fish matrix of SFFP equal or higher than the proportion of ice added per mass of fish in CFFP, will ensure with 99–100% certainty (almost certain) that the fish temperature of SFFP and the consequent increase of relevant hazards will be lower or equal to that of CFFP. In practice, the degree of superchilling can be estimated using the fish temperature after superchilling and its initial freezing point, which are subject to uncertainties. The tool can be used as part of ‘safety‐by‐design’ approach, with the reliability of its outcome being dependent on the accuracy of the input data. An evaluation of methods capable of detecting whether a previously frozen fish is commercially presented as ‘superchilled’ was carried out based on, amongst others, their applicability for different fish species, ability to differentiate fresh fish from fish frozen at different temperatures, use as a stand‐alone method, ease of use and classification performance. The methods that were considered ‘fit for purpose’ are Hydroxyacyl‐coenzyme A dehydrogenase (HADH) test, α‐glucosidase test, histology, ultraviolet–visible–near–infrared (UV‐VIS/NIR) spectroscopy and hyperspectral imaging. These methods would benefit from standardisation, including the establishment of threshold values or classification algorithms to provide a practical routine test.
Shifting from traditional hazard-based food safety management toward risk-based management requires statistical methods for evaluating intermediate targets in food production, such as microbiological ...criteria (MC), in terms of their effects on human risk of illness. A fully risk-based evaluation of MC involves several uncertainties that are related to both the underlying Quantitative Microbiological Risk Assessment (QMRA) model and the production-specific sample data on the prevalence and concentrations of microbes in production batches. We used Bayesian modeling for statistical inference and evidence synthesis of two sample data sets. Thus, parameter uncertainty was represented by a joint posterior distribution, which we then used to predict the risk and to evaluate the criteria for acceptance of production batches. We also applied the Bayesian model to compare alternative criteria, accounting for the statistical uncertainty of parameters, conditional on the data sets. Comparison of the posterior mean relative risk, E(RR|data) = E (P (illness| criterion is met)/P (illness|data), and relative posterior risk, RPR = P (illness|data, criterion is met)/P (illness|data), showed very similar results, but computing is more efficient for RPR. Based on the sample data, together with the QMRA model, one could achieve a relative risk of 0.4 by insisting that the default criterion be fulfilled for acceptance of each batch.
The qualified presumption of safety (QPS) approach was developed to provide a regularly updated generic pre‐evaluation of the safety of microorganisms, intended for use in the food or feed chains, to ...support the work of EFSA's Scientific Panels. The QPS approach is based on an assessment of published data for each agent, with respect to its taxonomic identity, the body of relevant knowledge, safety concerns and occurrence of antimicrobial resistance. Safety concerns identified for a taxonomic unit (TU) are, where possible, confirmed at the species/strain or product level and reflected by ‘qualifications’. In the period covered by this statement, no new information was found that would change the status of previously recommended QPS TUs. Of the 50 microorganisms notified to EFSA in October 2021 to March 2022 (inclusive), 41 were not evaluated: 10 filamentous fungi, 1 Enterococcus faecium, 1 Clostridium butyricum, 3 Escherichia coli and 1 Streptomyces spp. because are excluded from QPS evaluation, and 25 TUs that have already a QPS status. Nine notifications, corresponding to seven TUs were evaluated: four of these, Streptococcus salivarius, Companilactobacillus formosensis, Pseudonocardia autotrophica and Papiliotrema terrestris, being evaluated for the first time. The other three, Microbacterium foliorum, Pseudomonas fluorescens and Ensifer adhaerens were re‐assessed. None of these TUs were recommended for QPS status: Ensifer adhaerens, Microbacterium foliorum, Companilactobacillus formosensis and Papiliotrema terrestris due to a limited body of knowledge, Streptococcus salivarius due to its ability to cause bacteraemia and systemic infection that results in a variety of morbidities, Pseudonocardia autotrophica due to lack of body of knowledge and uncertainty on the safety of biologically active compounds which can be produced, and Pseudomonas fluorescens due to possible safety concerns.
Listeria monocytogenes (in the meat, fish and seafood, dairy and fruit and vegetable sectors), Salmonella enterica (in the feed, meat, egg and low moisture food sectors) and Cronobacter sakazakii (in ...the low moisture food sector) were identified 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 zoning 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 implementation 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. monocytogenes 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 research are provided.
The specific concentrations of tetracycline, chlortetracycline, oxytetracycline and doxycycline in non‐target feed for food‐producing animals, below which there would not be an effect on the ...emergence of, and/or selection for, resistance in bacteria relevant for human and animal health, as well as the specific antimicrobial concentrations in feed which have an effect in terms of growth promotion/increased yield were assessed by EFSA in collaboration with EMA. Details of the methodology used for this assessment, associated data gaps and uncertainties are presented in a separate document. To address antimicrobial resistance, the Feed Antimicrobial Resistance Selection Concentration (FARSC) model developed specifically for the assessment was applied. The FARSC for these four tetracyclines was estimated. To address growth promotion, data from scientific publications obtained from an extensive literature review were used. Levels in feed that showed to have an effect on growth promotion/increased yield were reported for tetracycline, chlortetracycline, oxytetracycline, whilst for doxycycline no suitable data for the assessment were available. Uncertainties and data gaps associated with the levels reported were addressed. It was recommended to perform further studies to supply more diverse and complete data related to the requirements for calculation of the FARSC for these antimicrobials.
Vibrio parahaemolyticus, Vibrio vulnificus and non‐O1/non‐O139 Vibrio cholerae are the Vibrio spp. of highest relevance for public health in the EU through seafood consumption. Infection with V. ...parahaemolyticus is associated with the haemolysins thermostable direct haemolysin (TDH) and TDH‐related haemolysin (TRH) and mainly leads to acute gastroenteritis. V. vulnificus infections can lead to sepsis and death in susceptible individuals. V. cholerae non‐O1/non‐O139 can cause mild gastroenteritis or lead to severe infections, including sepsis, in susceptible individuals. The pooled prevalence estimate in seafood is 19.6% (95% CI 13.7–27.4), 6.1% (95% CI 3.0–11.8) and 4.1% (95% CI 2.4–6.9) for V. parahaemolyticus, V. vulnificus and non‐choleragenic V. cholerae, respectively. Approximately one out of five V. parahaemolyticus‐positive samples contain pathogenic strains. A large spectrum of antimicrobial resistances, some of which are intrinsic, has been found in vibrios isolated from seafood or food‐borne infections in Europe. Genes conferring resistance to medically important antimicrobials and associated with mobile genetic elements are increasingly detected in vibrios. Temperature and salinity are the most relevant drivers for Vibrio abundance in the aquatic environment. It is anticipated that the occurrence and levels of the relevant Vibrio spp. in seafood will increase in response to coastal warming and extreme weather events, especially in low‐salinity/brackish waters. While some measures, like high‐pressure processing, irradiation or depuration reduce the levels of Vibrio spp. in seafood, maintaining the cold chain is important to prevent their growth. Available risk assessments addressed V. parahaemolyticus in various types of seafood and V. vulnificus in raw oysters and octopus. A quantitative microbiological risk assessment relevant in an EU context would be V. parahaemolyticus in bivalve molluscs (oysters), evaluating the effect of mitigations, especially in a climate change scenario. Knowledge gaps related to Vibrio spp. in seafood and aquatic environments are identified and future research needs are prioritised.
The qualified presumption of safety (QPS) process was developed to provide a safety assessment approach for microorganisms intended for use in food or feed chains. In the period covered by this ...statement, no new information was found that would change the status of previously recommended QPS TUs. The TUs in the QPS list were updated based on a verification, against their respective authoritative databases, of the correctness of the names and completeness of synonyms. A new procedure has been established to ensure the TUs are kept up to date in relation to recent taxonomical insights. Of 81 microorganisms notified to EFSA between October 2023 and March 2024 (45 as feed additives, 25 as food enzymes or additives, 11 as novel foods), 75 were not evaluated because: 15 were filamentous fungi, 1 was Enterococcus faecium, 10 were Escherichia coli, 1 was a Streptomyces (all excluded from the QPS evaluation) and 46 were TUs that already have a QPS status. Two of the other eight notifications were already evaluated for a possible QPS status in the previous Panel Statement: Heyndrickxia faecalis (previously Weizmannia faecalis) and Serratia marcescens. One was notified at genus level so could not be assessed for QPS status. The other five notifications belonging to five TUs were assessed for possible QPS status. Akkermansia muciniphila and Actinomadura roseirufa were still not recommended for QPS status due to safety concerns. Rhizobium radiobacter can be recommended for QPS status with the qualification for production purposes. Microbacterium arborescens and Burkholderia stagnalis cannot be included in the QPS list due to a lack of body of knowledge for its use in the food and feed chain and for B. stagnalis also due to safety concerns. A. roseirufa and B. stagnalis have been excluded from further QPS assessment.
Surveillance data published since 2010, although limited, showed that there is no evidence of zoonotic parasite infection in market quality Atlantic salmon, marine rainbow trout, gilthead seabream, ...turbot, meagre, Atlantic halibut, common carp and European catfish. No studies were found for greater amberjack, brown trout, African catfish, European eel and pikeperch. Anisakis pegreffii, A. simplex (s. s.) and Cryptocotyle lingua were found in European seabass, Atlantic bluefin tuna and/or cod, and Pseudamphistomum truncatum and Paracoenogonimus ovatus in tench, produced in open offshore cages or flow‐through ponds or tanks. It is almost certain that fish produced in closed recirculating aquaculture systems (RAS) or flow‐through facilities with filtered water intake and exclusively fed heat‐treated feed are free of zoonotic parasites. Since the last EFSA opinion, the UV‐press and artificial digestion methods have been developed into ISO standards to detect parasites in fish, while new UV‐scanning, optical, molecular and OMICs technologies and methodologies have been developed for the detection, visualisation, isolation and/or identification of zoonotic parasites in fish. Freezing and heating continue to be the most efficient methods to kill parasites in fishery products. High‐pressure processing may be suitable for some specific products. Pulsed electric field is a promising technology although further development is needed. Ultrasound treatments were not effective. Traditional dry salting of anchovies successfully inactivated Anisakis. Studies on other traditional processes – air‐drying and double salting (brine salting plus dry salting) – suggest that anisakids are successfully inactivated, but more data covering these and other parasites in more fish species and products is required to determine if these processes are always effective. Marinade combinations with anchovies have not effectively inactivated anisakids. Natural products, essential oils and plant extracts, may kill parasites but safety and organoleptic data are lacking. Advanced processing techniques for intelligent gutting and trimming are being developed to remove parasites from fish.
To reduce consumer health risks from foodborne diseases that result from improper domestic food handling, consumers need to know how to safely handle food. To realize improvements in public health, ...it is necessary to develop interventions that match the needs of individual consumers. Successful intervention strategies are therefore contingent on identifying not only the practices that are important for consumer protection, but also barriers that prevent consumers from responding to these interventions. A measure of food safety behavior is needed to assess the effectiveness of different intervention strategies across different groups of consumers. A nationally representative survey was conducted in the Netherlands to determine which practices are likely conducted by which consumers. Participants reported their behaviors with respect to 55 different food‐handling practices. The Rasch modeling technique was used to determine a general measure for the likelihood of an average consumer performing each food‐handling behavior. Simultaneously, an average performance measure was estimated for each consumer. These two measures can be combined to predict the likelihood that an individual consumer engages in a specific food‐handling behavior. A single “food safety” dimension was shown to underlie all items. Some potentially safe practices (e.g., use of meat thermometers) were reported as very difficult, while other safe practices were conducted by respondents more frequently (e.g., washing of fresh fruit and vegetables). A cluster analysis was applied to the resulting data set, and five segments of consumers were identified. Different behaviors may have different effects on microbial growth in food, and thus have different consequences for human health. Once the microbial relevance of the different consumer behaviors has been confirmed by experiments and modeling, the scale developed in the research reported here can be used to develop risk communication targeted to the needs of different consumer groups, as well as to measure the efficacy of different interventions.