1 DEPARTMENT OF HEALTH AND HUMAN SERVICES FOOD AND DRUG ADMINISTRATION CENTER FOR FOOD SAFETY AND APPLIED NUTRITION LISTERIA MONOCYTOGENES: RISK ASSESSMENT AND RISK MANAGEMENT Thursday, December 4, 2003 8:15 a.m. 5100 Paint Branch Parkway College Park, Maryland 2 C O N T E N T S Opening Remarks: Ms. Janice Oliver 4 Listeria monocytogenes: The Burden of Illness and Public Health Challenge: Dr. Robert Tauxe 6 Risk Assessment: Approach and Structure: Dr. Richard Whiting 21 Risk Assessment: Results and Conclusions: Dr. Robert Buchanan 38 Questions to Risk Assessment Panel: 61 Moderator: Dr. Robert Buchanan Panel: Mr. John Bowers, Dr. Mary Brandt, Ms. Janelle Kause, Dr. Richard Raybourne, Dr. Richard Whiting FDA/CDC Risk Management Action Plan: Dr. John Kvenberg 93 FSIS Risk Management Plan: Dr. Daniel Engeljohn 106 Question for Risk Management Panel 118 Moderator: John Kvenberg Panel: Dr. Joe Baca, Dr. Robert Brackett, Ms. Marjorie Davidson, Dr. Daniel Engeljohn, Dr. Gary German, Dr. Robert Tauxe, Dr. Richard Whiting Public Comments: Jenny Scott 160 Caroline Smith DeWaal 167 3 1 P R O C E E D I N G S 2 MS. DENNIS: Welcome to our public meeting 3 on Listeria monocytogenes. We will be covering 4 today risk assessment and risk management. 5 My name is Sherri Dennis. Before we get 6 started, I want to just go over a few housekeeping 7 details. In the packet you received when you 8 registered, there is a list of local restaurants. 9 There is also a cafe that is located just to your 10 left at the entrance to our building. We have 11 some courtesy phones that are located on the wall 12 in the hallway lobby where you registered this 13 morning. We would ask that you would please turn 14 off your cell phones when you are here in the 15 auditorium. You may find that your cell phone 16 probably doesn't even work in this building anyway. 17 You may not get good reception. 18 The rest rooms are located down the 19 hallway towards the security guard's desk and they 20 will be on your right. If you need anything today, 21 please let me know. I will be glad to try to help 22 you and I hope you enjoy the meeting. 23 Our first speaker will be Janice Oliver. 24 She is the Deputy Director for the Center for Food 25 Safety and Applied Nutrition. 4 1 Opening Remarks 2 MS. OLIVER: Thanks very much. I would 3 just like to say good morning and welcome you all 4 to this meeting where we will be discussing the 5 Listeria monocytogenes Risk Assessment among the 6 selected categories of ready-to-eat food and we 7 will also be discussing the Listeria monocytogenes 8 Risk Management Action Plan. 9 As we all know, Listeria is a harmful 10 bacterium and it is found in a variety of foods. 11 In pregnant illness, the illness attributed to 12 Listeria monocytogenes can result in miscarriage, 13 fetal deaths or severe illness or death of the 14 newborn. The elderly and those with weakened 15 immune systems are at elevated risk. 16 Most of us know that the Healthy People 17 2010 Goals called on the Federal Food Safety 18 Agencies to reduce the incidence of foodborne 19 listeriosis by 50 percent by 2005. We got off to a 20 good start in that, if you look at the preliminary 21 2001 FoodNet data, the incidence of foodborne 22 illness from Listeria monocytogenes between 1996 23 and 2001 went from 0.5 to 0.3 cases per 100,000. 24 But then what happened is it just sat 25 there at that plateau. So it is obvious that we 5 1 need to take some additional steps in order to 2 reach our Healthy People 2010 goal and reduce the 3 incidence of this illness. 4 In 2001, FDA and FSIS put out a Draft Risk 5 Assessment and the a Draft Risk Management plan. 6 We received a lot of comments. We held a public 7 meeting and, as a result of the comments that we 8 received and as a result of new data and some 9 updated modeling techniques that you will be 10 hearing about, we have revised the Risk Assessment 11 which we put out earlier this year and we have also 12 revised our Risk Management Plan which we put out 13 in November. 14 We are going to go through both of those 15 today and give you an opportunity to ask questions 16 about the Risk Assessment, how we conducted it, any 17 of the data in there as well as the Risk Management 18 Action Plan. 19 We look forward to your participation, 20 your active participation, in the meeting. I will 21 be here most of the day. Dr. Brackett will be here 22 most of the day. Dr. Buchanan will obviously be 23 here all day. We look forward to talking to you 24 all also at breaks. 25 Today, we are going to discuss a number of 6 1 things and we are going to start out with a 2 discussion of the public-health issues associated 3 with Listeria monocytogenes. The first speaker is 4 going to be Dr. Robert Tauxe from the Center for 5 Disease Control and Prevention. Most of you know 6 Dr. Tauxe. We have worked with him for years and 7 he has an excellent background in dealing foodborne 8 illness. 9 He is going to talk about the burden of 10 illness and the public-health challenge that is 11 associated with Listeria monocytogenes. I am going 12 to be going up to the back of the room and sitting 13 there for a little. I have to leave for about an 14 hour for a meeting, but then I will be back 15 Thank you very much. 16 Listeria monocytogenes 17 The Burden of Illness and the 18 Public Health Challenge 19 DR. TAUXE: Thank you very much, Janice 20 Oliver, and good morning to everyone to this fine 21 day I hope which lasts long enough to complete the 22 meeting before the snow starts. I am pleased to 23 spend a few minutes with you this morning 24 introducing the topic of Listeria monocytogenes, 25 the infections that it causes and the burden and 7 1 public-health challenge that it poses to all of us. 2 I am going to present a little bit of 3 background first on the organism, itself, and the 4 infection. I will describe some information we 5 have from our surveillance systems about the 6 current public-health burden, the trends and the 7 incidence of listeriosis, describe a few recent 8 outbreaks that highlight some of the issues that we 9 have faced and conclude. 10 Listeria monocytogenes, I am sure this is 11 not new information to anyone in this room, is an 12 interesting and curious bacterium. It is small and 13 rod-shaped. It lives in the soil and it can be 14 found in many locations. It thrives particularly 15 in the moist, cool locations in nature and in the 16 human environment and has the unusual and 17 unfortunate property of actually being able to 18 multiply slowly even at refrigerator temperatures, 19 or near refrigerator temperatures. 20 In the animal or the human being, 21 particularly the compromised human being, it can be 22 invasive. It can leave the intestine and enter 23 into the bloodstream and has a particular focus 24 that it can find its way into the lining, the 25 meninges, that surround the brain and spinal cord, 8 1 or to the fetus of the human is pregnant. It also 2 gets inside white cells and survives there which 3 makes it a little more difficult to treat with 4 antibiotics. 5 If the situation when an infection occurs 6 is in a pregnancy, that is a perinatal infection, 7 in late pregnancy, particularly, the pregnant woman 8 may have fever or may not have a defined illness at 9 all or may have a sense of a fluishness, a malaise 10 that is maybe difficult to characterize. 11 Nonetheless, the infection can spread to 12 the fetus and cause a serious infection there, a 13 sepsis or bloodstream infection or miscarriage or 14 stillbirth. If the infection occurs later or if 15 the transfer occurs to the baby as it is born, it 16 can, again, spread within the baby to the meninges 17 causing meningitis. These are very serious 18 infections in a newborn whose immune system is not 19 really mature yet. 20 Nonperinatal infections are typically in 21 compromised adults or possibly children whose 22 immune systems are not fully functional because of 23 malignancy, because of organ transplant, because 24 they are taking immunosuppressive medications, 25 because they have HIV or AIDS, or because they are 9 1 simply quite old. 2 In these compromised host, Listeria 3 monocytogenes can lead to an invasive infection and 4 disease; again, sepsis, meningitis or encephalitis, 5 a direct infection of the brain, itself. 6 The rest of the population also gets 7 infected from time to time. Most often, in the 8 normal host, it is an asymptomatic infection. We 9 have seen outbreaks where it has produced and 10 important diarrheal illness with fever or 11 gastroenteritis and, very rarely, it can lead to 12 invasive disease. 13 What we track, at CDC in public health, 14 are the invasive illnesses diagnosed because 15 someone's blood culture or spinal-fluid culture or 16 a culture of a newborn is positive. We have 17 estimates published in 1999 but which actually 18 refer to 1997, the overall burden of illness for a 19 number of different foodborne illnesses--I have 20 highlighted Listeria here to sort of put it on the 21 map with other foodborne infections that we track 22 and try to control. 23 This is estimated total number of 24 infections including those that are not diagnosed 25 or reported. We have estimated that there were 10 1 approximately 2,500 cases of listeriosis--invasive 2 listeriosis, this is--with about 500 deaths. That 3 is a case-fatality rate of 20 percent. The 4 proportion of illnesses leading to a fatal outcome 5 is 20 percent, which is, of course, much higher 6 than the other infections. 7 So, even though the number of cases is 8 relatively small, it accounts for a large 9 proportion of the deaths associated with the 10 foodborne illnesses. In fact, in that year, our 11 estimate was it was approximately 30 percent of the 12 deaths attributed to any of the known established 13 causes of foodborne infection. So it is the severe 14 illness that makes Listeria, otherwise a rare 15 infection--it is the severe illness that makes it 16 of public-health importance. 17 We have been tracking listeriosis with 18 active surveillance for some time. The impetus 19 began in 1985 when there was a very large outbreak 20 in California. There were 142 cases and 40 21 fatalities that were traced to a Mexican-style soft 22 cheese, a queso fresco. In the wake of that 23 outbreak, CDC, with collaboration from FDA, began 24 active surveillance in sentinel locations which 25 continued more or less until today. 11 1 Active surveillance means that the 2 public-health staff contacts the clinician 3 laboratories in a population area in order to 4 identify cases as they are diagnosed even if they 5 are not officially reported to the public-health 6 system. There may not be reporting locally. But, 7 if it is diagnosed, it will be identified and 8 tracked down. 9 In 1996, this active-surveillance system 10 was expanded to include other foodborne infections 11 and became what we call FoodNet with joint support 12 from USDA and FDA continuing to present. 13 FoodNet began in 1996 with these five 14 sites and a population of 14 million under 15 surveillance. Let me fast-forward quickly now and 16 add several sites more until 2003 when we welcomed 17 New Mexico, not yet reporting. This gives us ten 18 reporting sites and 13 percent of the U.S. 19 population. 20 We use the information from FoodNet to 21 answer a number of different questions. I will 22 just present a few highlights of information that 23 we have had. One question is has the mortality of 24 listeriosis changed over time? There have been new 25 antibiotics, new treatment, perhaps better 12 1 diagnosis. 2 The case-fatality data from the active 3 surveillance suggests we are not making a lot of 4 progress. Back in 1986, the first year of active 5 surveillance, the case-fatality rate was 31 6 percent. By '89, it was 24 percent and it has sort 7 of drifted down to 20 percent where it has really 8 stayed and has been pegged there since. So 9 advances in medical care are not going to solve 10 this problem. We are left with a high-mortality 11 disease. 12 We have also been able to look a little at 13 different population groups. If we look at the 14 aggregate period, 1996 through 2000, 4 per million, 15 the infection rate, then, among the population 16 under surveillance was 4 per million in the general 17 population. We can divide that out among different 18 groups and look at what are the relative rates of 19 infection among different groups. 20 Among white non-Hispanics, it was 2 per 21 million. Among African Americans, it was 2 per 22 million. Among Asians, 4 per million. In general, 23 among non-Hispanics, 2 per million. Among 24 Hispanics, 7 per million, an interesting difference 25 and one, I think, we need to be bearing in mind. 13 1 This difference was particularly important when we 2 look at the perinatal fraction of the problem. 3 Among infants, the incidence rate was twelve-fold 4 higher among Hispanics than non-Hispanics and a 5 similar rate among women of childbearing age. 6 The incidence of recorded cases of 7 listeriosis that we have been tracking with our 8 active surveillance since 1986 looks like this. 9 Here is the beginning, back in '86, and it, 10 perhaps, increased slightly or just varied up to 11 close to 8 per million in 1989. At that time, 12 there was a rather well-documented case, a single 13 case, of illness associated with a turkey frank 14 which crystallized a lot of concern and energy and 15 lead to new regulatory policies and renewed efforts 16 to reduce contamination in some ready-to-eat meat 17 plants. This was followed by a decline down to 4 18 or 5 per million where it remained relatively 19 steady. 20 Then, in the most recent phase, there have 21 been a new series of outbreaks--there were several 22 more outbreaks in here but there were several large 23 outbreaks, important outbreaks, here recently, and 24 there has been a further decline. 25 What happened recently has been really 14 1 conditioned by an advance in surveillance, another 2 advance in surveillance in addition to the active 3 surveillance, and that is molecular subtyping in 4 the service of public health, a network that we 5 call PulseNet. 6 What we did was equip first a few and then 7 subsequently all state public-health laboratories 8 with the ability to do pulse-field gel 9 electrophoresis on bacterial isolates. There is 10 also a good collaboration, I should say, with the 11 laboratories of FDA and USDA in this system. Those 12 began to look at subtyping of E. coli 157 first and 13 then, in that year, at the very end of '98 and '99, 14 began to look at Listeria where each state could 15 look at the pulse-field gel patterns of Listeria, 16 could compare them with what other states were 17 finding and with the national PulseNet database. 18 This system has evolved to include virtually all 19 states now doing E. coli 157 and, more and more, 20 doing Listeria and other pathogens as well, and 21 sharing their information with the national 22 database at CDC. 23 As that was just beginning, we had this 24 outbreak in the Northeast, 1998, 1999, a large 25 outbreak, again, 101 cases in 22 states, a rather 15 1 diffuse, spread-out, outbreak, all of whom were 2 infected with the strain that had the same 3 pulse-field gel pattern. There were 15 deaths and 4 six miscarriages as part of this. 5 Epidemiologic investigation implicated hot 6 dogs from a particular plan. This epidemic curves 7 shows the evolution, really, the first case with 8 that pattern being in August and going on, really, 9 until the New Year season following some events at 10 the plant that may possibly have been related. A 11 slow increase. The first realization that there 12 might have been a problem coming rather late, we 13 would say. We were just beginning to do subtyping 14 of strains and that happened late. Then the 15 recognition that there was a problem, investigation 16 and, ultimately the recall, and then a prompt drop 17 off in the cases after that. 18 A second outbreak, just to show you 19 another sort of problem, 2000, 2001, a 20 cheese-associated outbreak in North Carolina, again 21 a winter-season outbreak. There were 12 cases in 22 North Carolina, all in Hispanic people. Ten were 23 in pregnant women and there were five stillbirths 24 as a result. 25 Epidemiologic investigation, again, this 16 1 time indicated an association with eating queso 2 fresco. In this case, the queso fresco was 3 purchased from door-to-door vendors. It was a very 4 informal sort of cheese, a home-made queso fresco, 5 that was made in several locations using local raw 6 milk obtained from local dairies. 7 The strains, again, has the single 8 PulseNet pattern that was identified both in the 9 patients and the cheeses and in the raw milk. The 10 state, as a result, banned the sale of raw milk and 11 reinforced some other regulations and launched an 12 education program about the potential hazard of 13 unpasteurized cheeses among the Hispanic 14 population. 15 Finally, an outbreak I will mention that 16 just happened last year, 2002, affecting 54 17 patients in nine states, again detected only 18 because we have PulseNet observations going on. 19 Among those 54 patients, there were eight deaths 20 and three miscarriages, or still births. The 21 outbreak was caused by turkey deli meat. Again, as 22 in the past, post-processing contamination appeared 23 the likely source. 24 The FSIS part of USDA issued new 25 microbial-sampling policy in the wake of this 17 1 outbreak increasing environmental testing and 2 planning to base recall on environmental test 3 results. 4 Let's go back to this curve, now, and say 5 that here, with that first hot-dog outbreak, the 6 queso fresco outbreak, the deli-meat outbreak, a 7 lot of interest, attention and concern focused on 8 Listeria again and, concomitant with that, a 9 decline, as Janice Oliver mentioned, down to about 10 2.7 per million which went horizontal here for the 11 last two years. 12 Here is simply a picture of those 13 outbreaks over time, of outbreaks of listeriosis 14 that we have been investigating reported to us. 15 The yellow indicates outbreaks, like the North 16 Carolina outbreak, were just in a single state. 17 The red blocks indicate outbreaks that were 18 multi-state outbreaks. You will notice that, until 19 PulseNet really came in, no red blocks occurred. 20 There were no multistate outbreaks that were 21 identified. 22 I think they were occurring all in this 23 time. We simply had no surveillance system that 24 would pick them up. Introduction of a subtyping 25 network begins to identify them and, since, then, 18 1 we find them along with, perhaps, more of the 2 one-state outbreaks as well. 3 So this increase in surveillance 4 sensitivity led to an increase in the sorts of 5 outbreaks that were detected. Here is simply the 6 same information line-listed. You can see--again, 7 the colored ones represent the multi-state 8 events--that we have had an array of different food 9 vehicles implicated in outbreaks which is sort of 10 one very crude measure but a summary measure of at 11 least where that part of the problem is coming from 12 and includes turkey deli meats, raw milk, cheeses, 13 pate and imported cheese, the hot dogs and a couple 14 of outbreaks in earlier years from milk. 15 So, in general, at least the conclusion 16 that we can draw from the outbreak information is 17 that these outbreaks have been traced to processed 18 ready-to-eat meats, especially turkey and hot dogs. 19 Fresh soft cheeses made with raw milk have been a 20 recurrent problem and the contamination, when the 21 investigators go into the plant and try to 22 understand where did this come from, how did it 23 occur, it appears that there is often a 24 contamination event after an initial processing and 25 that the locus of contamination is in the 19 1 processing plant. 2 We should also say that, outside of the 3 outbreaks, we have conducted two sporadic 4 case-control studies looking at the potential 5 exposures of the sporadic cases. We have recently 6 wrapped up the data-collection phase of a third 7 study. In 1988, the first study, indicated an 8 association between listeriosis and unreheated hot 9 dogs and undercooked poultry. A second study 10 published in 1992 implicated foods from a deli 11 counter, which is a pretty broad category, and soft 12 cheeses. 13 So what we have learned from our 14 surveillance data, to summarize very quickly, is 15 that overall incidence has declined from a peak of 16 8 per million to 2.7 per million in the last 17 decade. The mortality remains high, at 20 percent. 18 The risk groups include immunocompromised persons, 19 the elderly, pregnant women, and, particularly with 20 the question about the Hispanic population, the 21 young women of the Hispanic population. 22 With PulseNet, our ability to detect and 23 investigate Listeria outbreaks has improved in the 24 past five years and we have seen that investigating 25 the outbreaks is often a joint investigation, 20 1 involving not just CDC but state health-department 2 officials, and FDA or USDA and the industry, 3 itself, leads to immediate and longer-term control 4 measures and suggest that an important fraction of 5 Listeria can be prevented. Efforts to reduce 6 contamination are followed by declines and the 7 incidence of human illness. 8 How to proceed, now, further with 9 listeriosis control is the subject of this meeting 10 and you will hear a great deal more about the 11 National Action Plan shortly and the goal to reduce 12 the incidence of reported cases to below 2.5 13 million per year which would be the current 14 population, approximately below 700 cases per year. 15 This is measured--our role in this at CDC 16 will be measuring one aspect of this, the 17 incidence, through active public-health 18 surveillance of invasive cases with FoodNet. In 19 addition, we are enhancing surveillance nationwide, 20 not just in FoodNet sites, with enhanced subtyping 21 with the PulseNet. We will be attempting to get 22 most, if not all, of the listeriosis case isolates 23 subtyped and into the PulseNet system and 24 conducting standardized patient interviews so that 25 outbreaks can be detected very quickly and 21 1 investigated and controlled. 2 Thank you very much. 3 MS. DENNIS: Thank you very much, Dr. 4 Tauxe. 5 Our next speaker, to kick off our 6 discussion about the HHS-USDA Listeria 7 monocytogenes risk assessment is Dr. Richard 8 Whiting. He is with the Center for Food Safety and 9 Applied Nutrition. 10 Risk Assessment: Approach and Structure 11 DR. WHITING: Thank you, Sherri. I guess 12 the first news is that it is out, finally. What I 13 want to do today is to quickly give a background 14 and an introduction to the science part of it. 15 This is our flow for the next two presentations 16 here. I want to talk about the background, say a 17 little bit about the risk-assessment process for 18 those of you might not be familiar with risk 19 assessments. 20 I will mention what some of the key 21 modifications between our 2001 draft and this 22 current draft are and then talk a little bit about 23 the data and the modeling that went into the Risk 24 Assessment. In the middle of this part, Dr. 25 Buchanan will then pick up the story and continue 22 1 with the data modeling and go on into the results 2 and, then, what we call "what if" scenarios, which 3 is a new thing to this version of the Risk 4 Assessment that I think you will find very 5 interesting; then, finally, the conclusions and 6 interpretations. 7 This slide gives a brief summary of sort 8 of the historical part of this project. It began 9 with the Federal Register Notice in 1999. We had 10 several public meetings. Our National Advisory 11 Committee was involved in looking at our data. We 12 had internal and external reviews of our data, 13 assumptions, modeling techniques and a draft 14 document at that time. 15 Then, in January of 2001, we came out with 16 our Draft Risk Assessment. We had a public meeting 17 which was this one at that time and we asked for 18 comments. And we did get many comments back from a 19 wide variety of people from industry, government, 20 academia. We got quite a bit of new data as well 21 in the time period since 2001 and we have improved 22 our modeling techniques. 23 So, with that, we have revised the 24 document and this is what are presenting today. It 25 is obviously a big team effort and I just want to 23 1 put a few of the names up there of the people who 2 managed the project and, in the middle, the team 3 leaders for some of the various sections of the 4 Risk Assessment and, on the bottom, some of the 5 team members who were very active with this last 6 couple years' effort. 7 I want to particularly point out that FSIS 8 has been very much in this Risk Assessment with us, 9 contributing very much to the Product Section. We 10 have also had many discussions over the telephone 11 with some of our colleagues down at the CDC, 12 particularly the epidemiology people. So we are 13 very thankful for their collaboration. 14 Our Risk Assessment, we did try to carry 15 it out in the evolving guidelines from Codex and 16 the National Advisory Committee and the ICMFS for 17 conducting a risk assessment, that we are trying to 18 be transparent. We are trying to have a broad 19 scientific and stakeholder input and have an 20 extensive peer review built into the process, and 21 also trying to meet our own and OMB's data-quality 22 guidelines. 23 I want to take a couple slides here just 24 to say a few comments about risk assessments, in 25 general, at this point. The first is to point out 24 1 that the Risk Assessment is a tool to help the risk 2 managers reach a decision. It is not something to 3 come up with the answer for them and it is also not 4 the only information that they should use in coming 5 up with their policy or whatever decisions that 6 they might want to make. 7 The Risk Assessment assembles and 8 organizes our current knowledge but it does not 9 create any new knowledge on itself. When we do 10 these modeling techniques and groupings and so on, 11 we have to realize these are all just 12 simplification of what actually happens in the very 13 complex world of our foods. These kinds of 14 calculations typically cannot be experimentally 15 verified and we call this a science-based process, 16 but there is a much interpretation and judgment of 17 the data that has to be done. So, it is 18 science-based but, perhaps, not strictly speaking, 19 science. 20 On the general approach to modeling, we 21 use a technique called stochastic modeling or 22 sometimes using Monte Carlo techniques. The key 23 thing in that we take a dataset and use the entire 24 dataset, not just focusing on using the average of 25 that dataset. I am going to show a couple of 25 1 simple slides here just to show you what we mean by 2 that. 3 In this particular example, we are just 4 taking two dice and saying what is the sum of 5 rolling these two dice. Each die is, in fact, a 6 distribution. You have six possible outcomes. 7 Each is equally likely to occur. So we have we say 8 a uniform distribution here, each, as I say, likely 9 to occur. 10 Essentially the question, then, is, when 11 we roll the dice, what is the sum of these two 12 equal distributions. We could make a guess that 13 the average is 3.5 for each die so the sum is 14 probably 7. But we know this is not going to 15 happen every time. So what we can do is what you 16 would probably think of immediately; you can roll 17 the dice many times, tabulate up the results and 18 get a distribution of the sum of these two die. 19 If I do that, this is, in fact, what you 20 get. The average is 7, as you would have expected, 21 but you see we actually get a distribution out of 22 here that is now triangle shape. What is very 23 important, also, in this process is we use the 24 entire distribution on the output as well, so, on a 25 problem like this, we might not be that concerned 26 1 with what the average of the two die is. We are 2 more interested in what fraction of the time might 3 the answer be 9 or above? These are the kinds of 4 results we can get out with this kind of analysis 5 that you normally can't with just working with 6 averages. 7 This is why, when we talk about variation 8 and uncertainty and they add up and you see results 9 with confidence intervals showing the uncertainty, 10 this is the kind of spread that you get as you go 11 through these calculations. 12 Returning now to our Listeria Risk 13 Assessment, the objective of this Risk Assessment 14 was to try to determine which foods are the ones 15 that are most likely to cause or lead to 16 listeriosis and then, secondly, to say a little bit 17 about what are the different human subpopulations 18 and what do we know about the difference 19 susceptibilities of these different subpopulations 20 to listeriosis. 21 It is also important to realize that this 22 Risk Assessment was not trying to determine how a 23 food became contaminated or try to look at the 24 particular processing process that went into that 25 resulted in the contamination level. We generally 27 1 start our risk assessment with data that was taken 2 at the retail level, so how the food became 3 contaminated at retail was not a part of Risk 4 Assessment. 5 Bob Tauxe has already gone over much of 6 this data on listeriosis, but I want to point out 7 two points here, that Listeria monocytogenes will 8 grow at refrigeration temperatures. We based our 9 study on the statistics of about 500 deaths and 10 2,000 additional cases per year. 11 We made extensive modifications in the 12 current draft from what we presented in 2001. The 13 cheeses were reorganized and based primarily on the 14 moisture content. Frankfurters were separated into 15 two different categories, those that were reheated 16 by the consumer or the food preparer before 17 consumption and that were not reheated. In the 18 previous Risk Assessment, the not-reheated ones 19 were just sort of a side or a sub-calculation. 20 After much debate, we decided that 21 coleslaw was more appropriately a deli salad than a 22 vegetable salad. We changed several of the small 23 parts of the contamination data and foods in it to 24 remove foods that were not a problem and, finally, 25 then, the miscellaneous dairy products were 28 1 separated into two categories, where, in this 2 version, the cultured dairy products, yogurt and 3 sour cream, are a separate category and the 4 remaining dairy products, then, are what we call 5 the high-fat products because they are primarily 6 cream and butter. 7 We also had much new contamination data to 8 put into this Risk Assessment. Over 40 studies 9 altogether, but two extensive databases, one 10 collected by the NFPA through JIFSAN and some data 11 from the IBFA on milk contamination. We also had a 12 couple of years of more recent data from FSIS on 13 meat products. 14 In the first Risk Assessment, we did not 15 have any information on the growth of Listeria in 16 deli salads. We since came upon one study, did 17 some research in this lab on our own and found 18 that, in fact, Listeria dies off in most deli 19 salads, so there is a fairly major change in that 20 category. 21 The American Meat Institute also did some 22 surveys that gave us some information on what 23 consumers do in the home for storage for 24 frankfurters and deli meats and also some data on 25 the percentage of frankfurters that are consumed 29 1 without reheating. 2 We had a lot of comments on the question 3 of using non-U.S. data and on the recency of the 4 contamination data that went into the first Risk 5 Assessment. So what we did in this version of it, 6 we have a weighting system where recent data is 7 given more weight than the older data. Also foods 8 that make contributions to the U.S. food supply are 9 given much more weight than data from countries 10 that do not contribute to the U.S. food supply. 11 So this would mean a product like shrimp, 12 which we do import from Southeast Asia, would be 13 part of the U.S. food supply but, say, milk from 14 Southeast Asia would not. So we had some food 15 categories where we did not have any recent data. 16 So, of course, we could not necessarily balance 17 categories that have recent data with categories 18 that do not have recent data so we also devised a 19 correction factor for some of these categories. 20 The general flow and approach that we used 21 in the Risk Assessment is on this slide. The first 22 step was to use the contamination and growth data 23 to try to predict the levels at the time of 24 consumption for each food category and then sum all 25 of those up for the entire ready-to-eat food 30 1 category, or the entire Risk Assessment. 2 With the contamination, then, consumption 3 data. And we had the epidemiology data from the 4 Centers for Disease Control. We then created a 5 dose-response model. This model was for the three 6 populations. Essentially, the model was fitted, if 7 you will. So, given the contamination, it would 8 predict the right number of cases. This was the 9 general approach. We used the contamination data 10 and the epidemiological data to create the 11 dose-response model. 12 Having the dose-response model, we could 13 then go back to each individual food categories and 14 predict the risk per serving for that food category 15 and what we thought was an estimate of the number 16 of cases from each of the food categories. 17 Then we went on, since we had the models 18 and the process available, to do what we call the 19 "what if" scenarios and some various other 20 uncertainty analyses to give us an idea of some 21 other types of analysis that we can get out of the 22 data. You will see examples of this later. Then, 23 finally, to compare our results against the 24 knowledge of epidemiological records and interpret 25 the data. 31 1 I would like now to move into the exposure 2 assessment part which is the calculations of the 3 number of Listeria that we think people are 4 consuming. We selected the food categories based 5 on their potential for Listeria contamination. 6 After looking at these, we decided that the 7 ready-to-eat food categories was where the problem 8 was coming from with the exception of frankfurters. 9 As I mentioned earlier, we have frankfurter that 10 are not reheated and frankfurters that are 11 reheated. 12 We picked foods that had the history of 13 causing listeriosis. We looked at the 14 food-contamination data and the consumption data. 15 When we got done with this, we ended up with 23 16 food categories. 17 This slide summarizes those, or lists all 18 of them. With seafood, we have the smoked seafood, 19 the raw seafood, which would be the sushi-type 20 products, preserved fish which is primarily pickled 21 and salted, and then the cooked, ready-to-eat 22 crustaceans. Fruits and vegetables; again, this 23 would be ready-to-eat produce not vegetables that 24 are cooked before consumption. 25 Under meats, we have the frankfurters that 32 1 are reheated, frankfurters that are not reheated by 2 the food preparer, the dry, semi-dry, fermented 3 sausages, and then the deli meats. The difference 4 between these two categories; of course, the dry, 5 fermented sausage is like salami, genoa, lebanon 6 baloney which have a fair degree of acidity and 7 generally a lower water activity compared to the 8 deli meats which would be primarily your 9 sliced-turkey products, baloney and ham. Then, 10 finally, pate and meat spreads. 11 In the dairy products, we have the fresh 12 soft cheeses. This is primarily the queso fresco 13 Hispanic-style cheese, the soft-unripened cheese, 14 which would be cottage cheese and ricottas, the 15 soft-ripened cheeses, which would your Brie, 16 Camembert. Semi-soft would be Gouda, brick, 17 Edam-type cheeses. Finally, the hard cheeses, 18 cheddar and swiss. 19 The processed cheeses, of course, again, 20 another low-growth type of product. We have 21 pasteurized milk, unpasteurized milk, and then the 22 high-fat and other dairy products with the butter 23 and cream, the cultured dairy products and then, 24 finally, the ice cream and frozen products. Our 25 last category is the deli-type salads. 33 1 In the exposure assessment, then we 2 attempt to estimate the number of Listeria that 3 people actually consume because, from an 4 epidemiological point of view, this is the 5 important number of Listeria. But, of course, most 6 of our data on Listeria contamination come from 7 samples taken at the retail level so a major part 8 of the exposure assessment, then, is to try to 9 calculate what are the changing numbers of Listeria 10 from the retail sampling to the time of 11 consumption. 12 To do this, we needed information on the 13 frequency of contamination of a food, not only was 14 it positive or not but, if it was positive, what 15 are the numbers of Listeria that are there; how 16 much could Listeria grow before consumption, 17 exponential growth rates, storage times, storage 18 temperatures and the maximum amount of growth; how 19 frequently was the food consumed and what was the 20 serving size. 21 As far as sources of some of the data, the 22 consumption surveys were the CSFII and the NHANES. 23 The contamination data came from the literature, 24 came from some of the special studies which I will 25 show you later, and also from some of the 34 1 government and other surveys. Growth, survival, 2 thermal-inactivation data primarily from the 3 scientific literature. 4 Altogether, on the exposure-assessment 5 part, somebody calculated, at one point, over half 6 a million datapoints went into this Risk 7 Assessment. 8 I would like to show you a few of the 9 datasets just to give you a little idea of what is 10 involved here and why some of the results came out 11 the way they did. This slide shows the number of 12 servings per year in the U.S. for some of the 23 of 13 the food categories. Notice here, this is now the 14 total number of ready-to-eat servings; we estimated 15 this to be about 340 billion servings. 16 When we looked at some of the individual 17 food categories, smoked seafood, for example, with 18 200 million, pates with 120 million, are some of 19 the relatively low number of servings. Yet, these 20 are two of the foods that traditionally have been 21 associated with causing listeriosis. 22 Contrast those number of servings with 23 products like deli meats, with 21 billion servings 24 a year; high-fat and dairy products also there, and 25 pasteurized milk up at 87 billion servings a year. 35 1 So you can see there is a very wide range, as I 2 look at it here, almost a 700-fold range between, 3 say, pate and pasteurized milk. 4 This is some of the data that was 5 collected within the last two years on what the 6 contamination of some of our foods like with 7 Listeria. This was collected by the NFPA and 8 published this summer, the Gombas, et al., article. 9 The pasteurized-milk data came from the IDFA. 10 We have the total number of samples for 11 some of the different food categories here. The 12 less than 0.4 meant the Listeria were undetectable 13 in that sample. And then, if the sample was 14 positive, then what was the number of Listeria CFU 15 per gram that was in the sample. 16 You can see, first of all, we have got a 17 very extensive dataset--I think over 30,000 samples 18 altogether. Most of them, about 98 percent, I 19 think, on total, were negative. You can see there 20 were a substantial number of samples that did 21 contain Listeria, but when they were positive, most 22 of them are in this 0.01 or barely detectable 23 levels of Listeria. 24 However, there are also a significant 25 number of samples here which do contain some fairly 36 1 high numbers of Listeria monocytogenes. I would 2 want to point out again that this is samples taken 3 at the retail level, so we have not allowed for 4 whatever changes might occur during the home 5 storage. 6 Some of them, like the smoked seafood, 7 deli meats and deli salads, you can see are quite 8 high. But then pasteurized milk, here, with 5,400 9 samples, we only had one sample that was positive. 10 This slide shows the study by Audits 11 International on home refrigerator temperatures. 12 Again, the average here, 39 to 41 degrees 13 Fahrenheit, which is very good. But, as I 14 mentioned earlier in the risk-assessment process, 15 we are looking at whole range of the data here. 16 What this shows is about 10 percent of the home 17 refrigerators are running above 45 degrees 18 Fahrenheit. Even more significantly, about 19 1.4 percent of the refrigerators are running above 20 51 degrees Fahrenheit. 21 We didn't have very much information on 22 what the home-storage times would be. So this is a 23 generic curve that we use. This one is for 24 semi-soft cheese indicating that most of the 25 products are consumed within 15 days of purchase. 37 1 But, again, there is a small number of servings 2 which are held for a fairly extensive period of 3 time before they are consumed. 4 When the AMI did their study for, this is 5 prepackaged deli meats and frankfurters, this is 6 the data they came up with. I think it does 7 reflect the general shape that we used. But, 8 again, you see most of the products consumed within 9 seven days or ten days of purchase. But, again, 10 here is 3 percent of the product that is stored for 11 a fairly extensive period of time. 12 When we then put this data in, put in the 13 growth rates, calculated the numbers of Listeria 14 that we thought were in the products at consumption 15 for each of the 23 food categories and then summed 16 up all 23 of the food categories, this slide, then, 17 is the composite of what we think the U.S. consumer 18 is exposed to for Listeria monocytogenes. 19 This would be the number of Listeria in a 20 serving. If there were no Listeria, that was 3.3 21 times 10-11 servings per year out of 3.4 times 1011 22 servings. If you divide that through by the U.S. 23 population, that comes out to about 1,300 servings 24 a year per person, so about three-and-a-half 25 servings per day. 38 1 If we look at the number of servings that 2 have 1 to 1,000 Listeria, it is 109 Listeria or 3 about 19 servings a day. More than a thousand. 4 Then, in the millions. And then in the billions. 5 Looking at the last two categories, summing the 6 million and billion, about 2 times 108 servings per 7 year, or approaching once per year that we are 8 estimating that the person would be exposed to a 9 serving of this number of Listeria. 10 So the key numbers to kind of think about 11 at this point here is we have 340 billion servings. 12 Only about 1 in 1,000 is heavily contaminated, but 13 that will then be paired up against our 14 epidemiological record of about 2,000 cases and 500 15 deaths per year. 16 Bob? Do you want to just flow right in? 17 Risk Assessment: Results and Conclusions 18 DR. BUCHANAN: Good morning, everyone. I 19 am going to pick up where Dick left off and go 20 through the second half of the presentation of the 21 Risk Assessment. I will be picking up with the 22 hazard characterization or the dose-response phase 23 of the work followed by the risk characterization 24 and the presentation of the results in a simplified 25 form. Then I will finish up the talk with some 39 1 interpretation and conclusions that we have reached 2 as a result of that. 3 So let's start off with the hazard 4 characterization. For those of you that are not 5 familiar with risk-assessment techniques, this is 6 the area where we take the exposure and then relate 7 it to a biological effect or the adverse effect in 8 the population. 9 I might note here, in putting together the 10 hazard characterization, we employed three age 11 groups in terms of potential cases. These are the 12 perinatal. These are a fetus from 16 weeks after 13 fertilization up to 30 days after birth. These 14 were all assumed to be exposed in utero; that is, 15 it was the mother that was consuming a food 16 contaminated with Listeria and it resulted in an 17 infection within the fetus. 18 The elderly; we used the CDC definition of 19 elderly which is 60 years of age. This allowed us 20 to compare directly to the FoodNet data. Then we 21 took the intermediate group which we considered 22 anyone over 30 days of age and less than 60 years. 23 It is important to note that this is actually a 24 mixed population because it includes both healthy 25 individuals and also some of the more susceptible 40 1 subpopulations--for example, people receiving 2 certain medical treatment. 3 Our approach to modeling the hazard 4 characterization, to go over what Dick introduced 5 and then provide it in a little bit more detail, 6 the biological endpoint that we were looking at was 7 mortality. We found that this was the most direct 8 measurable and least controversial measurement that 9 we had. It is a very clear-cut biological 10 endpoint. 11 The shape of the dose-response curve is 12 based on the serious of mouse-lethality studies 13 that were done earlier. It is important to note 14 this was simply to elicit the shape of the curve. 15 The actual position of the dose-response curve was 16 then fixed by anchoring it to annual disease 17 statistics and, in this case, it was basically the 18 FoodNet data. 19 Then we considered, within that process, 20 the variation in the virulence of Listeria 21 monocytogenes isolates based on a number of studies 22 that have been done. Then we also considered, as 23 part of our consideration for variation and 24 uncertainty, the variation in susceptibility within 25 age groups, the variation and susceptibility 41 1 between age groups and then, also, variation in the 2 fatality to hospitalization ratios; that is, there 3 was 20 percent plus or minus something in terms of 4 the number of cases that were hospitalized that led 5 to fatalities. 6 This is the mouse dose-response curve 7 fitted with a modeling technique. We use, as a 8 simple example, the exponential curve that is the 9 blue line there. The actual model process was 10 somewhat more sophisticated, but basically this 11 provided the shape. Basically, the curve was then 12 moved to the left or right in order to get it match 13 up with the appropriate annual disease statistics. 14 I might note the anchoring process. This 15 is the national projected annual CDC incidence of 16 Listeria based on the most recent FoodNet data and 17 this was the data to which this Risk Assessment was 18 anchored. 19 The resultant was that we were able to 20 produce three dose-response curves, one each for 21 the different groups. You will see that the 22 intermediate age, which is the least susceptible 23 population, is to the right and that is the 24 dose-response curve we use for it. The elderly is 25 then the next-most susceptible. Then, finally, the 42 1 neonatal is the most with this risk assessment. 2 I might note, we have simplified this. 3 There are uncertainty limits around each of these 4 curves which were taken into account when we were 5 using them. 6 So, Dick presented the exposure 7 assessment. I have provided you the background, a 8 little background, on the hazard characterization. 9 We then move into actually generating results from 10 the risk assessment in the process known as risk 11 characterization. The process that we used here 12 was to combine the exposure assessment and hazard 13 characterization to quantitatively determine risk 14 for each of the food categories. 15 We then used these quantitative risk 16 categories to directly establish the link. So if 17 we, at any point, talk about ranks, they are 18 quantitatively linked to the actual calculation of 19 risk. Then we used this information and modeling 20 techniques to also measure the variability and 21 uncertainty around those estimates, and I will 22 briefly introduce some of the techniques we used 23 there. 24 Finally, using those risk estimates, we 25 compared those risk estimates against the 43 1 qualitative data and the epidemiologic record that 2 we had to help us, one, interpret the results and, 3 two, the epidemiologic record also helped us to 4 validate the efficacy of the modeling techniques. 5 It is important to note, as I go through 6 the risk characterization phase that we use two 7 different metrics. There is a need to understand 8 these. The first is risk per serving. This is the 9 risk that would be faced by an individual consumer 10 if they picked up an individual serving and 11 consumed it. Typically, the metric used here was 12 the predicted number of cases per million serving, 13 or the probability that you would get an illness as 14 a result of consuming a million servings. 15 The second metric that we use is risk per 16 annum. This is based on taking the risk per 17 serving and then multiplying it by the number of 18 servings consumed within the United States. So 19 this can be conceived of as the risk to the country 20 or the total disease burden that would be 21 associated with any of the commodity groups. 22 So that brings us to the results and I 23 will be presenting the results, both in the format 24 of risk per serving and also the risk per annum. 25 This is a graphic representation of the risk per 44 1 serving for the total population. I might note we 2 have these all individually for each of the 3 subpopulations we considered. 4 Basically, what we are looking at here is 5 the total cases of listeriosis per serving on a log 6 scale, so we are really looking at a small 7 probability for any single serving in almost any of 8 the foods. You can see a very large variation in 9 the risk per serving. It ranges from a high for 10 deli meats on the left, on your left, in this chart 11 at a little under a risk of about 10-7 per serving 12 all the way down to the far right, which is the 13 last group there. It is hard cheeses which is down 14 around 10-14 cases per serving. 15 We do note--I have color-coded these 16 thanks to some input from the Risk Assessment Team, 17 into three colors. It is interesting here because 18 it provides some real insight to some of the things 19 that impact risk. The blue squares are those 20 products that support growth of Listeria 21 monocytogenes at refrigeration temperatures. 22 The yellow squares are foods where 23 Listeria monocytogenes either does not grow or dies 24 off slowly during storage. That is at the far 25 right. Then the group in the middle are foods that 45 1 may or may not support growth of Listeria 2 monocytogenes at refrigeration temperatures. You 3 can see that these work out very nicely, that the 4 products on a per-serving basis with the highest 5 degree of risk are those that support growth. 6 These are the same results now calculated 7 to take into account the total number of servings. 8 You will notice that the rankings among these 9 change rather dramatically in some cases because a 10 product that might be relatively risky but is only 11 consumed occasionally, their per-annum risk 12 actually drops. Conversely, a product that may 13 only be moderately risky but, if it is consumed in 14 large amounts very often, its per-annum risk would 15 go up. 16 So, again, we see this, the way to 17 interpret this, this is, again, on a log scale so, 18 in this case, the estimated number of cases per 19 year on the far left for deli meats is somewhere 20 just above 1,000 cases. If you go down to where 21 you start falling below zero--zero on the log scale 22 would be one case per year. When you go into a 23 lower value, a negative number, you interpret that 24 taking into account a number of years. So if you 25 go to the far right, for hard cheeses, the value 46 1 says that the prediction is that you would have a 2 case associated with hard cheese once every 10,000 3 years. 4 This is the overall data summed. This is 5 available in your packet out there and you can look 6 at the details. I don't expect any of you can read 7 it or you have much better eyes than I do. But it 8 is important to note that we have attempted to rank 9 these in part. This is our first attempt of 10 evaluating them both in terms of an objective 11 numerical ranking based on their quantitative 12 results and also some initial characterization we 13 did on a somewhat arbitrary setting of high, medium 14 or low risk for per-serving risk and also on the 15 basis of very high, high, medium and low on the 16 per-annum risk. 17 We did, as you will see and I will come 18 back to, attempt to find different ways of taking 19 these results and trying to combine them or 20 evaluate them in comparison to each other to 21 provide our risk managers with some interpretation 22 of the data. 23 I might note, also, as I have indicated 24 earlier, we did a substantial amount of examination 25 within the risk assessment of the uncertainty 47 1 associated with this. The risk model was every 2 time we ran it, it involved 30,000 iterations and 3 then we did that 4,000 times to get an estimate of 4 our uncertainty within the model. 5 We did this--I am showing you one of the 6 forms that we did. This is just based on the 7 ranking. This for deli meats. If you will 8 remember, this came out on the per serving as the 9 highest one, you will see, when we ran the model a 10 number of times where it came out in terms of its 11 ranking. This is a pretty tight distribution, one 12 that indicates that there is not a lot of 13 uncertainty associated with these results. 14 Conversely, that is not always true for 15 all of the categories. So, for example, with 16 fruits, which is a very diverse group, some fruits 17 supporting growth, others not supporting growth, we 18 actually got a substantial amount of variation and 19 uncertainty showing up. This probably 20 reflects--you can almost see two or three 21 modalities here in terms of the results focusing 22 largely at the bottom end. These probably 23 represent the fact that, in some fruits, the 24 organism dies off and, in others, it will actually 25 grow. 48 1 We also did the same process for the 2 per-annum results so we do have uncertainty 3 estimates in the Risk Assessment available for each 4 of these categories. This I put in pretty much as 5 an example of what most of the food categories look 6 like. The uncertainty was typically normally 7 distributed around the median value for the risk 8 ranking or the risk estimate, itself, and typically 9 the distribution was pretty tight, again indicating 10 a reasonable degree of uncertainty in terms of 11 making subsequent risk estimations. 12 One of the things that is very powerful 13 about doing a quantitative risk assessment of this 14 type, or actually what we did is actually 23 15 quantitative risk assessments, one for each food 16 category, is that it allows it to go back and 17 perform "what ifs" scenarios where you change some 18 of the data or you change some of the factors to 19 see what would happen. This is a very good tool 20 for examining what would be potential 21 risk-management options in terms of getting the 22 most control for the dollar. 23 We did examine a number of them within the 24 Risk Assessment, itself These are the list that we 25 considered within the Risk Assessment. We 49 1 continued to do these almost on the weekly basis as 2 additional people come up with questions about what 3 they would like us to consider. 4 So I am just going to go through a couple 5 of them to show you some of the power that we have 6 as a result of these techniques. 7 This is one, for example, where we 8 considered the impact of this home-storage time on 9 the number of cases predicted, in this case for 10 deli meats for the elderly subpopulation. What we 11 did is we started with the storage-time 12 distribution that we had used in the risk 13 assessment and then began to truncate it at 14 different times putting anything that had been 15 normally higher with the distribution along this 16 point; that is, everyone quickly realized that they 17 were at the limit of it and they went and ate all 18 the food to make sure that they didn't waste it. 19 What you can see here, we did pick two 20 arrows, one including the entire distribution. You 21 can see that annual mortality associated with this 22 on this estimate was about 230 cases and that, by 23 decreasing the storage time so that nothing was--no 24 serving was consumed after 15 days as opposed to 25 30, we got a reduction, a predicted reduction, in 50 1 number of cases, about 10 percent. You can see, 2 though, that the curve gives you actually the 3 results for any of the intended values. 4 We did the same thing for the same product 5 in the same subpopulation with temperature, again 6 truncating the temperature so that, if you could 7 get everyone's home refrigerator not to exceed a 8 certain temperature--that is, we cut off the end of 9 that distribution, the tail where you had the 10 refrigerators that were really out of sync, you can 11 see what that impact is. 12 You can see from this one, again, we used 13 two points just to dramatize it where we chopped 14 off the distribution so that no refrigerator was 15 above 7. This led to an estimated 50 to 70 percent 16 reduction in the predicted number of cases. 17 I might note that this is for one food 18 with a certain distribution of storage time. You 19 can have an even larger effect with some of the 20 other foods. So, for example, this is the same 21 temperature scenario for pasteurized milk for the 22 elderly. You can see here, by getting rid of the 23 refrigerators that are above the recommended 40 24 degrees Fahrenheit, or, in this case, 45 degrees 25 Fahrenheit, that there is essentially an 51 1 elimination of the case predicted for pasteurized 2 milk. 3 You can also use it to determine what 4 other options might be available. So, for example, 5 one of the areas of interest was what was the 6 predicted difference between raw-milk cheeses, 7 fresh soft cheeses made from pasteurized versus 8 unpasteurized milk. Most of the data that we had 9 on these cheeses were from commercial sites that 10 used heat-treated milk and then looked at what 11 would happen if you eliminated that. It becomes 12 very apparent that you get about a 40-fold increase 13 in risk by making cheeses from raw milk, these 14 fresh soft cheeses. 15 We have done the same thing looking at 16 potential interventions in terms of different 17 products, designing the products so that they also 18 have barriers to growth incorporated in them. In 19 this case, we examined the frankfurters not 20 reheated examining what would be the impact on 21 predicted number of cases if the product was 22 formulated so it did not support growth. Again, 23 this provides you with the type of estimates that 24 you can use in looking at the impact, potential 25 impact, of these interventions which I might note 52 1 here was quite substantial. 2 So this brings us to conclusions and 3 interpretations. What we have found in the Risk 4 Assessment is that it is very helpful for the Risk 5 Assessment Team to sit and work with the risk 6 managers to actually help through some of the 7 interpretation of the data in terms of telling a 8 story. 9 Certainly, one of the things that this has 10 emphasized, for those of you that have worked with 11 Listeria, is there are some things here that we 12 have identified that are not new but have certainly 13 reinforced, in a quantitative manner, some of the 14 lessons that we have learned in the past. 15 One, some of the broad themes that have 16 been reemphasized as a result of the Risk 17 Assessment is that this is a disease that primarily 18 impacts specific at-risk subpopulations. This 19 continues to account largely for--about 96 percent 20 of the cases seem to be attributed to people with 21 increased susceptibility; that the disease is rare 22 but, when it does occur, it is severe and that 23 there are substantial differences in the risk among 24 food categories. 25 As I have already indicated by that one 53 1 graph, there is actually about a 10-million-fold 2 difference in the risk associated with the highest 3 and the lowest in terms of per-annum risk. 4 There are certain factors that contribute 5 strongly to these and when you take them down to 6 some of the broad themes, a risk associated with 7 any one food category is really dependent on the 8 interaction of these five characteristics; the 9 amount and frequency of consumption, the more often 10 you eat a product that is contaminated, the more 11 your risk is; the frequency and levels of 12 contamination; the level of contamination, or the 13 more frequent the product is contaminated, again 14 that will increase your risk. 15 As you already saw from that per-serving 16 distribution, the ability of a food to support 17 growth has a tremendous impact on the risk 18 associated with a product. Likewise, and these are 19 interconnected, the storage temperature and the 20 storage time for those products that support growth 21 are going to strongly influence their risk. 22 We attempted to take the data, 23 particularly since we were using two metrics to 24 examine them in terms of their expressing their 25 risk, risk per serving and risk per annum. Within 54 1 the body of the Risk Assessment, we do go through 2 several different ways of evaluating the data, 3 statistical analyses. We did some criteria 4 setting, et cetera. One of the techniques that we 5 found most effective in this process is a technique 6 called cluster analysis. Looking at a specified 7 confidence level, this allowed us to more 8 objectively and repeatedly be able to put the 9 different food categories into different clusters 10 that sort of said, these product are alike. 11 We did this clustering technique for both 12 the per-serving risk results and the per-annum risk 13 results because each of them will have an impact on 14 how we would then subsequently manage the risk. To 15 try to provide these in terms of getting an overall 16 view, what we did is develop a two-dimensional 17 matrix that, then, laid out the cluster analysis in 18 two dimensions; the per-serving risk and the 19 per-annum risk. 20 When you do that, this actually sets up a 21 very useful array of products that is read on the 22 diagonal where we were able to subdivide the 23 products, based on the results of the Risk 24 Assessment, from very high risk down to very low 25 risk. I will go through each of these categories 55 1 and our interpretation of them in the next several 2 slides. 3 The risk designation of very high included 4 two products, deli meats and frankfurters not 5 reheated. These had the characteristics of having 6 high predicted relative risk per serving and high 7 relative risk per annum. These are products that 8 were characterized by reasonably high rates of 9 contamination. They are products that supported 10 the rapid growth of Listeria monocytogenes at 11 refrigeration temperatures and, usually very 12 quickly, at marginal abuse temperatures. These 13 were products that where at least a percentage were 14 stored for extended periods. 15 These are also products, as you can 16 anticipate from their high per-annum risk that are 17 consumed extensively and in usually large serving 18 sizes. These were interpreted as products that 19 really need immediate attention in either 20 developing new control strategies or in finding 21 ways to communicate these risks to the consumer. 22 The high-risk designation consisted of 23 high-fat and other dairy products, pasteurized 24 fluid milk, pate and meat spreads, soft unripened 25 cheeses, smoked seafood and unpasteurized fluid 56 1 milk. These were products that had either a high 2 predicted per-serving risk or a high per-annum 3 risk. It is important to note it is an "or" there. 4 The ability to support the growth; these 5 products all had the ability to support the growth 6 of Listeria monocytogenes. However, when we looked 7 into the category of these food categories, we 8 basically found that we needed to break them up 9 into two subgroups. One subgroup consisted of 10 pate, meat spreads, smoked seafood and 11 unpasteurized fluid milk. The second subgroup was 12 high-fat dairy products, pasteurized fluid milk and 13 soft unripened cheeses. 14 Just to go through those and why we broke 15 them up into two subcategories, the first 16 subcategory, the pate, smoked seafood and 17 unpasteurized fluid milk were products that had 18 moderate to high rates of contamination and had 19 high to moderate rates in terms of the levels of 20 contamination. These were products that were 21 typically consumed infrequently so none of these 22 would be high on the list of what would be 23 considered a staple commodity within the diet. 24 Then, finally, our interpretation of these 25 is that this is a priority for new control measures 57 1 or continued avoidance in the case of unpasteurized 2 fluid milk. 3 The second subcategory within the 4 high-risk designation again included the high-fat 5 dairy products, pasteurized fluid milk, soft 6 unripened cheeses. These were products that had 7 low to moderate contamination frequency and low 8 contamination levels. However, this was 9 counterbalanced by the fact that they either had 10 moderate or high consumption rates as; for example, 11 in the case of pasteurized fluid milk, it had the 12 highest consumption rate. 13 It is also noted here that, in these three 14 categories, there is a minimal amount of 15 quantitative data available. So we consider these 16 to be priority candidates for either advanced 17 epidemiologic studies or advanced scientific 18 investigations to better match the epidemiologic 19 record with the predicted risk estimates. 20 The moderate-risk designation included a 21 variety of foods. They are all included here. I 22 am not going to try and read them all off. 23 However, just to categorize them, this is a very 24 diverse group. Many of these products include a 25 bactericidal step or an inhibitor. The presence of 58 1 Listeria monocytogenes in these foods is primarily 2 associated with either recontamination of the 3 product after that bactericidal step in the case of 4 those products that support growth or poor initial 5 sanitation in those products that don't support 6 growth. 7 I might note here, these are well-known 8 products that have a history of control activities. 9 Safety in these products is primarily associated 10 with continued vigilance of proven control 11 measures, both HACCP-type measures and good 12 manufacturing practices. It is worth noting here 13 that several of the previously higher-risk products 14 that were identified in the Draft Risk Assessment 15 actually fell quite a bit as a result of new data. 16 In part, we interpret these as successful control 17 programs. An interesting one to note here is the 18 soft unripened cheeses where our most recent data 19 from commercial sources indicated a relatively low 20 contamination rate as opposed to then our 21 discussions with CDC in terms of some continuing 22 outbreaks. 23 What is interesting there is that those 24 products that continue to be associated with 25 outbreaks were primarily products that were 59 1 noncommercial or not part of the controls that we 2 would put in for interstate commerce. 3 The low-risk designation included two fish 4 products or seafood products, preserved fish and 5 raw seafood. These had low to moderate 6 contamination rates. They had low consumption 7 rates. At least there is some indication in the 8 literature that there may be some types of natural 9 barriers to Listeria monocytogenes growth in these 10 products. 11 There is a substantial amount of 12 uncertainty associated with these products; for 13 example, is there really a natural barrier to 14 growth. However, overall, based on the risk 15 assessment, these products seem to represent a low 16 level of risk. 17 Then this brings us to the final category, 18 the very low-risk designation. These included five 19 dairy products. These dairy products had the 20 following characteristics, that they had both low 21 per-serving and low per-annum risk, that these 22 products typically combine a bactericidal treatment 23 plus a condition that prevents growth or 24 inactivates the organism. 25 So, for example, ice cream was typically a 60 1 product that was pasteurized and then it is held at 2 frozen temperatures that would prevent growth, or 3 hard cheeses typically made from pasteurized or 4 thermized milk that reduces the level of Listeria, 5 and then the conditions as a result of aging where 6 the organism actually slowly dies off during 7 storage. In the interpretation of this, it is not 8 likely that these are significant sources of 9 foodborne Listeria. 10 Finally, I would like to end with a slide 11 that you have already seen. This was a tremendous 12 effort by a lot of dedicated people in multiple 13 agencies. This was a ground-breaking study that 14 demonstrated that you could do multiple commodities 15 at the same time and be able to develop a 16 quantitative Risk Assessment. I would personally 17 like to express my thanks for just the tremendous 18 amount of work that was done by Dick Whiting and 19 his team in putting this together, working closely 20 with a variety of federal agencies and many of the 21 people in industry to make this significant 22 contribution. 23 And, with that, I think we have now 24 presented the Risk Assessment. 25 MS. DENNIS: Thank you both Dick Whiting 61 1 and Bob Buchanan. Unless my watch has stopped, I 2 think we are very much ahead of schedule. So what 3 I suggest is that we take a short break but be back 4 here at 10:30 instead of 11:00. To remind you, 5 there is a little cafe. It is directly to your 6 left as you are existing this building. 7 At the end of the break, I would ask our 8 panel members to come up to the front of the 9 auditorium and be seated at the table. 10 Thank you. 11 (Break.) 12 Questions for Risk Assessment Panel 13 DR. BUCHANAN: What I would like to do is 14 introduce you to the distinguished panel that now 15 sits in front of you. There are a number of the 16 team leads and team members that participated 17 within the Risk Assessment Team that basically 18 dedicated a substantial chunk of their lives for 19 the last four years, getting this work done and 20 doing it in a transparent manner, a manner we hope 21 that communicated. 22 On the far end is John Bowers. John is 23 one of our very good modelers and a statistician. 24 We have Mary who is our expert in consumption 25 studies and surveys and really did a yeoman's job 62 1 in putting together the consumption patterns that 2 we use. Clark is the lead modeler for this. 3 Janelle is our primary contact for the FSIS team. 4 Rich headed up our writing up the scientific 5 background for the hazard characterization and is 6 one of our expert immunologists. And then you all 7 got to meet Dick earlier. He was our Team Lead on 8 the entire document and did a yeoman's job, again, 9 in getting this altogether. 10 What you don't see because she just 11 disappeared on me is Sherri Dennis who was both 12 the--oh; there you are. If you want to know where 13 all the skeletons are hidden, you can ask Sherri. 14 She was the Risk Assessment Coordinator whose job 15 was to make sure that all these people showed up 16 and did a great job in actually getting this out 17 the door on time this second time. So we really 18 want to give her a round of thanks. 19 But these are the people, along with the 20 rest of the team, that actually put this 21 altogether. If you have any burning questions of 22 how this was done or why it was done or what that 23 one datapoint meant, et cetera, this is your time. 24 For all of those of you that haven't gone through 25 and read the full technical text, you can get--no; 63 1 actually I asked that question and I found a number 2 of people that have read the technical text. So 3 this is your change to ask whatever burning 4 questions you have. 5 I do ask that you come down and use the 6 mikes because we are recording it. 7 MS. FRYE: I always have a question. Good 8 morning. Cary Frye with International Dairy Foods 9 Association. I commend the group here for an 10 excellent job in the Risk Assessment. I just have 11 one question, or actually would like some maybe 12 clarification about fluid pasteurized milk. When 13 we went over the data today, in one of the tables, 14 we saw the case per annum and the epidemiological 15 data, I think, predicted at 90.8 cases per annum 16 for pasteurized fluid milk. Oh; no. I'm sorry. 17 Yes. That was predicted, not from the 18 epidemiological data. It was predicted. 19 But when, earlier, we had seen the 20 epidemiological data for pasteurized milk, the 21 outbreak data showed that there was one outbreak in 22 1983 and one outbreak in 1994. So that is once 23 every eleven years and now we are into 2004. So 24 how do we reconcile, or could you explain a little 25 bit more about your thoughts of the differences 64 1 between the prediction of cases per annum and the 2 actual outbreak or epidemiological data. 3 Thank you. 4 DR. BUCHANAN: At some point, I may ask 5 Rob Tauxe to come up and talk a little bit about 6 the strengths and weaknesses of case-control 7 studies and what their limitations are just to help 8 in this process. I am giving you warning, Rob. 9 Who would like to take that up? Ray? 10 DR. RAYBOURNE: Yes; I guess I could take 11 a stab at it even though it is actually not exactly 12 the area that I was working mostly in. But I would 13 say that one possible answer might be that the 14 outbreaks are a little bit hard to reconcile with 15 the total number of cases. There is a great number 16 of sporadic listeriosis cases where we don't know 17 what the cause of it was at all. So there is a 18 little bit of a disconnect between using the 19 outbreaks to assign what the risk to a particular 20 food might be because, along with the outbreaks, 21 you also have a kind of a constant running of 22 sporadic cases as well, and so they may account for 23 some cases attributable to foods that don't show up 24 in the outbreaks. 25 I guess what Bob was referring to with the 65 1 case-control studies is when we get better data 2 from that, we may get a better handle on the 3 sources of all of these outbreaks because, like, 4 for example, in the FoodNet data, for the most 5 part, the FoodNet cases we don't really know what 6 the source of the person's infection was, really. 7 DR. BUCHANAN: Cary, also to expand on 8 that, if you will permit me. 9 DR. WHITING: If you look at these figures 10 where, particularly the one table that Bob showed 11 quickly that it said here is the estimated number 12 of cases per annum, now that was a median of a 13 distribution. In other words, it was kind of like 14 somewhat the weather report, maybe, or something, 15 probability of rain tomorrow. 16 But we have a range. That value is sort 17 of where the peak of it is. The mean is actually 18 slightly different but there is a range. It could 19 be one year you might get twenty cases and then 20 next year you don't get any. It is not necessarily 21 going to be a consistent sort of thing. But this 22 was sort of our estimate of where it should go. 23 The other thing I can think of in milk, 24 the pattern for milk, how it occurs. You saw the 25 figures where one sample out of 5,400 was 66 1 contaminated. Then you had the one slide from Bob 2 where he was looking at the refrigerator storage 3 temperatures and showed, when you had the full 4 range of refrigerator storage temperatures with the 5 high temperatures, I think that was for the elderly 6 population, I think had 13 deaths per year, as I 7 recall. 8 But when the high refrigerators were 9 eliminated, you essentially fell to virtually zero. 10 So what I think is happening with something like 11 milk is you have a lot of individual cartons, a 12 very rare contaminated one. But then, if that rare 13 contaminated one gets into one of those 14 high-temperature refrigerators, now you have got a 15 really good growth situation. 16 That is where something is going to 17 happen. And then, if a susceptible person should 18 happen to consume that carton--so this is kind of 19 the scenario that I think has to happen in a food 20 like milk. You are only going to get isolated 21 sporadic cases in something like that. I would not 22 expect to see outbreaks in milk very often. As you 23 say, there are a couple of documented ones but this 24 is the pattern. 25 I think we see, with all these foods, you 67 1 need a contaminated food. You need the opportunity 2 or the ability of the food to support the growth 3 and then the opportunity for growth. Generally, 4 all of that has to happen. 5 DR. BUCHANAN: If I can attempt to 6 summarize what was said in terms of the factors 7 and, again, I refer you back to the fact that we 8 use two metrics, a risk per serving and a risk per 9 annum. We have pasteurized fluid milk which had 10 among the lowest rates of contamination at the 11 retail level that we experienced. I think the only 12 one that got close to it was ice cream. So it is a 13 product that is not contaminated very often. 14 So, in the risk per serving, despite the 15 fact that it supports the growth of Listeria 16 growth, despite the fact that it will grow fairly 17 quickly, particularly in temperature-abused 18 products, there just aren't many--based on the best 19 results we can have, there aren't many contaminated 20 servings out there in the marketplace. 21 However, this is counter-balanced when you 22 start looking at risk per annum by the fact that it 23 is overwhelmingly the largest ready-to-eat product, 24 that is consumed the most frequently by the 25 greatest percentage of the population. So you just 68 1 have a sheer number of samples out there that, 2 based on the way we did the risk assessment, which 3 was to start with the contamination data and move 4 forward to consumption, that it would identify that 5 as the risk would be elevated in the per-annum 6 calculation. 7 Certainly, the way that we interpret this, 8 because there is a recognized disconnect, as I 9 indicate in my presentation, between the 10 epidemiologic record--certainly there have been 11 outbreaks identified, but the epidemiologic record 12 has not, through case-control studies, identified 13 pasteurized fluid milk as a vehicle. 14 That is why I sort of gave Rob a warning. 15 Most standard case-control studies would not be 16 able to detect or identify milk as a potential 17 vehicle for sporadic cases just because it is 18 consumed by too many people. So you could never 19 match up the controls. If you have further 20 explanation, Rob, now is the time to jump in. 21 But, certainly, the way we are 22 interpreting this is this is an important area 23 where we assume it is associated with sporadic 24 cases and it needs to be investigated. 25 DR. TAUXE: Rob Tauxe. I will be glad to 69 1 just make a comment or two. It is an important 2 question. I think, first of all, the record of the 3 outbreaks does show that, in perhaps unusual 4 circumstances, but it has happened. The best 5 documented outbreak and the more recent one is 6 actually an outbreak that affected normal people 7 and caused the febrile gastroenteritis version of 8 Listeria with just a very small number of invasive 9 cases that were associated and, at an exceedingly 10 high dose, we believe, in the milk that was 11 consumed. That was milk that was pasteurized but 12 then there was a problem in a holding tank used 13 after pasteurization with a defective jacket and 14 poor refrigeration and probably inadequate 15 disinfection. 16 There were a series of problems after the 17 pasteurization that meant contamination could 18 easily be imagined. Then there was temperature 19 abuse of the product after that. Then a group of 20 people consumed the milk. Under those 21 circumstances, it is possible to have an outbreak. 22 So I am very glad there are not more 23 outbreaks in our experience but, given that a 24 series of things could go wrong like that makes us 25 wary of the possibility we could see one in the 70 1 future. 2 The sporadic cases are a real 3 epidemiologic challenge, especially with 4 listeriosis where the person you want to talk to 5 has a 20 percent that they are not alive so you are 6 going to have to talk to a proxy. Then you are 7 going to have to ask, what have you had to consume 8 in the last month before the illness began. And 9 then, if it is a food that is unusual that people 10 might remember, your information is better. But if 11 it is a food that is very common and that is 12 consumed regularly by most of the population most 13 of the time, maybe daily, that is going to be very 14 difficult to say that this one person was more 15 likely to consume, let's say, milk than a range of 16 otherwise similar people. 17 I think our epidemiologic methods are--it 18 would be like saying in China, is the illness 19 associated with rice, in the sense everyone eats 20 rice with every meal every day. We will never have 21 an association with rice because it is always 22 consumed with whatever else you are eating. And 23 yet rice may be a problem, leftover rice, say. 24 So our concern is raised by the history of 25 the outbreaks and by the fact that it is possible 71 1 to have contamination that has been seen. But it 2 is something that I think does require a good deal 3 more study and assessment. 4 MS. FRYE: I just wanted to say I 5 appreciate the comments. It is not that I disagree 6 with the risk ranking at all, and the risk 7 characterization of this food. We know, from 8 history, that it can be risky and that improvements 9 have been made since the Dairy Product Safety 10 Initiatives in the 1980s to improve and lower 11 contamination rates. I think we showed that with 12 our retail sampling. But I think the Risk 13 Assessment--this information is important and I was 14 just interested in learning more about your 15 thoughts on the differences. 16 The dairy industry is committed to this 17 action plan that is going to be laid out this 18 afternoon. I think there are certainly 19 opportunities with temperature and possibly in the 20 future to work more closely with consumers in 21 education about handling and storage of milk and 22 additionally with processors in distribution and 23 storage temperature. 24 So I don't disagree at all. That wasn't 25 the reason I raised the question, but I just wanted 72 1 to learn more insight. So thank you. 2 DR. BUCHANAN: Thank you. Caroline? I 3 assume you are next to the microphone. 4 MS. SMITH DEWAAL: Thank you. Caroline 5 Smith DeWaal, Center for Science and the Public 6 Interest. In an ideal world, you would have a way 7 to double-check the results of your risk 8 assessment. With a pathogen that causes an 9 estimated 2,500 cases a year many of which are 10 sporadic and for which the food attribution data is 11 very much lacking, what are you using to 12 double-check your finding? 13 DR. BUCHANAN: Panel? 14 DR. WHITING: I guess I would have to say 15 there really is not a good dataset. In a classic, 16 maybe, approach like this what we would have would 17 be a series of outbreaks of which we could get 18 consumption of the bacteria, number of people ill, 19 and, from those outbreaks, we would create a 20 dose-response curve and do this analysis. 21 Then you would use the epi data to 22 cross-check how everything went. But, in this 23 particular case, since we felt there were really 24 only two outbreaks. The Los Angeles cheese 25 outbreak in '85 and then the butter outbreak in 73 1 Finland of a few years ago are the only two 2 outbreaks that have been really investigated very 3 thoroughly where we know very much about how many 4 people got sick and about their health status and 5 things like that. 6 So those two points were not enough to 7 create any sort of a curve, although we were quite 8 pleased to see, though, that when we went through 9 with ours and then went back to the Los Angeles 10 cheese outbreak and said, all right, now, where do 11 they fall on the curve, they really came very close 12 to where they should have. 13 So I think that part was quite reassuring 14 to us. But, overall, I think you are right. I 15 think maybe the ultimate test will be, over the 16 next couple of years, if we begin to see the 17 contamination rates go down and we begin to see the 18 case rates go down. Does a linking there make 19 sense? I think that will be the real ultimate 20 test. 21 MS. SMITH DEWAAL: Thank you. 22 DR. BUCHANAN: Can I follow up a little 23 bit on that Caroline because I did mention it in my 24 presentation, that it is very difficult to 25 "validate" these types of models. But one of the 74 1 things that you asked, when it is all done and you 2 look at the results, is did the results make sense. 3 When you go back and look at that, and if you go 4 back and look in the risk-characterization section 5 of the technical document, where we went over each 6 of the foods and then said, what is the comparison 7 of the results that have been predicted starting 8 with contamination and moving forward versus the 9 epidemiologic record. 10 I have to say that there were no 11 surprises. The products that wound up being high 12 risk, or in the higher-risk categories, within the 13 Risk Assessment, match up, for the most part, very 14 nicely with the epidemiologic record. So I use 15 that, at least in my own mind, as saying that is a 16 pretty good validation that we are on the right 17 track. 18 If it came out that hard cheese was 19 calculated to be the most risky product, I would 20 say, we have got a problem here. But, on the other 21 hand, you saw the blue dots, the black dots and the 22 yellow dots. If you go to the epidemiologic 23 record, it is the blue dots, the black dots and the 24 yellow dots match up with the epidemiologic record. 25 So that, in itself, is about the best we have right 75 1 now for validating the model. It turned out to be 2 pretty good. 3 MS. SMITH DEWAAL: I have two other 4 questions, one simple and one hard. Preserved fish 5 versus smoked seafood. Why is there such a 6 difference in the risk linked to those two 7 products? 8 DR. WHITING: Anyone else want to start? 9 Okay. Maybe starting with preserved fish. That is 10 your salted, pickled type product. So that would 11 be a product that would not support the growth of 12 Listeria. The data did show a reasonable amount of 13 contamination in those products but, because they 14 did not support the growth, the numbers do not get 15 up high enough. 16 Again, thinking from a statistical 17 probability point of view, you have some 18 contamination but the numbers do not get to the 19 high levels often enough that you have a good 20 probability of illness. Smoked seafood; again, 21 there is quite a bit of contamination there, but 22 those are foods that do support the growth and so 23 you can start reasonable high and get higher. 24 That would be thinking on the 25 case-per-serving mentality. Then when you, I 76 1 think, also think in total and, of course, I think 2 these preserved fish products would be a very minor 3 product in the whole diet where smoked fish, while 4 it is certainly not a major one, is much more 5 commonly eaten than the others. So that would then 6 make the total number of cases come out higher in 7 smoked fish. 8 But I think one thing that did come out 9 with some of these is that the ability of a food to 10 support growth really turned out to be a key factor 11 in where these foods would fall. 12 MS. SMITH DEWAAL: I will just observe, 13 being someone who watches seafood carefully, we 14 don't see the epidemiologic data yet with smoked 15 seafood. We know the hazard is out there and it is 16 certainly one consumers and especially elderly 17 consumers and the high-risk consumers need to be 18 made aware of, but I am not seeing the outbreaks 19 from smoked seafood yet. They may be happening as 20 sporadic cases. 21 DR. WHITING: Yes. Of course, I have said 22 to people if we had really a complete 23 epidemiological record, we wouldn't have to do the 24 risk assessment in the first place. I think what 25 we have of the record shows that Listeria is 77 1 overwhelmingly a sporadic-case situation. I mean, 2 it has been two years now since we had the chicken 3 deli meats up in Pennsylvania. So it has been two 4 years. 5 MS. SMITH DEWAAL: A year and a half. 6 DR. WHITING: We have had 5,000 cases 7 since then and we haven't really identified very 8 many foods. So that was the reason for it. But I 9 also think you can look at like seafood in 10 relatively small production lots. Even if a food 11 is contaminated, you looked at Bob's--the 12 dose-response curve. Even if you consume a million 13 Listeria, your risk is still 1 in 100,000, 1 in a 14 million. So, by the time you begin to put these 15 probabilities together, you could have a lot of 16 contaminated smoked seafood and a fair number of 17 susceptible people eat it. But there is still most 18 likely only one or two people would become ill. So 19 it just doesn't show up. 20 MS. SMITH DEWAAL: Now my hard question. 21 DR. WHITING: I thought that was the hard 22 one. 23 MS. SMITH DEWAAL: This actually may be 24 seen by you as being the easier question. I am 25 very interested in your risk per serving versus 78 1 your risk per annum and then your, what did you 2 call it, the graph at the end with the very high 3 risk, the high risk, and going down to low risk. 4 That division of the data in your findings I think 5 is very helpful to risk managers. 6 Did you see that there may be different 7 risk-management strategies for foods depending on 8 where they fall either on the per-annum or on the 9 per-serving standard. Bob, you may want to answer 10 this, but it is a very interesting way to look at 11 the data because, in my mind, the risk-management 12 strategies may be different even for the high-risk 13 foods that are high risk on a per-annum basis 14 versus high risk on a per-serving basis. It may 15 call for different risk-management strategies. 16 DR. BUCHANAN: I think you are going to 17 have to go back and look at each of these food 18 categories and determine what the risk-management 19 strategy is on a food-category basis. But, to 20 answer your question most directly and, in fact, we 21 did that in the presentation. If you will remember 22 the high-risk designation that we actually took and 23 divided it up into two categories, two 24 subcategories, based on the characteristics of the 25 food. 79 1 The one subcategory was products that had 2 a high per-annum risk and a somewhat lower risk per 3 serving. So there the strategies were different. 4 In fact, that is one where we said that we really 5 need to go back and match up the epidemiologic 6 record, as opposed to the second subcategory which 7 had products that had a high per-serving risk or a 8 higher per-serving risk but were products that 9 weren't consumed by very many people so they had a 10 lower per-annum risk. 11 The strategy there, the risk-management, 12 strategy, at least to me seems substantially 13 different for that type of category. So we 14 actually, interpreting that, have made that 15 distinction to our risk managers in terms of you 16 are going to have to deal with these as two 17 different subgroups in interpreting the risk 18 assessment. 19 Again, I would encourage anyone that moves 20 into the risk-management phase to go back in the 21 technical document and actually read the more 22 detailed discussion and characterization of each of 23 the food groups because that gets into a lot more 24 than we attempted to do with the interpretive 25 summary. 80 1 MS. MARCOTTE: Hi. I am Michelle Marcotte 2 from Marcotte Consulting and I am here on behalf of 3 IBA Guardion which is a company that operates 4 several radiation facilities in the United States 5 and the Association of International Industrial 6 Radiation. So I am one of these people that hasn't 7 read your technical document. 8 I am just after trying to assist a number 9 of the companies in the radiation-processing 10 industry to really sort of target where it is that 11 this technology might be of better assistance. 12 If we look at some of your food 13 categories, looking at smoked seafood, are you 14 primarily including in that category cold smoked 15 salmon or is hot smoked salmon, which was 16 essentially a cooked product, in there as well, or 17 how did you divide that out? 18 DR. WHITING: Do you want to answer that 19 question? 20 MS. MARCOTTE: I have got a couple of sort 21 of little questions. 22 DR. WHITING: Let's go one at a time here. 23 As far as hot and cold smoked salmon, they were 24 combined into one group. I think Mary was saying 25 that the ready-to-eat food categories and the 81 1 nutritional databases have about 640 entries into 2 it. So we obviously had to try to combine foods 3 that did make some sense with still keeping a 4 reasonable sized Risk Assessment. 5 The contamination data also suggest that 6 hot and cold smoked salmon were contaminated at 7 approximately the same level. Now, true, the hot 8 smoking should kill the Listeria during the 9 processing, but then they get recontaminated 10 afterwards so, in fact, there is not that much 11 difference. 12 So kind of on that basis, we felt we will 13 just keep the two together and go ahead. Also, I 14 think in the nutritional databases, oftentimes, 15 they would just say the person consumed smoked fish 16 or sometimes, even in the scientific literature, it 17 is surprising when you look up Listeria 18 contamination in the literature and somebody will 19 say they analyzed some cheese and found 3 out of 20 100 positive or something like that. 21 Likewise, often they will just say they 22 analyzed smoked fish and this is their result. So 23 the problem is getting enough specific data to 24 start splitting these into finer and finer 25 categories. 82 1 MS. MARCOTTE: Okay. You have got 2 frankfurters not reheated. Is this because the 3 consumer has made a choice not to reheat the hot 4 dogs? I thought that all, or most, hot dogs would 5 be labeled for slight or reheating or further 6 cooking. 7 MS. KAUSE: This is Janelle Kause. I will 8 speak to that issue. Actually, the American Meat 9 Institute went out and did a survey because we had 10 some issues. There are people who consume hot dogs 11 straight out of the package. 12 MS. MARCOTTE: So it is a consumer choice. 13 It is not a labeling problem? 14 MS. KAUSE: Right. The model is set up to 15 model what actually happens in terms of the 16 consumption pattern. So the person is actually 17 consuming the hot dog without boiling it, 18 microwaving it or something. So the survey found 19 about 7 percent. 20 MR. MARCOTTE: Further on the hot dogs, 21 when you looked at the storage time, which looked 22 really kind of long for me in some cases, were you 23 looking at just refrigerator storage or was that 24 including freezer storage of the hot dogs, or is 25 freezer storage very common with hot dogs in the 83 1 United States? 2 MS. KAUSE: I would have to go back and 3 look at the AMI study but I believe actually Dick 4 looks like he is ready to speak to this issue. So 5 I will let him go ahead. 6 DR. WHITING: On frankfurters, there was 7 some data, and I forget the percentages--I think 13 8 percent of people claim that they just take their 9 frankfurters and immediately put them in the 10 freezer. So those were set aside. Frankfurters 11 turned out to be a fairly complicated one for us to 12 deal with because of this. We separated those 13 frozen and then we would assume, if you kept them 14 frozen, you probably had to cook them in order to 15 thaw them and eat them. 16 So we assumed, then, any of those 13 17 percent that were frozen were also properly 18 reheated. So that is how we dealt with that. 19 MS. MARCOTTE: Yes; there is a lot of 20 interest in hot dogs and I am just trying to sort 21 out a lot of details on that. When we are looking 22 at pate and meat spreads, are those the backpack 23 products that you commonly see in retail or is that 24 category quite broad? Are you looking like fresh 25 refrigerator pates and meat spreads that are in 84 1 deli counters. It seems to me there is sort of a 2 crossover between like sort of deli-salad meat 3 spreads and that type of product. 4 DR. WHITING: Mary? 5 DR. BRANDT: I am checking here. For the 6 food-consumption data, I believe that there was 7 some of what you mentioned. Okay; liver paste or 8 pate. But then there were also meat spreads, 9 potted meat, chicken-salad spread, ham, deviled or 10 potted. So it is a combination from the 11 food-consumption standpoint. But I believe that, 12 with the contamination data, it is primarily pates 13 as opposed to the meat spreads. But there were not 14 enough eating occasions just to go with the pates, 15 so we had to try to get a handle on how much people 16 would eat of this type of food. 17 DR. WHITING: Although, if I can add one 18 more thing, if I would kind of interpret that 19 category, so if you had a canned pate, that would 20 obviously be a different situation. That kind of a 21 pate would not be the hazard that the fresh 22 refrigerated pate would be, at least until the can 23 is opened. 24 DR. BRANDT: And the food-consumption data 25 don't differentiate. It is just strictly pate or 85 1 salad or spread. 2 MS. MARCOTTE: Okay. And cooked, 3 ready-to-eat crustaceans; I assume that includes 4 cooked shrimp. Does this include the whole gamut 5 of cooked shrimp that enters the United States from 6 imported or are you only looking at products that 7 are cooked in the United States? 8 DR. BRANDT: From the food-consumption 9 standpoint, there were only three food types; 10 steamed crab, steamed or boiled shrimp, and then 11 shrimp cocktail. So, from the contamination, I am 12 not sure. 13 DR. WHITING: We basically put them all 14 together. 15 MS. MARCOTTE: So the imported shrimp is 16 in here, too, as well? 17 DR. WHITING: Which one? 18 MS. MARCOTTE: Imported, cooked? 19 DR. WHITING: Yes. 20 MS. MARCOTTE: All right. Thank you very 21 much. 22 DR. BUCHANAN: Just to follow up your 23 answer there, remember that the contamination data 24 that we used was at retail. So anything that had 25 happened before retail was just accumulated into 86 1 the dataset that we had. 2 DR. WHITING: So we would have had to then 3 assume that the different products that were picked 4 up at retail would be in proportion to all of these 5 various sources. So we are assuming, really, 6 within these categories that a lot of the 7 processing pathways and such have all converged and 8 we have just sampled in a purely random manner. 9 MR. GARFIELD: Bob Garfield, American 10 Frozen Food Institute. Refrigeration was an 11 important consideration in your "what if" 12 scenarios. Did you, or would you, consider 13 freezing of foods between servings also as a 14 potential "what if" scenario to reduce the cases of 15 listeriosis? 16 DR. WHITING: I would consider a food that 17 is frozen would now be a no-growth food. The 18 analysis and scenarios we have run show that 19 Listeria does not have a chance to grow. You are 20 just not that likely to get the high numbers; in 21 other words, starting maybe from the beginning. 22 The contamination levels, while they can 23 occasionally be high, as you saw, but, but, by and 24 large, they are not high enough to cause very many 25 cases of listeriosis. So they need that 87 1 opportunity to get that final growth to get to 2 those high numbers. 3 So if you had a food which, as Bob's 4 example with the frankfurters that were 5 reformulated to not support growth, you saw the 6 number of cases went down from 30.5 to less than 1 7 and the risk per serving went down about a 8 hundred-fold. 9 Now, I would consider, if you took a food 10 that was normally kept fresh and, instead, kept it 11 frozen during that equivalent time, you have now 12 made it a non-growth food during that time and 13 growth would be prevented and the numbers would be 14 lower. So, yes; that should have a quite dramatic 15 effect. 16 DR. BUCHANAN: Just to follow up a little 17 bit in terms of what was actually done in the Risk 18 Assessment, in those cases where we know that there 19 was a substantial portion of the product that, for 20 any duration, was going to be frozen from retail 21 on, that was considered a no-growth condition and 22 the growth models that were used underlining the 23 determination between what you had the consumer 24 actually eating were frozen likewise. We assumed 25 that it did not grow. The calculations reflected 88 1 that. 2 As Dick indicated, for that portion of the 3 frankfurters that were frozen, they were just--we 4 stopped the clock on those until they were then 5 thawed again. 6 DR. WHITING: And you could extend that 7 idea to any other food processing that would 8 prevent growth--I mean, perhaps, modified 9 atmosphere packaging at least until the extent 10 that--until the package is opened. While maybe not 11 completely stopping growth, it would certainly slow 12 growth. Again, we didn't particularly try to model 13 situations where growth was slowed. I think you 14 can look at it and make some estimates of what the 15 effect might be, though. Certainly, anything that 16 would slow growth would certainly have a beneficial 17 effect. 18 MS. SCOTT: Jenny Scott, NFPA. Just a 19 quick follow on with respect to the modeling. How 20 easy is it for us to take this model that you have 21 and manipulate it so that we could take a food and 22 have it stored for a certain period of time frozen 23 and then slacked that at retail and refrigerate it 24 and see the effect of that on a particular product? 25 I haven't tried that yet. 89 1 DR. WHITING: I think I would say right 2 now the model is probably rather difficult for 3 anybody to use. But we have been working with a 4 group which is represented here to try to make a 5 user-friendly version of this. You wouldn't 6 replicate our whole Risk Assessment. You could put 7 a single food in and, perhaps, change the 8 contamination or change the storage times or do 9 various things and get probably an estimate of the 10 risk per serving. I don't think we would go on to 11 the total number of cases. We would just do it on 12 a per-serving basis. 13 So we are in the process of trying to get 14 a software package developed that will--and we will 15 certainly distribute it when it is ready. 16 MS. SCOTT: Have you run any scenarios of 17 that sort? For example, I know that a lot of 18 smoked seafood is actually frozen and delivered to 19 retail where it is slacked out and sold, 20 refrigerated there. That is a significant portion 21 of the shelf life. Have you run anything like that 22 yet? 23 DR. WHITING: That would not have directly 24 come into the Risk Assessment because that 25 freezing, or not being frozen, happened before 90 1 retail. So if the smoked seafood might sort of 2 separate out into two groups; those that were 3 frozen with lower numbers and those that weren't 4 frozen at higher numbers. That would all be 5 encompassed within the retail contamination. So it 6 didn't really show up in the Risk Assessment but, 7 again, I think from just thinking through what the 8 Risk Assessment is saying in generalities, 9 certainly the smoked seafood that got to retail in 10 the frozen state would be much more likely to have 11 lower numbers and less likely to lead to illness. 12 DR. BUCHANAN: Any additional questions? 13 This is your chance to take apart the Risk 14 Assessment and ask all those questions when you 15 say, what did they do, and why? 16 Seeing none, I am going to close this 17 section and pass this back to Sherri to talk about 18 housekeeping items and modifying our schedule. 19 Sherri? 20 MS. DENNIS: Thanks, Bob. We are really 21 ahead of schedule. I don't think this has ever 22 happened before in a Risk Assessment meeting. We 23 will move up the schedule a little bit. We would 24 ask for you to be back from lunch at 1 o'clock 25 instead of 1:30. 91 1 There is a list of restaurants, local 2 restaurants, in your packet and there is also the 3 cafeteria that is directly to the left outside of 4 the building. So I will see you back here at 1 5 o'clock. 6 [Whereupon, at 11:15 a.m., the proceedings 7 were recessed to be resumed at 1 o'clock p.m.] 92 1 A F T E R N O O N P R O C E E D I N G S 2 (1:00 p.m.) 3 MR. KAUTTER: Good afternoon. Welcome 4 back. I hope your lunch was well. My name is Don 5 Kautter. I am with the Food and Drug 6 Administration. I will moderating this afternoon's 7 session. Reminders; please turn your cell phones 8 onto vibrate or off. 9 This afternoon, I just want to give you a 10 quick walk-through of what we are going to do. We 11 are going to shift from the assessment to the Risk 12 Management Action Plan. We are going to have both 13 the FDA and CDC perspective as well as the SSI 14 perspective. We will take a quick fifteen-minute 15 break in the afternoon. We will come back for 16 questions for the Risk Management panel and then we 17 will have some time for public comment. We have 18 four individuals who have signed up so far for 19 public comment. 20 With that, let me go ahead and get 21 started. Our first presentation is from Dr. John 22 Kvenberg here at the Center for Food Safety and 23 Applied Nutrition, Food and Drug Administration, 24 who will be discussing the FDA-CDC Risk Management 25 Action Plan. 93 1 John? 2 FDA/CDC Risk Management Action Plan 3 DR. KVENBERG: Good afternoon, everybody. 4 I am very happy to be here and talk to you about 5 FDA and CDC's piece of the Action Plan for Listeria 6 monocytogenes. 7 I would like to start off with the 8 strategic goals that we have envisioned for the 9 Action Plan for Listeria monocytogenes which are 10 matched up and consistent with FDA's overall 11 strategic goals. First off is efficient risk 12 management, the use of science-based efficient 13 risk-management strategies to target the highest 14 Listeria monocytogenes risk products and practices. 15 The second important point of our 16 strategic goal on this Action Plan is empowerment 17 of consumers through improving health through 18 better information; specifically, consumer messages 19 and information to reduce the exposure to Listeria 20 monocytogenes, particularly by the vulnerable 21 groups such as elderly and pregnant women and to 22 improve consumer safety through reducing the 23 exposure of Listeria monocytogenes to the 24 population with improved food manufacturing and 25 distribution of foods. 94 1 To take you back into history, back in 2 2001 when the Draft Risk Assessment and the initial 3 Action Plan was published in January, two years 4 ago, now, almost, for public comment, the scope was 5 and remains that ready-to-eat foods identified in 6 the risk assessments warrant control and warrant 7 additional information of data at that time. 8 There were eight action areas, short and 9 long-term implementation time lines. 10 To achieve the overall goal that we set 11 for ourself--just to recall, it is now, as it was 12 then, to accomplish a 50 percent reduction in 13 Listeria monocytogenes illnesses by the target year 14 2005. This was an acceleration of the initial 15 projection put forward by the Healthy People 2010 16 objective for this issue and, specifically, the 17 target was the reduction of 50 percent between 1996 18 and 2005 from 0.5 to 0.25 cases of the population 19 of 100,000 people. 20 I think you heard earlier today that the 21 projections were down to 0.3 cases in 2001. 22 Although this is preliminary data and you can't 23 trust too much in trends over time because things 24 will go up and they will go down, I would like to 25 point out that preliminary data from the 2002 95 1 FoodNet data indicates 0.27 cases per year which is 2 really approaching the goal that we set for ourself 3 in 2005, and we are moving in that direction. 4 The Action Plan, as updated, has six goals 5 and I would just briefly--I will go through these 6 in detail but just to touch the highlights on them. 7 The first goal is really a consolidation of what 8 was two goals. It is appropriate that we did this 9 because retail and food-service establishments are 10 linked with processors and manufacturers relative 11 to development and guidance on the Listeria 12 monocytogenes control programs. 13 The second goal, likewise, is to develop 14 and deliver training, both to the industry and to 15 regulators. These are linked goals and the 16 messages should be consistent so those were the two 17 linkages and explains why we have six goals when we 18 had eight previously. 19 Enhancing consumer and healthcare provider 20 information in educational efforts, redirecting and 21 revising our enforcement and regulatory strategies 22 remains the same. We continue and have upgraded, 23 as you heard this morning from Dr. Rob Tauxe, our 24 disease surveillance and outbreak response 25 information. We continue to work with coordinating 96 1 research to refine the risk assessment, its 2 preventive controls in support of the regulatory 3 enforcement and educational activities that are 4 discussed elsewhere in the strategy. 5 In 2003, on the first objective, to 6 develop and revise guidance for processors, 7 manufacturers and retail establishments, the first 8 item I would like to point out on this lists is 9 controlling Listeria monocytogenes, a finalized and 10 published guidance document. That is not out 11 currently, but it is our goal to get that out in 12 the near future. It is in its final stages of 13 preparation and will be moved forward, hopefully, 14 in the next several months because we are, I 15 believe, relatively close to publishing our 16 guidance document. 17 We continue to work with the framework on 18 the international front through Codex Alimentarius 19 Commission and the Food Hygiene Commission. I had 20 the privilege of working with Dr. Bob Buchanan this 21 fall and others from the United States on a draft 22 guidelines for control of Listeria monocytogenes 23 with foods that is moving along the same track and 24 same lines that our national effort of guidances is 25 going. So I think I am encouraged that, again, if 97 1 there are going to be consistencies in measuring 2 how you are going to produce safe food, this kind 3 of is a good thing from the standpoint that advice 4 on controlling for the industry is going to be a 5 consistent message on both foods produced 6 domestically and those produced in imports. 7 We also will continue to work with the 8 University of Michigan in the retail arena to study 9 transfer rates between foods contaminated in 10 Listeria monocytogenes and contact surfaces where 11 you can come up with better systems to identify 12 areas that cross-contamination can occur and 13 present itself with a risk. 14 What is new this year, in addition to 15 information that I had discussed a moment ago in 16 the general overview, is additional efforts to 17 draft guidance in specific areas that were 18 identified in this morning's discussion relative 19 the risk-assessment process. This includes 20 guidance on enhancing the safety of production in 21 milk and milk products, guidance on enhancing the 22 safety of production of fresh-cut produce, the 23 design of the project for smoked-seafood operations 24 to pilot a Listeria monocytogenes control-measures 25 effort and to go through a review of the Food Code 98 1 for retain and food-service operations, provisions 2 that exist, that may address Listeria monocytogenes 3 control measures with an emphasis toward the 4 at-risk populations. 5 The second action item is the delivery of 6 training and technical assistance to both industry 7 and to the food-regulatory employees, both at the 8 federal and state level. I think the important 9 point to mention is not only classroom training but 10 long-distance technical teaching instruments that 11 have evolved and are getting much better to include 12 website trainings and downlink interactive type 13 programs that can really disseminate educational 14 materials better. 15 With the publication of the final Listeria 16 Risk Assessment document and this Action Plan that 17 we are moving ahead, review the existing training 18 programs and revise them as necessary on the 19 findings of the Risk Assessment, itself, so that 20 the materials that are currently in the system can 21 be updated and be more focused. 22 What's new here in specific areas under 23 Training would include providing specific technical 24 assistance to small and very small dairy 25 facilities--that would include facilities that are 99 1 producing dairy products, specifically those of the 2 higher-risk category as well as working with those 3 that are producing fluid milk, per se--and 4 development of training segments on auditing the 5 effectiveness of Listeria monocytogenes controls 6 that are currently in practice. 7 The third item under our Action Plan this 8 year is the enhancement of healthcare-provider and 9 consumer information and, in general, educational 10 efforts. The first area is to focus consumer 11 messages on Listeria monocytogenes to what actions 12 the consumer can take that would be most effective 13 and to continue to inform and educate medical 14 guidance for healthcare professionals about the 15 diagnosis, treatment and prevention of Listeria 16 monocytogenes illnesses; also to achieve the 17 maximum outreach to women and women's healthcare 18 professionals with the information and collaborate 19 with other public and private organizations on 20 healthcare-provider and consumer information. 21 What's new in the area of this goal is 22 emphasis on the storage of foods at 40 degrees or 23 lower for short periods of time. I think there is 24 one dramatic thing you heard this morning relative 25 to the risk-assessment process is that keeping 100 1 temperatures cold, at 40 degrees, is one of the 2 most effective things that we can accomplish 3 relative to providing information and empowering 4 people to reduce the risk of Listeria monocytogenes 5 from refrigerated ready-to-eat foods. 6 We also plan to work with the consumer on 7 refrigeration monitoring of these cold temperatures 8 with a dialogue that we have not yet established 9 with the refrigerator manufacturing operators to 10 look at the current designs of what is available 11 and what consumers can do to assure the adequate 12 refrigeration of their foods and, specifically, to 13 enhance targeted education campaigns, as you heard 14 from Dr. Rob Tauxe this morning, on specific 15 segments of the population to include Hispanic 16 women of childbearing age about the importance of 17 eating fresh soft cheeses that we can be assured 18 that were made under good conditions using 19 pasteurized milk to reduce their risks. 20 The fourth action area is the review and 21 redirection of our enforcement and regulatory 22 strategies to include our microbial product-testing 23 sampling efforts. Under domestic inspections, it 24 is our intention to increase the inspections of 25 firms that produce the higher risks of ready-to-eat 101 1 foods and shift resources away from those very low 2 to low-risk foods so that we can focus better our 3 resources to the moderate and high-risk foods that 4 were identified in the Risk Assessment, and further 5 to focus both domestic and import surveillance 6 sampling programs on the higher-risk foods for 7 Listeria monocytogenes testing as well. 8 What's new under the enforcement and 9 regulatory action item is to develop an action plan 10 to address the unlawful importation of raw-milk 11 soft cheeses into the United States because that 12 is, I think, one of the identified factors that we 13 have had experience with; to work with state 14 government to eliminate the production and sale of 15 raw-milk soft cheeses. I think you heard again 16 that this is a potential area where outbreaks can 17 occur and for states that need assistance in 18 passing appropriate legislation to reduce or 19 eliminate this product, we will do what we can to 20 work with state governments for the intrastate 21 regulation of these type products. 22 Also, to develop a model statute that will 23 prohibit the sale of raw milk for dissemination to 24 those states which still permit its sale. If they 25 ask for assistance on that, we will be actively 102 1 working with them to be able to accomplish that 2 goal as well; and to review our currently Good 3 Manufacturing Practice regulations and evaluate 4 whether to revise these regulations based on the 5 Risk Assessment that was conducted for Listeria 6 monocytogenes, and that GMP internal review of our 7 GMP regulations is ongoing at this time. 8 What's new relative to our enforcement 9 provisions is we are currently commissioning the 10 National Advisory Committee for Microbiological 11 Criteria on Foods on the scientific basis for 12 redefining the term "pasteurization." I think some 13 of you may be aware that legislation was passed in 14 the Farm Bill which became public law that allows 15 for the labeling of pasteurization other than by 16 conventional time-temperature procedures, 17 techniques such as the utilization of high 18 pressure, various other types of electrical-pulse 19 irradiation sources, and such, basically are now in 20 for consideration and we are trying to look at the 21 scientific basis for how you could address the 22 adequate safety of pasteurized products and how 23 they could be applied to ready-to-eat foods through 24 the National Advisory Committee. 25 We are considering revising dairy 103 1 standards to allow the use of bactericidants which 2 would aid in the prevention of the outgrowth of 3 Listeria in products and to seek also advise from 4 the National Advisory Committee on the scientific 5 basis for establishing safety-based use-by dates 6 for the consumer so that the shelf-life date 7 labeling of refrigerated ready-to-eat foods is 8 based on a use-by date that will afford adequate 9 safety. 10 The fifth item is basically the CDC's 11 portion as the lead on enhancing disease 12 surveillance and outbreak response. The FDA does 13 move forward with outbreak response in 14 collaboration with CDC when food is involved. I 15 think it is very useful and quite exciting that we 16 are accomplishing PulseNet, looking at a nationwide 17 system to be able to better identify specific 18 strains so that we can make a quicker connection to 19 foods that may be causing illness in the United 20 States by our surveillance type systems. 21 CDC will continue to work with state and 22 local health departments to improve the detection 23 and reporting of Listeria monocytogenes illnesses 24 and enhance the monitoring of illness to identify 25 and evaluate trends in the disease occurrence so we 104 1 can better define our risk-management strategy, and 2 continue to work on two projects to attribute the 3 percentage of foodborne-illness cases from Listeria 4 monocytogenes to various vehicles based on the 5 epidemiological data. So we are moving forth in 6 that regard. 7 Specifically, what's new; CDC will 8 complete and publish a comprehensive case-control 9 study that has currently been under development and 10 CDC will work with FDA to develop improved methods 11 for the use of surveillance data in assessing the 12 public-health impact of regulatory and outreach 13 programs. 14 The sixth and final Action Plan item is 15 the coordination of research activities that 16 stretches across the entire Action Plan to redefine 17 the risk assessment and to enhance preventive 18 control measures that we have in place. Our 19 Research Program will continue to coordinate and 20 support and conduct research areas to evaluate the 21 effectiveness of commercial treatments, 22 preservatives, bactericidants, irradiations, other 23 types of thermal destructions and pasteurization or 24 treatments that may eliminate or reduce Listeria 25 monocytogenes. 105 1 Continue to seek data on the frequency and 2 concentration of Listeria monocytogenes in a 3 variety of ready-to-eat foods to enhance our 4 already good database on exposure assessments and 5 continue, within CFSAN, our center, for detection 6 and quantification-methods research for the 7 organism. 8 What's new specifically in the area of 9 research is additional Listeria monocytogenes 10 product-pathway simulation studies to evaluate 11 variables in manufacturer and distribution systems 12 that can be used to provide guidance on specific 13 control strategies. FDA and CDC will seek means 14 for readily and objectively assessing the immune 15 status of susceptible populations and initiate 16 research to determine the growth rate of Listeria 17 monocytogenes in specific foods to provide 18 information and guidance on food-safety-related 19 date marking and practical means for consumers to 20 control Listeria. 21 So this, in conclusion, is the Action Plan 22 in its format, six points to it, as designed by FDA 23 and CDC to meet the goal pronounced by meeting our 24 target of 50 percent in 2005. 25 Thank you. 106 1 MR. KAUTTER: Thank you, John. 2 Our next presentation is by Dr. Daniel 3 Engeljohn from the United States Department of 4 Agriculture, Food Safety and Inspection Service. 5 FSIS Risk Management Plan 6 DR. ENGELJOHN: Good afternoon. I am very 7 happy to be here and identify the Risk Management 8 Plan that FSIS has with regard to Listeria 9 monocytogenes. The work that I am presenting today 10 is, in fact, reflective of the joint efforts that 11 FSIS has had with FDA and CDC on identifying 12 strategies for how to take care of this particular 13 pathogen which we all have great concern about. 14 FSIS, in terms of following up on the 2001 15 Joint Action Plan has identified a number of 16 accomplishments from that point and then what I 17 will concentrate on today is the items for which we 18 intend to take specific action on this coming year 19 with regards to some of the activities that have 20 occurred, in fact, earlier this year. 21 In February of 2001, after the release of 22 the Joint Action Plan, FSIS issued a proposed rule 23 that had specific regulatory implications with 24 regards to the control of Listeria monocytogenes in 25 post-lethality-exposed products. It was that 107 1 proposed rule that identified the sanitary control 2 and the documentation of its effectiveness that 3 were believed to be one way to address Listeria 4 monocytogenes in an effective manner. 5 We also identified that we didn't have 6 good information about the practicality or the 7 effectiveness of verification testing of 8 food-contact surfaces. 9 As a consequence, the agency held three 10 public meetings over the course of the next year 11 and a half, specifically designed to identify and 12 clarify what research was available on the 13 effective control of Listeria, particularly in the 14 environment, and to gather new information as we 15 could about the research that was, in fact, being 16 done to address how effective controls with regard 17 to environmental measures were with regard to this 18 pathogen. 19 At the same time, the agency issued a new 20 directive last December that used the findings of 21 the preliminary risk ranking and risk assessment 22 that FDA and FSIS worked on to identify high-risk 23 products. At that time, we were concentrating our 24 efforts on looking at deli-meat products and hot 25 dogs. So we identified new inspection tasks and, 108 1 for the first time, specifically identified that 2 the agency would, in fact, begin doing 3 environmental testing as a verification measure on 4 a routine basis in the establishments that we 5 regulated. 6 At the same time, the agency put forward 7 what it thought was an incentive to industry to 8 begin sharing data in a more effective manner. The 9 agency had always held that we believe that data 10 was, in fact, appropriate to be shared with the 11 agency but, in fact, wasn't being done in a routine 12 manner. So the efforts that were put in place last 13 December I think had a significant impact on the 14 relationship between the industry and the agency in 15 terms of being able to look at how the industry was 16 controlling this pathogen. 17 Consequently, then, in this year, 2003, 18 FSIS then followed up with an implant risk 19 assessment specifically designed to address the 20 mitigations that we believe may have a dramatic 21 impact on the control measures for Listeria and, as 22 identified this morning by the quantitative and 23 qualitative risk ranking Risk Assessment that was 24 presented by FDA, we specifically looked at growth 25 inhibition and lethality treatments as ways to have 109 1 effective control measures. 2 We issues and Interim Final Rule in June 3 of this year. We identified it as an interim final 4 rule for the specific purpose that the agency would 5 assess the effectiveness of the regulatory 6 requirements that we put in place. I am going to 7 identify a number of the items that we view will be 8 part of that effectiveness check in the slides that 9 I am going to present this afternoon. 10 As we issued this regulation which went 11 into effect in October, the agency held five 12 workshops across the country that specifically were 13 designed to answer questions for small and very 14 small businesses as to how they can meet the intent 15 of the regulation. 16 We issued new compliance guidelines that 17 specifically addressed how industry could, in fact, 18 take the available research and use it as 19 validating materials for their control measures 20 without having to do additional validation within 21 their operations. And we identified new inspection 22 procedures for our inspectors, again specifically 23 directing our inspectors to be looking in a 24 risk-based fashion at higher-risk products versus 25 low-risk products. 110 1 In terms of looking at the effectiveness 2 of the Interim Final Rule, we will be doing that 3 throughout the course of this year. The Interim 4 Final Rule period for implementation was through 5 December of 2004. Some of the things that the 6 agency is now beginning to develop procedures for 7 actually measuring the effectiveness of the 8 procedures we have put in place include the changes 9 in industry practices. The rule was designed to 10 encourage industry to go from strictly sanitary 11 control measures to incorporation of post-lethality 12 treatments as well as antimicrobial-growth 13 inhibitors. 14 So the agency will be looking to see how 15 industry changes its practices with regards to 16 those control alternatives. We also were 17 specifically interested in whether or not the rigor 18 associated with the post-lethality treatments, the 19 amount of growth suppression and the effectiveness 20 of the sanitary controls are, in fact, changed as a 21 consequence of this rule-making. 22 The agency is more clearly articulating 23 what it means by risk-based inspection and 24 verification in which our intention, much like what 25 FDA and Dr. Kvenberg just identified in the Action 111 1 Plan that they are working on, will be to focus on 2 high-risk products and operations, not ignoring the 3 lower-risk products but, in fact, refocusing our 4 inspection activities so that we are concentrating 5 on those products that present the greatest risk 6 for Listeria. 7 In addition, we have set up a mechanism to 8 assess the training of our inspection personnel to 9 see what training needs that they have to 10 understand this new approach that we have with 11 regard to risk-based verification and the 12 information that they have about how they do their 13 task on a daily basis. 14 We are also looking to identify 15 specifically within the various product groups that 16 we regulate the changes in prevalence in terms of 17 the presence of Listeria monocytogenes and the 18 level of Listeria monocytogenes in those products, 19 particularly in the high-risk groups versus those 20 that have lower risk. So we will be changing, in 21 part, our microbiological-sampling program to 22 assess that type of prevalence change. 23 Also the agency is specifically interested 24 in the issue of labeling claims as one way to 25 provide consumers with information about how to 112 1 handle their products and how to select for 2 products that may, in fact, present lower risk 3 because they are formulated in particular ways. 4 The Compliance Guidelines that the agency 5 has presented have been updated with input that we 6 received from the five workshops. We have a number 7 of questions and answers that have been added. 8 Over 100 questions, actually, were submitted to the 9 agency as a consequence of the rule-making. We 10 have provided answers to those questions. So we 11 are, in fact, evaluating the effectiveness of the 12 Compliance Guidelines that we specifically designed 13 for our small and very small plants to see if, in 14 fact, they are effective in helping them address 15 the control measures in the proposed rule and, in 16 particular, helping them to be able to operate 17 within one of the more controlled alternatives such 18 as the use with antimicrobial growth inhibitors and 19 post-lethality treatments. 20 Then the agency is particularly interested 21 in working with FDA and CDC on the attribution of 22 illnesses for meat, poultry and processed egg 23 products and, in particular, the FoodNet 24 case-control study. It is really critical for us 25 to have a better understanding about the 113 1 contribution of the products that we regulate in 2 terms of foodborne illness. 3 That, then, raises some additional 4 concerns that we have in terms of the distribution 5 control for the products that we regulate. This 6 particularly relates to products at retail and the 7 growth of the organism throughout the time that the 8 product has a shelf life. 9 The agency has traditionally held that it 10 would focus its in-plant inspection activity within 11 the plants, themselves, that we regulate. We have 12 taken a broader view of what we need to do to 13 impact public health. We do have jurisdiction for 14 our products as they travel throughout the 15 distribution chain and, in particular, at retail. 16 The agency is looking into what it can do to effect 17 more of a change in the control of Listeria at 18 retail. 19 That leads, then, to the state programs. 20 Many states, approximately 50 percent of the 21 states, have their own state program in terms of 22 production of meat and poultry products. The 23 agency is particularly interested in how effective 24 the states are operating with regard to the 25 control, that they are providing for the plants 114 1 that they regulate as well as to the prevalence of 2 Listeria in the products that are coming out of 3 those facilities. So we will have an increased 4 focus on the state-program activities as well. 5 Then, finally, we have education and 6 outreach as a specific goal that we want to measure 7 in terms of improvements being made as a 8 consequence of this rule-making. We believe that 9 we have a necessary need to provide increased 10 understanding about this particular organism, about 11 the chemicals and processes that may, in fact, 12 provide additional safety or enhanced safety to the 13 products and that the consumer needs additional 14 information about how to select for products that 15 may, in fact, be safe in terms of providing 16 enhanced safety over the products that don't have 17 formulated antimicrobials or other actions that 18 may, in fact, reduce the risk of Listeria being in 19 the products. 20 With that, then, we move into our 21 risk-communication activities in which we are 22 specifically looking at the FDA-FSIS risk ranking 23 and risk assessments that have been done to inform 24 our communication activities. We have worked 25 closely with FDA to ensure that the messages for 115 1 consumers are, in fact, consistent. 2 Based on the scientific information 3 available today, the consumer advice on Listeria 4 monocytogenes that we provide to at-risk 5 populations will remain the same. That advice, 6 then, that we have in place today is that you would 7 not eat hot dogs, luncheon meats or deli meats 8 unless they are reheated until steaming hot, and do 9 not eat refrigerated pate or meat spreads except 10 that the stable pate or meat spreads can be eaten. 11 Additionally, the consumer advice that the 12 agency is concentrating on is to follow up on the 13 findings from the most recent iteration of the 14 FDA-FSIS risk-ranking assessment about the issue of 15 the organism being able to grow at temperatures 40 16 degrees and above, in particular, and to ensure 17 that consumers have information about refrigeration 18 temperatures. So we join FDA in advising all 19 consumers to use perishable items that are 20 precooked or ready to eat as soon as possible, to 21 clean the refrigerators regularly and to use a 22 refrigerator thermometer to make sure that the 23 refrigerator always stays at 40 degrees or below. 24 In addition, with our outreach activities, 25 the agency will continue to target the at-risk 116 1 populations with its consumer messages including 2 the elderly, pregnant women and individuals with 3 weakened immune systems. We brochures on 4 listeriosis in pregnancy as an example. 5 Those brochures are remaining the same at 6 this time. However, FSIS is increasing its 7 outreach to these groups in a variety of ways. We 8 are providing food-safety messages in 9 patient-education videos that are shown in doctors' 10 waiting rooms and we are expanding our efforts to 11 reach medical centers that treat organ-transplant 12 patients. 13 We also are using our food-safety mobile 14 to get out additional information around the 15 country, particularly with message about 16 listeriosis. 17 As industry makes further strides in 18 reducing the levels of Listeria monocytogenes in 19 ready-to-eat products, and as more information is 20 gathered regarding consumer behavior, FSIS will 21 reevaluate whether changes in consumer advice is, 22 in fact, warranted. The agency is particularly 23 interested in moving away from the advice that we 24 presently have for particularly at-risk 25 populations in that they should not eat segments of 117 1 the products that we regulate or, if they do choose 2 to eat them, that they should cook them. 3 Therefore, the agency expects to refine 4 its risk-based verification activities to enhance 5 public-health protection at the points in the 6 farm-to-table continuum where mitigation is most 7 effective. Again, this will be an increased focus 8 on how effective our regulation is in terms of 9 providing an incentive to the industry to enhance 10 the controls that they already have in place with 11 regard to Listeria and, also, to look at the 12 messages that are on those products and the 13 understanding the consumer has about how to 14 properly handle the meat and poultry products that 15 we regulate. 16 Then, we will continue to work closely 17 with FDA and CDC to reduce the burden of human 18 illnesses caused by Listeria monocytogenes. These 19 will be the two primary focuses that we will have 20 this coming year to measure the success of where we 21 are with regards to the accomplishments we have put 22 in place for Listeria control. Then we will either 23 modify the regulation that we have in place or 24 leave it as it is and enhance the information that 25 is contained within it and the guidance that we 118 1 have available to our inspection personnel and to 2 industry. 3 Thank you very much. 4 MR. KAUTTER: Thank you, Dan. We are, 5 again, ahead of schedule. Let's give you about a 6 twenty-minute quick break, come back at 2:00. I 7 will assemble my panel and be ready to go for the 8 Risk Management Panel and then we will jump right 9 into the Public Comments after that. 10 (Break.) 11 MR. KAUTTER: To complete out the public 12 meeting and this afternoon, we are going to have a 13 question and answer session with the panel up 14 front. Dr. John Kvenberg will be moderating it and 15 then we will go into Public Comments to close out 16 the afternoon. 17 John? 18 Questions for Risk Management Panel 19 DR. KVENBERG: Thank you, Don. It is nice 20 to see everybody back again. So I guess this is 21 the point in time when we appreciate getting 22 questions. We will give folks a few more minutes 23 to get into the room, especially Marjorie, and then 24 I will introduce the panel. 25 It gives me great pleasure to introduce 119 1 that panel. I am really appreciative of the people 2 who are on the panel, frankly. For those of you 3 who don't know, I will just briefly go down the 4 row. Dr. Robert Brackett currently heads our Food 5 Safety staff and, as of January 1, will be running 6 this place as the Director for Center for Food 7 Safety and Applied Nutrition. 8 Next to him is Mr. Joe Baca, head of our 9 Office of Compliance; Dan Engeljohn from FSIS who 10 spoke previously; Rob Tauxe from CDC who likewise 11 spoke previously; Dick Whiting, again who spoke 12 previously. We have Gary German. Gary German is 13 with our Office of ORA and is our Chief Training 14 Officer and the Chancellor, I guess I would call 15 you, Gary, of ORAU and our educational effort; and 16 Marjorie Davidson who heads our consumer 17 information effort here at CFSAN. 18 With that, any questions you have about 19 the Action Plans for FDA, CDC or FSIS? Jenny 20 Scott, NFPA. 21 MS. SCOTT: Thank you. I am Jenny Scott 22 from NFPA. I have a question for Dan Engeljohn. 23 Dan, you said, as one of your action steps, that 24 you were going to look at prevalence changes and 25 the levels of Listeria monocytogenes groups of 120 1 products. So you are going to start enumerating 2 Listeria now? 3 DR. ENGELJOHN: Yes. I think that, in the 4 design of the verification-testing program that we 5 are contemplating putting in place this coming 6 year--and, again, it is dependent, in part, upon 7 knowing who is producing what. So the regulation 8 that we have put together asks to have the industry 9 inform us as to who is producing what and what they 10 are doing and that type of thing. 11 That information, then, would be used to 12 segregate products into various sampling schemes 13 based on the types of products and then some random 14 type of sampling, much like a statistically based 15 baseline that we would do to get general prevalence 16 information. So the intention would be to get not 17 only just positive or negative, but to do some 18 quantitative analysis as well. 19 MS. SCOTT: You also talked about doing 20 some--looking at retail. What, exactly, do you 21 have in mind with where you are going to go with 22 retail? 23 DR. ENGELJOHN: Generally speaking, the 24 agency does have jurisdiction at retail with regard 25 to products labeled. So the meat and poultry 121 1 products that are there are ones for which we have 2 traditionally looked at in a concerted way. I 3 don't think we have actually put together a 4 complete plan as to what all that would entail. 5 I think we are well aware of many of the 6 activities that FDA has under way with themselves 7 as well as many of the state programs and to begin 8 assessing what, in fact, is being done at retail 9 and what more we could gain by either doing some 10 additional studies as to the types of sanitation 11 programs and the types of documentation, that type 12 of thing. We haven't ruled out doing sampling at 13 retail in terms of food-contact surfaces, but we 14 also recognize what implications that would have. 15 So it would be something that we would 16 have a well-defined strategy for before we started 17 doing something in that area. Our goal, at this 18 point in time, though, is to actually do a 19 substantial change in our thought process and 20 actually think about going outside of the implant 21 activity. So I think that being said, just looking 22 into the feasibility of doing more things, that 23 distribution is a big step for the agency and how 24 we are planning. 25 MS. SCOTT: Will you have some discussions 122 1 with industry before you roll out this program? 2 DR. ENGELJOHN: I would imagine that the 3 agency certainly would have a plan that it would 4 get public input on, although none of that has been 5 decided yet. But I would guess that there would be 6 the opportunity to have those kinds of discussions. 7 I think it is something that we need to be working 8 together on, quite frankly. 9 MS. SCOTT: Thank you. 10 DR. KVENBERG: Thank you, Jenny. 11 Any additional questions or comments? 12 MS. MARCOTTE: Michelle Marcotte from 13 Marcotte Consulting. I have a suggestion for your 14 Action Plan. In February, 2002, the FDA filed a 15 petition asking, or requesting, approval for the 16 irradiation of the number of different ready-to-eat 17 products from the National Food Processors 18 Association on behalf of a coalition of food 19 irradiation that included a number of different 20 producer and processor associations and some 21 companies that included some of my clients. 22 We are coming on February, 2004 soon and 23 there have been a lot of things that could have 24 been prevented if we had had that irradiation 25 approval in place. There are a lot of companies 123 1 that are waiting for it, waiting for that signal to 2 come from FDA of the publication of a proposed 3 regulation so at least we know things are coming. 4 So I suggest that that regulatory approval 5 and allowing irradiation of the ready-to-eat foods 6 be put in place, that it would assist in the Action 7 Plan. It is something that is real that you know 8 is going to be effective for a number of the 9 products that you have identified as high-risk 10 products. So if you have any news to report on 11 that petition, I would be very interested in 12 hearing it, but that is a suggestion, that we get a 13 move on and get that approval in place. 14 DR. KVENBERG: Thank you. Any comments 15 from the panel on that particular point? Thank you 16 for the comment. Additional comments, questions? 17 MR. GARFIELD: Bob Garfield, American 18 Frozen Food Institute. John, in your presentation, 19 you talk about the importance of healthcare 20 providers being educated on Listeria and prevention 21 for their patients. What, explicitly, has the 22 agency been doing to get the information out to 23 healthcare providers and what kinds of things are 24 you planning for the future, if you have any plan 25 that is different from what you have done in the 124 1 past? 2 I know that the IPIC survey that came out 3 about a year or two ago showed that there was 4 really inadequate education of healthcare 5 professionals, especially those that are dealing 6 with immunocompromised individuals that they don't 7 know about Listeria and the impact of Listeria. 8 But I would be curious to know what, if anything, 9 you have got in plans for the future. 10 DR. KVENBERG: Thank you. This is a joint 11 plan. Rob, if I could impose on you from CDC's 12 point of view, to make a comment relative to that 13 as an initial response. 14 DR. TAUXE: The first step, I think, was a 15 collaboration of the American Medical Association 16 putting together a package of information about 17 foodborne diseases in general with case vignettes, 18 one of which was a Listeria case. That was sent 19 out--it was made available through the MMWR, 20 Morbidity-Mortality Weekly Work, our attractively 21 named weekly publication from CDC and also as a 22 continuing medical-education kit that was shipped, 23 I think, to a large number of physicians in the 24 country. 25 I think we are now in the process of 125 1 updating that but it did include a listeriosis case 2 as one of the case vignettes. I think there is 3 more that needs to be done and we are reaching out 4 to the professional groups, like the obstetrical 5 group is a specific targeted group that would be 6 very good. 7 DR. KVENBERG: Thank you. Marjorie or 8 Bob? 9 DR. DAVIDSON: We have also been working 10 closely with the healthcare providers that work 11 directly in education programs with the at-risk 12 groups. We have mailed to all the obstetricians 13 and gynecologists with information about Listeria 14 to keep in their office waiting rooms, if you will. 15 We also are going to be launching a 16 pregnant women's program for women, or for health 17 professionals that are carrying out education 18 programs for women and they are pregnant, and it 19 will contain Listeria advice as well. Those kinds 20 of operations are going on. Also work with the 21 elderly in the same regard. 22 MS. HOLLINGSWORTH: Jill Hollingsworth 23 with the Food Marketing Institute. I actually have 24 more of a comment than a question. The retail 25 industry has been working very diligently on 126 1 implementing active managerial controls at the 2 retail level and particularly focusing in on 3 Listeria. We were sponsors, or cosponsors, of the 4 NFTA study, of the survey that tested for Listeria 5 in ready-to-eat products. 6 We want to particularly thank FDA for 7 including that data in this revised Risk 8 Assessment. We think that it was a significant 9 contribution and we appreciate FDA's willingness 10 and openness to use that kind of data to constantly 11 look at the Risk Assessment and keep it current. 12 The food industry, in general, we believe, 13 has done a very good job in addressing the Healthy 14 People 2010 challenge. The information from CDC, 15 for example, about the over 40 percent decrease in 16 the foodborne LM infections over five years and 17 also USDA's recent release of data for the past 18 year showing the 25 percent drop in the positive LM 19 samples in products that FSIS regulates we think 20 are all really positive signs that we are heading 21 in the right direction. 22 But we also know that we can't just stop 23 there, that more has to be done. We are very 24 committed as retailers to working with the 25 government to improve the kinds of actions that are 127 1 in place in reducing listeriosis cases. 2 To that end, the retail industry recently 3 put together a Listeria Working Committee. That 4 committee outlined six objectives that we laid out 5 for the retailers to look at listeriosis and 6 participate in its reduction. It was quite 7 interesting. On Monday, when we first saw the 8 revised updated FDA Management Plan, that there was 9 a tremendous amount of similarity in the six 10 objectives that we had already defined and the ones 11 that came out in the updated Action Plan. 12 But, to be honest with you, one of our 13 concerns was that there did seem to be sort of a 14 "let's do everything and rush to get everything 15 done" approach. We would like to ask both FDA and 16 USDA to work with us and the rest of the food 17 industry in looking at a really targeted approach 18 to reducing listeriosis and meeting the goals that 19 we have set for ourselves. 20 We believe that one of the most important 21 take-away messages from the Risk Assessment was 22 that it is this kind of targeted approach that is 23 going to be necessary to bring down the level of 24 listeriosis even further. So, within our 25 objectives, we look at those kinds of targets; 128 1 risky foods, risky behaviors, susceptible 2 populations. 3 We would like to offer, at this time, to 4 work closely with FDA and CDC and USDA to further 5 refine the Action Plan and to be sure that we are 6 targeting actions that would produce measurable 7 results. 8 The Food Code states that food safety 9 should be a shared responsibility between the 10 industry and the government and we are very 11 committed to that and we hope that we can work with 12 you in the future to develop plans that will be 13 targeted to reduce Listeria. 14 The other thing that we wanted to share 15 with you is one of the approaches that the retail 16 industry has taken is to look at ways of enhancing 17 our in-store programs. Just this week, we released 18 a Total Food Safety Management Guide for dealing 19 with handling the parent and selling fresh-cut 20 produce at retail. This is the second guide we 21 produced. We did one not too long ago on ground 22 beef and we have already begun the work and are 23 well underway on developing a Best Retail Practices 24 Guide for the deli department. We would like to 25 work with you on those kinds of projects. 129 1 Again, thanks for the opportunity to be 2 here and we look forward to working with you. 3 Thank you. 4 DR. KVENBERG: Thank you. 5 MS. SMITH DEWAAL: Is this the Public 6 Comment period or the Question and Answer period. 7 DR. KVENBERG: This is the Question and 8 Answer period. You will have plenty of 9 opportunity--I was thinking, too, if people had 10 comments about the presentation but it is the 11 Question-- 12 MS. SMITH DEWAAL: I have lots of 13 comments, but I was just going to ask some 14 questions. 15 DR. KVENBERG: That would be fine. We can 16 do both--in order. 17 MS. SMITH DEWAAL: My first question--I 18 think I will start with Bob Brackett at FDA. Bob, 19 you rolled out an interesting Risk Management Plan 20 this afternoon. I will have some comments on that 21 later, but I want to know what your new level of 22 testing and inspection is for your high-risk 23 plants. 24 DR. BRACKETT: The answer to that--I don't 25 know the specific numbers, but John, you said you 130 1 had something on that? 2 DR. KVENBERG: I think the issue that we 3 are looking at, within our resources, is 4 retargeting our sampling procedures against those 5 commodities that were identified as being at higher 6 risk. 7 MS. SMITH DEWAAL: Okay. That's great. 8 So let's take a hypothetical. I am running a 9 smoked-salmon plant. How often are you going to 10 come and inspect me and how often are you going to 11 test my environment and my products for Listeria 12 now that you have got this great Risk Assessment. 13 DR. KVENBERG: Let me answer it in two 14 parts. Number one, we don't have the resources 15 within FDA to meet all smoked-fish plants equally 16 at all times. But we do have state partnerships. 17 I think one of the things that we will be able to 18 do with a focussed Risk Assessment is go at those 19 high-risk and moderate-risk commodities a focused 20 effort and have not only FDA's resources but 21 cooperate with state resources who are also working 22 in the area such as seafood for inspection and 23 sample analysis. 24 DR. KVENBERG: Joe Baca? 25 DR. BACA: I think we can safely say that, 131 1 at a minimum, we would be in every high-risk plant 2 annually. If we encounter a problem, however, we 3 will continue to go back and do additional work and 4 bring the regulatory resources necessary to address 5 the issues. 6 Also, like John says, we will work with 7 the states to also bring whatever resources they 8 have to bear on that problem. 9 MS. SMITH DEWAAL: How often are you going 10 to actually test in my plant? 11 DR. BACA: If we were to go into a plant 12 and they are complying with HACCP and we don't find 13 any major issues, we may collect a sample. We 14 don't necessarily have to collect a sample. On the 15 other hand, if we look at see that there appears to 16 be problems, then we will sample more stringently. 17 There is no formula. It depends on the situation 18 as we encounter it. 19 MS. SMITH DEWAAL: As you know, in the 20 Seafood HACCP Evaluation that was released, I don't 21 know, maybe a year or so ago, they found that 22 Listeria remains a serious problem even, what, four 23 or five years after Seafood HACCP implementation. 24 So I would hope you would conduct a sample in every 25 plant. 132 1 But you are saying that annually, you 2 think, for all high-risk plants that produce 3 products identified as high-risk for Listeria, you 4 would have someone in there once a year. 5 DR. KVENBERG: Correct. But if I could 6 explain just a bit on smoked seafood because you 7 touched on that specific area, and this goes back 8 to the science and research of what we are doing, 9 smoked fish is maybe a subset of the larger 10 industry. But, if on finding a problem and then 11 finding a reoccurring problem, I guess one of the 12 things we are going to try to do using a 13 science-based risk approach in our inspectional 14 process is try to be a little more--I am searching 15 for the correct word--scientific, I guess, or 16 precise in our work. 17 We have tools now that we didn't have 18 before. What I am getting at is that in 19 situations, especially in small processing plants, 20 not to pick on seafood as opposed to cheese or 21 anybody else, part of the area of what we have 22 learned and what is becoming part of the 23 international knowledge is you may have reoccurring 24 problems in a plant due to sanitation or good 25 manufacturing issues within that plant that cause 133 1 the problem to reoccur. 2 I think, obviously, you do have a problem 3 with Listeria coming in on raw materials from a 4 source. That is one area. But part of our 5 directed effort is going to be to try to go in and 6 do some in-depth work. We have had problem areas 7 in certain plants and have previously worked 8 through consent decrees, injunctions, these kinds 9 of procedures. But I think, with the tools that we 10 have now, maybe the industry can work and we can 11 work, as well, and so can the states with better 12 tools to eliminate potential sources of 13 contamination, thereby reducing the risk. 14 So it not merely a sample collection and 15 gathering more baseline data. Let's use the 16 knowledge we have and the tools can get very 17 precise. If you have the same kind of Listeria, by 18 strain, by PulseNet type, until it reoccurs within 19 a same plant, now let's go find where that thing is 20 coming from if it is not coming in from the 21 outside. 22 Those are the kinds of things we were 23 putting into our strategy we didn't really cover 24 this afternoon, but we are going to try to be using 25 our head and a science-based approach to get at 134 1 these issues that we have in plants. 2 MS. SMITH DEWAAL: I would support that 3 but you need to be testing to know that you have 4 got the same strains showing up multiple times. So 5 I think testing the environment of the plant is 6 critical to your having an effective inspection 7 program in this area. 8 In addition, although I am not entirely 9 happy with the way they did it, FSIS has at least 10 recognized the utility of having industry testing 11 as a requirement. Now, they didn't require nearly 12 enough industry testing for Listeria monocytogenes 13 in ready-to-eat meat plants, but they have a 14 requirement in place which you don't have today. 15 So you go in and you find it in a 16 smoked-seafood plant or a plant producing soft 17 cheeses, what action are you going to take? When 18 are you closing the plant down? How many positives 19 do I need to have as a plant owner before I get 20 shut down? 21 DR. KVENBERG: Joe, do you want to take 22 the initial-- 23 DR. BACA: In my experience, and what we 24 have done, is if we go into a plant and we sample a 25 product, if that product is distributed before we 135 1 come back with a positive response, the industry, 2 in my experience, has been very responsible about 3 recalling. So that addresses that issue. 4 Now, if we go back in and we find, 5 routinely find it, or we find it in the environment 6 of the plant, then we are going to seek other 7 remedies. We will work with the plant 8 cooperatively as far as it goes. If that doesn't 9 address the issues, then we will seek injunction. 10 There are a number of seafood plants that are under 11 injunction right now. 12 MS. SMITH DEWAAL: So you rely, though, 13 just on recalls to get the product back? 14 DR. BACA: Right. If we identify the 15 problem in a product, we would expect that firm to 16 recall. They have--in my experience, firms have 17 recalled the product. They have acted fairly 18 responsibly in that area. 19 MS. SMITH DEWAAL: With a voluntary recall 20 system. 21 DR. BACA: Yes. 22 MS. SMITH DEWAAL: Now I get to talk to 23 Dan Engeljohn. You said that--in one of your 24 slides, you made an interesting point about 25 industry sharing of plant-generated data. What do 136 1 you mean and I may have a few follow-up questions. 2 DR. ENGELJOHN: When we issued our 3 directive last December, part of the design of that 4 directive, and this is pre-regulation--part of the 5 design of that directive was to provide an 6 incentive to industry to share the data that they 7 had available with regards to the implant-testing 8 that they were doing. 9 I would say that it was a general belief 10 on industry's part that they did not have to share 11 that type of data with us. I would contend that 12 the agency has always held that that data is part 13 of their decision-making with regards to the hazard 14 analysis and HACCP systems that they should have in 15 place. 16 The agency was not uniformly enforcing 17 that interpretation and, in fact, wasn't enforcing 18 it well at all at that time. But, as a consequence 19 of directing our verification activities in plants 20 that were not sharing the data that we knew that 21 they were, in fact, collecting, I would say that 22 overwhelmingly the establishments began making that 23 available as part of their hazard analysis and 24 decision-making documents. 25 As part of the Final Regulation that we 137 1 issued then in June and went into effect in 2 October, it made it mandatory--it made it clear 3 that the agency's expectation was that any of the 4 data that is used in terms of decision making about 5 their food-safety system is, in fact, applicable to 6 the regulations that we have and have to be made 7 available to the agency. 8 MS. SMITH DEWAAL: There is this whole 9 issue of industry having data regarding either 10 Listeria monocytogenes or E. coli 015787 which then 11 becomes highly relevant during the time that we 12 have an outbreak going on and we have the 13 PulseField, the PFGE results coming in on the human 14 health side. Is there any mechanism or any even 15 voluntary system right now to gather that industry 16 data, maybe by a third party, a nongovernmental 17 party, so that that data is available, the 18 plant-generated testing data is available to match 19 up to the PulseNet data and, perhaps, give the 20 industry an early warning that their products need 21 to be recalled and giving consumers greater 22 protection because the products could actually be 23 taken off the market much faster than waiting for 24 the government to go in and run their tests to 25 actually identify plants with the strain of Listeria. 138 1 Is there any effort in this direction? 2 DR. ENGELJOHN: I am not aware of a 3 specific effort on the issue you raised about 4 making the fingerprint information available ahead 5 of time, although there are a number of 6 initiatives, I know, underway in terms of industry 7 collectively trying to find ways to make 8 information available to the agency and greater 9 information about sharing information as well. 10 I think we are all looking at ways to 11 improve our ability to anticipate when there is the 12 potential for an outbreak to occur. I think we 13 have reason to believe that there are indicators 14 that we should be tapping into that we haven't 15 necessarily found the mechanisms and have 16 identified yet. 17 I would hold that that is an interesting 18 area which we need to pursue. But I am not aware 19 of anything specifically right now. 20 MS. SMITH DEWAAL: Rob, are you aware of 21 anything on this, because you mentioned, also, in 22 your slides this idea of industry sharing their 23 data or getting the PFGE results out of PulseNet 24 faster so we can actually get industry data 25 matching up and getting an early-warning system in 139 1 place so the products get off the market faster? 2 DR. TAUXE: Our efforts are focused at 3 encouraging the clinical labs to send the Listeria 4 isolate from a human into the state health 5 department as quickly as they can without delay and 6 then having the PFG at the state, again without 7 delay. 8 I am not aware of a specific plan to make 9 industry isolates available but we did recently 10 hold, I think, with--I think it was with NFPA, was 11 it not--a training course--I'm sorry; ILSI--a 12 training course for industry microbiologists to use 13 the PulseNet techniques so that they have the 14 capacity now to use the same standard method and 15 generate results that could be compared. 16 DR. ENGELJOHN: I did want to just follow 17 up. I don't think quickly enough sometimes. But I 18 would say that there are efforts underway with 19 regards to the sharing of isolates which is 20 something we hadn't considered before. So there is 21 active discussion underway as to what we can do and 22 should be doing about that particular issue, which 23 sort of gets at moving forward in that area. 24 DR. KVENBERG: Additional comment? 25 Question? 140 1 MS. KOSTY: I am Lynn Kosty with the 2 American Meat Institute. I actually had two 3 questions. The first one, I might have just missed 4 the answer because I got a little lost trying to 5 find 1500 Paint Branch Avenue today. So I am a 6 little bit late. Sorry about that. But this 7 probably goes to Dr. Tauxe, and that is, in the FDA 8 objectives, it said that CDC was conducting a 9 comprehensive case-control study on Listeria and 10 the factors that make people more susceptible to 11 the disease. 12 It says that that is going to be completed 13 this year but I was wondering are there going to be 14 the results released this year or do you have a 15 more definitive time frame on that? 16 DR. TAUXE: Yes. This is our three-year 17 Listeria case-control study that has been underway 18 in FoodNet sites. We have concluded the 19 data-collection phase and so now we are into the 20 analysis phase. The results will not be released 21 this calendar year, but we hope to be able to 22 present them in this coming spring, early spring, 23 at least the preliminary results of that. 24 It was intended to be a two-year 25 case-control study. I should point out, that is a 141 1 long time to collect data, three years. It was 2 intended and powered to be a two-year case-control 3 study but, as the rate of listeriosis has 4 decreased, the rate of which new cases could be 5 enrolled was less than we had anticipated and we 6 had to extend it an additional year. 7 MS. KOSTY: Thank you. Then my second 8 question is just kind of to FDA and John and that 9 has to do with refining the risk assessment. We 10 heard Jill Hollingsworth say earlier that--she 11 referred to the market-basket study that was done 12 by NFTA and FMI in conjunction to look at 13 differences be retail sliced deli meats versus 14 those that are sliced in federal establishments and 15 prepackaged and sold at retail. 16 I was just wondering if, in the refined 17 Risk Assessment, whenever that is going to be put 18 together, whether there would be some kind of 19 separation of those different categories of deli 20 products and if they would take into consideration 21 the work that has been done recently while the risk 22 assessment has been going on that shows the 23 increase in formulated products to inhibit 24 Listeria. I think 70 percent in hot dogs and 50 25 percent in formulated deli-sliced products. 142 1 DR. KVENBERG: If I could, because the 2 risk assessors were very good in doing Roberts 3 Rules of Order and not getting into risk-management 4 questions. Bob, you or Dick would like to comment? 5 Dick? This is because this is, I think, a 6 risk-assessment question and, Dick, you have been 7 involved in that. 8 DR. WHITING: I don't know. There are a 9 lot of kind of comments in there. You are correct. 10 I think the study did show that deli-sliced and 11 packaged samples did have a higher contamination 12 rate, although I think the highest contamination 13 level was in a manufactured processed one, as I 14 recall. 15 Again, when you start splitting things up, 16 your database becomes even smaller than what you 17 saw on the screen. But I don't think we have any 18 plans at this point to follow up on that with a 19 risk assessment, if that is what one part of your 20 question was asking about. 21 I think if the data is collected and so 22 on, it will pretty much stand on its own. I don't 23 think we will need a full analysis to do it. 24 MS. KOSTY: That was generally my 25 question, since you are going to be refining the 143 1 Risk Assessment, does that mean that you are going 2 to be putting out an entire revised version in 3 another year or two years or are you just going to 4 be working off what you have. I guess I am 5 wondering because of the ranking of the products 6 and, as you are getting more information from 7 industry that they have been collecting and from 8 the agencies, from your own collection of data on 9 these types of products, are you going to reissue a 10 new Risk Assessment that might separate out this 11 category into more categories or are you just going 12 to keep the Risk Assessment the way it is and 13 modify your management strategies based on that, 14 regardless of what the rankings does. 15 DR. KVENBERG: Could I ask Bob Buchanan to 16 come forward because I think he could assess the 17 future. 18 DR. BUCHANAN: We find the quantitative 19 risk assessments of the types we have done to be 20 two things; extremely powerful tools that can 21 provide a great deal of useful information and 22 extremely labor-intensive activities that have to 23 be balanced against the rest of the risk-assessment 24 needs that we have within the agency. 25 We will be looking at ways to continue--we 144 1 have the models in place. We have the databases in 2 place. They are not going to disappear. We think 3 that they will be very powerful tools particularly 4 going back and looking at the effectiveness of our 5 programs every so many years. We do not think that 6 we are going to be in the continuous business of 7 updating the Listeria Risk Assessment every year. 8 However, we can take parts of it and respond to 9 very specific questions as we update the material. 10 Likewise, as we get into questions related 11 to specific product groups, there is actually a 12 different form of the Risk Assessment that is more 13 appropriate. It is called a product-pathogen 14 pathway analysis. We are planning on doing those 15 for some of the FDA-associated products and FSIS 16 has already started one on some of theirs. 17 So we will go back and look at the 18 appropriate risk assessment based on the questions 19 that are being posed and for which there needs to 20 be an answer. But, like I said, we are certainly 21 not going to take this Risk Assessment and throw it 22 out the window and start all over. But, on the 23 other hand, we also have to balance the repeating 24 of this on a routine basis against our other needs. 25 We do have a number of other risk assessment that 145 1 are already planned on related subjects. 2 That is sort of a non-answer. 3 MS. KOSTY: Thank you. 4 MR. WEIGNER: Tim Weigner with the Food 5 Marketing Institute. I have actually two 6 questions. The first one is going to be the longer 7 question and the second one, hopefully, will be the 8 shorter question. 9 Several years ago, the FDA, working 10 through the Conference for Food Protection, 11 released the retail program standards targeted for 12 regulatory agencies to monitor at the state and 13 local and the county level critical risk factors 14 that contribute to the different types of foodborne 15 illnesses. 16 This year, the FDA, working through the 17 industry, has allowed industry to take this 18 baseline tool and allowed the industry to monitor 19 those particular critical risk factors and how we 20 can control that. How does that particular 21 standard and that baseline tool play into this 22 Action Plan? 23 DR. KVENBERG: I will take it. As you 24 know, Tim, I think that one of the things that I 25 have always admired about the retail program and 146 1 the interaction with both the industry and the 2 states is it has been kind of forward-thinking in 3 how it is doing. 4 I have looked at the Listeria Action Plan 5 as kind of being a vanguard in basically 6 implementing a risk-management program. So, 7 likewise, I think there is something to be learned 8 from the efforts that we have done at retail to 9 take under consideration for general-management 10 plans. 11 So, yes; I understand the basis of your 12 question and I think there is an applicability, 13 especially in the retail sector, to look at the 14 work that is ongoing with the conference and with 15 our retail folks as it would address Listeria. 16 Does that help? 17 MR. WEIGNER: It does a little bit. The 18 second question, and hopefully this one will be a 19 little bit easier, you have six different 20 objectives. Are there any discussions to talk 21 about measurements of each one of those objectives 22 as we move into the Year 2005. You have all these 23 ideas out here using the consumers as one point. 24 How do you quantify that there was an actual change 25 in behavior in the consumers that reflects what 147 1 your goals are in your objectives? 2 DR. KVENBERG: That is a very good 3 question. If nobody else has one, I at least have 4 one to come from the forefront on. I think that it 5 is beholding of us to have measurable quantitative 6 measurements, not only for our own evaluation but 7 also to OMB and to the Executive Branch and through 8 Congress and the GAO to have measurable outcomes 9 under the requirements that we have under PART and 10 under GIPRA for how we are doing things. 11 So we are very aware of establishing 12 baselines and measurable goals or programs. The 13 overall objective, as we have said all along, was 14 to reduce the incidence of Listeria monocytogenes 15 by 50 percent by the Year 2005. Part of what we 16 have had internally under consideration is using 17 the Listeria model for the evaluation procedures so 18 that we can come to where we have contributed to 19 that goal or where we have accomplished that goal 20 through various efforts. So we are looking at 21 measures and how we would measure this Action Plan 22 against the goal and then try to document that 23 relative to not only that we achieved the goal but 24 the measurements of how these actions played in 25 that reduction. 148 1 So that is part of the evaluation process 2 that I think we are going to be looking forward to 3 on this Action Plan when it is over. 4 DR. DAVIDSON: We, also--FDA has the 5 consumer survey, a national consumer survey, that 6 looks at attitudes, knowledge and behavior on the 7 part of consumers. We will be looking at that 8 regarding Listeria. 9 DR. KVENBERG: Additional questions? 10 AUDIENCE: Thank you. This is an 11 excellent nexus for my question, and I apologize 12 for my voice. But, on the issue of consumer 13 education and, if I may, I have a comment and one 14 question, Dr. Kvenberg. 15 DR. KVENBERG: Sure. 16 AUDIENCE: The FSIS Interim Final Rule on 17 the control of LM in ready-to-eat meat and poultry 18 products allows an establishment to declare the use 19 of a post-lethality treatment or an antimicrobial 20 agent or process on the label. 21 We all recognize that if the label 22 statement is to be effective, it must be both 23 understood and interpreted correctly by consumers. 24 I realize FDA and FSIS have this understanding, 25 whether it is with labeling as well as advisories. 149 1 To this end, message testing is an absolute 2 necessity. In order to obtain such data, the 3 industry has undertaken a web-based survey of a 4 number of label statements to try and assess what 5 message is conveyed to the consumer. 6 Messages are important but it is critical 7 their ultimate result be a positive behavioral 8 change so, Dr. Kvenberg, you can measure positive 9 results. The survey results that we have reviewed 10 are still preliminary and time does not permit me 11 to provide the details of the survey today. 12 However, the information gathered by the survey, 13 when completed, will be submitted to both FDA and 14 FSIS for consideration as you focus on 15 risk-management strategies as well as 16 risk-communication strategies. 17 But, based on the preliminary results, it 18 appears that considerable variability exists in the 19 consumer's ease of understanding of the statements 20 and in how the statements tested affected whether 21 or not consumer would purchase cold cuts or hot 22 dogs bearing the statement. 23 I know I am taking this out of context 24 because you don't have the--I don't have the time 25 and we don't the time at this point in this 150 1 discussion to give you a full scale. But all 2 statements tested in some way appeared to promote 3 questionable handling practices by consumers, such 4 as storing products for long periods of time or 5 they had the effect of steering consumers away from 6 products that were enhanced for food safety. 7 It is important to point out that messages 8 and labels are not substitutions for education. I 9 know I am speaking to the choir here. Messages and 10 labels are tools for food-safety education and not 11 final solutions in and of themselves. 12 Recognizing that our collect goal is to 13 ensure the continued advancement of the public 14 health, and this includes education, using tools 15 such as labels and messages. Because our 16 preliminary findings suggest that none of the 17 statements tested were able to positively and 18 clearly communicate the safety of the product 19 without promoting questionable food-handling 20 practices and, in fact, may encourage, as we have 21 discovered unsafe food-handling practices in some 22 instances, we believe it is critical that further 23 study be undertaken. 24 My question to you is, and I will open it 25 to the panel, what is the proper process for making 151 1 sure the agencies, both FSIS and FDA, factor such 2 research into their decision-making process as it 3 relates to risk management and as it relates to 4 risk communication. 5 DR. BRACKETT: I am going to answer the 6 first part of that, and that is knowing what to put 7 on the label in the first place. This morning, and 8 this afternoon, perhaps, it was mentioned about the 9 National Advisory Committee's study of shelf-life 10 based on safety, or dating for shelf-life based on 11 safety. The information, hopefully, that will come 12 out of that evaluation I think is going to be 13 critical for information that will be on the labels 14 of specific classes of foods. But, in addition, 15 there are also, and the part, I think, that you are 16 concerned about which is focus groups. Marjorie 17 talked about that. 18 DR. DAVIDSON: Yes; that is one of a 19 variety of ways that we use here to look at 20 messages and how people will react and understand 21 them. I totally agree with you on the need for 22 that. It is of no value if it is not being 23 understood and acted upon. 24 DR. ENGELJOHN: I will just address the 25 issue. In part, I would say that it is important 152 1 to recognize there is a difference between a 2 truthful label and one that, in fact, would have 3 some positive impact on consumer behavior. I think 4 that is the issue that we are talking about here. 5 To that extent, I would also say that we 6 clearly recognize the need for some type of an 7 educational campaign to be joined together with the 8 labeling and that type of thing, I think would be a 9 positive way to demonstrate whether or not there is 10 an effective change that could result from it. 11 The agency made it clear within the final 12 regulation that we issued that this is something 13 that could be done today and we were looking to see 14 ways to push forward incentives for industry and 15 consumers and government to work together to, in 16 fact, address the issue of listeriosis. So, to 17 that end, I certainly believe that we do have to 18 have in place some effective education tied to a 19 labeling campaign such as that. 20 DR. KVENBERG: Are there any additional 21 questions? 22 MS. FRYE: Cary Frye, International Dairy 23 Foods Association. I just have two quick, fairly 24 simple, questions. One has to do with the first 25 action item about guidance for processors. You 153 1 alluded to that, within the next few months, 2 guidance would be coming out. Also, there would be 3 specific guidance to enhancing the safety and 4 production of milk and milk products. 5 I just wondered if you could comment at 6 this time. Will that be coming out as an interim 7 guidance that will allow for comments and industry 8 input because I think the industry does have a lot 9 of valuable experience in this area and I would 10 urge you to work together with the industry on this 11 guidance. 12 So I don't know if you can comment at this 13 time. 14 DR. KVENBERG: I can certainly take it 15 from my standpoint. Others can comment as well. I 16 think that we are at such a stage, and your remarks 17 are well-taken. I think that cooperation on 18 guidance putting forward on where we are and where 19 the risk assessment takes us on dairy. That is 20 what you are speaking to in general. We are well 21 along in the general Guidance Document. It would 22 be welcome. Dialogue is always good on how we would 23 move ahead with such guidance. I think we are at 24 much earlier stages on specific guidances that were 25 identified as a result of conducting a risk 154 1 assessment on both fresh-cut produce and on dairy 2 products that it would be useful to collect 3 information in the process of doing such guidance. 4 We are in the early phases. 5 MS. FRYE: On the general guidance, will 6 that come out as a final guidance or will there be 7 a comment period that will be allowed? 8 DR. KVENBERG: I don't know. Does anyone 9 know? I don't have the answer to it. Normally, I 10 guess, when we publish guidance--it is Level 1. 11 MS. FRYE: It is something to consider, I 12 guess. The other has to do with technical 13 assistance for the industry. You talked about 14 having training that would be for regulatory and 15 industry and then, specifically, about training 16 modules for Listeria monocytogenes that would be 17 maybe satellite-based computer-distance training. 18 I know that ORA University has expanded 19 and had quite a bit of resources to expand the 20 training. But it was my understanding, at least in 21 the milk program, that ORA University training 22 through the satellite base and computer was only 23 available to state regulators. 24 So I would urge that this training be 25 supported and I would urge that it also be readily 155 1 available for the industry because the industry 2 could benefit from using that at many different 3 levels. They could take it to the employee level, 4 if it were appropriate, or the management level. 5 So I am hoping that it is a format--or maybe you 6 could talk about what kind of format is planned. 7 Thank you. 8 DR. GERMAN: The courses on ORAU are made 9 available to the FDA and to the state and local 10 folks, just like you are describing. But they are 11 also available to industry. They are available 12 through our partner. We have a business partner up 13 in New Jersey and the way we have created this 14 university is through a cooperative research and 15 development agreement. 16 So, in fact, those courses are available 17 to industry. So that is one tool in the toolbox 18 through ORAU. After we hear this total package of 19 all the issues and the things that we have to 20 deliver, we have a lot of courses going on. We 21 have initiatives with the Milk Shippers Conference. 22 We have things going on with the Conference of Food 23 Protection. We have things going on with APTO, 24 with all these associations. 25 We will include--I guess we will tweak the 156 1 existing training courses to highlight and to raise 2 the awareness of the Listeria concern here. But 3 then, also, we talked about having this auditing 4 process. 5 So my answer to you is that we don't know 6 what all the training programs will look like, but 7 getting them available to you in the same format or 8 tools, positing it on the web so you can pull 9 things down so you can use it and we can use it, we 10 are right there with you. 11 DR. KVENBERG: If there are no additional 12 questions--one question. 13 MS. HOLLINGSWORTH: Now that I have given 14 you time to think about my comments, I thought I 15 would come back and ask a question. In the 16 Executive Summary of the Action Plan, you talk 17 about the need for targeted actions, and that is 18 really where we are thinking at retail is through 19 targeted actions. But, within the plan, itself, I 20 believe there are 49 or 50 bullet points. That is 21 just from FDA and CDC and then we also have a 22 series of action items and bullet points presented 23 by USDA. 24 The question is, is there going to be a 25 process to go through all of those proposed or 157 1 potential actions and try to prioritize them or 2 identify which ones are likely to give the most 3 benefit, which ones are the most likely to be 4 measurable, or is it going to be sort of this 5 "attack the whole world" with all 50 different 6 actions? 7 We are looking to try to find a way to 8 focus those actions from the 50 to the ones that 9 will have the most impact and we are wondering if 10 there is any plan or any discussion to look at 11 those 50 and try to focus on them. 12 DR. KVENBERG: If there are no comments 13 from the panel, I guess I would address it. Number 14 one is--I think it will go through the evaluation 15 process, quite frankly. The Action Plan, at least 16 from FDA and CDC's perspective, is a continuation 17 of ongoing activities. 18 We feel some of the--at least if you 19 followed me through the "what's new" sections, 20 perhaps, I think that was our thinking relative to 21 those action areas where we are now believing that 22 a pointed focus makes sense, target certain 23 activities such as specific guidances, such as 24 shifting priorities and inspectional activities 25 such as targeting educational messages to Hispanic 158 1 women relative to avoiding certain kinds of cheese 2 that came from unpasteurized milk. 3 We kind of delineated in this updated 4 Action Plan a combing, if you will, and that is a 5 refinement from the Action Plan of 2001. 6 The second part of your question was are 7 we going to evaluate. I think we have to go--these 8 are ongoing processes in these bullets that you 9 have seen. I am a firm believer in evaluation of 10 where we are. Number one, I am really hoping we 11 are going to make that goal of 50 percent by the 12 end of 2005. 13 At the same time, I think, as I said 14 earlier, it behooves us to go through the process 15 of evaluating what worked and what didn't work. 16 These are the ongoing activities that we have 17 identified. 18 Bob, do you have a comment you would like 19 to put forward on the Action Plan. 20 DR. BUCHANAN: That we go through, each 21 year--in fact we just finished our end of that 22 process for this past year--a process of publicly 23 announcing what our tentative priorities are for 24 each year. We go through a rather rigorous 25 informal process of evaluating all of the things 159 1 that are on our plates and what we think or the 2 priority items for each year. 3 At the end of the year, we then put out a 4 report card on what we actually accomplished. As 5 part of that process, we do seek public input on 6 where we should be going in terms of our 7 priorities. So I would recommend strongly, for 8 those of you that, one, feel that Listeria should 9 be considered high on our priority list, consider 10 those comments and if there is specific action, 11 there is a mechanism of identifying that for us as 12 we set our next-year's processes and priorities 13 into place. 14 DR. KVENBERG: Seeing no additional 15 questions at this time, I would like to do too 16 things; number one, thank the panel. And I would 17 like to get a chair and join them and turn the 18 podium back to Don Kautter for the Comment Period 19 of the program. 20 Thank you, again. Thank you, panel, very 21 much. 22 Public Comments 23 MR. KAUTTER: To conclude this public 24 meeting, we have four individuals who have signed 25 up to give public comment. I have them listed. If 160 1 you guys could just go in that order. Try to keep 2 your comments to under ten minutes. 3 I have got Michelle Marcotte going first. 4 Then Jenny Scott. Caroline Smith DeWaal. And then 5 Jill Hollingsworth, if she has anything--you are 6 done? You are done, too? Okay. 7 Why don't we start with Jenny and then we 8 can move on to Caroline. 9 MS. SCOTT: I have lots of comments to 10 make here. I have good things to say so you are 11 not going to cut me off, I hope, Don. First of 12 all, I think you all know that NFPA has long been 13 an advocate of using sound science as a basis for 14 policy decisions by FDA and FSIS. So we think that 15 the Risk Assessment is an appropriate tool for 16 prioritizing resources and focusing attention on 17 those areas where there is evidence there is a true 18 risk. 19 So I want to compliment the agencies on 20 conducting a Risk Assessment with such a broad 21 scope. Actually, this was 23 separate risk 22 assessments. Industry does recognize the magnitude 23 of the effort that went into this. 24 We think we have an excellent step forward 25 here in the process of assessing the risk for 161 1 microorganisms in foods and I think we have 2 consolidated a wealth of information here on 3 Listeria monocytogenes. It does give us a 4 tremendous insight into the appropriate control 5 measures that we can apply and in areas for future 6 risk assessments. 7 We are glad to see that the FDA is going 8 to do these product-pathway simulations. We would 9 like to offer the industry's assistance as they go 10 forward there in providing whatever data we can and 11 helping you make sure that the pathways that you 12 are using are those that are actually applied by 13 the industry today. 14 We are pleased that NFTA was able to 15 contribute significant data to improve the Draft 16 Risk Assessment. As I say, we do want to continue 17 to help you as you go forward. 18 Bob Listed the five broad factors that can 19 impact on reducing risk. One of those and, 20 probably, I think one of the most important ones is 21 the potential for the food to support growth. The 22 two others, of course, were refrigerated storage 23 temperature and the amount of refrigerated storage 24 time. Of course, those last two are only factors 25 if the food supports growth. 162 1 One area that did not get incorporated 2 into the Risk Assessment that Lynn Kosty touched on 3 and that is the risk reduction due to the 4 incorporation of inhibitors such as lactates and 5 di-acetates in hot dogs and deli meats. A ranking 6 of deli meats and non-reheated frankfurters is very 7 high in this Risk Assessment may be true for those 8 products that support growth but it is not the case 9 for those that do not. 10 I think it is very important that we make 11 clear that many deli meats and hot dogs are now 12 formulated with inhibitors that minimize the growth 13 and hence significantly reduce the risk from these 14 products. 15 This Risk Assessment, because it did 16 address that, can be used as a benchmark so that we 17 can compare the risk-reduction impact of industry's 18 proactive efforts in reformulating products. Of 19 course, reformulation was listed as one of the 20 control strategies that could be used in reducing 21 listeriosis. 22 Another control strategy is to encourage 23 consumers to keep refrigerator temperatures at 40 24 degrees Fahrenheit or below and for retailers and 25 distributors to maintain foods in the high-risk 163 1 category as cold as possible. 2 For many years, NFPA has focused on having 3 the agency try and address their resources towards 4 those foods that support growth of Listeria 5 monocytogenes. So we are glad to see that, with 6 this revised Action Plan, this is where the focus 7 is going. 8 The Risk Assessment does indicate that 9 most cases of listeriosis result from consuming 10 high levels of Listeria monocytogenes from foods 11 that permit growth. So it is not surprising that 12 the Risk Assessment indicated the importance of 13 refrigeration storage and temperature. But what I 14 think is really surprising from this is the 15 magnitude of that effect. Look at the results that 16 Bob presented earlier, that, by keeping the 17 refrigerator temperature at 45 degrees Fahrenheit, 18 he can reduce the numbers of cases of listeriosis 19 by 69 percent. 20 If you go further and get those 21 temperatures down to 41 degrees constantly, then 22 the number of cases could be reduced by over 98 23 percent. One control strategy alone can have that 24 much of an impact. I think that is really amazing. 25 For foods that support growth above some 164 1 minimum concentration, the risk is largely 2 determined by growth that occurs subsequent to 3 purchase. That is indicated in the Risk 4 Assessment. Now, while we are not trying to shift 5 the blame or the burden for preventing listeriosis 6 to the consumer alone, I think this Risk Assessment 7 serves to underscore that prevention and control 8 are most effective if the entire food chain is 9 involved and that the consumer can play a highly 10 significant role in reducing risk. 11 I want to touch a little bit upon the 12 impact of numbers of Listeria in listeriosis. NFTA 13 did this retail survey. We did a risk assessment 14 based on the results and we demonstrated that 15 control strategies that reduce the level of the 16 Listeria monocytogenes in a ready-to-eat food can 17 have a greater impact on public health in reducing 18 the prevalence in all ready-to-eat foods. 19 In this Risk Assessment, if you go through 20 that big tome there, you will find the discussion 21 about reducing numbers of Listeria. Despite 22 ten-fold increases in contamination levels, there 23 were only small changes in risk, reducing levels 24 from 103 or 102 or even down to the 100 CFUs per 25 gram didn't reduce the risk very much. This is 165 1 primarily a consequence of the fact that these 2 products would support growth and the numbers would 3 subsequently grow when the consumer took it home. 4 FDA implied, or stated that this implied 5 that the foods that support growth of LM for these 6 foods reducing contamination to some specified 7 level but not zero was not adequate by itself to 8 control the risk of listeriosis. I think it is 9 very important to note that this also implies that, 10 in foods that do not support growth, there is 11 little public-health benefit in reducing the levels 12 of Listeria monocytogenes when they are low. 13 The Risk Assessment demonstrated the 14 effectiveness of heating and reducing risk. I 15 think that this is very important to recognize that 16 foods that may be considered ready-to-eat but they 17 are still going to undergo some heating step by the 18 consumer really aren't products that have a high 19 risk. If it is a product in which growth is 20 controlled or limited such as by freezing and then 21 reheated, the risk is virtually nil. 22 So we have recommended that the agencies 23 prioritize their risk-management strategies for 24 food based on whether or not the food supports the 25 growth of Listeria monocytogenes and whether or not 166 1 it will be stored refrigerated for an extended 2 period of time. This is something we have been 3 saying for a number of years and I think that this 4 is where we are going now with the recent Action 5 Plan. 6 For low and very-low-risk products, there 7 is really no need for industry or the government to 8 expend valuable resources in extensive control 9 measures to further reduce the contamination level 10 and there is really not much need to sample those 11 products unless you go into a facility and you find 12 that they are not controlling Listeria in the 13 environment. Then I think it would be appropriate 14 to focus some attention on foods that don't support 15 growth. 16 So we are really glad to see that the 17 risk-management strategy is going to focus 18 inspectional resources away from categories that 19 were identified in the Risk Assessment as being low 20 and very low risk. 21 We support the guidance that you are 22 developing for industry and I think it is very 23 important that you work with industry if you 24 develop this guidance. That is going to make sure 25 that it is more effective. 167 1 We are certainly happy to be a part of the 2 project where we are devising the Listeria control 3 strategies for smoked-seafood operations. We are 4 glad to see that in the Action Plan because I think 5 now it means that project is likely to go forward. 6 We have been wondering about that for a while. 7 So I want to finish by, once again, 8 offering kudos to the Risk Assessment Team. I want 9 to thank them for listening to our concerns when 10 the draft came out supporting data gathering to 11 fill in the gaps and to approve the final product. 12 I think that now we have a lot more information on 13 Listeria and how to control it. I think we are 14 going to be able to get to that goal of 0.25 cases 15 per 100,000 by 2005. 16 Thank you. 17 MR. KAUTTER: Thanks, Jenny. 18 MS. SMITH DEWAAL: Does this mean I get 19 the last word? This morning, I felt like we came 20 out and you rolled out the Cadillac of risk 21 assessments. It is beautiful. It is polished. I 22 am sure I could poke a few holes in it but, 23 generally, it really is an impressive effort after 24 years of work. 25 This afternoon, I thought we came from 168 1 lunch and you rolled out the new Bug of risk 2 management. It is cute. I like the flower holder 3 but I am not sure I trust the safety of my family 4 in it. That is my concern as I look at what you 5 are talking about this afternoon. 6 I will start talking about guidance 7 documents. I know they are popular, especially in 8 this Bush Administration. They are quick. They 9 are easy. They are embraced, not surprisingly, by 10 the industry and the OMB. But they have not 11 minimum level of consumer protection. Simply 12 having a guidance document doesn't do much for me. 13 It doesn't do much to protect my family, my parents 14 or the consumers that I represent. 15 You have no minimum level of protection. 16 If you walk in a plant and they are not abiding by 17 your guidance, so what? You can't close them down. 18 You can't stop them from producing food. You can't 19 stop them from selling products and you are not 20 testing a whole lot of product as it is. 21 So I think this concept of moving forward 22 with guidance documents, though clearly popular in 23 this room of industry trade representatives, is not 24 giving consumers what they need. 25 I want to note, also, that the 169 1 guidance--before USDA published its final Interim 2 Listeria Regulation, they had guidance in place. 3 They had had guidance in place for a number of 4 years and it was not sufficient to prevent the 5 Northeastern Listeria monocytogenes outbreak of 6 about fifteen, eighteen months ago. 7 So, having guidance in place doesn't 8 prevent the kind of outbreaks that I know you all 9 want to avoid. So I would like to urge FDA--get 10 your guidance documents done but you need to move 11 forward with regulations or potentially other 12 steps, additions, for example, to your seafood 13 HACCP regulations, to require high-risk processes 14 to test both their plant environments and their 15 products for Listeria monocytogenes. 16 It is not just the government's job to 17 find the problem. It is critically important that 18 the industry have a role and that they be testing 19 their own plants and their own products and, when 20 you come in to do your HACCP inspections, they can 21 give you that data and show you what they are 22 doing. 23 Prevention is going to be more effective 24 can consumer education. I am all for consumer 25 education but if we can prevent the products from 170 1 getting out the door with Listeria monocytogenes, 2 we can prevent a lot of problems. So I would urge 3 you, in your regulatory agenda, to move forward 4 with regulations or other steps that would put 5 these testing programs in place for the industry. 6 USDA has already done this--not well enough, I 7 might add. But they have put a mandate in place. 8 I think the industry and us are still 9 trying to figure out what it means, but the 10 industry is supposed to be testing their products. 11 I will move on to the consumer messages 12 which is another big part of your Action Plan. It 13 is critical they be clear and direct. I was 14 actually impressed with what Dan Engeljohn put up, 15 the "do not eat" messages used by FSIS. Those are 16 clear and direct. I didn't have a whole lot to 17 pick apart on those. 18 But what about smoked seafood? Why are we 19 telling consumers, pregnant women and elderly 20 consumers, similar messages about smoked seafood. 21 The second element, aside from clear and direct, is 22 that these messages need to be easy to find. 23 Now, when FDA published a regulation a 24 couple of years ago on unpasteurized juices, they 25 didn't rely on consumer education to tell 171 1 consumers. They put a message right on the label 2 of unpasteurized juice bottles. When FDA published 3 a regulation dealing with eggs, they had a 4 safe-handling label on the egg carton. Now, it is 5 not big enough to read, but it is supposed to be 6 there. 7 Similarly, with FSIS. They have had 8 safe-handling labels on meat. They don't just rely 9 on consumers to get the message through their 10 doctors or through their women's magazines or 11 through their AARP membership. 12 Food labeling is the best way to reach 13 consumers with safety messages. So we would also 14 urge you to consider the use of labeling, which you 15 have done with unpasteurized juices and you have 16 done with eggs and with meat products, generally, 17 to get this message out. 18 We respect any consumer education. I 19 agree with the speaker earlier who urged you to 20 measure the effectiveness. If you devote a whole 21 lot of resources to doctors only to find out, as 22 IPIC did when the did their surveys, that no matter 23 what you give doctors, they are not going to spend 24 their ten minutes with the patient talking about 25 what foods they shouldn't eat. 172 1 Then all that work towards educating the 2 doctors isn't actually getting to the people who it 3 needs to get to. So you need to be measuring the 4 effectiveness of your efforts. 5 Thank you. 6 MR. KAUTTER: Thank you, Caroline. 7 With that, we have come to the conclusion 8 of the public meeting. I want to thank everybody, 9 all the presenters, the panel members. Everyone 10 have a nice evening and holiday season. 11 (Whereupon, at 3:00 p.m., the meeting was 12 adjourned.) 13 - - -