Development of a quantitative risk assessment model for Salmonella enteritidis in pasteurized liquid eggs

Modelling the low temperature growth boundaries of Salmonella Enteritidis in raw and pasteurized egg yolk, egg white and liquid whole egg: Influence of the initial concentration

Salmonella is the most frequently reported cause of foodborne outbreaks with known origin in Europe, with eggs and egg products standing out as the most frequent food source (when it was known). The growth and survival of Salmonella in eggs and egg products have been extensively studied and, recently, it has been reported that factors such as the initial concentration and thermal history of the egg product can also influence its growth capability. Therefore, the objective of this study was to define the boundary zones of the growth/no growth domain of Salmonella Enteritidis (4 strains) as a function of temperature (low temperature boundary) and the initial concentration in different egg products. A series of polynomial logistic regression equations were successfully adjusted, allowing the study of these factors and their interaction on the probability of growth of S. Enteritidis in these products. Results obtained indicate that the minimum growth temperatures of Salmonella Enteritidis are higher in egg white (9.5-18.3 °C) than in egg yolk (7.1-7.8 °C) or liquid whole egg (7.2-7.9 °C). Results also demonstrate that in raw liquid whole egg and raw and pasteurized egg white, the minimum growth temperature of Salmonella Enteritidis does depend on the initial concentration. Similarly, the previous thermal history of the egg product only influenced the minimum growth temperature in some of them. On the other hand, large differences in the minimum growth temperatures among strains were observed in some products (up to approx. 6 °C in egg white). Finally, it should be noted that none of the strains grew at 5 °C under any of the conditions assayed. Therefore, storage of egg products (particularly whole liquid egg and egg yolk) below this temperature might be regarded/proposed as a good management approach. Our experimental approach has allowed us to provide a more accurate prediction of S. Enteritidis minimum growth temperatures in egg products by taking into account additional factors (initial concentration and thermal history) while also providing a quantification of the intra-specie variability. This would be of high relevance for improving the safety of egg products.

Effect of nisin, EDTA, and abuse temperature on the growth of Salmonella Typhimurium in liquid whole egg during refrigerated storage

Salmonella spp. can be present in pasteurized liquid egg products because of its heat resistance or post-processing contamination, thereby representing a food safety risk. The effect of 1000 IU nisin/ml plus 20 mM disodium ethylenediaminetetraacetic acid (EDTA), two refrigerated temperatures (7 °C and 10 °C), and two inoculation levels (103 and 105 cfu/ml) were studied in the growth of S. Typhimurium in pasteurized liquid whole egg (LWE). Two mathematical models were used to fit the microbial curves. Physicochemical characteristics of LWE, such as pH and color, were assessed for 31 days at the two storage temperatures, and no significant changes (p < 0.05) were observed for most of the samples. Results showed the significant impact of temperature on microbial growth. Samples kept at 7 °C showed the decay of microbial cells during storage; meanwhile, the effect at 10 °C was shown as fast growth. The combination of nisin plus EDTA and 7 °C accelerated the decay of microbial cells during the storage showing a synergistic effect. The Weibull model described the decline of cells during the shelf-life. Meanwhile, the logistic model fitted the growth of Salmonella in LWE at 10 °C. Adding these additives to LWE, combined with the correct temperature during pasteurization and adequate conditions during the cold chain, can minimize the food safety risk related to Salmonella.

Salmonella Risk Assessment in Poultry Meat from Farm to Consumer in Korea

This study predicted Salmonella outbreak risk from eating cooked poultry in various methods. The incidence of Salmonella in poultry meat and the environment from farm to home for consumption was investigated. To develop the predictive models, Salmonella growth data were collected at 4-25 °C during storage and fitted with the Baranyi model. The effects of cooking on cell counts in poultry meat were investigated. Temperature, duration, and consumption patterns were all searched. A simulation in @Risk was run using these data to estimate the probability of foodborne Salmonella disease. In farm, Salmonella was detected from only fecal samples (8.5%; 56/660). In slaughterhouses, Salmonella was detected from feces 16.0% (38/237) for chicken and 19.5% (82/420) for duck) and from carcasses of each step (scalding, defeathering, and chilling) by cross contamination. In chicken (n = 270) and duck (n = 205), Salmonella was detected in 5 chicken (1.9%) and 16 duck meat samples (7.8%). Salmonella contamination levels were initially estimated to be -3.1 Log CFU/g and -2.5 Log CFU/g, respectively. With R² values between 0.862 and 0.924, the predictive models were suitable for describing the fate of Salmonella in poultry meat with of 0.862 and 0.924. The Salmonella was not detected when poultry meat cooks completely. However, if poultry meat contaminated with Salmonella were cooked incompletely, Salmonella remained on the food surface. The risk of foodborne Salmonella disease from poultry consumption after cooking was 3.0 × 10^-10/person/day and 8.8 × 10^-11/person/day in South Korea, indicating a low risk.

Monte Carlo Simulation Model for Predicting Salmonella Contamination of Chicken Liver as a Function of Serving Size for Use in Quantitative Microbial Risk Assessment

ABSTRACT

The first step in quantitative microbial risk assessment (QMRA) is to determine the distribution of pathogen contamination among servings of the food in question at some point in the farm-to-table chain. In the present study, the distribution of Salmonella contamination among servings of chicken liver for use in the QMRA was determined at meal preparation. Salmonella prevalence (P), most probable number (MPN, N), and serotype for different serving sizes were determined by use of a combination of five methods: (i) whole sample enrichment; (ii) quantitative PCR; (iii) culture isolation; (iv) serotyping; and (v) Monte Carlo simulation. Epidemiological data also were used to convert serotype data to virulence (V) values for use in the QMRA. A Monte Carlo simulation model based in Excel and simulated with @Risk predicted Salmonella P, N, serotype, and V as a function of a serving size of one (58 g) to eight (464 g) chicken livers. Salmonella P of chicken livers was 72.5% (58 of 80) per 58 g. Four Salmonella serotypes were isolated from chicken livers: (i) Infantis (P = 28%, V = 4.5); (ii) Enteritidis (P = 15%, V = 5); (iii) Typhimurium (P = 15%, V = 4.8); and (iv) Kentucky (P = 15%, V = 0.8). Salmonella N was 1.76 log MPN/58 g (median) with a range of 0 to 4.67 log MPN/58 g, and the median Salmonella N was not affected (P > 0.05) by serotype. The model predicted a nonlinear increase (P ≤ 0.05) of Salmonella P from 72.5%/58 g to 100%/464 g, a minimum N of 0 log MPN/58 g to 1.28 log MPN/464 g, and a median N from 1.76 log MPN/58 g to 3.22 log MPN/464 g. Regardless of serving size, predicted maximum N was 4.74 log MPN per serving, mean V was 3.9 per serving, and total N was 6.65 log MPN per lot (10,000 chicken livers). The data acquired and modeled in this study address an important data gap in the QMRA for Salmonella and whole chicken liver.

HIGHLIGHTS

Development of predictive models evaluating the spoilage-delaying effect of a bioprotective culture on different yeast species in yogurt

Yeast spoilage of fermented dairy products causes challenges for the dairy industry, including economic losses due to wasted product. Food cultures with bioprotective effects are becoming more widely used to help ensure product quality throughout product shelf life. To assist the dairy industry when evaluating product quality throughout shelf life and the effect of bioprotective cultures, we aimed to build stochastic models that provide reliable predictions of yeast spoilage in yogurt with and without bioprotective culture. Growth characterizations of Debaryomyces hansenii, Yarrowia lipolytica, Saccharomyces cerevisiae, and Kluyveromyces marxianus at storage temperatures of 7, 12, and 16°C during a 30-d storage period were conducted in yogurt with and without a bioprotective culture containing Lacticaseibacillus rhamnosus strains. The kinetic growth parameters were calculated using the Buchanan growth model, and these parameters were used as baseline values in Monte Carlo models to translate the yeast growth into spoilage levels. The models were developed using 100,000 simulations and they predicted yeast spoilage levels in yogurt by the 4 yeast types. Each modeled yogurt batch was set to be contaminated with yeast at a concentration drawn from a normal distribution with a mean of 1 log₁₀ cfu/mL and standard deviation of 1 log₁₀ cfu/mL and stored for 30 d at a temperature drawn from a normal distribution with a mean of 6.1°C and a standard deviation of 2.8°C. Considering a spoilage level of 5 log₁₀ cfu/mL, the predicted number of spoiled samples was reduced 3-fold during the first 10 d and by 2-fold at the end of shelf life when a bioprotective culture was added to the yogurt. The models were evaluated by sensitivity analyses, where the main effect factors were maximum yeast population, storage temperature, and yeast strain. The models were validated by comparing the model output to actual observed spoilage data from a European dairy using the bioprotective culture. When the model prediction, based on a mixture of the 4 specific yeast strains, was compared with spoilage data from the European dairy, the observed effect of bioprotective cultures was considerably higher than predicted, potentially influenced by the presence of contaminating strains more sensitive to a bioprotective culture than those characterized here. The developed Monte Carlo models can predict yeast spoilage levels in yogurt at specific production settings and how this may be affected by various parameters and addition of bioprotective cultures.

ANTIBACTERIAL AND ANTIFUNGAL EFFECT OF ACHILLEA MILLEFOLIUM ESSENTIAL OIL DURING SHELF LIFE OF MAYONNAISE

The importance of food-borne disease and consumer demands for avoiding synthetic food preservatives shifted the research interest to natural food preservatives such as essential oils which have antimicrobial activity. Also, spoilage of foods by fungi is a major problem, especially in developing countries. The aim of this study was to evaluate the efficiency of Achillea millefolium essential oil as natural food preservative in high fat and low fat mayonnaise kept at 4°C for 6 months. Mayonnaise samples were divided into four experimental treatments, namely: EO (essential oil in concentrations of 0.45-7.2 mg/ml), BS (sodium benzoate and potassium sorbate in concentration of 0.75 mg/ml), Cmo (control: no preservative with added microorganisms) and C (control: no preservative and no added microorganisms). The results showed that of essential oil of Achillea millefolium had influence against all of the tested microorganisms in mayonnaise and all of the pathogens and fungi did not grow in mayonnaise, whereas in the control samples all of the microorganisms grew. The maximum cell counts of bacteria and fungus in low fat mayonnaise were approximately lower than the high fat mayonnaise or resistance to inactivation of microorganisms appeared to be greater in high fat mayonnaise than in low fat mayonnaise (p< 0.05). Also, BS samples exhibited antimicrobial properties against tested species during storage. In conclusion the essential oil of Achillea millefolium would lead to control food pathogenic organisms and food spoilage organisms and therefore, it can be used as natural preservative in food industry such as mayonnaise.  

Quantitative Risk Assessment Model for Salmonellosis in Chicken Skewers from Street Food Vendors in South Korea

HIGHLIGHTS

Extended holding time increases likelihood of illness in warmer than it does in cooler months. Split heating is less effective than continuous heating for inactivation of microbes. Longer holding times make meat more vulnerable to microbes.

Risk Assessment for Salmonellosis in Chicken in South Korea: The Effect of Salmonella Concentration in Chicken at Retail

Salmonellosis caused by chicken consumption has been a critical issue in food safety worldwide, including in Korea. The probability of illness from consumption of chicken was simulated in study, based on the recipe of Dakgalbi, a commonly eaten chicken dish in Korea. Additionally, the processing stage at slaughterhouses to decrease Salmonella concentration in broilers was modeled to explore its effect on the likelihood of illness. A Monte Carlo simulation model was created using @RISK. Prevalence of Salmonella in chickens at the retail stage was found to be predominantly important in determining the probability of illness. Other than the prevalence, cooking temperature was found to have the largest impact on the probability of illness. The results also demonstrated that, although chlorination is a powerful tool for decreasing the Salmonella concentration in chicken, this effect did not last long and was negated by the following stages. This study analyzes the effects of variables of the retail-to-table pathway on the likelihood of salmonellosis in broiler consumption, and also evaluates the processing step used to decrease the contamination level of Salmonella in broilers at slaughterhouses. According to the results, we suggest that methods to decrease the contamination level of Salmonella such as chlorination had little effect on decreasing the probability of illness. Overall, these results suggest that preventing contamination of broiler with Salmonella must be a top priority and that methods to reduce the concentration of Salmonella in broilers at slaughterhouses hardly contribute to safe consumption of Salmonella-contaminated chicken.

Development and Validation of Predictive Model for Salmonella Growth in Unpasteurized Liquid Eggs

Liquid egg products can be contaminated with Salmonella spp. during processing. A predictive model for the growth of Salmonella spp. in unpasteurized liquid eggs was developed and validated. Liquid whole egg, liquid yolk, and liquid egg white samples were prepared and inoculated with Salmonella mixture (approximately 3 Log CFU/mL) containing five serovars (S. Bareilly, S. Richmond, S. Typhimurium monophasic, S. Enteritidis, and S. Gallinarum). Salmonella growth data at isothermal temperatures (5, 10, 15, 20, 25, 30, 35, and 40°C) was collected by 960 h. The population of Salmonella in liquid whole egg and egg yolk increased at above 10°C, while Salmonella in egg white did not proliferate at all temperature. These results demonstrate that there is a difference in the growth of Salmonella depending on the types of liquid eggs (egg yolk, egg white, liquid whole egg) and storage temperature. To fit the growth data of Salmonella in liquid whole egg and egg yolk, Baranyi model was used as the primary model and the maximum growth rate and lag phase duration for each temperature were determined. A secondary model was developed with maximum growth rate as a function of temperature. The model performance measures, bias factor (B _f , 0.96-0.99) and r² (0.96-0.99) indicated good fit for both primary and secondary models. In conclusion, it is thought that the growth model can be used usefully to predict Salmonella spp. growth in various types of unpasteurized liquid eggs when those are exposed to various temperature and time conditions during the processing.

Risk of Illness with Salmonella due to Consumption of Raw Unwashed Vegetables Irrigated with Water from the Bogotá River

The Bogotá River receives untreated wastewater from the city of Bogotá and many other towns. Downstream from Bogotá, water from the river is used for irrigation of crops. Concentrations of indicator organisms in the river are high, which is consistent with fecal contamination. To investigate the probability of illness due to exposure to enteric pathogens from the river, specifically Salmonella, we took water samples from the Bogotá River at six sampling locations in an area where untreated water from the river is used for irrigation of lettuce, broccoli, and cabbage. Salmonella concentrations were quantified by direct isolation and qPCR. Concentrations differed, depending on the quantification technique used, ranging between 107.7 and 109.9 number of copies of gene invA per L and 105.3 and 108.4 CFU/L, for qPCR and direct isolation, respectively. A quantitative microbial risk assessment model that estimates the daily risk of illness with Salmonella resulting from consuming raw unwashed vegetables irrigated with water from the Bogotá River was constructed using the Salmonella concentration data. The daily probability of illness from eating raw unwashed vegetables ranged between 0.62 and 0.85, 0.64 and 0.86, and 0.64 and 0.85 based on concentrations estimated by qPCR (0.47-0.85, 0.47-0.86, and 0.41-0.85 based on concentrations estimated by direct isolation) for lettuce, cabbage, and broccoli, respectively, which are all above the commonly propounded benchmark of 10-4 per year. Results obtained in this study highlight the necessity for appropriate wastewater treatment in the region, and emphasize the importance of postharvest practices, such as washing, disinfecting, and cooking.

Isolation, Identification, and Characterization of Foodborne Pathogens Isolated from Egg Internal Contents in China

Eggs continue to be significant in terms of the world economy and human nutrition. The aim of this study was to estimate the prevalence of foodborne pathogens in a cross section of egg production types in China and to examine the virulence features of the isolated pathogens. Three hundred eggs from three provinces were tested for Salmonella , Escherichia coli , Staphylococcus aureus , and Campylobacter . Two eggs were positive for Salmonella , one was positive for both E. coli and S. aureus , and none were positive for Campylobacter . When comparing different egg laying systems, there were no significant differences in findings between packaged and unpackaged eggs and between battery cage and cage-free eggs, but there were significant differences in the findings among pathogens and among provinces. Other potential foodborne pathogens were identified in addition to the targeted pathogens. The virulence of Serratia marcescens was higher than that of the isolated Salmonella Typhimurium, and the virulence of Staphylococcus saprophyticus was comparable to that of the isolated S. aureus . These findings are important for local risk assessments concerning possible human foodborne infections via cross-contamination of eggs.

Kinetic Behavior of Salmonella on Low NaNO2 Sausages during Aerobic and Vacuum Storage

This study evaluated the growth kinetics of Salmonella spp. in processed meat products formulated with low sodium nitrite (NaNO2). A 5-strain mixture of Salmonella spp. was inoculated on 25-g samples of sausages formulated with sodium chloride (NaCl) (1.0%, 1.25%, and 1.5%) and NaNO2 (0 and 10 ppm) followed by aerobic or vacuum storage at 10℃ and 15℃ for up to 816 h or 408 h, respectively. The bacterial cell counts were enumerated on xylose lysine deoxycholate agar, and the modified Gompertz model was fitted to the Salmonella cell counts to calculate the kinetic parameters as a function of NaCl concentration on the growth rate (GR; Log CFU/g/h) and lag phase duration (LPD; h). A linear equation was then fitted to the parameters to evaluate the effect of NaCl concentration on the kinetic parameters. The GR values of Salmonella on sausages were higher (p<0.05) with 10 ppm NaNO2 concentration than with 0 ppm NaNO2. The GR values of Salmonella decreased (p<0.05) as NaCl concentration increased, especially at 10℃. This result indicates that 10 ppm NaNO2 may increase Salmonella growth at low NaCl concentrations, and that NaCl plays an important role in inhibiting Salmonella growth in sausages with low NaNO2.

Application of Molecular Approaches for Understanding Foodborne Salmonella Establishment in Poultry Production

Salmonellosis in the United States is one of the most costly foodborne diseases. Given that Salmonella can originate from a wide variety of environments, reduction of this organism at all stages of poultry production is critical. Salmonella species can encounter various environmental stress conditions which can dramatically influence their survival and colonization. Current knowledge of Salmonella species metabolism and physiology in relation to colonization is traditionally based on studies conducted primarily with tissue culture and animal infection models. Consequently, while there is some information about environmental signals that control Salmonella growth and colonization, much still remains unknown. Genetic tools for comprehensive functional genomic analysis of Salmonella offer new opportunities for not only achieving a better understanding of Salmonella pathogens but also designing more effective intervention strategies. Now the function(s) of each single gene in the Salmonella genome can be directly assessed and previously unknown genetic factors that are required for Salmonella growth and survival in the poultry production cycle can be elucidated. In particular, delineating the host-pathogen relationships involving Salmonella is becoming very helpful for identifying optimal targeted gene mutagenesis strategies to generate improved vaccine strains. This represents an opportunity for development of novel vaccine approaches for limiting Salmonella establishment in early phases of poultry production. In this review, an overview of Salmonella issues in poultry, a general description of functional genomic technologies, and their specific application to poultry vaccine developments are discussed.

Assessment of the risk of salmonellosis from internally contaminated shell eggs following initial storage at 18 °C (65 °F), compared with 7 °C (45 °F)

In the U.S., chicken-breeder farms that supply hatcheries typically store and transport eggs intended for broiler production at a temperature of 18.3 °C (65 °F). However, in case of surplus, some of these eggs may be diverted to human consumption. According to the U.S. Food and Drug Administration's 'Egg Safety Final Rule,' shell eggs intended for human consumption are required to be held or transported at or below 7.2 °C (45 °F) ambient temperature beginning 36 h after time of lay. We adapted a risk assessment model developed by the U.S. Department of Agriculture's Food Safety Inspection Service, to quantify human exposure to Salmonella Enteritidis and the risk of human salmonellosis if eggs are held and transported at 18.3 °C for up to 5.5 days after time of lay, as has been observed when hatchery eggs are diverted to human consumption, rather than held and transported at 7.2 °C within 36 h after time of lay. Storage at 18.3 °C leads to considerable bacterial growth in internally contaminated eggs. The model predicted that more than 10% of internally contaminated eggs would remain contaminated after in-shell pasteurization resulting in a 5-log10 reduction, and that some bacteria would survive after home-cooking. The model predicted that, alternatively, eggs stored at 7.2 °C after lay would have limited bacterial growth prior to pasteurization, and Salmonella would be very unlikely to be present after pasteurization. The predicted risk of salmonellosis from the consumption of eggs held and transported at 18.3 °C and subsequently diverted to human consumption is 25 times higher than the risk when eggs are held and transported at 7.2 °C.

Avian influenza: public health and food safety concerns

Avian influenza (AI) is a disease or asymptomatic infection caused by Influenzavirus A. AI viruses are species specific and rarely cross the species barrier. However, subtypes H5, H7, and H9 have caused sporadic infections in humans, mostly as a result of direct contact with infected birds. H5N1 high pathogenicity avian influenza (HPAI) virus causes a rapid onset of severe viral pneumonia and is highly fatal (60% mortality). Outbreaks of AI could have a severe economic and social impact on the poultry industry, trade, and public health. Surveillance data revealed that H5N1 HPAI has been detected in imported frozen duck meat from Asia, and on the surface and in contaminated eggs. However, there is no direct evidence that AI viruses can be transmitted to humans via the consumption of contaminated poultry products. Implementing management practices that incorporate biosecurity principles, personal hygiene, and cleaning and disinfection protocols, as well as cooking and processing standards, are effective means of controlling the spread of the AI viruses.

Expanded Fermi solution for estimating the survival of ingested pathogenic and probiotic microbial cells and spores

The expanded Fermi solution was originally developed for estimating the number of food-poisoning victims when information concerning the circumstances of exposure is scarce. The method has been modified for estimating the initial number of pathogenic or probiotic cells or spores so that enough of them will survive the food preparation and digestive tract's obstacles to reach or colonize the gut in sufficient numbers to have an effect. The method is based on identifying the relevant obstacles and assigning each a survival probability range. The assumed number of needed survivors is also specified as a range. The initial number is then estimated to be the ratio of the number of survivors to the product of the survival probabilities. Assuming that the values of the number of survivors and the survival probabilities are uniformly distributed over their respective ranges, the sought initial number is construed as a random variable with a probability distribution whose parameters are explicitly determined by the individual factors' ranges. The distribution of the initial number is often approximately lognormal, and its mode is taken to be the best estimate of the initial number. The distribution also provides a credible interval for this estimated initial number. The best estimate and credible interval are shown to be robust against small perturbations of the ranges and therefore can help assessors achieve consensus where hard knowledge is scant. The calculation procedure has been automated and made freely downloadable as a Wolfram Demonstration.

Modelling the contamination of lettuce with Escherichia coli O157:H7 from manure-amended soil and the effect of intervention strategies

AIMS

A growing number of foodborne illnesses has been associated with the consumption of fresh produce. In this study, the probability of lettuce contamination with Escherichia coli O157:H7 from manure-amended soil and the effect of intervention strategies was determined.

METHODS AND RESULTS

Pathogen prevalence and densities were modelled probabilistically through the primary production chain of lettuce (manure, manure-amended soil and lettuce). The model estimated an average of 0.34 contaminated heads per hectare. A minimum manure storage time of 30 days and a minimum fertilization-to-planting interval of 60 days was most successful in reducing the risk. Some specific organic farming practices concerning manure and soil management were found to be risk reducing.

CONCLUSIONS

Certain specific organic farming practices reduced the likelihood of contamination. This cannot be generalized to organic production as a whole. However, the conclusion is relevant for areas like the Netherlands where there is high use of manure in both organic and conventional vegetable production.

SIGNIFICANCE AND IMPACT OF THE STUDY

Recent vegetable-associated disease outbreaks stress the importance of a safe vegetable production chain. The present study contributed to this by providing a first estimate of the likelihood of lettuce contamination with E. coli O157:H7 and the effectiveness of risk mitigation strategies.

Applying the food safety objective and related standards to thermal inactivation of Salmonella in poultry meat

The objective of this study was to investigate the practicality of designing a heat treatment process in a food manufacturing operation for a product governed by a Food Safety Objective (FSO). Salmonella in cooked poultry meat was taken as the working example. Although there is no FSO for this product in current legislation, this may change in the (near) future. Four different process design calculations were explored by means of deterministic and probabilistic approaches to mathematical data handling and modeling. It was found that the probabilistic approach was a more objective, transparent, and quantifiable approach to establish the stringency of food safety management systems. It also allowed the introduction of specific prevalence rates. The key input analyzed in this study was the minimum time required for the heat treatment at a fixed temperature to produce a product that complied with the criterion for product safety, i.e., the FSO. By means of the four alternative process design calculations, the minimum time requirement at 70 degrees C was established and ranged from 0.26 to 0.43 min. This is comparable to the U.S. regulation recommendations and significantly less than that of 2 min at 70 degrees C used, for instance, in the United Kingdom regulation concerning vegetative microorganisms in ready-to-eat foods. However, the objective of this study was not to challenge existing regulations but to provide an illustration of how an FSO established by a competent authority can guide decisions on safe product and process designs in practical operation; it hopefully contributes to the collaborative work between regulators, academia, and industries that need to continue learning and gaining experience from each other in order to translate risk-based concepts such as the FSO into everyday operational practice.

Number of Salmonella on chicken breast filet at retail level and its implications for public health risk

This study aimed to characterize the number of Salmonella on chicken breast filet at the retail level and to evaluate if this number affects the risk of salmonellosis. From October to December 2005, 220 chilled raw filets (without skin) were collected from five local retail outlets in The Netherlands. Filet rinses that were positive after enrichment were enumerated with a three-tube most-probable-number (MPN) assay. Nineteen filets (8.6%) were contaminated above the detection limit of the MPN method (10 Salmonella per filet). The number of Salmonella on positive filets varied from 1 to 3.81 log MPN per filet. The obtained enumeration data were applied in a risk assessment model. The model considered possible growth during domestic storage, cross-contamination from filet via a cutting board to lettuce, and possible illness due to consumption of the prepared lettuce. A screening analysis with expected-case and worst-case estimates for the input values of the model showed that variability in the inputs was of relevance. Therefore, a Monte Carlo simulation with probability distributions for the inputs was carried out to predict the annual number of illnesses. Remarkably, over two-thirds of annual predicted illnesses were caused by the small fraction of filets containing more than 3 log Salmonella at retail (0.8% of all filets). The enumeration results can be used to confirm this hypothesis in a more elaborate risk assessment. Modeling of the supply chain can provide insight for possible intervention strategies to reduce the incidence of rare, but extreme levels. Reduction seems feasible within current practices, because the retail market study indicated a significant difference between suppliers.

An expanded Fermi solution for microbial risk assessment

'Fermi solution' refers to an estimate of a quantity of interest derived from a sequence of guesses about factors of which detailed knowledge is unavailable. When one makes such guesses, it is unlikely that the large majority of them will be either too high or too low. Most probably, some of the overestimates will be offset by some of the underestimates, and the final result will be often close to the correct value. The method has been popularized as recreational physics but it has also been applied in risk assessment, where the factors involved, but not their exact magnitudes, are known. The concept has potential application in certain types of food poisoning risk assessments, and in estimating the number victims of a bioterrorist attack on the food or water supply, where some guessing is inevitable because of the absence of accurate relevant data. We consider a version of the method in which ranges instead of single values are entered as the factors' estimates. For simplicity, the risk to be assessed is taken to be the product of the factors, and their true values are regarded as being uniformly distributed over their respective ranges. The risk itself is therefore construed as a random variable with a probability distribution whose parameters are explicitly determined by the individual factors' ranges and which can often be approximated by a lognormal distribution. The mode of this lognormal distribution is taken to be the "best guess" of the risk, and a credible interval is constructed with a specified level of "confidence". The best guess and credible interval are shown to be robust against small perturbations of the ranges. Thus, even if the ranges are misspecified to some degree, assessments based on the best guess or credible interval will not be substantially altered. This can help to achieve consensus among assessors in situations where very little hard knowledge exists. The calculation procedure has been automated in software that has been made freely available over the Internet. The concept is demonstrated with two hypothetical problems: predicting the number of persons who would come down with acute food poisoning after consuming a contaminated dish, and estimating the number of daily salmonellosis cases in a large metropolitan area.

Validation of a tertiary model for predicting variation of Salmonella typhimurium DT104 (ATCC 700408) growth from a low initial density on ground chicken breast meat with a competitive microflora

Growth of a multiple antibiotic-resistant strain (ATCC 700408) of Salmonella Typhimurium definitive phage type 104 (DT104) from a low initial density (10(0.6) most probable number [MPN] or CFU/g) on ground chicken breast meat with a competitive microflora was investigated and modeled as a function of time and temperature (10 to 40 degrees C). MPN and viable counts (CFU) on a selective medium with four antibiotics enumerated the pathogen. Data from five replicate challenge studies per temperature were combined and fit to a primary model to determine maximum specific growth rate (micro), maximum population density (Nmax), and the 95% prediction interval (PI). Nonlinear regression was used to obtain secondary models as a function of temperature for micro, Nmax, and PI, which ranged from 0.04 to 0.4 h(-1), 1.6 to 9.4 log MPN or CFU/g, and 1.4 to 2.4 log MPN or CFU/g, respectively. Secondary models were combined with the primary model to create a tertiary model for predicting variation (95% PI) of pathogen growth among batches of ground chicken breast meat with a competitive microflora. The criterion for acceptable model performance was that 90% of observed MPN or CFU data had to be in the 95% PI predicted by the tertiary model. For data (n=344) used in model development, 93% of observed MPN and CFU data were in the 95% PI predicted by the tertiary model, whereas for data (n=236) not used in model development but collected using the same methods, 94% of observed MPN and CFU data were in the 95% PI predicted by the tertiary model. Thus, the tertiary model was successfully verified against dependent data and validated against independent data for predicting variation of Salmonella Typhimurium DT104 growth among batches of ground chicken breast meat with a competitive microflora and from a low initial density.

Observations on Salmonella contamination of eggs from infected commercial laying flocks where vaccination for Salmonella enterica serovar Enteritidis had been used

Eggs were collected monthly from 12 cage-layer flocks on four farms where Salmonella Enteritidis was present in vaccinated flocks despite vaccination with an S. Enteritidis bacterin. Where possible, hens were also taken for culture at the end of the laying period, and faecal and environmental samples were taken from the laying houses before and after cleaning and disinfection. Twenty-four batches of six egg shells from the 13 652 tested (0.18% [0.11 to 0.26 CI(95)] single egg equivalent) were positive for S. Enteritidis and 54 (0.40% [0.30 to 0.52 CI(95)] single egg equivalent) for other serovars. Six batches of 13 640 (0.04% [0.02 to 0.10 CI95] single egg equivalent) egg contents, bulked in six egg pools, contained S. Enteritidis and three batches contained other serovars. In addition three further batches contained S. Enteritidis in both contents and shells, and two other batches contained other serovars in both. The total level of contamination by S. Enteritidis of both contents and shells found in vaccinated flocks was therefore 33 batches/13 682 eggs(0.24% [0.17 to 0.34 CI(95)] single egg equivalent). The total of contamination for any Salmonella serovar was 92 batches/13 682 eggs (0.68% [0.55 to 0.84 CI(95)] single egg equivalent). These results contrast with the findings of testing of eggs from three unvaccinated flocks prior to this study where 21 batches of egg shells from a total of 2101 eggs (1.0% [0.63 to 1.56 CI(95)] single egg equivalent) and six batches of contents from 2051 eggs (0.29% [0.11 to 0.64 CI95] single egg equivalent) were contaminated with S. Enteritidis. S. Enteritidis was found in 67/699 (9.6%) of vaccinated spent hens and 64/562 (11.4%) of bulked fresh faecal samples taken from laying houses. Failure to adequately clean and disinfect laying houses and to control mice appeared to be a common feature on the farms.

Consumer-phase Salmonella enterica serovar enteritidis risk assessment for egg-containing food products

We describe a one-dimensional probabilistic model of the role of domestic food handling behaviors on salmonellosis risk associated with the consumption of eggs and egg-containing foods. Six categories of egg-containing foods were defined based on the amount of egg contained in the food, whether eggs are pooled, and the degree of cooking practiced by consumers. We used bootstrap simulation to quantify uncertainty in risk estimates due to sampling error, and sensitivity analysis to identify key sources of variability and uncertainty in the model. Because of typical model characteristics such as nonlinearity, interaction between inputs, thresholds, and saturation points, Sobol's method, a novel sensitivity analysis approach, was used to identify key sources of variability. Based on the mean probability of illness, examples of foods from the food categories ranked from most to least risk of illness were: (1) home-made salad dressings/ice cream; (2) fried eggs/boiled eggs; (3) omelettes; and (4) baked foods/breads. For food categories that may include uncooked eggs (e.g., home-made salad dressings/ice cream), consumer handling conditions such as storage time and temperature after food preparation were the key sources of variability. In contrast, for food categories associated with undercooked eggs (e.g., fried/soft-boiled eggs), the initial level of Salmonella contamination and the log10 reduction due to cooking were the key sources of variability. Important sources of uncertainty varied with both the risk percentile and the food category under consideration. This work adds to previous risk assessments focused on egg production and storage practices, and provides a science-based approach to inform consumer risk communications regarding safe egg handling practices.

Detection and enumeration of Salmonella enteritidis in homemade ice cream associated with an outbreak: comparison of conventional and real-time PCR methods

Salmonellosis caused by Salmonella Enteritidis (SE) is a significant cause of foodborne illnesses in the United States. Consumption of undercooked eggs and egg-containing products has been the primary risk factor for the disease. The importance of the bacterial enumeration technique has been enormously stressed because of the quantitative risk analysis of SE in shell eggs. Traditional enumeration methods mainly depend on slow and tedious most-probable-number (MPN) methods. Therefore, specific, sensitive, and rapid methods for SE quantitation are needed to collect sufficient data for risk assessment and food safety policy development. We previously developed a real-time quantitative PCR assay for the direct detection and enumeration of SE and, in this study, applied it to naturally contaminated ice cream samples with and without enrichment. The detection limit of the real-time PCR assay was determined with artificially inoculated ice cream. When applied to the direct detection and quantification of SE in ice cream, the real-time PCR assay was as sensitive as the conventional plate count method in frequency of detection. However, populations of SE derived from real-time quantitative PCR were approximately 1 log higher than provided by MPN and CFU values obtained by conventional culture methods. The detection and enumeration of SE in naturally contaminated ice cream can be completed in 3 h by this real-time PCR method, whereas the cultural enrichment method requires 5 to 7 days. A commercial immunoassay for the specific detection of SE was also included in the study. The real-time PCR assay proved to be a valuable tool that may be useful to the food industry in monitoring its processes to improve product quality and safety.

Neutrophils in the war against Staphylococcus aureus: predator-prey models to the rescue

To address the question of whether a minimum concentration of blood neutrophils is necessary to decrease Staphylococcus aureus concentration in mastitic milk, literature was searched for studies in which neutrophils were incubated with Staph. aureus. Different mathematical models that describe the changes in Staph. aureus population as a function of neutrophilic concentrations were applied to the collected data. The best fitted model established (1) that the rate of bacterial killing depended on the ratio of neutrophils to bacteria with neutrophilic attack rate accelerating at first before decelerating as the ratio increases, and (2) that neutrophil concentration should be within a limited range to trigger a decline in the bacterial population. Outcomes of this model are supported by what is known about neutrophilic functions and laboratory findings in bovine and human neutrophils. These results may be of assistance in setting selection goals for a better resilience to Staph. aureus mastitis in dairy cattle. Indeed, an optimal neutrophilic concentration appears to exist for successful clearance of Staph. aureus infection, which is neither the lowest nor the highest one.

Uncertainty distribution associated with estimating a proportion in microbial risk assessment

The uncertainty associated with estimates should be taken into account in quantitative risk assessment. Each input's uncertainty can be characterized through a probabilistic distribution for use under Monte Carlo simulations. In this study, the sampling uncertainty associated with estimating a low proportion on the basis of a small sample size was considered. A common application in microbial risk assessment is the estimation of a prevalence, proportion of contaminated food products, on the basis of few tested units. Three Bayesian approaches (based on beta(0, 0), beta(1/2, 1/2), and beta(l, 1)) and one frequentist approach (based on the frequentist confidence distribution) were compared and evaluated on the basis of simulations. For small samples, we demonstrated some differences between the four tested methods. We concluded that the better method depends on the true proportion of contaminated products, which is by definition unknown in common practice. When no prior information is available, we recommend the beta (1/2, 1/2) prior or the confidence distribution. To illustrate the importance of these differences, the four methods were used in an applied example. We performed two-dimensional Monte Carlo simulations to estimate the proportion of cold smoked salmon packs contaminated by Listeria monocytogenes, one dimension representing within-factory uncertainty, modeled by each of the four studied methods, and the other dimension representing variability between companies.

Monte Carlo simulation of pathogen behavior during the sprout production process

Food-borne disease outbreaks linked to the consumption of raw sprouts have become a concern over the past decade. A Monte Carlo simulation model of the sprout production process was created to determine the most-effective points for pathogen control. Published literature was reviewed, and relevant data were compiled. Appropriate statistical distributions were determined and used to create the Monte Carlo model with Analytica software. Factors modeled included initial pathogen concentration and prevalence, seed disinfection effectiveness, and sampling of seeds prior to sprouting, sampling of irrigation water, or sampling of the finished product. Pathogen concentration and uniformity of seed contamination had a large effect on the fraction of contaminated batches predicted by the simulation. The model predicted that sprout sampling and irrigation water sampling at the end of the sprouting process would be more effective in pathogen detection than seed sampling prior to production. Day of sampling and type of sample (sprout or water) taken had a minimal effect on rate of detection. Seed disinfection reduced the proportion of contaminated batches, but in some cases it also reduced the ability to detect the pathogen when it was present, because cell numbers were reduced below the detection limit. Both the amount sampled and the pathogen detection limit were shown to be important variables in determining sampling effectiveness. This simulation can also be used to guide further research and compare the levels of effectiveness of different risk reduction strategies.

An epidemiologic critique of current microbial risk assessment practices: the importance of prevalence and test accuracy data

Data deficiencies are impeding the development and validation of microbial risk assessment models. One such deficiency is the failure to adjust test-based (apparent) prevalence estimates to true prevalence estimates by correcting for the imperfect accuracy of tests that are used. Such adjustments will facilitate comparability of data from different populations and from the same population over time as tests change and the unbiased quantification of effects of mitigation strategies. True prevalence can be estimated from apparent prevalence using frequentist and Bayesian methods, but the latter are more flexible and can incorporate uncertainty in test accuracy and prior prevalence data. Both approaches can be used for single or multiple populations, but the Bayesian approach can better deal with clustered data, inferences for rare events, and uncertainty in multiple variables. Examples of prevalence inferences based on results of Salmonella culture are presented. The opportunity to adjust test-based prevalence estimates is predicated on the availability of sensitivity and specificity estimates. These estimates can be obtained from studies using archived gold standard (reference) samples, by screening with the new test and follow-up of test-positive and test-negative samples with a gold standard test, and by use of latent class methods, which make no assumptions about the true status of each sampling unit. Latent class analysis can be done with maximum likelihood and Bayesian methods, and an example of their use in the evaluation of tests for Toxoplasma gondii in pigs is presented. Guidelines are proposed for more transparent incorporation of test data into microbial risk assessments.

Reduction of Salmonella by two commercial egg white pasteurization methods

The effect of pH, processing temperatures, and preheating steps in two commercial egg white pasteurization procedures (Armour and Standard Brands methods) were evaluated using a five-strain cocktail of Salmonella. We devised a benchtop pasteurization system that would more closely resemble the two commercial processes than could the traditional capillary tube method. The pasteurization methods both require hydrogen peroxide to be metered into the egg white stream between a required initial preheat step and the main heating regimen. Both processes were evaluated at three pH levels (pH 8.2, 8.6, 9.0), at four temperatures (51.7 degrees C/125 degrees F, 53.1 degrees C/127.5 degrees F, 54.4 degrees C/130 degrees F, 55.8 degrees C/132.5 degrees F), and over four residence times to allow calculation of D-values at each temperature. When compared at the minimum allowable time and temperatures for each process, our results showed at least a 1-log greater log reduction (P < 0.05) for the Standard Brands method than the Armour method in 10 of 12 of the pH and temperature combinations tested. Almost all runs at any given temperature showed more reduction at pH 9.0 than at pH 8.2 except for the Standard Brands method at 54.4 degrees C and 55.8 degrees C, which showed the most consistent reduction across all three pH levels tested. Analysis of the preheat portion of the two methods showed that there was no contribution (P > 0.05) toward Salmonella reduction when compared with the identical process without the preheating step. We generally observed a greater reduction of Salmonella with egg white at pH 9.0 that is typical of older, off-line processing than with low pH egg white (i.e., 8.2) that is typical of modern in-line processing facilities. This difference was as much as 3.5 log cycles depending on the processing conditions. The data has been used to make recommendations for minimum processing conditions for hydrogen peroxide-based egg white pasteurization.

A quantitative risk assessment model for Salmonella and whole chickens

Existing data and predictive models were used to define the input settings of a previously developed but modified quantitative risk assessment model (QRAM) for Salmonella and whole chickens. The QRAM was constructed in an Excel spreadsheet and was simulated using @Risk. The retail-to-table pathway was modeled as a series of unit operations and associated pathogen events that included initial contamination at retail, growth during consumer transport, thermal inactivation during cooking, cross-contamination during serving, and dose response after consumption. Published data as well as predictive models for growth and thermal inactivation of Salmonella were used to establish input settings. Noncontaminated chickens were simulated so that the QRAM could predict changes in the incidence of Salmonella contamination. The incidence of Salmonella contamination changed from 30% at retail to 0.16% after cooking to 4% at consumption. Salmonella growth on chickens during consumer transport was the only pathogen event that did not impact the risk of salmonellosis. For the scenario simulated, the QRAM predicted 0.44 cases of salmonellosis per 100,000 consumers, which was consistent with recent epidemiological data that indicate a rate of 0.66-0.88 cases of salmonellosis per 100,000 consumers of chicken. Although the QRAM was in agreement with the epidemiological data, surrogate data and models were used, assumptions were made, and potentially important unit operations and pathogen events were not included because of data gaps and thus, further refinement of the QRAM is needed.

A Bayesian approach to quantify the contribution of animal-food sources to human salmonellosis

Based on the data from the integrated Danish Salmonella surveillance in 1999, we developed a mathematical model for quantifying the contribution of each of the major animal-food sources to human salmonellosis. The model was set up to calculate the number of domestic and sporadic cases caused by different Salmonella sero and phage types as a function of the prevalence of these Salmonella types in the animal-food sources and the amount of food source consumed. A multiparameter prior accounting for the presumed but unknown differences between serotypes and food sources with respect to causing human salmonellosis was also included. The joint posterior distribution was estimated by fitting the model to the reported number of domestic and sporadic cases per Salmonella type in a Bayesian framework using Markov Chain Monte Carlo simulation. The number of domestic and sporadic cases was obtained by subtracting the estimated number of travel- and outbreak-associated cases from the total number of reported cases, i.e., the observed data. The most important food sources were found to be table eggs and domestically produced pork comprising 47.1% (95% credibility interval, CI: 43.3-50.8%) and 9% (95% CI: 7.8-10.4%) of the cases, respectively. Taken together, imported foods were estimated to account for 11.8% (95% CI: 5.0-19.0%) of the cases. Other food sources considered had only a minor impact, whereas 25% of the cases could not be associated with any source. This approach of quantifying the contribution of the various sources to human salmonellosis has proved to be a valuable tool in risk management in Denmark and provides an example of how to integrate quantitative risk assessment and zoonotic disease surveillance.

Estimating the probability of recontamination via the air using Monte Carlo simulations

Recontamination of food products can cause foodborne illnesses or spoilage of foods. It is therefore useful to quantify this recontamination so that it can be incorporated in microbiological risk assessments (MRA). This paper describes a first attempt to quantify one of the recontamination routes: via the air. Data on the number of airborne microorganisms were collected from literature and industries. The settling velocities of different microorganisms were calculated for different products by combining the data on aerial concentrations with sedimentation counts assuming that settling is under the influence of gravity only. Air movement is not explicitly considered in this study. Statistical analyses were performed to clarify the effect of different products and seasons on the number of airborne microorganisms and the settling velocity. For both bacteria and moulds, three significantly different product categories with regard to the level of airborne organisms were identified. The statistical distribution in these categories was described by a lognormal distribution. The settling velocity did not depend on the product, the season of sampling or the type of microorganism, and had a geometrical mean value of 2.7 mm/s. The statistical distribution of the settling velocity was described by a lognormal distribution as well. The probability of recontamination via the air was estimated by the product of the number of bacteria in the air, the settling velocity, and the exposed area and time of the product. For three example products, the contamination level as a result of airborne recontamination was estimated using Monte Carlo simulations. What-if scenarios were used to exemplify determination of design criteria to control a specified contamination level.