USDA, ARS Workshop Poultry Food Assess Risk Model (Poultry FARM)
A quantitative microbial risk assessment model (QMRA) for Listeria, Salmonella, Campylobacter and chicken meat.
Predicts the public health impact of chicken meat destined for specific distribution channels and consumer populations
Packaging Consumption Distribution Channel Cooking Safe Unsafe To maximize the public health benefit of chicken by ensuring its safety & consumption
Hazard Identification Exposure Assessment Hazard Characterization Risk Characterization How many people will get sick and die? Holistic
Salmonella (Se) Campylobacter (Cj) Initial distribution of pathogens among servings Listeria (Lm)
Pathogen levels on chicken meat (mean log MPN/carcass) Plant APlant BPlant CPlant DPlant E Lm Se Cj Waldroup et al. (1992) J. Appl. Poultry Res. 1: Cj levels are higher than Lm and Se, which are similar
Relative differences among pathogens are simulated in Poultry FARM
Plant Packaging Contamination Distribution Abuse Growth/Death Preparation Contamination Transfer Cooling Abuse Growth/Death Cooking Under-cooking Survival Serving Contamination Transfer Table Consumption Dose-response Predicts how pathogen levels change from farm-to-table Key: Unit Operation _ Human Action _ Pathogen Event
Physiological Differences Se grows, whereas Cj dies at ambient temperatures Burnette and Yoon (2004) Food Sci. Biotechnol. 13:
Predictive models can be developed for each pathogen event
No response InfectionMild illness Illness Determines whether or not an illness occurs
Depends on the outcome of the interaction between the pathogen, food and host Disease Triangle PathogenHost Food
IllnessHospitalDeath Determines the severity of illness
There are important differences in severity among pathogens OutcomeListeriaSalmonellaCampylobacter Hospital92.2%22.1%10.2% Death20.0%0.78%0.1% Mead et al. (1999)
Illness (C 1 ) Hospital (C 2 ) Death (C 3 ) Severity = C 1 + 2C C 3 Weight factor Cases
Foodborne illness is a random event
Monte Carlo simulation is a good method for modeling foodborne illness
A + B = C
Foodborne illness is a rare event
Iteration : 10,000 Discrete 1 0 : 0 Pert (0,2,4) : 2.2 Antilog : 0 Round 63 0 : 0 IncidenceExtent Pathogen Number =IF(RiskDiscrete=0,0,RiskPert) Poultry FARM simulates pathogen-free servings
Poultry FARM Tour
Lot ALot B Which is higher risk?
InputPathogenABCDE Q1 Q2 Q3 Lm Se Cj 0.0% 16.7% 78.1% 0.0% 28.1% 53.1% 29.2% 47.9% 90.6% 18.8% 24.0% 100.0% 0.0% 9.4% 82.3% Waldroup et al (1992) J. Appl. Poultry Res. 1:
QuestionABCDE Q45% Q510% Q615% 5%15%10% Q720% 10%20%10% Q825% 15%25%15% Q920%10% 30%20% Q1035%10%15%35% Q1125%5%15%25%15% Q1220%60%20%10%80%
QMRA Model = Poultry FARM 3.0 Iterations = 10,000 servings Simulations = 100 Sampling = Latin Hypercube Random Number Generator Seed = Random Selection
Each random number generator seed produces a unique outcome of the scenario
Listeria monocytogenesABCDELmSeCj0%16.7%78.1%0%28.1%53.1%29.2%47.9%90.6%18.8%24%100%0%9.4%82.3%
Salmonella entericaABCDELmSeCj0%16.7%78.1%0%28.1%53.1%29.2%47.9%90.6%18.8%24%100%0%9.4%82.3%
Campylobacter jejuniABCDELmSeCj0%16.7%78.1%0%28.1%53.1%29.2%47.9%90.6%18.8%24%100%0%9.4%82.3%
Lm + Se + CjABCDELmSeCj0%16.7%78.1%0%28.1%53.1%29.2%47.9%90.6%18.8%24%100%0%9.4%82.3%
It is important to consider multiple pathogens and post-process risk factors when assessing food safety