The Role of Predictive Microbiology in Microbial Risk Assessment Robert L. Buchanan U.S. DHHS Food and Drug Administration Center for Food Safety and Applied Nutrition
Microbiological Risk Assessment
Microbial Risk Assessment Dramatic changes are occurring in the way regulatory issues associated with microbiological foods safety concerns are evaluated as a result of the emergence of quantitative microbial risk assessment techniques during the past 8 years
Microbial Risk Assessment Prior to 1995: Generally considered that microbiological food safety concerns were too complex to be amenable to formal risk assessment techniques 1994-1996: First research attempts to conduct microbiological risk assessments are published 1998: First formal assessment conducted by a U.S. regulatory agency published
Barriers to Conducting Microbial Risk Assessments Large variation in virulence of strains Large variation in susceptibility of hosts Microbial populations change rapidly due to growth and inactivation Large variation in growth and survival characteristics of microorganisms Microbial contamination sporadic Substantial amount of data on adverse effects, but dose-response relations generally not established
Examples of Early Research Microbial Risk Assessments Drinking Water & Shellfish: U.S. EPA; 1988-1996 (Rose, Gerba, Haas) Salmonella Enteritidis - Eggs: USDA; 1996-1998 (Buchanan & Whiting; Baker et al.) Listeria monocytogenes: FDA, USDA, ICMSF; 1994-1999 (Peeler & Bunning; Miller et al.; Buchanan et al.; Whiting et al.) E. coli O157:H7 in Ground Beef: Health Canada, USDA; 1998-1999 (Cassin et al.; Marks et al.)
U.S. Microbiological Risk Assessments Listeria monocytogenes in ready-to-eat foods Salmonella Enteritidis in Eggs and Egg Products Vibrio parahaemolyticus in Oysters Listeria monocytogenes in Deli Meats Enterohemorrhagic Escherichia coli in Ground Beef Fluoroquinolone Resistance in Campylobacter
FAO/WHO Microbial Risk Assessments Listeria monocytogenes ready-to-eat foods Salmonella in broilers Salmonella Enteritidis in eggs Campylobacter in broilers Vibrio parahaemolyticus in fish and shellfish Enterohemorrhagic Escherichia coli in ground beef and produce
What Is Microbial Risk Assessment? Quantitative microbial food safety risk assessment is a group of modeling techniques that allow one to describe the relationships between the presence of a hazard in a food and the likelihood that the hazard will lead to an adverse public health consequence
What Is Microbial Risk Assessment? Risk assessments are a means of systematically arraying and analyzing scientific data that are pertinent to a risk management question: Structured to clearly tell what we know Descriptive to characterize how well we know it Transparent to reveal any bias
What Is Microbial Risk Assessment? The end products of a risk assessment are: a statement of probability or likelihood a statement of uncertainty.
Microbial Risk Assessment Qualitative Probability and uncertainty are expressed using descriptor (e.g., low risk, high risk) Used when either Insufficient data to describe risk quantitatively Insufficient time, resources, or need to do a full risk assessment Many qualitative risk assessment are actually semi-quantitative
Microbial Risk Assessment Quantitative Provides an quantitative estimate of risk Single point estimate Probability profile Done using mathematical modeling Simple as possible and still answer the risk management question adequately
Microbial Risk Assessment Flexible tool: Different types Risk Ranking priority setting Product/Pathogen Pathway Analysis intervention strategies Risk-Risk examination of consequences Geographical emergence Address all or part of the food chain
Microbial Risk Assessment While specific techniques will vary with the agent be examined, risk assessments are typically divided into four phases Hazard Identification Exposure Assessment Hazard Characterization (Dose- Response Relation + Severity Assessment) Risk Characterization
Why Dramatic Change? Internationally WTO SPS and TBT Agreements Codex Alimentarius Nationally (U.S.) Executive Order 1994 USDA Reorganization Act
Why Dramatic Change? Availability of personal computers Availability of PC-based software for simulation modeling Monte Carlo Techniques
Why Dramatic Change? 10-year worldwide investment in research in predictive microbiology Modeling techniques Models Expertise Training Application software
Predictive Microbiology
Predictive Microbiology The use of mathematics to describe the behavior of microorganisms
Predictive Microbiology Emergence during the 1930 s as a means of describing the inactivation of microorganisms during thermal processing Log # D = -1/slope Thermal Death Time Curve Time
Predictive Microbiology Expanded greatly during the 1980 s and 1990 s to include modeling of Growth Survival Inactivation Competition Food Unit Operations
Predictive Microbiology The availability of predictive microbiology models are critical to quantitative microbiology risk assessment Usually hidden Mathematical expressions are imbedded in the risk assessment model Need to look at the details of the risk assessment
Understanding Exposure Assessments Risk is based on what goes into the consumers mouth
Understanding Exposure Assessments Essentially all microbiological data on the frequency and extent of microbiological contamination is collected at sites earlier in food chain Predictive microbiology to estimate the actual level ingested by the consumer
FDA Listeria monocytogenes Risk Assessment Most data available on the various foods considered were at retail, with some critical databases at point of manufacturer Need to take retail data and convert to time of consumption data
Listeria monocytogenes Exposure Assessment Frequency of contamination of food Number of Lm when contaminated Growth before consumption Frequency of consumption Amount food consumed Number of Lm consumed with a serving
FDA Listeria monocytogenes Risk Assessment To estimate exposure needed to have data and models for Food consumption Food contamination Growth, survival and thermal inactivation -- refrigeration, storage, and cooking/reheating
Growth Model SqRT EGR = a ( T - To ) 0.4 Sq Rt EGR (log/d) 0.3 0.2 0.1 0.0-5 0 5 10 Temperature ( C) Ratkowsky et al., 1982
Dealing with Diversity 30 25 20 15 10 5 0 <32 33-35 36-38 39-41 42-44 45-57 48-50 51-53 54-56 57-59 60-63 Home Refrigerator Temperatures ( F)
Dealing with Diversity Semi-Soft Cheese 0.04 0.035 0.03 Probability Density 0.025 0.02 0.015 0.01 0.005 0-0.005 0 5 10 15 20 25 30 Storage (Days)
Dealing with Diversity Multiple Studies with Growth Data Adjusted to 5 C Cumulative Frequency 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0 0.2 0.4 0.6 Growth rate (log/d)
Mathematical Modeling Use Monte Carlo simulations to increase the accuracy of the predictions Considers parameters as distributions instead of single values Based on the Theory of Large Numbers
Monte Carlo Simulation What number would I get if I rolled three dice?
Monte Carlo Simulation 3 3 1 = 7
Monte Carlo Simulation 6 6 3 = 15
Monte Carlo Simulation 30 25 Frequency 20 15 10 5 0 0 5 10 15 20 Mean = 10.5 Sum
FDA Listeria monocytogenes Risk Assessment Example 1: Deli (Processed) meats that support growth Manufacturer to Retail: Growth rate model - temperature Retail to Consumption Growth rate model temperature Maximum growth model - temperature
FDA Listeria monocytogenes Risk Assessment Example 2: Frankfurters Manufacturer to Retail: Growth rate model - temperature Retail to Consumption Growth rate model temperature Maximum growth rate model temperature Food preparation: Thermal inaction models Percentage eaten without reheating
FDA Listeria monocytogenes Risk Assessment Example 3: Cultured Dairy Products Retail to Consumption Survival model temperature and ph
Impact of Risk Assessment on Predictive Microbiology L o g (B a c te ria l C o u n t) The lessons learned about variability and uncertainty from risk assessment has greatly expanded our knowledge, techniques, and tools that we use in predictive microbiology 3.5 3.0 2.5 2.0 0 5 10 Incubation Time (hr) 15
Summary Predictive microbiology and microbial risk assessment are intimately interconnected Cannot do risk assessment with predictive microbiology models Cannot understand predictive microbiology models without understanding of variability and uncertainty An investment in one will aid the advancement of the other Part of an overall evolution in food safety microbiology toward being more quantitative