Managing Food Safety Practices from Farm to Table: Workshop Summary (2009)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

2 Recent Outbreaks in Food Products: Lessons Learned from Past Experience Session moderator Henry Chin opened this first session of the work- shop, Recent Outbreaks in Various Food Products: Lessons Learned, by reiterating a point that Taylor had made during his keynote presentation: outbreaks, like the Salmonella Saintpaul outbreak, provide an opportunity to learn and improve food safety. This session focused not only on past outbreaks but also other experiences in a range of foods (from minimally processed raw produce to highly thermally processed canned goods), and under a range of regulatory frameworks (from the use of GAPs to the use of the very tight low acid can food regulations), as well as lessons learned from research on consumer behavior. This chapter provides summaries of the four presentations in the session and the discussion that followed. The session began with Natural Selection Foods’ Will Daniels describ- ing the sequence of events during and following the 2006 E. coli O157:H7 outbreak in bagged spinach (which traced back to spinach packed for Dole by California-based Natural Selection Foods, best known for its Earth- bound Farm brand of organic salads and produce). One of the key messages of his talk, Next Generation Food Safety in Fresh Produce: An Industry Perspective, was that while Natural Selection Foods has significantly im- proved its food safety system since the outbreak, there are still a multitude of food safety problems that can arise after a product leaves the farm and which are not being appropriately addressed. Stakeholders operating at all points along the farm-to-table continuum need to bear responsibility. Dan-   Henry Chin, PhD, is Senior Director of Scientific and Regulatory Affairs at the Coca-Cola Company, Atlanta, GA. 15 

16 MANAGING FOOD SAFETY PRACTICES FROM FARM TO TABLE iels reiterated one of Taylor’s keynote presentation points: that too many institutions involved with food safety are operating with a “stovepipe” mentality and that there needs to be more communication and collabora- tion, particularly among institutions operating at different points along the farm-to-table continuum. Daniels commented on the role of audits, arguing that while audits serve an important verification and validation purpose, they are not preventive measures and therefore should not be treated as such. Buyers (including distributors, retailers, and consumers) need to bear some preventive responsibility as well. Daniels also commented on the importance of food safety science and the need to generate trust between industry and regulators so that company-generated data could be shared and utilized in the effort to improve food safety. Daniels’s talk was followed with a presentation on Risk Management for Thermally-Processed Foods by Donald Zink of the FDA’s Center for Food Safety and Applied Nutrition (CFSAN). Zink used thermally pro- cessed canned foods to illustrate what could be accomplished with food safety when “all of the pieces are in place.” Canned foods are among the safest processed foods for three reasons: the science is complete, the packag- ing is well-developed, and consumers are well-educated. Zink identified the first of these factors—the science—as the most important, suggesting that most other processed foods are not as safe as canned foods because the sci- ence is still lacking. Not only do we not have a complete understanding of how contamination occurs in many cases, Zink argued, we do not always have the tools necessary for interrupting those contamination events. In the third presentation of this session, Lessons Learned in the Meat Industry: Control of Listeria in RTE Meat and Poultry Products, Randall Huffman, President of the American Meat Institute (AMI) Foundation pro- vided an overview of the history of Listeria control in ready-to-eat (RTE) meat products and the U.S. meat industry and U.S. governmental responses. He described the AMI’s recognition of a “Cycle of Control,” a four-stage cycle involving (1) awareness and detection of the problem/pathogen, (2) enlightenment and the beginning of an understanding of the problem/patho- gen, (3) prevention and the implementation of interventions and, finally, (4) predictive measurement of the impact of those interventions and continued learning about which interventions are most effective. Currently, with re- spect to Listeria control in RTE meat products, the U.S. meat industry is in the final state of the cycle, with the number of Listeria recalls and the preva- lence of Listeria in RTE meat products both showing significant downward trends. Huffman described in detail many of the specific steps that the meat industry has taken over these past 20 years in its effort to control Listeria contamination, emphasizing the critically important role of using science and data to inform decision making. In the fourth and final presentation of the session, Consumer Behav- 

RECENT OUTBREAKS IN FOOD PRODUCTS 17 ior in Managing Food Safety Risks, Christine Bruhn of the University of California, Davis, described results of recent research on consumer behav- ior around safe food handling. Most of the workshop discussion up until this point focused on industry activities in food safety and the challenges. Bruhn’s presentation was the only presentation of the day that focused exclusively on the consumer population, specifically on what research has elucidated about how consumers manage, or do not manage, food safety issues. A key message of her presentation was that even when consumers know what constitutes safe food handling or behavior, they do not always adhere to the recommendations. Therefore, while consumers do have a re- sponsibility to be aware of food safety risks and to take appropriate actions, the U.S. food safety system cannot rely on consumer education alone. The focus should be on making food as safe as possible. Next Generation Food Safety in Fresh Produce: An Industry Perspective Presenter: Will Daniels Will Daniels began his presentation with a brief description of Natural Selection Foods. With its Earthbound Farm brand, it is the Nation’s larg- est grower, packer, and shipper of organic produce, with products in 80 percent of all grocery stores across the country. The company produces a total of about 2.2 million pounds of leafy green fresh salads every week and distributes its produce internationally and as far away as Asia. On what Daniels described as a “fateful” September 14, 2006, the California Department of Health Services (CDHS) informed Natural Se- lection Foods that the company had been implicated in a nationwide out- break associated with fresh cut bagged spinach. The next day, with little to no information except that it had been implicated (among several other companies at that point) Natural Selection Foods issued a voluntary recall. Prior to September 14, the government had told the company only that the leafy greens industry in general was on high alert and that the government planned to act early and quickly in the event of any foodborne illness asso- ciated with bagged greens. Indeed, they acted quickly. When Natural Selec- tion Foods was implicated, thousands of samples were collected in both the field and the production facility as part of both government and company investigations. Matches were traced to a field that had supplied spinach on the production day associated with consumer infections. However, the source of the contamination was about a mile downwind of the field, with   Will Daniels, is President of Quality, Food Safety and Organic Integrity at Natural Selec- tion Foods, San Juan Bautista, CA. 

18 MANAGING FOOD SAFETY PRACTICES FROM FARM TO TABLE Figure 2-1  Natural Selection Foods’ new multi-hurdle approach to food safety. no clear indication of how the contamination had reached the field itself. Not having that crucial information and without a kill step in their process, 2-1 fixed image Natural Selection Foods was forced to examine the entire spectrum of food safety risk(s) and begin developing a multi-hurdle approach (Figure 2-1) to enhancing food safety which, today, represents a significant improvement in Natural Selection Foods’ safety management system. Daniels provided an overview of five of the major programs that Natu- ral Selection Foods has established over these past two years as part of its new multi-hurdle approach: •  eed to Harvest: Natural Selection Foods has developed a new plan S for enhanced GAPs, including: o onducting pre-season ranch assessments of the history, to- c pography, adjacent land use, and other relevant features of all potential ranches;  This is in addition to the 2007 California Leafy Green Products Handler Marketing Agreement. 

RECENT OUTBREAKS IN FOOD PRODUCTS 19 o  testing all seeds, fertilizers, water and other inputs for O157:H7 enterohemorrhagic E. coli (EHEC) and Salmonella and pre- venting all inputs from entering the field without certification that they have tested negative; o uring planting, testing water at shorter intervals than required d by the California Leafy Green Products Handler Marketing Agreement; and o racticing regular GAP harvest audits and ensuring compliance p in workers’ practices. •  aw Material Firewall: The company instituted a raw product R test and hold program about two weeks after the September 14, 2006, outbreak, whereby 100 percent of all leafy greens are now sampled from and prevented from entering the process stream un- til negative results are achieved. In addition to testing for EHEC and Salmonella, Natural Selection Foods recently added a Shigella screen. •  n the Plant: The company has made several enhancements in its I packing facilities, for example the use of laser sorters in an effort to reduce foreign materials. The laser sorters identify materials that do not contain chlorophyll and reject them from the process stream. This not only allows the wash systems to focus on the salad itself and not extraneous material, it also reduces consumer complaints about foreign materials in their salads. During the first month of having laser sorter across all processing lines, Earthbound Farm saw a 70 percent reduction in such complaints, which was remark- able given that foreign material complaints comprise about half of all complaints. Now, mostly just weeds still pass through. Ad- ditionally, in the wash system, the company has added dunk reels for increased contact time, enhanced filtration time, and added an extra wash step. •  inished Product Firewall: The company added a finished good test F and hold program in February 2007, whereby all finished goods are sampled for select pathogens and not released until negative results are achieved. •  aining Deeper Knowledge: Natural Selection Foods is now us- G ing its test data (e.g., from its Raw Material Firewall and Finished Product Firewall programs) to develop a deeper understanding of what is needed to prevent outbreaks. Importantly, however, Dan- iels emphasized that testing is not the answer to food safety. The Raw Material and Finished Product firewalls, for example, are not intended to provide a solution to food safety, rather as a means to “get to the answers” more quickly. 

20 MANAGING FOOD SAFETY PRACTICES FROM FARM TO TABLE While this multi-hurdle approach to food safety represents a fantastic step forward for Natural Selection Foods, Daniels emphasized that it does not address the multitude of problems that arise after a product leaves the farm and processing facilities—that is, during distribution, storage, retail- ing and consumption. Food safety does not stop at the farm, nor does it stop in the processing facility. In fact, what happens after products leave the farm probably amounts to more than 50 percent of the food safety continuum, Daniels said, which raises questions about what additional hurdles could be put in place further downstream with the ultimate goal of consumer protection. Daniels reiterated one of Taylor’s key messages: that “real collaboration” among different stakeholders responsible for food safety at various points along this food-to-table continuum is lacking. Too many institutions are still operating like “stovepipes.” In order to achieve progress and reduce risks, these different sectors need to start collaborating. He argued, “Collaboration, rather than competition, among those striving for improved food safety could achieve real results.” Right now, audits serve as one of the primary mechanisms for com- munication between different segments of the farm-to-table spectrum (e.g., buyers rely on audits to guarantee that what they are purchasing is safe). But as Daniels asked, “Can audits ensure safety?” He argued that while food safety audits serve as good verification that operators have func- tional food safety programs, it is really up to the individual operator to ensure that their food safety systems are functioning well. Likewise, fur- ther downstream, individual distributors, retailers and consumers need to be responsible for food safety at their respective positions along the food safety continuum. Audits, like other tests, are not preventive measures. Not only are audits not preventive, they are costly. Natural Selection Foods goes through multiple audits yearly, often repetitively for different buyers. Mul- tiple audits cost resources and time, which may be better spent on finding new answers to food safety issues. Daniels remarked that the same “stovepipe” mentality is at work among research institutions as well. While several universities and trade as- sociations are devoting many resources in an attempt to solve some of these issues, many of these efforts are duplicated. More national-level coordina- tion is needed to reduce the redundancy so that we can find answers more quickly and effectively. Moreover, most of the research is being conducted on foodborne illness in the United States, even though many of these issues are not unique to this country. Nor are they unique to the fresh produce industry. Research results often have more wide-ranging implications than are realized. Finally, Daniels commented on the large quantity of industry-generated data and how these data have been used against companies in the past (e.g., to implicate a company in an outbreak with which they might not have 

RECENT OUTBREAKS IN FOOD PRODUCTS 21 Figure 2-2  Initial reactive (IR) and molecular confirmation (MC) data presented by Daniels. Daniels emphasized the spike in contamination events during the sum- mer months (i.e., when operations are moved to San Juan Bautista [SJB], California, from Yuma, Arizona). otherwise been2-2 fixed image Figure associated). Government needs to take the lead in creating a trusted, safe environment where data sharing and research collabora- tion would quicken the pace of concrete, applicable advancements in food safety. “I would love it if we could have an opportunity to share more of the data that we’ve generated through this Test and Hold Program,” Daniels said. “I think that there could be efforts made to protect that data and al- low it to be used early on.” Daniels then gave two examples of the type of safety data being gener- ated at Natural Selection Foods: 1.  shown in Figure 2-2, a spike in both initial reactives (Raw IR) As and molecular confirmation (Raw MC) positives of raw product during the summer months, when operations are conducted in the Salinas, California, area (as opposed to the winter months, when operations are moved down to Yuma, Arizona). Daniels remarked that this spike raises questions about whether the company should be spending more of their test and hold monies during the summer months.   Daniels did not comment on the other data spikes (e.g., the IR spikes in November 2007 and December 2007 and the MC spike in August 2007). 

22 MANAGING FOOD SAFETY PRACTICES FROM FARM TO TABLE Figure 2-3  Data presented by Daniels showing the difference between contami- nation incidence (number of molecular confirmation positives [MC Incidents]) and prevalence (the number of positives per million of pounds received [MC/M Lbs Rec]). Figure 2-3 fixed image 2.  shown in Figure 2-3, a low level of molecular confirmation As positives of organic spinach when measured per million of pounds received, or MC/M Lbs Rec (as opposed to molecular confirmation, or MC, incidents). Daniels explained that the low normalized level for spinach is important given that spinach is Natural Selection Foods’ single largest individual ingredient, comprising about 30 percent of the company’s received products (i.e., if spinach was not such a large volume product, the high number of incidents, without taking into account volume, would be more alarming than it is). Daniels also pointed to the high rates for organic cilantro, parsley, and dill, noting that herbs are known to carry a higher risk and that the company has taken additional measures to protect them. He said that the spike for conventional freesia was probably an anomaly and that it was a very low-volume item at the time. When these data were generated, Natural Selection Foods was a mixed operation (i.e., mixed organically and conventionally grown prod- ucts). Today, the company is 100 percent organic. Daniels then moved on to the topic of government transparency, com- menting on how an FDA official had expressed “surprise” early on during FDA’s investigation of the more recent Salmonella Saintpaul outbreak that Florida tomatoes could end up on the West coast. Upon hearing that re- mark, it dawned on Daniels that the food safety continuum needs not only 

RECENT OUTBREAKS IN FOOD PRODUCTS 23 better collaboration but also greater transparency. In particular, government needs an industry group that they can really turn to for quick answers dur- ing outbreak situations in order to relieve some of the extreme pressure they are under to protect consumers. Too often, outbreaks generate mis- information and media frenzy, which in turn, cause a decline in consumer confidence. This type of misinformation also often leads to the allocation of resources to the wrong places. Having a mechanism in place for government agencies to find quick answers would also alleviate some of the pressure that industry is under during outbreak situations. During the Salmonella Saintpaul outbreak, Natural Selection Foods spent an inordinate amount of time reassuring customers that they were distributing neither tomatoes, cilantro, nor jalapeño peppers. Likewise, efforts need to be directed toward helping industry understand government actions and processes better (e.g., during tracebacks). Daniels concluded by posing the question, “How do we move for- ward?” He highlighted four necessary steps: 1. Recognize that pathogens exist in our environment and that, in fact our food safety systems control them “99.9 percent of the time.” Natural Selection Foods processed fresh produce for 24 years with- out an incident. 2. Realize that we are all in this together and that one segment’s ef- forts alone are not going to solve the problem or reduce consumer concern. 3. Develop national standards and standardized audits (with one audit verifying compliance for all buyers). Standardizing does not mean adopting a one-size-fits-all approach, rather implement- ing good agricultural and manufacturing processes and having a hazard analysis and food safety plan in place for that particular operation. 4. Collaborate! Companies need to work together as an industry to make our produce safer; they need to create a secure environment where research efforts can be coordinated, applied, and improved; industry and government need to work together such that coor- dinated efforts can result in quicker conclusions; and we need to develop better trust by establishing transparency and an open dialogue between industry and government. 

24 MANAGING FOOD SAFETY PRACTICES FROM FARM TO TABLE Risk Management for Thermally Processed Foods Presenter: Donald Zink Don Zink began by commenting that thermally processed foods are essentially and technically “completely safe” and, as such, illustrates what can be accomplished with respect to food safety and risk management when “all of the pieces are in place.” When failures occur, they are due to mechanical or human errors, not incomplete knowledge about how to manage the risks. Zink clarified that he would be focusing primarily on commercially and completely sterile products, which comprise just one end of a range of thermally processed foods. Complete sterility is the more extreme: It involves killing even the most heat-resistant thermophilic organisms. Com- mercial sterilization is not as extreme: It does not necessarily mean that there are no viable microorganisms in the canned product and, in fact, some thermophilic microorganisms usually survive the commercial sterilization process. Rather, it simply means that under the intended conditions of stor- age and consumer use, those microorganisms are no longer going to con- tinue to grow. Commercial sterility is the standard used today for low-acid canned foods. At the other end of the scale are blanched and pasteurized foods. Blanching is a very mild heat treatment primarily used in the frozen foods industry to fix color and inactivate enzymes that cause undesirable oxidation reactions; blanching may or may not kill some pathogens, de- pending on how it is done. Pasteurization is usually interpreted to mean that a heat or other form of treatment has been used to destroy harmful microorganisms. Completely sterile and commercially sterile foods arguably have the best safety record of any category of processed foods. It is difficult to put a number on their safety record, Zink said, given variability in safety and how foodborne disease is monitored around the world. That said the likeli- hood of contracting botulism or another serious foodborne illness from this type of product is probably less than 1 in 100 billion. Zink identified three major factors that contribute to the safety of canned foods: 1. The underlying science is defined and complete. While there are still some unknown things, we do understand and have understood for a long time the nature of several critical processes: a. The physics of heat transfer. Methods to measure the rate of heating among various foods were developed more than   DonaldZink, PhD, is Acting Senior Science Advisor and Senior Food Scientist at the FDA’s Center for Food Safety and Applied Nutrition, College Park, MD. 

RECENT OUTBREAKS IN FOOD PRODUCTS 25 50 years ago, so we know how foods heat, how different foods heat differently (e.g., dried beans heat differently than diced carrots) and how heat is transferred into a container from a piece of processing equipment. We also have well- developed methods for verifying the adequacy and uniformity of heating. b. Thermal bacteriology. While there is always some worry that a new, heat-resistant “superbug” is going to emerge, we un- derstand the nature of the bacterial flora that exists now. We have identified all of the important target bacteria for heat processing, we have studied their heat resistance and inactiva- tion kinetics, and we have established target values for what kind of inactivation we must have in order to make products safe. Interestingly, all of the heat-resistant microbial organisms of public health concern are actually less heat-resistant than spoilage organisms, which means that foods need to be heated even more than necessary (from a heat processing perspective) in order to kill the spoilage organisms. c. The engineering of heat processing equipment. We know how to design systems that heat food, we have well-established standards for the design and construction of retorts, and we have excellent process control capabilities. d. Development of tools for accurately and continuously measur- ing the process. While new microwave and other technologies are being developed for this purpose, we have other tools avail- able that give us an excellent fundamental understanding of what is going on. In fact, the basic critical variables in thermal processing are relatively easy to measure and document (i.e., temperature, time, flow rate, pH, water activity [aw], viscosity, thermal diffusivity, heat and temperature distributions). 2.  We have well-developed packaging systems. Zink remarked that it doesn’t do any good to produce a safe food if you can’t keep it safe before it is in consumer hands. a. We have a very long history with the metal can and glass jar, with the former dating back to Napoleonic times. So the packaging technology is well worked out—so well worked out that it is possible to request from a vendor a certain type of container for a certain amount of material, longevity, thermal resistance, etc., and the vendor can meet that request. For ex- ample, if you wanted a metal can that lasted 20 years, we could do that. Today, most metal cans typically last up to about four years. Similarly, if you wanted a glass jar that could survive any amount of trauma, again we could do that, although it would 

26 MANAGING FOOD SAFETY PRACTICES FROM FARM TO TABLE be heavy and expensive. We have had substantial experience with flexible pouches and cardboard laminates as well. While they may not date back to Napoleonic times, these types of packages have been on the market for decades. b.  have well-established durability standards for all types of We common packaging, including standardized testing protocols that could be used to determine whether your product has a reasonable chance of holding up in the marketplace. 3.  have well-educated consumers. We a. Nearly every consumer is familiar with the metal can, with the vast majority of consumers understanding that swelling or leakage means that there is some kind of problem. Most con- sumers are also aware that a badly dented can is suspect. There is some concern, however, that there may be a downward trend in this area, with fewer consumers being aware of potential problems. Not only is there a lot of new packaging entering the marketplace (and taking the place of the metal can), there is less food safety education occurring in the home. b. That said, consumers seem to be transferring what they know about the metal can to some of these other forms of packag- ing. If a consumer sees a swollen pouch, for example, he or she might suspect that something is wrong. c. Either way—with both canned and other types of packaged foods—mechanical or human errors are to some extent caught by consumers. In conclusion, Zink re-emphasized that the main value of considering the food safety record of thermally processed foods is to contrast it with other types of foods (e.g., fresh produce), and identify and understand those components that give the former a much better safety record. One of the most important components is science. Technically, complete food safety requires a complete understanding of how all the variables contribute to potential contamination problems and a complete understanding of how to build the tools needed for interrupting those routes of contamination (as well as the tools needed for measuring and controlling the process). 

RECENT OUTBREAKS IN FOOD PRODUCTS 27 Lessons Learned in the Meat Industry: Control of Listeria in RTE Meat and Poultry Products Presenter: Randall Huffman “The battle with this organism [Listeria] has caused more change for producers of RTE deli meat products than any one single factor or event in the last 30 years. Our scars are numerous and deep.” —John Butts, Vice President of Research, Land O’Frost Randall Huffman began his presentation by remarking that he would be focusing on the control of Listeria monocytogenes in RTE meat prod- ucts (which, he remarked, fall within the mild heat/pasteurization category on Zink’s thermal heat treatment continuum). Over the last 30 years, the food industry has made remarkable strides in not only understanding the Listeria problem and how to control it but also implementing those con- trol measures in its large processing plants. Also in his opening remarks, Huffman reiterated what Taylor had emphasized in his keynote address: that industry plays the primary role in producing food and producing that food safely. After all, not only is selling safe food good for consumers, but it is also good for business. Efforts to control Listeria in RTE products began in the mid-1980s and early 1990s, following an outbreak in California associated with Mexican- style cheeses. Some additional major outbreaks in the late 1990s further emphasized the importance of Listeria control in the processing environ- ment. Following a major outbreak and large recall of sliced deli meats in 1998, the industry experienced a one percent reduction in sales. So not only was the implicated company hit, but the rest of the industry as well. This and other similar experiences not only led to a recognition around that time that this is indeed a serious problem, but it also prompted the indus- try to take a collaborative approach in an effort to solve the problem. At that time, the American Meat Institute (AMI) and other trade associations developed the industry’s first Listeria guidelines and best practices, which have evolved since then. The U.S. government contributed to the effort as well, beginning with its initiation of surveillance by the CDC and product sampling of RTE   Randall Huffman, PhD, is Chief Food Safety Officer of Maple Leaf Foods, Inc., To- ronto, CA. At the time of this presentation he was President of the American Meat Institute Foundation.   Huffman noted that while Listeria exists in the home, in our kitchens and refrigerators, and while some of what he would address during this workshop could be applied there, the focus of his presentation would be on what industry has done to control the problem in the large processing plants. 

28 MANAGING FOOD SAFETY PRACTICES FROM FARM TO TABLE Cycle of Control Awareness: Detect Problem Predictive: Enlightenment: Measure Collect data to impact of understand the interventions problem Preventative: Implement Preventative: Interventions ID appropriate and Best points of Practices control Share Best Practices across industry Figure 2-4  The four stages of the AMI’s “Cycle of Control,” a data-informed approach toward dealing with Listeria and other pathogens. Figure 2-4 meats by USDA/FSIS (both in the late 1980s). Later, the USDA/FSIS ratch- eted up both RTE product and plant testing, and the CDC initiated two new programs (PulseNet, a system for rapidly identifying large multi-state outbreaks; and FoodNet, an active foodborne illness surveillance system). In the early 2000s, the USDA and FDA completed a major risk assessment and found that deli meats and non-reheated hot dogs topped the list among high-risk foods with respect to listeriosis. More recently, FSIS has begun conducting not-for-cause L. monocytogenes verification testing, which in- volves very aggressive sampling of plant facilities. The American Meat Institute Foundation recognizes four stages of what it refers to as the “Cycle of Control” with respect to controlling Lis- teria and similar pathogens (see Figure 2-4): 1. Awareness and detection of the problem; 2.  nlightenment and the beginning of an understanding of the E problem; 3.  reventative identification of the best points of control in the pro- P cess and the implementation of interventions, along with a sharing of best practices across the industry; and 

RECENT OUTBREAKS IN FOOD PRODUCTS 29 4. Predictive measurement of the impact of these interventions and continued learning about which interventions are most effective. Huffman spent most of his presentation time elaborating on each of these stages of the “Cycle of Control” with respect to industry actions in response to Listeria in RTE meat products. He provided a list of actions taken by industry at each of the four stages: 1. Awareness in the early 1990s •  he meat industry recognized the environmental nature of T the problem (i.e., that the pathogen existed in the refrigerated environment of the processing facility) and started sampling not just products but also contact surfaces in the processing equipment that could potentially serve as growth niches for the bacteria. •  n important growth niche was discovered in the hollow roll- A ers on the conveyors transporting products. In fact, hollow rollers have since been identified as a major growth niche for other microbes as well (i.e., hollow rollers that are not disas- sembled, cleaned, and heat-treated or otherwise sanitized so as to eliminate bacteria and other pathogens). As it turns out, during pressure washing of the equipment every night, organic matter can be blown into the center of a hollow conveyor roller and create an opportunity for Listeria to become established and grow and subsequently potentially contaminate product. •  here was also a recognition of the benefits and needs for T separation of the RTE area from the non-RTE or raw area of the facility. As a preliminary solution, the processing facility floors were painted with yellow lines to keep workers on one side or the other. This approach of using lines to demarcate RTE from raw areas had limited effectiveness and was difficult to maintain over time. 2. Enlightenment in the mid-1990s •  quipment teardown and a “Seek & Destroy” approach to E positive test results became common practices. •  teps were taken to redesign equipment and re-engineer equip- S ment to eliminate areas of equipment that were potential growth niches, such as hollow rollers. As a preliminary solu- tion, this re-engineering was an internal process conducted by company maintenance personnel, not equipment suppliers. •  quipment suppliers were informed of design problems. Ini- E tially, this was the source of a lot of tension and confrontation, 

30 MANAGING FOOD SAFETY PRACTICES FROM FARM TO TABLE with equipment suppliers hearing many mixed messages from companies about what was needed. •  loor problems persisted. Today there is an expectation that F all drains are Listeria-free. Back then, the expectation was that all drains were a source of contamination. No method of floor cleaning seemed to work. •  ompanies started persistent deep cleaning of their equipment, C including parts that were not normally disassembled during routine cleaning. •  here was recognition that many facility areas, including the T walls and absorbent materials in refrigerator doors, were a source of contamination. As with the hollow roller, this is an area where equipment suppliers were not initially but would eventually become involved in the effort to re-engineer materi- als that would not harbor Listeria. •  here was a realization that mid-shift cleanups were contribut- T ing to, not helping, the problem. Back then, mid-shift cleanups were common practice—until we realized that bringing water into the processing environment during processing could result in the spreading of contamination. 3. Preventative phase of the late 1990s •  elated to the last bullet point, there was a realization that R keeping the floor and other parts of the processing environ- ment dry during processing gave a much better chance of controlling the bacteria. •  ooking/pasteurization of equipment became commonplace. C Today, many companies are actually applying heat or steam to their slicing and other stainless steel equipment for more effective sanitization in addition to regular cleaning and sani- tization. Depending on the type of equipment, companies are using steam injected under a tented area, or even placing the entire piece of equipment in a smokehouse. •  arge area sampling of food contact and environmental sur- L faces became commonplace (instead of the small grid sampling or the use of small swabs which had been regular practice previously). •  here was a growing understanding of how organisms spread T from growth niches to transfer points (e.g., worker hands or bins used on multiple lines) and how the latter could be used effectively as an indicator site in an environmental testing program. •  ompanies began engineering a physical separation of RTE C areas, rather than simply painting yellow lines on the floor. 

RECENT OUTBREAKS IN FOOD PRODUCTS 31 4. Predictive phase from 2000 until today •  MI began a series of best practices workshops, based on A the large amount work done in the industry to achieve con- sensus on how to control Listeria, with the first one held in June 2000. AMI has conducted 15 of these workshops over the years, reaching more than 1,000 industry employees. The workshops are organized around AMI’s six “Strategies for Control”: i.  Prevent Listeria growth in a niche or other site that can lead to RTE product contamination. ii. Implement appropriate post-lethality technology to elimi- nate, reduce, or prevent the growth of Listeria. iii. Implement a Listeria sampling plan to assess in a timely manner whether the processing area is “under control.” iv.  Respond to each positive product contact sample as rap- idly and effectively as possible. v.  Verify that the problem was corrected. vi.  Review and analyze data to ensure the Listeria control program is working. •  n October 2001, the AMI declared that food safety was a I non-competitive issue, which contributed to companies sharing information and best practices more widely. •  oday, the industry has Listeria under control in most large T processors and many mid- to small-size processors, with most companies able to identify and control specific growth niches and some companies using very aggressive sampling meth- ods and early warning systems. In fact, many companies are now focusing their sampling efforts in zone 4 (e.g., employee welfare areas), indicating that good control has already been achieved in the critical zones. Several companies have achieved new levels of control (e.g., one year without having a drain test positive). •  asteurization of slicing logs has become commonplace (i.e., P after a product goes through the normal thermal processing step, it is removed from its container and the surface is pasteur- ized again before going onto the slicing line). •  he use of DNA analysis as a way to identify sources of growth T niches and the degree of the diversity in strains in a given facil- ity has become more commonplace. •  he supplier industry is now using AMI’s set of 11 principles T for the sanitary design of equipment, not just during the en- gineering phase of their product development but also during marketing as a way to promote their equipment. When you 

32 MANAGING FOOD SAFETY PRACTICES FROM FARM TO TABLE open up a meat industry trade magazine, Huffman noted, you’ll see these ads. •  ikewise, companies responsible for building RTE processing L facilities utilize AMI’s set of 10 principles for sanitary facility design. •  actate and diacetate are recognized as ingredients useful in the L control of Listeria growth over the shelf life of a product. •  ccording to officials at the USDA/FSIS, there have been no A foodborne illness-related Listeria investigations in RTE meat products for six years. All recalls of RTE meat products for Listeria over the past five years have exclusively resulted from products testing positive after not being held by the manufac- turer prior to being shipped. •  he sanitation control of growth niches is much better under- T stood today (i.e., the necessary degree of equipment disassem- bly, chemical sanitizer treatment, hand scrubbing of contact surfaces, heat treatment, non-daily scheduled sanitation, and effective Good Manufacturing Practices (GMPs) after flooding of sanitizer). These efforts are very aggressive, time-consuming, and expensive, but they do work. •  inally, the nature of high-risk situations is understood much F better today than it was 30 years ago. Usually, when a product in the marketplace tests positive, one or more of these high- risk situations is at the root of the problem (which means that eliminating these is key to Listeria control): drain backup, the use of high pressure water or air on the floor or in a drain, movement or significant modification of a packaging line, an equipment breakdown, the interchangeable use of personnel between raw and cooked products, construction in or adja- cent to the cooked product area, a warm room, a wet area or process, a water-retaining crack in the floor or cleaning of equipment while on the floor. In addition to all of the above listed changes, new knowledge gained over the past few decades has also led to the “myth-busting” of several mis- conceptions about Listeria. Myth-busting, Huffman argued, is important in food safety. Otherwise, misconceptions become dogma and a lot of time and resources are wasted chasing insignificant issues. Huffman listed a few examples of misconceptions around Listeria control in RTE processing: •  yth: Listeria is airborne. Fact: Listeria aerosolizes and moves M around during high-pressure water cleaning, but it is not airborne. 

RECENT OUTBREAKS IN FOOD PRODUCTS 33 •  yth: All raw meat is positive for Listeria. Fact: Thermal process- M ing kills Listeria. •  yth: Listeria cannot be removed from the processing environment. M Fact: Listeria can be removed from the processing environment. •  yth: Drains will always be positive. Fact: This is no longer true. M In conclusion, Huffman summarized lessons learned from the meat industry’s experience with Listeria: •  e need to rely on science and data to inform our decision-making W and guide this “Cycle of Control.” • Sampling and testing must be used strategically and aggressively. •  igilant and constant re-evaluation of the risk management system V is critical. • Industry sharing of best practices is important. •  yth-busting is important. Use the data and avoid misconceptions. M • A flexible regulatory approach enables companies to do their job. The meat industry has put these lessons to good use. Huffman showed data on the number of Listeria recalls since 2003 and the prevalence of Listeria in RTE meat and poultry products since 1990. Since 2003, there have been 73 recalls, but none of these recalls has resulted from an illness investigation. Huffman noted that all have been due to products testing positive but nonetheless being shipped into commerce, a problem that the industry continues to work on (i.e., encouraging companies to wait to ship product until test results have been returned). With respect to the prevalence of contamination in RTE products, there has been tremendous progress since 1990, from about 4.5 percent positive test results in 1990 to about 0.3 percent positive today. Similarly, based on CDC FoodNet data, we have seen a remarkable decline in the incidence of foodborne ill- ness from Listeria over the past 10 or more years, with the incidence per 100,000 at about 0.5 in 1996 and just slightly over 0.25 today (0.25 is a U.S. Healthy People 2010 objective). Most of that improvement occurred in 1999 and 2000, however, which raises the question: Why hasn’t there been any further improvement, especially given that RTE meat products are among the highest risk foods and that the prevalence of foodborne illness associated with this product has declined sharply during this timeframe? 

34 MANAGING FOOD SAFETY PRACTICES FROM FARM TO TABLE Consumer Behavior in Managing Food Safety Risks Presenter: Christine Bruhn Bruhn began her talk by commenting on a number of factors that have led to the growing proportion of the U.S. population at increased risk for a foodborne illness: •  hanging population demographics, for example the growing pro- C portion of people who are 65 years of age and older and the grow- ing number of people who are diabetic. Older people and people with diabetes are both at a higher risk of foodborne illness. •  hanging food preferences, such as the growing number of people C who consume raw fruits and vegetables and other minimally pro- cessed foods. Raw products do not have a kill step, which means that any pathogens that might be present pose a risk. This includes raw products that would otherwise be safe with pasteurization (e.g., raw milk). Bruhn pointed to a growing interest across the United States in raw milk and the growing number of testimoni- als as to the miraculous health-enhancing properties of raw milk, despite any scientific backing of these claims. She noted that many states are considering changing their regulations to enhance avail- ability for the public. Bruhn said that this is a growing area of con- cern because of the increased risk in foodborne illness associated with raw milk consumption. Last year in California, for example, a number of children who had consumed raw milk ended up in the hospital with HUS (hemolytic uremic syndrome). • Changing food safety knowledge and behavior. Bruhn remarked that the focus of her presentation would be on the last factor: consumer knowledge and behavior around food safety. Bruhn proceeded to summarize key findings from several recent research studies on consumer knowledge and behavior, beginning with a study showing that most people do not know who is at the highest risk for foodborne illness (Byrd-Bredbenner et al., 2007). Fewer than 6 percent of people surveyed were able to identify high-risk groups (i.e., older persons, youth, those with certain medical conditions, pregnant women, and people who are diabetic). The problem with this alarmingly low number, Bruhn explained, is that we have fallen into a rut of doing things that might not be so risky when   Christine M. Bruhn, PhD, is a Food Science Marketing Specialist and Director of the Center for Consumer Research at the University of California, Davis.   Bryd-Bredbenner, C., J. Maurer, V. Wheately, D. Schaffne, C. Bruhn, and L. Blalock. 2007. Journal of Food Protection 70(8):1917-1926. 

RECENT OUTBREAKS IN FOOD PRODUCTS 35 we are healthy and then continuing those practices even when we enter a riskier state or are involved with preparing food for people who are in a high-risk group. Even those who prepare food for high-risk individuals are not always aware that their audiences are at high risk. The same study showed that community volunteers who serve meals to senior centers and elsewhere (e.g., community and youth groups, Rotary and other service groups, church groups) do not recognize that some of those they serve may be among the highest risk groups for foodborne illness: 21 percent of those surveyed did not realize that seniors are at increased risk for foodborne illness; 26 percent did not realize that youth were an increased risk; 26 percent did not realize that people with certain medical conditions were at an increased risk; 48 percent did not realize that pregnant women were at an increased risk; and over half (56 percent) of those surveyed didn’t realize that people with diabetes were at an increased risk. Because they did not know that these people were at an increased risk for foodborne illness, the volunteers did not know that they should be taking even more than normal precautions when serving these audiences. In another study designed to see if educational programs have made a difference in attitudes and knowledge, little change was observed between 1999 and 2002 with respect to the percentage of people who had not followed certain food safety guidelines (Cody and Hogue, 2003).10 For example, in 1999, 9 percent of survey respondents acknowledged that they had forgotten to wash their hands before cooking, compared to 10 percent in 2002. While these percentages are relatively low, the lack of any improve- ment between 1999 and 2002 (i.e., before and after initiation of a nation- wide food safety campaign to educate consumers about key food safety messages) suggests that while food safety educational programs are very important and necessary, they have not had as major an impact as hoped. As another example, in both 1999 and 2002, 29 percent reported not hav- ing changed their kitchen cleaning cloth or sponge at least weekly. The great percentage of people who are not changing their dishcloths at least weekly (and some experts would recommend that dishcloths be changed every day or even every meal) suggests that people do not recognize that moist food- laden sponges are environments that bacteria love. As a final example, 22 percent of respondents reported using a meat thermometer in 1999, com- pared to 25 percent in 2003. So there was some slight improvement. Again, most people are using their thermometers for large pieces of meat, like roasts, and not for smaller pieces of meat that are cooked more frequently 10  Cody, M. M., and M. A. Hogue. 2003 (September). Results of the Home Food Safety— It’s in Your Hands 2002 survey: Comparisons to the 1999 benchmark survey and Healthy People 2010 food safety behaviors objective. Journal of the American Dietetic Association 103(9):1115-1125. 

36 MANAGING FOOD SAFETY PRACTICES FROM FARM TO TABLE and are more frequently associated with food illnesses (e.g., hamburgers). These numbers indicate that there has not been much progress over time and there is still a long way to go. With respect to food storage, again, a number of people are unaware of the value of making foods cold as quickly as possible. A survey by Cody and Hogue (2003) (see footnote 10) found that 16 percent of respondents thought it was acceptable to store cooked meat at room temperature, and 50 percent thought it was necessary to cool food to room temperature before refrigerating. Among the food preparation volunteers surveyed in the United States, only 56 percent knew the recommended refrigerator temperature (and thought that it was okay to set it higher than 40 degrees). Nationwide, 40 to 56 percent do not know the recommended refrigera- tor temperature. The same study showed that people confuse safety with spoilage and believe that food that looks, smells or tastes differently is con- taminated. Again, among the same volunteer survey population mentioned previously, 95 percent of survey respondents thought that they could tell if food was contaminated by how it smelled or tasted. Of course, that is not the case. Research also shows that people also often overstate their compli- ance and practices. For example, a review of the literature (Redmond and Griffith, 2003)11 showed that while 82 to 100 percent of people indicated that they knew that it is appropriate to wash their hands after handling meat and poultry, 75 to 100 percent failed to actually do so after handling raw chicken. Bruhn mentioned that some of her current research involves using video cameras to watch study participants prepare food in their own homes to see whether they do what they say they do. She emphasized that there is a large difference between knowledge and behavior. Sometimes people are confused about date labeling, which is an area where Bruhn suggested that industry could provide some assistance. Some- times finding the date is a challenge, particularly for an aging popula- tion—sometimes the lettering is too small or not clearly visible on the package. Additionally, consumers are often confused by “manufactured by,” “sell by,” “use by,” and “best used by” dates. A recent study by the Food Marketing Institute (FMI)12 showed that more people are concerned with the “best used by” date than the “use by” date even though the former is intended for quality, not safety. The differences among these dates have not been communicated to the public. The good news is that people are concerned about pathogens, as they 11  Redmond E. C., and C. J. Griffith. 2003. Consumer food handling in the home: A review of food safety studies. Journal of Food Protection 65(1):130-161. 12  Food Marketing Institute Research Department. 2008. Consumers and Food Safety. In U.S. Grocery Shopper Trends. Arlington, VA: Food Marketing Institute. Pp. 71-77. 

RECENT OUTBREAKS IN FOOD PRODUCTS 37 Consumer Perception - Health Risks Risk) (Serious Risk) Biotech/GM 20 Irradiated 21 Antibiotic/hormone 27 a e r e Qi c T i e ™ a n a u k m d e c mp r s s r o e o d n e det os e t i pi t e . d h s c u r Avian Influenza 38 Pesticides 43 Terrorists 44 Tampering 46 Bacteria 53 0 10 20 30 40 50 60 F MI Trends 2008 Figure 2-5  Data from the FMI show that people appropriately view bacteria as the number one food risk (SOURCE: Food Marketing Institute Research Depart- Figure 2-5, editable ment. 2008. Consumers and Food Safety. In U.S. Grocery Shopper Trends. Arling- ton, VA: Food Marketing Institute. Pp. 71-77). should be, according to other FMI data. As shown in Figure 2-5, more people are concerned about pathogens than they are pesticides, additives, irradiation, and genetic engineering of their food products. So people are viewing risks appropriately for the most part. The problem, Bruhn empha- sized, is the large difference between knowledge and behavior with respect to food handling and consumption. Research shows that many people do not adhere to safe food handling recommendations, suggesting that the focus should remain on making food as safe as possible. Bruhn noted that it is interesting that people are so concerned with irradiation and genetic engineering, two techniques that could be used to overcome both the patho- gen and pesticide residue issues. Bruhn posed the question, “How do people respond when there is a safety concern?” Again, according to FMI data, when spinach was associ- ated with a food safety risk in 2007, about 74 percent of consumers said that they stopped purchasing spinach. Of course, they didn’t have a lot of choice since it was withdrawn from the market. But, Bruhn asked, why not 100 percent? In 2008, when peanut butter was associated with a food safety risk, 22 percent of consumers stopped buying peanut butter. Again, why wasn’t it higher? Bruhn suggested that perhaps people bought other brands 

38 MANAGING FOOD SAFETY PRACTICES FROM FARM TO TABLE (i.e., brands not implicated in the outbreak). Or perhaps they believe that “it can’t happen to me,” or maybe they are simply unaware of recalls. That some people are unaware of recalls raises the question, where do people get their information about recalls and other food safety issues? FMI data (see above) have shown that most people get their information about recalls and other food safety issues from, first, the TV (slightly over 50 percent of consumers get their information from the TV) and, second, the Internet (about 45 percent)—then newspapers (about 35 percent), friends (about 25 percent) and magazines (20 percent). Although none of these sources are trusted very much, with the Internet and TV being trusted al- most equally (i.e., less than 20 percent of respondents trust either source) and the newspaper, friends, and magazines trusted by less than 10 percent of consumers. How do people respond to food recalls when they do know about them? FMI data show that about 81 percent of people check the food in their home. Bruhn questions that percentage, however, given the low per- centage of people that actually return food to their supermarket. Half of consumers (50 percent) rely on their supermarket to either offer safe food in the first place or get it off the shelves before they can buy it. Only about 40 percent of consumers are sufficiently concerned that they would be willing to sign up to receive an email alert in the event of a food recall. So basically, with only 40 percent willing to take the effort to be told when a food recall is in effect, food safety is not a high priority for most consumers. Finally, getting back to the issue of changing population demograph- ics, one might expect people in higher risk groups to manage their own food safety risks more efficiently. But data show that even among people with HIV/AIDS who have received food safety education information, still many people are not following appropriate food safety guidelines. In one study (Hoffman et al., 2005),13 only about half of those surveyed reported that they were currently washing their hands before preparing food, with another 45 percent saying that they would definitely or probably wash their hands in the future. While that total—95 percent—is pretty good, why isn’t it 100 percent? And that is as good as it gets. With respect to avoiding rare ground beef and raw shellfish, only about 80 percent of respondents said that they were currently or would definitely be willing to avoid rare ground beef in the future. But with soft cheeses and unheated luncheon meats, only about 60 percent indicated that they were currently or would definitely be willing to avoid these products in the future. So this represents a group of 13  Hoffman, E. W. , V. Bergmann, J. Armstrong Shultz, P. Kendall, L. C. Medeiros, and V. N. Hillers. 2005. Application of a five-step message development model for food safety educa- tion materials targeting people with HIV/AIDS. Journal of the American Dietetic Association 105(10):1597-1604. 

RECENT OUTBREAKS IN FOOD PRODUCTS 39 people who would be mostly likely to manage their food safety risks by ac- cepting and following certain recommendations, and yet they are not doing so. This suggests that current food safety education is not sufficient even for those who one would think would be in greatest adherence. This last finding raises the question, why are people one would expect to adhere to recommendations not doing so? Bruhn listed several possible reasons: •  ptimistic biases (i.e., people think that “it won’t happen to me”) O • People are too busy • It is inconvenient (e.g., to use a meat thermometer) •  t is not necessary (i.e., “I’ve been eating this way all my life, and I I haven’t become ill”) •  aste preferences, which often override food safety concerns (e.g., T people like the taste of rare ground beef) Bruhn concluded with a summary of five key points: 1. Foodborne diseases are likely to increase in the future because of changing demographics and population preferences. 2. While at least some consumers stop buying products implicated in an outbreak and are interested in hearing about outbreaks and re- calls when they occur, most consumers do not respond to outbreaks or recalls. 3. With the greatest number of people expressing concern about haz- ardous bacteria (e.g., as opposed to irradiation), people are con- cerned about the right thing (i.e., the most hazardous risk). 4. Despite this knowledge, many people do not follow food safety recommendations and guidelines. 5.  cannot rely on consumer education to keep people safe, because We they are not following the rules. The focus must be on making food as safe as possible, which Bruhn said should include a kill step from the processor before food reaches the consumer’s kitchen. open discussion14 The first question of the discussion was directed to Bruhn, asking whether she had seen any positive trends in terms of consumer acceptance of irradiation as a means of making food safer. Bruhn said that over the last 20 years or so, there have been some people, probably about 10 percent, 14  This section is a paraphrased summary of the discussion that followed the four presentations. 

40 MANAGING FOOD SAFETY PRACTICES FROM FARM TO TABLE who have read about food irradiation and wonder why they can’t buy ir- radiated products in their supermarkets yet. At the other extreme, another 10 percent say they “would not touch [irradiated products] with a ten- foot pole.” The vast majority of the people are somewhere in the middle. They are still curious about what it is and why it is used. Bruhn’s research suggests that when questions about irradiation are answered with science- based information about the safety and potential benefits, as well as who endorses irradiation, people are willing to approve that additional protec- tion for their families. The one area where data are lacking is in consumer attitudes toward the application of irradiation to leafy greens, like spinach and iceberg lettuce. Bruhn mentioned that she is hoping to conduct a study on this area in the near future and to develop the appropriate educational messages that would allow people to understand the process and how it would affect their families. Bruhn was also asked (by Doug Podolsky, health editor at Consumer Reports) what kinds of pathogens are typically found in raw milk. Bruhn responded, “You name it.” She short-listed tuberculosis, Q fever, E. coli O157:H7, Salmonella, and Campylobacter. Bruhn then praised Consumer Reports for pressuring legal action a few years ago requiring that raw milk sold in California carry a warning label. She also mentioned a bill cur- rently before the California legislature that would initiate a HACCP plan for raw milk producers and which would generally enhance the safety of these products. The next question was directed to Huffman with regards to the recent Listeria outbreak in Canada. Caroline Smith DeWaal of the Center for Sci- ence in the Public Interest (CSPI) asked whether Canada endorses the same safety standards and systems that Huffman described in his presentation. Huffman responded that if the products currently in question are associated with illness, then no, the company in question is not implementing the same control procedures outlined in his presentation, or at least not to the degree that is necessary. Additional steps need to be taken. Huffman mentioned that while companies exporting to the United States should have equivalent systems and the same types of control methods in place, he was unsure of the Canadian regulations for domestic products. Huffman mentioned that, coincidentally, he would be heading up to Canada the week after this work- shop to co-lead a Listeria workshop at the Canadian Meat Council (CMC) annual technical conference, where he and his U.S. colleagues would be sharing best practices. Zink was asked whether any data indicate that flexible packaging has been linked to any foodborne illnesses. Zink said “not specifically.” Flexible packaging does not have the same durability and integrity as glass jars and metal cans do, however, it does have slightly higher leakage and spoilage rates. 

RECENT OUTBREAKS IN FOOD PRODUCTS 41 A series of traceback questions were asked of Daniels, beginning with a question about the type of ranches from which Natural Selection Foods gets its produce. Daniels explained that Natural Selection Foods has a variety of relationships with a number of ranches, from partner to contract growers. When asked about what kind of oversight Natural Selection Foods has over these ranches, Daniels explained that they require third party audits for each ranch, as well as compliance with Natural Selection Foods’ own GAP guidelines and verification audits to ensure that this is happening. When asked if the company requires any physical sampling of the ranch environ- ment, Daniels said, “typically not.” He explained that one of the benefits of their multi-hurdle approach is that if, for example, positive results are obtained for raw materials, those results are obtained quickly enough that Natural Selection Foods can return in real-time (within 16 hours of harvest) to the field from which that crop was harvested for a thorough investiga- tion. At that point, the investigation would include physical samples. The questioner then asked if Natural Selection Foods did any co-packing for other companies and, if so, did they put out other products under different labels? Daniels responded, “yes.” Finally, the questioner commented that a key problem with tracebacks in produce is that co-packing, relying on multiple sources for foods, etc.; all create delays, which in turn lead to a loss in public trust. He asked: Is there anything that industry is doing in an effort to organize the farm-to-table distribution in order to avoid this kind of delay during a traceback? Daniels said that while he understands the need to improve tracebacks for commodities, the system for bag salads is in fact pretty tight. The com- pany’s traceback system can trace a finished bag to a ranch within a matter of minutes, with verification of that traceback requiring only two hours. He commented that he was unsure why co-packing should be an issue, since it simply puts more responsibility in the processor’s hands. He agreed that recall announcements can be confusing for consumers, with multiple brands listed, but argued that the use of multiple brands is a trend that is not going to disappear. It really is the processor’s responsibility to ensure that they have a sound recall plan in place—one that has been practiced and with as many holes poked as possible to ensure that you are ready in the event of a recall. Chin agreed with Daniels, reiterating that the responsibility falls on the processors and producers to have those systems in place. Our food supply system is global, and foods are going to continue to come from multiple sources—it is a trend that is not going to change any time soon. Another participant, Nancy Donley of S.T.O.P. (Safe Tables Our Priority), followed up on this last line of questioning, commenting on the fact that the amount of recalled product actually recovered is typically very small. Hopefully that will change now that USDA/FSIS will be identifying 

42 MANAGING FOOD SAFETY PRACTICES FROM FARM TO TABLE retailers in their press releases, making it easier for consumers to identify recalled products in their homes. And hopefully FDA will go down this same path, Donley said, “Because we should just have one way of doing business in this country and we should all be on the same playing field.” Donley then asked Daniels about the spike in pathogen incidence that oc- curs in the summer months and whether Natural Selection Foods is doing anything to take extra precautions during that time—not just with respect to testing but also with respect to working the company’s preventive process harder during those months. Daniels responded that the company follows the same standards year-round and that these data are fairly new and that, yes, perhaps there should be a greater focus on the summer months while nonetheless staying vigilant year-round. Daniels also re-iterated that testing is not the answer and that, despite best preventive efforts, pathogens are ubiquitous. Even the strongest preventive programs need to be continu- ally enhanced and improved. Finally, Donley applauded Natural Selection Foods’ aggressive testing for pathogens other than O157 and expressed disappointment that a recent USDA/FSIS position on declaring other EHEC strains as adulterants in meat and poultry products did not advance. She mentioned that one of the excuses given is that the technology for detect- ing other EHEC pathogens does not exist. She hopes that Natural Selection Foods will share its testing standards with the rest of the industry and with the USDA. Daniels agreed and noted that many of his colleagues have heard him repeat his message “loud and clear” that testing should encompass “the gamut of harmful pathogens.” He mentioned that Norwalk virus is also emerging as a pathogen of concern for fresh produce.