Janell Kause (FSIS) Régis Pouillot (FDA) Dan Gallagher (VA Tech) Sherri Dennis (FDA)

Transcription

1 Janell Kause (FSIS) Régis Pouillot (FDA) Dan Gallagher (VA Tech) Sherri Dennis (FDA) 2015 JIFSAN-GMA Workshop: Evaluation of Risk Factors for Foodborne Listeriosis --

2 Background, Scope, Purpose Description of the Virtual deli model What can we learn from the model? General behavior Variations according to baseline conditions Scenarios Conclusions 2

3 Why would retail-sliced products be more contaminated than prepackaged ones? Very little was known about the transmission of Lm in retail delis and how cross contamination occurs Contributing factors: Retail environment/sanitary conditions More than one product prepared at a given time Meat, Poultry, Vegetables, Seafood, More than one process at a given time and place Slicing, Cutting, Mixing, Other factors: Incoming contamination Temperature control 3

4 Cumulative Fraction Hypothesis: at retail Additional cross-contaminations? Temperature abuses???? retail sliced prepackaged Questions: What are the key processes that lead to additional Lm contamination / higher levels at retail? How much is the relative risk/serving reduced according to specific risk management options? log10(lm concentration, MPN/g) 4

5 Collaboration Federal Agencies: FDA, FSIS, CDC (EHSNet) Academia: VA Tech, U. MD, Cornell University, and others New approach: Retail cross-contamination model Discrete event model Develop data specifically for the risk assessment model Stakeholder participation throughout the process Initial public meeting (June 2009) Risk management questions provided Partnership in research (e.g., extension of FSIS-funded retail contamination study with Cornell University and Purdue University; FMI and AMI) 5

6 Academics Data needs Scientific Meetings Data RM questions Scientific Community Risk Assessment Team Risk Management Team Literature Peer review Data needs (call for data) Information (meetings) Data Comments Questions Results Stakeholders 6

7 What is the exposure to Listeria monocytogenes from consuming ready-to-eat foods prepared in retail facilities? What are the key processes that increase ready-to-eat foods contamination at retails? How much is the relative risk per serving reduced according to specific risk management options? 7

8 Further refined; a list of proposed what if scenarios to evaluate Sanitation Worker behavior Growth inhibition Cross contamination Storage temperature & duration Examples What is the public health impact of temperature abuse in deli cases? What would be the impact of separated slicers/counters for growth versus non-growth products? What is the impact of the use of gloves in the retail environment? 8

9 Risk mapping Expert opinion to validate where Lm occurs and which transfers likely (FSIS Contract) Hoelzer K, et al. (2012) Risk analysis, 32(7): Transfer coefficients and Slicer Meta analysis of literature data (In-house study) Hoelzer K, et al. (2012) International J of Food Microbiology, 157: Persistent strains in deli departments Longitudinal study, Cornell University (FSIS Contract) Simmons C, et al.(2014) J Food Protection, 77(11): Potential transfer during specific events Mock deli using GloGerm, Virginia Tech (FSIS Contract) Maitland J, et al. (2013) J Food Protection, 76 (2): Food worker behavior Observational study (FDA JIFSAN UMD collaboration) Lubran MB, et al. (2010) J Food Protection, 73 (10):

10 Food Meat Cheese Salad Food workers Behavior Events Listeria Sites Slicers Cases Food Contact Surfaces Non Food Contact Surfaces Utensils Niches Slicers Cases 10

11 11

12 Wipe Slicer Removes some Listeria from the slicer (if any) 12

13 Wipe Slicer Wash hands & change gloves Removes some Listeria from hands (if any). No bacteria on the gloves 13

14 Wipe Slicer Wash hands & change gloves Open case, remove chub, close case Potential cross contamination between gloves and case 14

15 Wipe Slicer Wash hands & change gloves Open case, remove chub, close case Slice on gloves Potential cross contamination among gloves, slicer, chub and potential contamination of the product sold 15

16 Wipe Slicer Wash hands & change gloves Open case, remove chub, close case Slice on gloves Touch scale 16 Potential cross contamination between gloves and scale 16

17 Wipe Slicer Wash hands & change gloves Open case, remove chub, close case Slice on gloves Touch scale 17 Rewrap chub Potential cross contamination between the chub and the food contact surface 17

18 Wipe Slicer Wash hands & change gloves Open case, remove chub, close case Slice on gloves Touch scale 18 Rewrap chub Open case, replace chub, close case Potential cross contamination between the gloves and the case 18

19 Listeria growth on products 19

20 20

21 log 10 Prob illness Contamination when sold Home Storage (bacterial growth) Contamination when eaten Serving Size Number of Lm ingested Dose Response Model Probability of Illness log 10 dose (FAO/WHO, 2004) 21

22 Ad-hoc studies Food worker behavior Risk mapping Meta-analyses of Transfer coefficients and Slicer coefficients Persistent strains in deli departments Additional Data/Model Growth model Mejholm and Dalgaard model Temperature in deli case Ecosure 2007 data Time and temperature during transport and at home Ecosure 2007 data Potential transfer during specific events Consumption data NHANES Data Dose-response model FAO/WHO

23 Discrete Event model + Few bacteria in the system = Slow convergence Each simulation: 100 Stores; 1,000,000 Servings Currently: 22 Scenarios tested, for 6 Baseline Conditions Parameters Excel Data File R model Parallel computing CSV Outputs Blue Meadow cluster 2,016 cores, 21 TFlops Available through the Office of Science and Engineering Laboratories, 23 CDRH - FDA

24 General correspondence between the frequency of contamination vs. observed one (Cornell longitudinal study) Important sources of contamination considered in the model (Risk mapping, Mock Deli study) Control of the mass balance Correspondence between the simulated bacterial density distribution vs. the observed one (NAFSS, 2008) (this graph) 24

25 June 18, 2013 IRAC/NCAC-SRA/USDA-ORACBA Washington, DC 25

26 FROM From Transfers: log10(t0.baseline + 1) Growth Gloves Growth RawFood Niche Sold NFCS FCS Slicer UtensilHandle Case Scale Handle Sink Floor Glove Hand Salad Cheese Meat Meat Cheese Salad Hand Glove Floor Sink Handle Scale TO To Case UtensilHandle Slicer FCS Sold Trash and Wash NFCS Sold Washed Trash Initial 14 contamination, growth, 12 inactivation and transfers from 10 sites to sites Large number associated with gloves/hand Large impact of growth Large number of bacteria inactivated Some are sold 26

27 84925-American Cured Turkey Non susceptible Susceptible American Potato w GI Uncured Ham Cured Turkey w GI Pepperoni Cured Bologna Cured Ham Cured Ham w GI Cured Turkey w GI American Cured Bologna w GI Sequential Sales Potato Cured Ham Uncured Turkey Cured Ham Uncured Ham Uncured Ham Uncured Turkey Salami Monterey Jack Cured Ham Salami Cured Ham w GI Cured Bologna Cured Turkey Protein w GI American Potato w GI Uncured Turkey Potato w GI Cured Turkey x e-4 2e-4 3e-4 Chub concentration, cfu/g Sale Total cfu 0 5x Serving Risk Dose, cfu 27

28 Contaminated product at time of sale (incoming or cross contamination) Growth supporting RTE product Consumer mishandling (time / temperature) Susceptible consumer Retail cross contamination leads to sporadic cases, not outbreaks. 28

29 June 18, 2013 IRAC/NCAC-SRA/USDA-ORACBA Washington, DC 29

30 We evaluated a range of retail delicatessen conditions Approach Define some baseline conditions Baseline #1: regular environmental contaminations occur in the stores Baseline #2: no environmental contamination occurs in stores Evaluate various scenarios within these baseline conditions 30

31 Baseline conditions characterize a retail deli and the RTE food it serves at different times over the course of operations. A retail deli with: Multiple niches or environmental transfers that regularly release LM to food contact surfaces No niches or environmental LM transfer No niche and incoming RTE product highly contaminated and supports growth No niche and incoming RTE product highly contaminated and does not support growth Multiple niches and compliant with FDA Food Code for temperature control No niches and compliant with FDA Food Code for temperature control 31

32 1.8e-6 Mean Susceptible Risk per Serving 1.6e-6 1.4e-6 4.0e-7 2.0e No Niche Multiple Niche 100W Temp. Control Incoming Non growth Chub Incoming Growth Chub Niche & Temp. Control Baselines 32

33 7.0e-7 Niche Contaminated Product Mean Susceptible Risk per Serving 6.0e-7 5.0e-7 4.0e-7 3.0e-7 2.0e-7 1.0e S-Baseline Q50 (30) S-Niche Slicer 100-W S-Niche Slicer 1000W S-Niche Slicer 100D S-Multiple Niches 100W S-Multiple Niches 1000W S-Multiple Niches 100D S-IncLm growth (-9.2) S-IncLm growth (-8.9) S-IncLm growth (-7) S-IncLm growth (-5) S-IncLm growth (-3) S-IncLm no growth (-9.2) S-IncLm no growth (-8.9) S-IncLm no growth (-7) S-IncLm no growth (-5) S-IncLm no growth (-3) S-IncLm Salad (-5) Contaminated Scenarios product sales excluded from risk. 33

34 June 18, 2013 IRAC/NCAC-SRA/USDA-ORACBA Washington, DC 34

35 Within each of the 6 Baseline conditions 22 Scenarios Answer the question (example): Given that there is a niche in a retail deli, what are the best mitigation strategies? and not What is the probability that there is a niche in the store? 35

36 Sanitation Related Scenarios Some NFCS cleaned as frequently as FCS, Increase the effectiveness of cleaning, No sanitation, Worker Behavior Related Scenarios No glove, No contact glove-case, Preslice products in the morning, Do not slice product on gloves Growth Inhibitor Related Scenarios All products with GI, No product with GI Cross contamination Related Scenarios Separate slicers, No cross contamination Storage Temperature and Duration Related Scenarios Temperature in compliance with FDA food code, Temperature set so that no growth can occur 36

37 Observed sanitation practices critical in reducing risk. Stopping sanitation increased risk across all baselines. Additional sanitation (more effective cleaning, more frequent cleaning, ) generally not significant. Relative Susceptible Risk to Baseline (%) No Sanitation No Niche Multiple Niche 100W Incoming Growth Chub Temperature Control Niche & Temperature Control Incoming Non-Growth Chub Baseline 37

38 a. all GI Growth inhibitors prevented growth both at retail and at home. Broad growth inhibitor use led to dramatic reduction in risk. Relative Ssuceptible Risk to Baseline (%) b. no GI Multiple Niche 100W No Niche Incoming Growth Chub Temperature Control Niche & Temperature Control Incoming Non-Growth Chub Baseline 38

39 Reducing incoming mean concentrations by factor of 2 reduced risk across all baselines except incoming growth chub. Relative Susceptible risk to Baseline (%) a. Reduce Level -60 No Niche Multiple Niche 100W Niche & Temperature Control Incoming Non-Growth Chub Incoming Growth Chub Temperature Control Baseline 39

40 60 40 a. Transfers to 0 20 Eliminating cross contamination reduced risk across all baselines, especially incoming non growth chub Slicer is primary nexus for cross contamination. Relative Susceptible Risk to Baseline (%) b. Transfers and Slicers to No Niche Multiple Niche 100W Incoming Growth Chub Temperature Control Niche & Temperature Control Incoming Non-Growth Chub Baseline 40

41 No Niche Multiple Niches niches Incoming non growth chub Baseline No Contaminated Sales Cross contamination Cross contamination 2.3 No Cross 1.1 Contamination Susceptible Risk per Serving x Susceptible Risk per Serving x Cross contamination Susceptible Risk per Serving x 10-7 Necessary conditions: Cross contamination from any source then growth 41

42 Incoming non growth chub Incoming growth chub Baseline No Contaminated Sales No Cross Contamination Sales contaminated product Cross contamination Susceptible Risk per Serving x Susceptible Risk per Serving x 10-7 Necessary conditions: Growth plus additional cross contamination 42

43 Major increases in risks to consumers arise from Increased Lm entering deli department from incoming product Whether they support growth of Lm or not Increased Lm entering deli department from environment / niches Lack of adequate temperature control Low-hanging fruit Lack of growth inhibitor use (growth at retail and at home) Use of growth inhibitors whenever possible Lack of adequate sanitation Importance of the Slicer Washington, DC 43

44 Technical report and Interpretive Summary available on FSIS and FDA websites Model description: Journal of Food Protection, 2015, 78(1),

45 Stakeholders Consumer Science in the Public Interest Food Marketing Institute American Meat Institute AFDO EHSNet Other contributors as part of the Interagency Listeria monocytogenes in retail risk assessment work group: Dare Akingbade Nathan Bauer Karin Hoelzer Meryl Silverman Jia Tang Academia Cornell University Virginia Tech University of Maryland Purdue University 45

46 Multiple Niche 100W No Niche Incoming Growth Chub Incoming Non-growth Chub Temp. Control Niche & Temp. Control Predicted risk per serving, susceptible population Sanitation Related Scenarios: Percent Change Relative to Baseline Wash & Sanitize: Increase the effectiveness of cleaning from simply washing to washing and sanitizing * Clean 8 Sporadic: Double the number of sites cleaned from 4 to * No Sanitation: No wiping, washing, or sanitizing 41.3 * 7.9 * 2.9 * 23.5 * 11.9 * 50.2 * No Sporadic Cleaning: Clean as required by the 2009 FDA Food Code, but no additional sporadic cleanings NFCS As FCS: Workers clean deli NFCSs at same rate as FCSs * 0.9 Worker Behavior Related Scenarios: No Glove: Workers do not use gloves when serving customers 5.1 * * 6.0 * 7.0 * Gloves Every Serving: Workers change gloves before every sale No Contact Glove Case: Workers do not use their hands to open the deli case (e.g. if a floor switch is used) Pre-slice: Workers pre-slice RTE products in the morning, after cleaning 6.0 * 24.9 * 49.5 * * 19.2 * 1.0 Separate Slicer: Workers use a separate slicer for RTE products that support growth of L. monocytogenes -6.3 * * 22.7 * Do Not Slice On Gloves: Workers collect the slices of RTE products on tissue paper rather than on his/her gloved hand * Growth Inhibitor Related Scenarios: All GI: Reformulate all RTE products sold at the retail deli that would otherwise support L. monocytogenes growth to include growth inhibitors * * * * * * No GI: Reformulate all RTE products that support L. monocytogenes growth that are sold at the retail deli to not include GI to restrict L. monocytogenes growth * * 35.1 * * * * Cross Contamination Related Scenarios: Transfers to 0: Cross contamination would only result from the deli slicer Transfers and Slicer to 0: No cross contamination in the retail deli * * -9.5 * * * * Reduce Level: Mean incoming L. monocytogenes concentration in all RTE products lowered from -9.2 to -9.5 log 10 cfu/g * * * * * Separate Slicer Case: Workers use a separate slicer and a separate deli case for RTE products that support the growth of L. monocytogenes * * Lower Env Cont: Reduce transfer of L. monocytogenes among RTE products, FCSs, and NFCs (i.e., reduce transfer coefficients by 50%) * Storage Temperature and Duration Control Related Scenarios: Temp = 5 C: Set the retail deli case temperature to 5 C (41 F) (i.e., in compliance with the 2009 FDA Food Code) for all delis, instead of using the deli case temperatures reported by Ecosure * -8.1 * -2.8 NA NA No Growth (T=-5 C): At this temperature, no L. monocytogenes growth will occur * * * -5.7 * NA NA Temp 5 C: Use only the retail deli case temperatures observed in the Ecosure dataset at or below 5 C (41 F) * * * -8.2 * NA NA Shorten Time in Retail Delis: Reduce the length of time RTE products are held before they are sold or disposed from 7 to 4 days