Previous Page Table of Contents Next Page



FAO and WHO undertook a risk assessment of Salmonella in eggs and broiler chickens in response to requests for expert advice on this issue from their member countries and from the Codex Alimentarius Commission. Guidance on this issue is needed, as salmonellosis is a leading cause of foodborne illness in many countries, with eggs and poultry being important vehicles of transmission.

The risk assessment had several objectives.

1. To develop a resource document of all currently available information relevant to risk assessment of Salmonella in eggs and broiler chickens and also to identify the current gaps in the data that need to be filled in order to more completely address this issue.

2. To develop an example risk assessment framework and model for worldwide application.

3. To use this risk assessment work to consider the efficacy of some risk management interventions for addressing the problems associated with Salmonella in eggs and broiler chickens.

This document could be used as a resource document that includes currently available information relevant to risk assessment of Salmonella in eggs and broiler chickens. Although a cost-benefit analysis of potential mitigations would assist risk managers in determining which mitigations to implement, it was not within the scope of this work and is not considered here.

In order to develop the model, the risk assessment was divided into two risk assessments with a shared hazard identification and hazard characterization. These two risk assessments included the four steps of risk assessment: hazard identification, hazard characterization, exposure assessment, and risk characterization.

One hazard identification and one hazard characterization, including a dose-response model, and two exposure assessment models - one of Salmonella in eggs and one of Salmonella in broiler chickens - were elaborated. For Salmonella in eggs, the risk characterization estimates the probability of human illness due to Salmonella following the ingestion of a single food serving of internally contaminated shell eggs, consumed as either whole eggs, egg meals or as ingredients in more complex food (e.g. cake). This work addressed selected aspects of egg production on farms; further processing of eggs into egg products; retail and consumer egg handling; and meal preparation practices. For Salmonella in broiler chickens, the risk characterization estimates the probability of illness in a year due to the ingestion of Salmonella on fresh whole broiler chicken carcasses with the skin intact, and which are cooked in the domestic kitchen for immediate consumption. This work commenced at the conclusion of slaughterhouse processing and considers in-home handling and cooking practices. The effects of pre-slaughter interventions and the slaughter process are not currently included in this model.

The inputs for this risk assessment were obtained from a variety of sources. Information was compiled from published literature, national reports and from unpublished data submitted to FAO/WHO by various interested parties.

The main outputs from the risk assessment are summarized below. It should also be noted that, in the course of the work, efforts were made to identify features that have an impact on the acceptability of findings and the appropriateness of extrapolating findings to scenarios not explicitly investigated in the risk assessments, and these are identified in the risk assessment document.


During the past two decades, Salmonella Enteritidis has emerged as a leading cause of human infections in many countries, with hen eggs being a principal source of the pathogen. This has been attributed to this serovar's unusual ability to colonize ovarian tissue of hens and be present within the contents of intact shell eggs. Broiler chicken is the main type of chicken consumed as poultry in many countries. Large percentages are colonized by salmonellae during grow-out and the skin and meat of carcasses are frequently contaminated by the pathogen during slaughter and processing. Considering the major role eggs and poultry have as vehicles of human cases of salmonellosis, an assessment of different factors affecting the prevalence, growth and transmission of Salmonella in eggs and on broiler chicken carcasses and the related risk of human illness would be useful to risk managers in identifying the intervention strategies that would have the greatest impact on reducing human infections.


The hazard characterization provides a description of the public health outcomes, pathogen characteristics, host characteristics, and food-related factors that may affect the survival of Salmonella through the stomach. It also presents a review of information on relevant dose-response models describing the mathematical relationship between an ingested dose of Salmonella and the probability of human illness. An extensive review of available outbreak data was also conducted. From these data, a new dose-response model was derived using a re-sampling approach, and this was used in both risk characterizations in preference to existing models that are defined within this component of the risk assessment. Finally, an attempt was made to discern whether separate dose-response curves could be justified for different human sub-populations defined on the basis of age and "susceptibility", and whether a dose-response for S. Enteritidis was distinguishable from a dose-responses for other Salmonella.

Three existing dose-response models for Salmonella were identified:

1. Fazil, 1996, using the Beta-Poisson model (Haas, 1983) fitted to the naive human data from Salmonella feeding trials (McCullough and Eisele, 1951a, b, c).

2. United States Salmonella Enteritidis Risk Assessment (US SE RA) (USDA-FSIS, 1998), based on the use of human feeding trial data for a surrogate pathogen (Shigella dysenteriae) with illness as the measured endpoint to describe the dose-response relationship.

3. Salmonella Enteritidis Risk Assessment conducted by Health Canada (2000, but unpublished) based on a Weibull-Gamma dose-response relationship. The model uses data from many different pathogen-feeding trials and combines the information with key Salmonella outbreak data, using a Bayesian relationship.

These dose-response models for S. Enteritidis and Salmonella were found to inadequately characterize the dose-response relationship observed in the outbreak data. A new dose-response model was developed in the course of this work. It was derived from outbreak data and was considered to be the most appropriate estimate for the probability of illness upon ingestion of a dose of Salmonella. The model was based on observed real world data, and as such was not subject to some of the flaws inherent in using purely experimental data. Nevertheless, the current outbreak data also have uncertainties associated with them and some of the outbreak data points required assumptions to be made. The outbreak data are also from a limited number of developed countries and may not be applicable to other regions.

From the outbreak data used to examine the dose-response relationship, it could not be concluded that S. Enteritidis has a different likelihood from other serovars of producing illness. In addition, comparing the attack rates of Salmonella for children less than five years of age, against those for the rest of the population in the outbreak database, did not reveal an overall trend of increased risk for this subpopulation. Although some indication for a difference in attack rates for the two populations had been noted in two of the outbreaks examined, the database of outbreak information might lack the potential to reveal the existence of any true differences. Severity of illness as a function of patient age, Salmonella serovar or pathogen dose were not evaluated, although severity could potentially be influenced by these factors and by pathogenicity. However, the current database of information was insufficient to derive a quantitative estimate for these factors.


The exposure assessment section for S. Enteritidis in eggs compares and contrasts previously completed models. It describes the general framework of these models, the data used, and the analysis completed for modelling analysis. Generally, these models comprise a production module, a module for the processing and distribution of shell eggs, a module for the processing of egg products, and a module for preparation and consumption. The production module predicts the likelihood of a S. Enteritidis-contaminated egg occurring. This depends on the flock prevalence, within-flock prevalence, and the frequency that infected hens lay contaminated eggs. The flock prevalence (i.e. the likelihood of a flock containing one or more infected hens) further depends on factors that serve to introduce S. Enteritidis into flocks (e.g. replacement pullets, environmental carryover from previously infected flocks, food contamination, etc.). The shell egg processing and distribution, and preparation and consumption modules predict the likelihood of human exposures to various doses of S. Enteritidis from contaminated eggs. The dose consumed in an egg-containing meal depends on the amount of S. Enteritidis growth between the time the egg was laid and when it was prepared, as well as how the egg was prepared and cooked. Growth of S. Enteritidis in contaminated eggs is a function of storage time and temperature. The output of the exposure assessment, in general, feeds into the hazard characterization to produce the risk characterization output. This output is the probability of human illness per serving of an egg-containing meal.

The exposure assessment included consideration of yolk-contaminated eggs and growth of S. Enteritidis in eggs prior to processing for egg products. These issues have not been previously addressed by exposure assessments of S. Enteritidis in eggs. Yolk-contaminated eggs might allow more rapid growth of S. Enteritidis inside such eggs compared with eggs that are not yolk-contaminated.

This risk characterization of S. Enteritidis in eggs was intentionally developed so as not to be representative of any specific country or region. However, some model inputs are based on evidence or assumptions derived from specific national situations. Caution is therefore required when extrapolating from this model to other countries.

Key findings

The risk of human illness from S. Enteritidis in eggs varies according to the different input assumptions in the model. The risk of illness per serving increases as flock prevalence increases. However, uncertainty regarding the predicted risk also increases as flock prevalence increases. Reducing flock prevalence results in a directly proportional reduction in human health risk. For example, reducing flock prevalence from 50% to 25% results in a halving of the mean probability of illness per serving. Reducing prevalence within infected flocks also results in a directly proportional reduction in human health risk. For example, risk of illness per serving generated from eggs produced by a flock with 1% within-flock prevalence is one-tenth that of a flock with 10% within-flock prevalence.

Adjusting both egg storage time and temperature profiles for eggs from production to consumption was associated with large effects on the predicted risk of human illness. The risk of human illness per serving appears to be insensitive to the number of Salmonella Enteritidis in contaminated eggs across the range considered at the time of lay. For example, whether it is assumed that all contaminated eggs had an initial number of 10 or 100 S. Enteritidis organisms, the predicted risk of illness per serving was similar. This may be because the effect of S. Enteritidis growth is greater than the initial contamination level in eggs.

As an example of how the efficacy of interventions aimed at reducing flock prevalence may be assessed the risk assessment examined the effect of a "test and divert" programme. Two protocols were assumed, with either one (at the beginning of egg production) or three (beginning of egg production, four months later & just before flock depopulation) tests administered to the entire population of egg production flocks and their effectiveness was estimated over a four-year period. Testing three times per year for four years reduced the risk of human illness from shell eggs by more than 90% (i.e. >1 log). Testing once a year for four years reduced risk by over 70%.

Other potential interventions evaluated included vaccination and refrigeration. To evaluate the effectiveness of vaccination against S. Enteritidis a single test, or two tests four months apart, with 90 faecal samples per test, was considered. The vaccine was assumed to be capable of reducing the frequency of contaminated eggs by approximately 75%. The effects of time and temperature restrictions were evaluated assuming a flock prevalence of 25%. Restricting shelf-life to less than 14 days reduced the predicted risk of illness per serving by a negligible amount (~1%). However, keeping retail storage temperature at no more than 7.7°C reduced risk of illness per serving by about 60%. Were shelf-life to be reduced to 7 days, risk per serving would also be reduced by about 60%.


The available data on which this risk assessment was based was limited. For example, evidence regarding enumeration of the organism within eggs was based on only 63 S. Enteritidis-contaminated eggs, and in part on estimates of the concentration of the organism in contaminated eggs. It is difficult to represent uncertainty and variability with such limited data. Apparently, there is a lot of uncertainty and it is difficult to quantify. In addition, statistical or model uncertainty was not fully explored.

Much uncertainty attends the effectiveness of various management interventions for controlling S. Enteritidis. The magnitudes of uncertainty regarding test sensitivity, effectiveness of cleaning and disinfecting, and vaccination efficacy have not been measured. Some data were available to describe these inputs, but the data may not be relevant to all regions or countries where such interventions might be applied.

Statistical or model uncertainty was not fully explored in this risk characterization. For example, alternative distributions to the lognormal for within-flock prevalence were not considered. In addition, the predictive microbiology used in this model was dependent on very limited data pertaining to S. Enteritidis growth inside eggs. Alternative functional specifications for S. Enteritidis growth equations were not pursued in this analysis.


The risk assessment model is defined in terms of a number of parameters that describe the processes of broiler chicken carcass distribution and storage, preparation, cooking and consumption. Some of these parameters can be considered general in that they can be used to describe the situation in many countries. At the same time, some parameters are country specific, such as the prevalence of carcasses contaminated with Salmonella at the completion of processing. Predictions of risk for a particular country are best obtained from data relevant to that country.

The exposure assessment of Salmonella in broiler chickens mimics the movement of Salmonella-contaminated chickens through the food chain, commencing at the point of completion of the slaughter process. For each iteration of the model, a chicken carcass was randomly allocated an infection status and those carcasses identified as contaminated were randomly assigned a number of Salmonella organisms. From this point until consumption, changes in the size of the Salmonella population on each contaminated chicken were modelled using equations for growth and death. The growth of Salmonella was predicted using random inputs for storage time at retail stores, transport time, storage time in homes, and the temperatures the carcass was exposed to during each of these periods. Death of Salmonella during cooking was predicted using random inputs describing the probability that a carcass was not adequately cooked, the proportion of Salmonella organisms attached to areas of the carcass that were protected from heat, the temperature of exposure of protected bacteria, and the time for which such exposure occurs. The number of Salmonella consumed were then derived using a random input defining the weight of chicken meat consumed per serving and the numbers of Salmonella cells in meat as defined from the various growth and death processes. Finally, in the risk characterization, the probability of illness was derived by combining the number of organisms ingested (from the exposure assessment) with information on the dose-response relationship (hazard characterization).

Key findings

The Salmonella in broiler chickens risk assessment does not consider all parts of the production-to-consumption continuum, and this limits the range of control options that can be assessed. This is primarily due to the lack of representative data to analyse how much change in either the prevalence or level of Salmonella in poultry could be attributable to any specific treatment or action. However, the establishment of a baseline model provided a means to compare the effects on risk when prevalence and cell numbers were changed. The model parameters can be modified to evaluate the efficacy of risk mitigation strategies that target those parameters. For example, the parameter describing prevalence of Salmonella-contaminated broiler chickens exiting processing can be modified to evaluate the effectiveness of a processing measure such as chlorination of the chilling water to reduce the prevalence of Salmonella-contaminated carcasses.

Reduction in the prevalence of Salmonella-contaminated chicken was associated with a reduction in the risk of illness. A one-to-one relationship was estimated, with a percentage change in prevalence, assuming everything else remains constant, reducing the expected risk by a similar percentage. For instance, a 50% reduction in the prevalence of contaminated poultry (20% to 10%) produced a 50% reduction in the expected risk of illness per serving. Similarly, a large reduction in prevalence (20% to 0.05%) would produce a 99.75% reduction in the expected risk of illness. If management strategies are implemented that affect the level of contamination, i.e. the numbers of Salmonella on chickens, the relationship to risk of illness is estimated to be greater than a one-to-one relationship. A shift in the distribution of Salmonella cell numbers on broiler chickens exiting the chill tank at the end of processing, such that the mean number of cells is reduced by 40% on the non-log scale, reduces the expected risk of illness per serving by approximately 65%.

A small reduction in the frequency of undercooking and the magnitude of the undercooking event results in a marked reduction of the expected risk of illness per serving. The important caveat here is that altering cooking practices does not address the risk of illness through the cross-contamination pathway. The strategy of changing the consumer's cooking practices needs to be tempered by the fact that cross-contamination may in fact be the predominant source of risk of illness, and the nature of cross-contamination in the home is still a highly uncertain phenomenon.

Limitations and caveats

It was not possible to provide a perfect representation of growth of Salmonella in raw poultry and seasonal variations in ambient temperature were not accounted for. The model adopted also assumed that ambient temperature had no impact on the rate of change for storage temperatures used for predicting growth, and this is intuitively inappropriate in some circumstances. Similarly, limitations were present in the way the model predicts the death of Salmonella in broiler chicken carcasses during the cooking process.

At several steps, reliance was placed on expert opinion to estimate the value of model inputs. While often easily accessible and sometimes sufficiently accurate, occasionally, expert opinion might reduce transparency and introduce an unacceptable bias that may not be detected by the risk assessors.

Surveillance data from some countries often show a marked seasonality in the number of notifications of human salmonellosis, with peak incidence occurring in the warmer months and the current model cannot account for or explain this important phenomenon.

A lack of detailed understanding of all aspects of cross-contamination in the home hampered the ability of the risk assessment to address this process. While the uncertainty associated with several parameters in the consumption portion of the risk assessment was accounted for, a full analysis of statistical and model uncertainty was not done. Thus, the influence of uncertainty in the cross-contamination pathway was not explored.


This Salmonella risk assessment provides information that should be useful in determining the impact intervention strategies may have on reducing cases of salmonellosis from contaminated eggs and poultry. In the risk assessment of Salmonella in broiler chickens, for example, it was determined that there is a relationship between changing the prevalence of Salmonella on the broiler chickens and reducing the risk of illness per serving. In the risk assessment of S. Enteritidis in eggs, reducing the prevalence of S. Enteritidis in poultry flocks was directly proportional to the reduction in risk to human health. The model can also be used to estimate the change in risk of human illness from changing storage times or temperature of eggs. However, comparison of effects of intervention measures, i.e. sensitivity analysis, cannot be done because this risk assessment is not conducted for a specific region or country, or for global settings. Data was collected from different countries for different input parameters. If those data were changed reflecting a specific national situation, the impact of a measure would also be changed. Therefore, caution would be needed in interpreting the results of this risk assessment in Codex activities.


Fazil, A.M. 1996. A quantitative risk assessment model for salmonella. Drexel University, Philadelphia PA. [Dissertation].

Haas, C.N. 1983. Estimation of risk due to low doses of microorganisms: a comparison of alternative methodologies. American Journal of Epidemiology, 118: 573-582.

Health Canada. [2000]. Risk assessment model for Salmonella Enteritidis. Unpublished Health Canada Document.

McCullough, N.B., & Eisele, C.W. 1951a. Experimental human salmonellosis. I. Pathogenicity of strains of Salmonella Meleagridis and Salmonella anatum obtained from spray-dried whole egg. Journal of Infectious Diseases, 88: 278-289.

McCullough, N.B., & Eisele, C.W. 1951b. Experimental human salmonellosis. II. Immunity studies following experimental illness with Salmonella Meleagridis and Salmonella anatum. Journal of Immunology, 66: 595-608.

McCullough, N.B., & Eisele, C.W. 1951c. Experimental human salmonellosis. III. Pathogenicity of strains of Salmonella Newport, Salmonella derby, and Salmonella Bareilly obtained from spray dried whole egg. Journal of Infectious Diseases, 89: 209-213.

USDA-FSIS. 1998. Salmonella Enteritidis Risk Assessment. Shell Eggs and Egg Products. Final Report. Prepared for FSIS by the Salmonella Enteritidis Risk Assessment Team. 268 pp. Available on Internet as PDF document at:

Previous Page Top of Page Next Page