Globalization has increased the volume and diversity of food traded internationally, highlighting the need for robust, risk-based monitoring strategies for imported foods. Traditional inspection systems tend to focus on pathogens that are regulated, cause outbreaks or lead to small numbers of severe individual cases of disease. As a result, they often overlook the broader disease burden associated with pathogens that are not regulated, mostly cause sporadic cases or cause large numbers of mild symptoms Here, we present a novel approach to prioritize public health risks in imported food using Disability-Adjusted Life Years (DALYs) as a metric. Our criterion integrates worldwide available epidemiological incidence estimations, food attribution estimates, consumption data in exporting countries, import volumes, pathogen inactivation during frozen transport and DALY per case in the importing country. As an illustration, we apply this DALY criterion to major meat categories and pathogens relevant to Dutch public health, enabling the ranking of unique combinations of pathogen, food group, mode of transport, and geographical regions in terms of risk. Results reveal that poultry, beef and meat of small ruminants imports are the most important, mainly due to Salmonella, Campylobacter and Toxoplasma. The approach is intended for comparative, screening-level prioritization and provides actionable insights to complement traditional import inspection systems. Key assumptions and limitations are discussed, including possible differences in food safety practices, preparation styles and immunity across countries. The DALY criterion offers a transparent and flexible tool to guide resource allocation in food import monitoring using public health considerations.

Quantitative Microbiological Risk assessment (QMRA) models are essential tools for setting up mitigation strategies. Traditional QMRA modelling approaches do not account for the correlation between genetic traits and variability among pathogens, potentially leading to over- or underestimation of microbial exposure and associated risks. We aimed to integrate genomic data into QMRA to propagate bacterial strain variability and update the existing framework of QMRA, following a Next Generation Risk Assessment (NGRA) approach. We used a benchmark QMRA model describing the prevalence and concentration of Campylobacter jejuni on chicken in all stages from farm-to-fork, to model the risk of infection and illness related to consumption of chicken meat. We integrated extended the storage step, to account for genetic variability in cold inactivation by incorporating gene-level genomic data associated with cold tolerance, derived from literature and a large C. jejuni genomic dataset, into the traditional QMRA model by setting up cold inactivation curves from existing data to map the relationship between the number of cold tolerance genes and temperature-dependent inactivation. The predicted number of cases was 8822 human cases/year in the benchmark QMRA model. The contamination of meat with C. jejuni strains having lower cold tolerance genes can reduce the expected number of human campylobacteriosis cases up to 100%; on the other hand, higher number of cold tolerance genes resulted in an increase up to 335.8% on the expected number of cases. Although our results are based on simulations, we show a potential implementation of the genetic information into QMRA, linking risk estimates with whole-genome sequencing data. More research is needed to understand how genetic features shape phenotypical characteristics, which is one of the main uncertainties in the current NGRA model, and to further explore the implications for risk management.

Extended-spectrum beta-lactamase (ESBL)-producing Escherichia coli (E. coli) in animals are considered a human health threat, because this type of bacteria can serve as a reservoir of antibiotic resistant genes and act as a continuous threat of the emergence of new resistant bacteria, in addition to the direct effect of making infection untreatable. Although the prevalence of ESBL producing bacteria in broilers was drastically reduced in the Netherlands, chicken meat still has the highest prevalence among meat products. Therefore, further control of the ESBL-producing E. coli in the broiler production chain is important to reduce public health risks. The main objectives of this study were to evaluate the effectiveness of intervention scenarios to reduce the transmission of ESBL-producing E. coli in the broiler production chain and to quantitatively estimate the risk to public health. In this study, we developed two different types of transmission models that described the observed time-related decline in prevalence during a production round: one with time-dependent decline in susceptibility and one with partial immunity to phylogenetic groups. Both models incorporated the environmental contamination effect between production rounds and within flocks. The parameter values, including transmission rate and recovery rate, were estimated by Approximate Bayesian computation (ABC) method using data from a longitudinal study in a Dutch organic broiler farm. We applied the models to the three production stages in the broiler production chain, beginning from the Parent Stock (PS) farms, the hatcheries, and to the broiler farms. In our models, eggs were collected from different parent stock farms and transported to the hatchery and from there to a broiler farm.The size of a flock and the number of farms were adjusted to the Dutch situation. Both models were able to describe the observed dynamics within and between the production stages equally well, with estimated ESBL-producing E. coli prevalence of 8.98% and 11.47% in broilers at slaughter and 0.12% and 0.15% in humans due to chicken consumption. Both models indicated that improving farm management to eliminate the bacteria from the environment was the most effective intervention, making this outcome robust. Although chicken meat consumption is not a major risk factor for human carriage of the bacteria according to our models, reducing the bacteria in the PS and broiler farm environment to at least one percent can further decrease the prevalence in humans.

A quantitative microbiological risk assessment model for the cross-contamination transmission route in the kitchen (KCC) is presented. Bacteria are transmitted from contaminated (chicken) meat to hands, kitchen utensils, and other surfaces, subsequently contaminating a salad. The model aims to estimate the fraction of bacteria on the meat that is ingested due to cross-contamination, determine the importance of the different transmission routes, and assess the effect of scenarios (interventions) on the fraction ingested. The cross-contamination routes defined, bacterial source-to-recipient transfer fractions as available and derived from literature, and important characteristics (e.g., washing in cold water vs. hot water with soap) shaped the KCC model. With this model, 32 scenarios of an eight-step preparation of a “meat and salad” meal in a domestic kitchen were stochastically simulated. The “cutting board-salad” route proved dominant and the salad plays a major role in the final exposure. A realistic scenario (washing hands, cutting board, and knife with cold water after cutting the meat) estimates that a mean fraction of 3.2E − 3 of the bacteria on the meat is ingested. In the case of “hand washing with hot water and soap” and “cutting board and knife replacement,” the mean fraction ingested is 3.6E − 6. For a subsequent meal, where the contaminated sources were kitchen fomites, the estimated mean fraction is 4.3E − 4. In case of hamburger, part of the bacteria is unavailable for cross-contamination, resulting in a mean fraction ingested of about 5.4E − 5. The role of the dishcloth in cross-contamination transmission proved to be minor.

A quantitative microbiological risk assessment model for the cross-contamination transmission route in the kitchen (KCC) is presented. Bacteria are transmitted from contaminated (chicken) meat to hands, kitchen utensils, and other surfaces, subsequently contaminating a salad. The model aims to estimate the fraction of bacteria on the meat that is ingested due to cross-contamination, determine the importance of the different transmission routes, and assess the effect of scenarios (interventions) on the fraction ingested. The cross-contamination routes defined, bacterial source-to-recipient transfer fractions as available and derived from literature, and important characteristics (e.g., washing in cold water vs. hot water with soap) shaped the KCC model. With this model, 32 scenarios of an eight-step preparation of a "meat and salad" meal in a domestic kitchen were stochastically simulated. The "cutting board-salad" route proved dominant and the salad plays a major role in the final exposure. A realistic scenario (washing hands, cutting board, and knife with cold water after cutting the meat) estimates that a mean fraction of 3.2E - 3 of the bacteria on the meat is ingested. In the case of "hand washing with hot water and soap" and "cutting board and knife replacement," the mean fraction ingested is 3.6E - 6. For a subsequent meal, where the contaminated sources were kitchen fomites, the estimated mean fraction is 4.3E - 4. In case of hamburger, part of the bacteria is unavailable for cross-contamination, resulting in a mean fraction ingested of about 5.4E - 5. The role of the dishcloth in cross-contamination transmission proved to be minor.

Quantitative microbiological risk assessment (QMRA) methodology aims to estimate and describe the transmission of pathogenic microorganisms from animals and food to humans. In microbiological literature, the availability of whole genome sequencing (WGS) data is rapidly increasing, and incorporating this data into QMRA has the potential to enhance the reliability of risk estimates. This study provides insight into which are the key pathogen properties for incorporating WGS data to enhance risk estimation, through examination of example risk assessments for important foodborne pathogens: Listeria monocytogenes (Lm), Salmonella, Campylobacter and Shiga toxin-producing Escherichia coli. By investigating the relationship between phenotypic pathogen properties and genetic traits, a better understanding was gained regarding their impact on risk assessment. Virulence of Lm was identified as a promising property for associating different symptoms observed in humans with specific genotypes. Data from a genome-wide association study were used to correlate lineages, serotypes, sequence types, clonal complexes and the presence or absence of virulence genes of each strain with patient's symptoms. We also investigated the effect of incorporating WGS data into a QMRA model including relevant genomic traits of Lm, focusing on the dose-response phase of the risk assessment model, as described with the case/exposure ratio. The results highlighted that WGS studies which include phenotypic information must be encouraged, so as to enhance the accuracy of QMRA models. This study also underscores the importance of executing more risk assessments that consider the ongoing advancements in OMICS technologies, thus allowing for a closer investigation of different bacterial subtypes relevant to human health.

Extended-spectrum beta-lactamase (ESBL)-producing Escherichia coli (E. coli) in animals are considered a human health threat, because this type of bacteria can serve as a reservoir of antibiotic resistant genes and act as a continuous threat of the emergence of new resistant bacteria, in addition to the direct effect of making infection untreatable. Although the prevalence in broilers was drastically reduced, chicken meat still has the highest prevalence among meat products. Therefore, further control of the ESBL-producing E. coli in the broiler production chain is important to reduce public health risks. The main objectives of this study were to evaluate the effectiveness of intervention scenarios to reduce the transmission of ESBL-producing E. coli in the broiler production chain and to quantitatively estimate the risk to public health. In this study, we developed two different types of transmission models that described the observed time-related decline in prevalence during a production round: one with time-dependent decline in susceptibility and one with partial immunity to phylogenetic groups. Both models incorporated the environmental contamination effect between production rounds and within flocks. The parameter values, including transmission rate and recovery rate, were estimated by Approximate Bayesian computation (ABC) method using data from a longitudinal study in a Dutch broiler farm. We applied the models to the broiler production chain and further added the effect of mixing eggs and chicks from different origins and set the size of a flock and the number of farms according to the Dutch situation. Both models were able to describe the observed dynamics within and between the production stages equally well and estimated the outcome of interventions quantitatively. Both models indicated that improving farm management to eliminate the bacteria from the environment was the most effective intervention, making this outcome robust. According to our models, chicken meat consumption was not a major risk factor for human carriage of the bacteria.

Numerous source attribution studies for foodborne pathogens based on epidemiological and microbiological methods are available. These studies provide empirical data for modelling frameworks that synthetize the quantitative evidence at our disposal and reduce reliance on expert elicitations. Here, we develop a statistical model within a Bayesian estimation framework to integrate attribution estimates from expert elicitations with estimates from microbial subtyping and case-control studies for sporadic infections with four major bacterial zoonotic pathogens in the Netherlands (Campylobacter, Salmonella, Shiga toxin-producing E. coli [STEC] O157 and Listeria). For each pathogen, we pooled the published fractions of human cases attributable to each animal reservoir from the microbial subtyping studies, accounting for the uncertainty arising from the different typing methods, attribution models, and year(s) of data collection. We then combined the population attributable fractions (PAFs) from the case-control studies according to five transmission pathways (domestic food, environment, direct animal contact, human-human transmission and travel) and 11 groups within the foodborne pathway (beef/lamb, pork, poultry meat, eggs, dairy, fish/shellfish, fruit/vegetables, beverages, grains, composite foods and food handlers/vermin). The attribution estimates were biologically plausible, allowing the human cases to be attributed in several ways according to reservoirs, transmission pathways and food groups. All pathogens were predominantly foodborne, with Campylobacter being mostly attributable to the chicken reservoir, Salmonella to pigs (albeit closely followed by layers), and Listeria and STEC O157 to cattle. Food-wise, the attributions reflected those at the reservoir level in terms of ranking. We provided a modelling solution to reach consensus attribution estimates reflecting the empirical evidence in the literature that is particularly useful for policy-making and is extensible to other pathogens and domains.

Background Extended-spectrum beta-lactamase-producing Escherichia coli (ESBL-EC), plasmid-mediated AmpCproducing E coli (pAmpC-EC), and other bacteria are resistant to important beta-lactam antibiotics. ESBL-EC and pAmpC-EC are increasingly reported in animals, food, the environment, and community-acquired and health-careassociated human infections. These infections are usually preceded by asymptomatic carriage, for which attributions to animal, food, environmental, and human sources remain unquantified. Methods In this population-based modelling study, we collected ESBL and pAmpC gene data on the Netherlands population for 2005-17 from published datasets of gene occurrences in E coli isolates from different sources, and from partners of the ESBL Attribution Consortium and the Dutch National Antimicrobial Surveillance System. Using these data, we applied an established source attribution model based on ESBL-EC and pAmpC-EC prevalence and gene data for humans, including high-risk populations (ie, returning travellers, clinical patients, farmers), farm and companion animals, food, surface freshwater, and wild birds, and human exposure data, to quantify the overall and gene-specific attributable sources of community-acquired ESBL-EC and pAmpC-EC intestinal carriage. We also used a simple transmission model to determine the basic reproduction number (R-0) in the open community. Findings We identified 1220 occurrences of ESBL-EC and pAmpC-EC genes in humans, of which 478 were in clinical patients, 454 were from asymptomatic carriers in the open community, 103 were in poultry and pig farmers, and 185 were in people who had travelled out of the region. We also identified 6275 occurrences in non-human sources, including 479 in companion animals, 4026 in farm animals, 66 in wild birds, 1430 from food products, and 274 from surface freshwater. Most community-acquired ESBL-EC and pAmpC-EC carriage was attributed to human-to-human transmission within or between households in the open community (60 center dot 1%, 95% credible interval 40 center dot 0-73 center dot 5), and to secondary transmission from high-risk groups (6.9%, 4.1-9.2). Food accounted for 18.9% (7.0-38.3) of carriage, companion animals for 7.9% (1.4-19.9), farm animals (non-occupational contact) for 3.6% (0.6-9.9), and swimming in freshwater and wild birds (ie, environmental contact) for 2.6% (0.2-8.7). We derived an R-0 of 0.63 (95% CI 0.42-0.77) for intracommunity transmission. Interpretation Although humans are the main source of community-acquired ESBL-EC and pAmpC-EC carriage, the attributable non-human sources underpin the need for longitudinal studies and continuous monitoring, because intracommunity ESBL-EC and pAmpC-EC spread alone is unlikely to be self-maintaining without transmission to and from non-human sources. Funding 1Health4Food, Dutch Ministry of Economic Affairs, and the EU's Horizon-2020 through One-Health European Joint Programme. Copyright (c) 2019 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY-NC-ND 4.0 license.

Microbial survival of heating and cross-contamination are the two transmission routes during food preparation in the consumers' kitchen that are relevant for QMRA (Quantitative Microbial Risk Assessment). The aim of the present study was to extend the limited amount of data on microbial survival during real-life preparation of meat and meat products and to obtain accessory temperature data that allow for a more general (product unspecific) approach. Therefore survival data were combined with extensive measurements of time- and location dependent temperature using an infrared camera for the surface and buttons for the inside of the product, supplemented with interpolation modelling. We investigated the survival of heating of Escherichia coli O111:H2 in beefsteak, hamburgers (beef and 50% beef 50% pork (HH)), meatballs (beef and HH) and crumbs (HH).

EFSA regards the household as a stage in the food chain that is important for the final number of food-borne infections. The fate of a pathogen in the private kitchen largely depends on consumer hygiene during preparation of food and on its proper cooking, especially in the case of meat. Unfortunately, detailed information on the microbiological survival in meat products after heating in the consumer kitchen is lacking. The aim of the study was to improve the estimation of the inactivating effect on pathogens by heating meat or a meat product by the consumer in the kitchen. On that account, artificially contaminated meat and meat products were cooked according to several degrees of doneness and simulating real world conditions, and bacterial survival was measured. Heat camera pictures and button temperature loggers inserted into the food matrix served to record time and the temperature of heating. Temperature, time and the microbial survival ratio observed served to inform a mathematical model able to explain the thermal inactivation of meat or a meat product in home settings. The results of the study would help to improve microbiological comparative exposure assessments of pathogens in food, as an attribution tool and as a supportive tool for risk-based sampling in monitoring and surveillance.

A methodology is presented to optimize a sampling plan for retail products to monitor the prevalence of foodborne pathogens in relation to disease burden. The optimization procedure links an exposure assessment, a quantitative measure for disease burden (i.e. Disability Adjusted Life Years), number of samples to be analyzed (based on uncertainty of the prevalence estimate) and costs for sample analysis. The methodology attributes DALY's on ‘pathogen-matrix level’ to ‘pathogen-food products’ using the exposure assessment. The subsequent procedure includes the number of samples that need to be analyzed per retail product such that the prevalences can be monitored within a preset uncertainty bound, and the costs per sample. The final optimization step sorts pathogen-product combinations using the costs per DALY criterion which results in a monitoring program with a maximum number of DALY's given a certain amount of money. An optimized sampling plan was established for four foodborne pathogens on meat products: Campylobacter in pork and poultry meat, Salmonella in pork, Toxoplasma in pork and Shiga-toxin producing Escherichia coli (STEC) O157 in beef, veal, and mutton/lamb. Results show that Campylobacter on poultry products and one Toxoplasma – pork combination consitute the top 10 in the proposed Dutch public health risk meat monitoring program. This optimized sampling plan monitors 98% of the total amount of DALY's attributed to the considered pathogen-animal species combinations. At the same time, the procedure gives insight in how the preset optimization criteria leads to the proposed set of pathogen-product combinations. An iterative implementation of updated model input (prevalences, DALY estimates) will lead to an up-to-date optimized risk based monitoring program.

[This corrects the article DOI: 10.1371/journal.pone.0193834.].

We developed an improved simplified Quantitative Microbiological Risk Assessment (QMRA) model and tool with reduced data need, applicable to any pathogen - food product combination and in addition suitable for basic QMRA education. The swift QMRA (sQMRA2) model follows pathogen numbers through part of the food chain, starting at the retail phase, and ends with the estimated number of human cases of illness. The accompanying tool was implemented in Excel/@Risk. Relative risk (compared to other pathogen-food product combinations) rather than absolute risk was considered the most useful model output. The model includes storage at home (categories: room/fridge/freezer), cross contamination (yes/no) and heating (done/undercooked/raw) during preparation in the kitchen and a dose response relationship (Binomial/Beta-Binomial). The model also includes variability, e.g. of pathogen concentration and food product heating (time, temperature) in the kitchen. The general setup of the sQMRA2 tool consists of 14 consecutive (sets of) questions for values of parameters and per phase detailed intermediate model output broken down into categories. On a separate sheet, attribution of storage, cross-contamination and heating transmission routes in terms of exposure (probability of a contaminated portion, number of cfu) and number of human cases are presented. Further, intermediate exposures (number of contaminated portions, number of cfu) and final risks (number of human cases, DALYs, cost of illness), relative as well as absolute, are given.sQMRA2 is useful for quickly obtaining relative public health risk QMRA estimates of multiple pathogen - food combinations, which can be directly useful for risk management in terms of attribution or for the selection of high risk candidates for the application of extensive QMRA. It is also useful for educational purposes because of the insightful presentation of intermediate and final model output. As an example, sQMRA2 calculations were given for Campylobacter and Salmonella in chicken fillet, filet americain and table eggs. (C) 2016 Elsevier Ltd. All rights reserved.

The presence of extended-spectrum β-lactamase (ESBL) and plasmidic AmpC (pAmpC) producing Escherichia coli (EEC) in food animals, especially broilers, has become a major public health concern. The aim of the present study was to quantify the EEC exposure of humans in The Netherlands through the consumption of meat from different food animals. Calculations were done with a simplified Quantitative Microbiological Risk Assessment (QMRA) model. The model took the effect of pre-retail processing, storage at the consumers home and preparation in the kitchen (cross-contamination and heating) on EEC numbers on/in the raw meat products into account. The contribution of beef products (78%) to the total EEC exposure of the Dutch population through the consumption of meat was much higher than for chicken (18%), pork (4.5%), veal (0.1%) and lamb (0%). After slaughter, chicken meat accounted for 97% of total EEC load on meat, but chicken meat experienced a relatively large effect of heating during food preparation. Exposure via consumption of filet americain (a minced beef product consumed raw) was predicted to be highest (61% of total EEC exposure), followed by chicken fillet (13%). It was estimated that only 18% of EEC exposure occurred via cross-contamination during preparation in the kitchen, which was the only route by which EEC survived for surface-contaminated products. Sensitivity analysis showed that model output is not sensitive for most parameters. However, EEC concentration on meat other than chicken meat was an important data gap. In conclusion, the model assessed that consumption of beef products led to a higher exposure to EEC than chicken products, although the prevalence of EEC on raw chicken meat was much higher than on beef. The (relative) risk of this exposure for public health is yet unknown given the lack of a modelling framework and of exposure studies for other potential transmission routes.

In this article we present a model for Salmonella contamination of pig carcasses in the slaughterhouse. This model forms part of a larger QMRA (quantitative microbial risk assessment) on Salmonella in slaughter and breeder pigs, which uses a generic model framework that can be parameterized for European member states, to describe the entire chain from farm‐to‐consumption and the resultant human illness. We focus on model construction, giving mathematical formulae to describe Salmonella concentrations on individual pigs and slaughter equipment at different stages of the slaughter process. Variability among individual pigs and over slaughterhouses is incorporated using statistical distributions, and simulated by Monte Carlo iteration. We present the results over the various slaughter stages and show that such a framework is especially suitable to investigate the effect of various interventions. In this article we present the results of the slaughterhouse module for two case study member states. The model outcome represents an increase in average prevalence of Salmonella contamination and Salmonella numbers at dehairing and a decrease of Salmonella numbers at scalding. These results show good agreement when compared to several other QMRAs and microbiological studies.

A farm-to-consumption quantitative microbiological risk assessment (QMRA) for Salmonella in pigs in the European Union has been developed for the European Food Safety Authority. The primary aim of the QMRA was to assess the impact of hypothetical reductions of slaughter-pig prevalence and the impact of control measures on the risk of human Salmonella infection. A key consideration during the QMRA development was the characterization of variability between E.U. Member States (MSs), and therefore a generic MS model was developed that accounts for differences in pig production, slaughterhouse practices, and consumption patterns. To demonstrate the parameterization of the model, four case study MSs were selected that illustrate the variability in production of pork meat and products across MSs. For the case study MSs the average probability of illness was estimated to be between 1 in 100,000 and 1 in 10 million servings given consumption of one of the three product types considered (pork cuts, minced meat, and fermented ready-to-eat sausages). Further analyses of the farm-to-consumption QMRA suggest that the vast majority of human risk derives from infected pigs with a high concentration of Salmonella in their feces (≥10(4) CFU/g). Therefore, it is concluded that interventions should be focused on either decreasing the level of Salmonella in the feces of infected pigs, the introduction of a control step at the abattoir to reduce the transfer of feces to the exterior of the pig, or a control step to reduce the level of Salmonella on the carcass post-evisceration.

The disease burden of pathogens as estimated by QMRA (quantitative microbial risk assessment) and EA (epidemiological analysis) often differs considerably. This is an unsatisfactory situation for policymakers and scientists. We explored methods to obtain a unified estimate using campylobacteriosis in the Netherlands as an example, where previous work resulted in estimates of 4.9 million (QMRA) and 90,600 (EA) cases per year. Using the maximum likelihood approach and considering EA the gold standard, the QMRA model could produce the original EA estimate by adjusting mainly the dose-infection relationship. Considering QMRA the gold standard, the EA model could produce the original QMRA estimate by adjusting mainly the probability that a gastroenteritis case is caused by Campylobacter. A joint analysis of QMRA and EA data and models assuming identical outcomes, using a frequentist or Bayesian approach (using vague priors), resulted in estimates of 102,000 or 123,000 campylobacteriosis cases per year, respectively. These were close to the original EA estimate, and this will be related to the dissimilarity in data availability. The Bayesian approach further showed that attenuating the condition of equal outcomes immediately resulted in very different estimates of the number of campylobacteriosis cases per year and that using more informative priors had little effect on the results. In conclusion, EA was dominant in estimating the burden of campylobacteriosis in the Netherlands. However, it must be noted that only statistical uncertainties were taken into account here. Taking all, usually difficult to quantify, uncertainties into account might lead to a different conclusion.