NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Institute of Medicine (US) Food Forum. Managing Food Safety Practices from Farm to Table: Workshop Summary. Washington (DC): National Academies Press (US); 2009.

Cover of Managing Food Safety Practices from Farm to Table

Managing Food Safety Practices from Farm to Table: Workshop Summary.

Show details

4The Way Forward: Varying Perspectives

The third and final session of the day, “Panel Discussion: Where Do We Go From Here?,” was a four-person panel discussion. Each speaker was asked to provide some insights into the many and varied issues that were raised during the earlier sessions and propose options for moving forward. Moderator Michael Doyle1 opened the session by introducing Steve Sundlof of the FDA’s Center for Food Safety and Applied Nutrition (CFSAN).

Sundlof noted that the FDA has received the message “loud and clear” that there is a need for more cooperation and coordination between industry and government and that the FDA is truly committed to making this happen within the confines of what its mandate allows. Sundlof described the direction the FDA has been heading over the past year or so with respect to food safety, including the November 2007 launch of its Food Protection Plan, which calls for a range of domestic and international initiatives and activities.

The second panelist was Caroline Smith DeWaal of the Center for Science in the Public Interest. After citing some of the positive developments in the area of food safety that have occurred over the past 10 years, DeWaal described some fundamental differences between how the U.S. federal government’s two food regulatory agencies—the FDA and USDA—approach food safety. She encouraged a move away from both approaches (i.e., the FDA’s enforcement-based approach and the USDA’s recall-based approach) and toward a more unified regulatory structure. DeWaal urged moving forward even if and when the science is incomplete, which she noted will always be the case. She also commented on the need for more work in the area of risk communication at both the national and international levels.

Mike Robach of Cargill, Inc., spoke after DeWaal, providing an industry perspective on some of the issues. In particular, he argued the importance of recognizing the multitude and range of food systems worldwide and the need to reach consensus on key criteria that impact food safety and public health regardless of where that food is produced, processed, distributed, or consumed. The principles underlying these criteria exist, Robach argued, and are already standardized in multiple ways, from GAPs to HACCP, but food companies, inter-governmental agencies, regulatory agencies, consumer groups, academia have yet to make a concerted effort to harmonize these criteria.

The fourth and final panelist of the session, Richard Raymond, Under Secretary for Food Safety at the USDA, presented a USDA perspective on three issues: (1) risk communication and how USDA’s recently established “retail rules” will help alleviate some risk communication problems; (2) how USDA’s new information system, which is currently in development, will strengthen the government’s capacity to conduct risk-based inspections; and (3) USDA’s involvement in a range of international regulatory activities.

A lengthy discussion on several key topics followed the panelist presentations:

  • The need to improve communication between industry and government
  • The global nature of the food supply chain: challenges and opportunities
  • Divergent opinions about the need for more government oversight on farms
  • Controversy around the notion of a single unified food agency
  • Water quality and waste management: who is responsible?
  • Food safety research: some salient topics
  • Unanswered questions about food safety trends
  • Traceability: room for improvement
  • Challenges with microwaveable foods
  • The need for an international perspective on the issues addressed during this workshop
  • Sustainability and other future food safety issues


Presenter: Stephen Sundlof2

Sundlof commented on the importance of government and industry cooperation and coordination while moving forward toward establishing a more effective food safety system. It is a familiar theme that has been repeated in many recent meetings involving both government and industry representatives at the national, state, and local levels. The FDA has received the message “loud and clear,” he said, and is truly committed to trying to make it happen within the confines of what government can do.

Sundlof then briefly described the direction that the FDA has taken over the past year or so with respect to food safety. Many recent events, including the E. coli outbreak in spinach, the peanut butter outbreak, the melamine issue and, most recently, Salmonella in peppers, have forced the FDA to realize that it cannot continue doing things the way it has been doing them in the past; and that the agency needs to make a major shift in how it addresses food safety. In November 2007, FDA launched its Food Protection Plan, based on three core elements, or pillars: prevention, intervention, and response:

  1. Prevention: This is where the FDA will be putting the greatest emphasis, Sundlof noted. Now, when a food safety incident occurs, FDA typically takes a corrective action. The agency wants to start moving toward a more preventive mode. This aspect of the plan encompasses several initiatives and activities:
    1. Opening new posts in five geographic areas (India, China, Europe, Latin America, and the Middle East), through FDA’s Beyond Our Borders Initiative and as a way of increasing regulatory presence oversees and ensuring the safety of food imports—not just through inspections but also capacity-building (e.g., making sure our foreign partners understand what is required to meet U.S. standards and providing technical assistance where necessary).
    2. Building a database of information received from foreign partners about food inspections and food quality (e.g., results of inspections, where problems are detected) and using the database to better target the agency’s inspection and import surveillance programs.
    3. Strengthening FDA’s capacity to collect and interpret data necessary for risk-based prevention, so that the agency can better identify food vulnerabilities and assess food risks.
    4. Conducting research to improve risk-based prevention strategies.
    5. Assessing areas of vulnerability (i.e., vulnerability to intentional contamination of food), identifying the most readily available targets that will have the highest impact, and making sure that the food industry is aware of these targets and that they are conducting their own vulnerability assessments.
    6. Developing and validating rapid detection tools to detect food contamination (e.g., through genetic fingerprinting and the use of other molecular tools) and improving the speed and accuracy of both microbiological and chemical contamination detection (i.e., achieving higher throughput so that test results are provided more quickly during ongoing outbreaks).
  2. Intervention: This component of the Food Protection Plan relies on targeted risk-based inspections and again encompasses several initiatives and activities:
    1. Conducting testing to verify that preventive controls are in place and are running adequately and inspecting the highest-risk firms on a more frequent basis.
    2. Hiring additional investigators so that the extent of domestic and particularly foreign inspections could be expanded.
    3. Developing and using new tools to conduct enhanced risk analysis (i.e., analyses that consider a multitude of factors that contribute to risk, like economic fraud, which was clearly an important risk factor with melamine).
    4. Improving FDA’s ability to integrate and assimilate risk-based information into data systems that allow for early signal detection during contamination problems.
  3. Response: The goal of this third component of the Food Protection Plan is to improve FDA’s response capability. Again, it encompasses a broad range of initiatives and activities:
    1. Reducing the length of time between detecting and containing a foodborne illness.
    2. Working with various organizations to better understand traceability, identify best practices with respect to traceability, and develop ways to implement those practices (e.g., using the fresh produce industry’s own tracing systems to build better ones). For example, FDA will be conducting a workshop in 2009 to review a $15 million traceability study that has been underway in the European Union since 2007.
    3. Creating a health hazard alert system, to quickly alert the public about outbreaks and illnesses.


Presenter: Caroline Smith DeWaal3

DeWaal reflected on her participation in an IOM Food and Nutrition Board workshop held in this very room about 10 years ago. The focus of that workshop was on the structure of the U.S. food safety system. She noted that the National Academy’s Ensuring Safe Food: From Production to Consumption (1998), which stemmed in part from that workshop, added tremendously to the literature and our scientific understanding of the hurdles created by that structure in improving food safety. We have come a long way since then, she said, with many positive developments:

  • We have widened our understanding of food safety problems and have moved beyond problem identification and putting tools in place to a point at which we are ready to act.
  • Several new coalitions have formed over the past several years with the goal of educating Congress about food safety. One of the most successful of these is the Alliance for a Stronger FDA, where the food industry, consumer groups, the drug and medical device industries, and patient groups have banded together around a single message: that the FDA cannot possibly carry out its mission with its current budget.
  • There has been considerable development around the concept of risk assessment; and we have seen the formal development and even impartial implementation of an international risk analysis framework.
  • With respect to risk management, in the United States, there has been a rise in the use of microbial risk assessments and adoption of HACCP. The USDA, for example, has adopted HACCP for its entire regulatory industry. FDA, on the other hand, utilizes HACCP only for seafood and juice products, relying on a commodity-by-commodity approach elsewhere.

DeWaal expanded on this last bullet point, noting that adoption of HACCP by the USDA versus the commodity-by-commodity, or crisis-by-crisis, approach of the FDA reflects important and fundamental differences in the regulatory approaches of these two agencies.

The FDA follows an “enforcement model.” A great example of this, she explained, was when Commissioner David Kessler4 seized processed orange juice because it was labeled “fresh,” with little or no warning to the industry. He simply enforced the law. Still today, the Office of Regulatory Affairs (i.e., the inspection arm of the FDA, as opposed to the policy-setting arm) conducts their inspections by developing evidence in order to bring action. While that action might be only a warning letter, their level of review and analysis is such that the action would be supported in a court of law if necessary. They follow a very high standard. A shortcoming with that approach, however, is that because FDA inspectors are not in the food plants very often, they are not looking for systematic problems and are not necessarily finding solutions. They are largely enforcing the law, in accordance with their statutory framework (i.e., the enforcement-based Federal Food Drug and Cosmetic Act).

USDA, on the other hand, relies on somewhat newer “recall model,” whereby a high-level regulatory official declares “zero tolerance” for high-risk pathogens or other ingredients and then, when violations in the law are detected, recalls are issued. While these are largely voluntary recalls, the industry largely complies. The recalls can be massive and can extend to products with very, very low relative risks, with the goal being to remove all contaminated or potentially contaminated product from the market. While an enforcement action linked to that recall may follow, the focus usually remains on removal of the contaminated product from the market.

DeWaal then offered advice to the next administration (i.e., the Obama administration): “To take us away from what we’ve got today, with all its warts and problems, into a modern food safety structure.” She called for a unified structure with unified food laws (both the 1906 and 1957 laws are antiquated, she said) and a move away from both the enforcement model (FDA’s current approach) and the recall model (USDA’s current approach). While both of those approaches could be used as last resorts, they should not be the first regulatory step taken when problems arise.

DeWaal also called for the next administration to develop policies that would put the United States into an international consumer leadership position. While we already play an important role at Codex meetings, DeWaal would like that role to be only part of the U.S. leadership role with respect to consumer protection. For example, she would like to see greater expertise applied to the important role of risk communication, so that risk communication triggers the appropriate consumer responses. Just over the past year, we have had massive recalls and many examples of risk communication at work. The reality is that many people do not always respond appropriately and, at a certain point, many consumers simply get tired of the messages and stop paying attention altogether. So, while risk assessment and risk management have both received a lot of attention and are both moving forward, risk communication has not received the same level of attention at either the national or international level. It is time.

Finally, DeWaal commented on Caswell’s notion of developing a generic joint public–private approach toward food-safety risk management. Importantly, DeWaal emphasized, any structure we devise has to be based on imperfect science because “we are always going to be operating with imperfect science.” Doing nothing because the science is incomplete (e.g., not appropriately responding to a foodborne illness outbreak because we are not sure how the pathogen entered the food supply) is unacceptable and “not good consumer protection.”


Presenter: Michael Robach5

Robach reiterated DeWaal’s comment that we have learned a lot about food safety over the past 10–15 years. He commented, however, that we would be learning even more if we could effectively share our collective experiences and use those as a basis for moving forward. Too often we reinvent the wheel when solutions to particular problems may already exist, he said, if not here in the United States then in other parts of the world. Robach argued that: We need to take a step back and get a global view of the wide range of existing food systems, from Natural Selection Foods’ organic system (which Daniels described earlier in the workshop) to Cargill’s worldwide operations. Importantly, while Cargill conducts business around the world (in 66 countries) many of its businesses are focused on local marketplaces. In Central America, for example, products from poultry and meat processing and feed operations are sold locally. Others, like the large operations in Brazil and throughout North America, export worldwide. So even within this one company, there are many different food systems. We must be cognizant of this range of food systems in our efforts to achieve better food safety. The same system that works in one place, for example in the United States, does not necessarily work in, say, Paraguay or Indonesia.

  • We need to more effectively take into account our entire supply chain, linking what is happening on the farm, whether in agriculture or animal husbandry, with what is happening inside the processing plant (i.e., during production and packaging) and with what is happening after a product leaves the process plant (i.e., during distribution and consumption).
  • We need to reach consensus on key criteria that impact food safety and public health regardless of where that food is produced, processed, distributed or consumed.

Robach elaborated on the last point, arguing that in fact these criteria already exist and are already being utilized but not in a harmonized, consistent way. For example, we have GAPs and other (botanical) plant-related criteria established by the International Plant Protection Convention (IPPC), all of which were vetted and agreed upon. Likewise, we know what the right criteria are for good animal handling practices (i.e., through the work of the World Organization for Animal Health, or OIE), and we have an established set of principles for good hygiene (i.e., through Codex) and, in many places, GMPs. Similarly, the principles of HACCP were vetted through Codex, as well as through the National Advisory Committee on Microbiological Criteria for Food. We also have a standardized risk- and science-based hazard analysis, including an understanding of what the likely hazards are, where they will likely enter the process and what control measures can and should be put into place. So these principles exist. Now, we just need all stakeholders at the table making an effort to position these standards and ensure that our supply chain is safe. Food companies around the world need to work together with national governments, intergovernmental organizations (i.e., FAO, Codex, WHO, OIE), regulatory agencies, consumer groups, and academia to assemble food safety programs operating against the same set of principles, or criteria, and with the available technologies. The process needs to be transparent, and it needs to stay focused on the desired outcome: to have safe food around the world.

Robach commented on the proliferation of private audits, arguing that they are conducted primarily out of frustration and because of the disharmony that exists around the international regulatory infrastructure. For example, it is very difficult to be working in a poultry plant in Brazil while trying to meet the regulatory requirements of seven different countries with the same product. We need to improve that way of conducting business. Over the next 40 years, we will be going from six-plus billion people to nine billion. We are going to have a lot more mouths to feed, and we will need to be a lot more efficient and smarter with food production.

Finally, Robach expressed the need to adopt a more global mindset and recognize that there are no “bad” countries with respect to food safety. For example, while China has received a lot of negative attention, in fact there are some really good processing plants in China that do outstanding jobs. Likewise in the United States—there are good and bad plants. Instead of looking at country-level food systems, we need to think of the food supply chain as the global food system.

Robach also stated that the industry needs more holistically trained individuals. Too often, new college graduates focus on animal nutrition, food microbiology or some other discipline and are not seeing the whole picture with respect to the way the different components of the supply chain are linked together.


Presenter: Richard Raymond6

Raymond began by echoing other comments on the international nature of the food supply chain, pointing to ground hamburger as an example: you don’t know whether the ground beef you are eating came from Uruguay, New Zealand, Australia, Canada, or elsewhere. Then, he remarked on three specific issues from a USDA perspective: risk communication, the use of risk-based USDA inspections in the future, and USDA’s role in ensuring a safe global food supply:

  1. Risk communication. Raymond remarked that the “retail rules” passed and published in July 2008, which now allow USDA to notify consumers which stores sold recalled products, will be a tremendous boost for the consuming public and will help alleviate some risk communication problems. Raymond agreed with DeWaal that risk communication is still a major problem and that there needs to be a better way of teaching people how, for example, to use thermometers when they cook poultry, ground beef or other USDA-regulated products. He noted that improving risk communication is not, however, just a matter of getting more information out there. It is also a matter of removing the “bad” information that is already out there and dispelling rumors that feed misinformation.
  2. Risk-based inspections of the future. Raymond briefly described a new “21st century” database under development that he said is “short of fantastic.” With the press of a button and within a few seconds, answers can be retrieved regarding, for example, which plant(s) has the highest risk of being responsible for a food borne disease outbreak based on pathogen testing conducted over the past two years. The database is not operational yet. It will be launched in the third quarter of 2009. When fully operational, it will make USDA “nimbler and quicker,” he said, giving the agency the capacity to not only respond more quickly to recalls but also prevent recalls through predictive analysis. USDA has organized a new group, DAIG (Data Analysis Integration Group), which along with the Data Coordinating Council (comprised of members from all parts of the FSIS), will determine how to coordinate and make best use of the data.
  3. Raymond explained how a past Freedom of Information Act request cost thousands of person-hours. You couldn’t just search “tongue” or “tonsil,” for example, you had to search all of the different possible spellings (or misspellings) of each word. This new information system will not only alleviate some of this type of data-mining frustration, it will also allow for risk-based inspections of both processing and, eventually, slaughter plants. Right now, as bound by law, USDA spends exactly the same amount of time inspecting a ground beef plant as it does a plant that processes chicken noodle soup. This is true despite the fact that the former have involved more recalls and suspended inspections. You just don’t see children dying from eating vegetable-beef soup, Raymond said, but you do see children dying from eating contaminated meats. Yet, USDA is not allowed to conduct different levels of inspection among these different types of plants; at least not yet. USDA’s new information system will allow for a risk-based inspection system not possible until now.
  4. The international nature of the food supply. By law, every exporting country must have an equivalent food-safety system for meat and poultry products shipped to the United States. There are several steps that must be taken to determine equivalency. First, USDA conducts a paper audit to ensure that the country has the rules, regulations, policies, and finances in place, such as the daily presence of federal inspectors in continuous processing and slaughter plants, etc. If on paper, a country has an equivalent food safety system USDA conducts on-site inspections of selected establishments, laboratories, etc. If based on those on-site inspections, USDA determines that the country has an equivalent food safety system in place; then the lengthy rule-making process can begin. After that initial audit, USDA conducts yearly audits and on-site inspections to ensure that equivalency is being maintained. Also 10 percent of all import boxes are opened and the shipped products themselves are inspected, and 5 percent are sampled for pathogens and chemical residues.

To date, the USDA has determined that 34 countries have equivalent food-safety systems for meat and poultry. Only 29 of those countries currently export to the United States. The other 5 do not export primarily for economic but also other reasons. Compared to what the FDA must manage, Raymond said that 29 are “pretty easy.” But, there are some vague areas and issues that USDA is still contending, for example how to define equivalency and which is more important: inspecting processing or inspecting the final product (i.e., whether it is more important to conduct on-site inspections of processing or slaughter plants and determine whether tests for E. coli in Australia, for example, are comparable to tests for E. coli in the United States or inspect products after they have been shipped by opening and sampling boxes). And should countries with perfect records (e.g., a country that has never had problems with mislabeling, damaged products, or pathogen issues upon sampling of import boxes) be audited just as frequently as countries that have had problems (e.g., a country that has been de-listed twice in the last three years)? International inspections and equivalency-testing need to be just as risk-based as domestic food safety considerations, he said.

Open Discussion7

Improving Communication Between Industry and Government

The discussion started with a question by Doyle about communication between industry and government. Given that so many speakers today had recognized the importance of more open and frequent communication between industry and government with respect to food safety, Doyle asked, how can this be achieved? Usually when the FDA and industry meet to discuss food safety, it is during an outbreak situation or in regards to a particular inspection. Is there a way to facilitate a regular dialogue?

Sundlof was the first to respond. He noted that much of the lack of communication is “cultural.” Even among the various regulatory agencies, there is a tendency not to share information. The CDC, for example, works very closely with state-level agencies and is sometimes reluctant to share that state-level information with the FDA in order to avoid having particular states implicated in outbreak situations. The FDA in turn, is often reluctant to share information, which if released, would be harmful to industry. We need to create a culture where information is shared more openly while also being cognizant of the fact that some of that information could be harmful and needs to be protected to the extent possible. To that end, the FDA is currently seeking legal solutions to some of the current problems with the Federal Advisory Committee Act (FACA). As an illustration of those problems, at a recent meeting with executives from the produce industry, the FDA could not ask advice or seek consensus because of FACA and was limited to going around the room and asking the attendees to simply share their experiences. Having a mechanism in place to allow for more information sharing would help the FDA help industry.

Raymond agreed that there needs to be more open and regular communication among the regulatory authorities, industry, and consumers. He noted that the USDA meets with industry representatives and consumers on a monthly basis and, in fact, to the extent that sometimes questions are raised about “how cozy” the relationship is given USDA’s regulatory role.

Robach agreed that there must be more communication between industry and government. He relayed the story of a ground beef recall in October of 2008 and the lack of communication between the USDA, public health authorities, and industry. The recall was because of an E. coli O157:H7 illness, which had occurred in Minnesota but was associated with product coming from a Cargill plant in Wisconsin. Robach didn’t even know that the Cargill plant had been implicated until he received a call from the Wisconsin plant informing him that some USDA compliance officers were at the door. That phone call, in turn, prompted Robach to call the Minnesota Departments of Health and Agriculture. Coincidentally, when he made those calls, both agencies were in the middle of a conference call discussing the outbreak and a potential recall. It was “maddening and frustrating” not to have been notified or included in that conference call, Robach said. He urged more open communication, admitting that sometimes industry may push back. More often than not, however, “people are going to really roll up their sleeves and find out what the issue is.” When that E. coli O157:H7 illness was traced to a product processed in Cargill’s Wisconsin plant, that product was off the market that afternoon. Communication needs to begin early on, Robach emphasized, at the first hint of a problem and when the collective use of both government and industry resources can be leveraged to offer the best protection possible for public health—not after the regulatory agency has already crossed all its Ts and dotted all its Is.

DeWaal noted that while much of the discussion has been focused on the need for more communication between government and industry, the FDA also needs to improve its risk communication with the consumer community. The USDA has done a good job at keeping consumers informed about food safety policy and has made efforts to improve risk communication, she said. Noting that the FDA has far fewer personnel available to work on communication, DeWaal encouraged the FDA to “work harder” with respect to communicating with the consumer population—both with respect to food safety policy and crisis communication.

The Global Nature of the Food Supply Chain

Doyle identified another common theme throughout much of the day’s discussion: the rapid increase in food imports and the global nature of the food supply chain. He asked how the FDA was addressing issues around unsanitary practices currently used in many foreign countries, such as the use of raw human sewage in agricultural irrigation and the use of untreated animal manure in tilapia and shrimp ponds. These practices are at the root of many of the food contamination issues here in the United States, he argued.

Sundlof replied that the FDA obviously strongly advocates GAPs and, in many outbreak situations, has provided technical assistance to other countries in an effort to establish GAPs. This includes everything from teaching farmers how to conduct environmental sampling to teaching them to recognize the importance of have sanitary conditions for their workers. He noted the existence of international third parties that conduct GAP inspections. These efforts aside, Sundlof agreed that is an issue that needs to be addressed more fully, especially now that we are seeing new types of outbreaks (e.g., Salmonella in peppers, which has never occurred before). But it is a monumental problem and one that will require an international, not just FDA, solution. While the USDA has identified 34 food safety equivalent countries, of which 29 export to the United States, the FDA deals with 150 countries—that is the majority of the 190-plus countries that exist. Conditions are highly variable among countries even with the same commodity, which makes the development of comprehensive standards applicable to all countries extremely complex and difficult.

Robach agreed that the problem is not one that the U.S. government will be able to solve by itself. They do not have the resources, nor should they be expected to be responsible for solving this problem. Robach reiterated that the private sector is ultimately responsible for the safety of its supply chain. Cargill, for example, assumes responsibility for assuring the origination of products entering its processing plants. Still, the private sector needs to partner with inter-governmental organizations and national governments so that standard, appropriate criteria can be developed. There are many good examples of well-vetted, consistent supply chains, with most companies having assembled them for their own use. He argued that the criteria for standards exist—now, the private and public sectors need to sit down and agree that these criteria exist so that we can move forward. While Robach warned not to “go overboard and start imposing our western ideals on developing countries,” he emphasized that there are some essential elements of food safety that need to be employed by all countries participating in the global marketplace.

Raymond identified two ways that U.S. federal government efforts have impacted the global food supply chain by decreasing the proportion of U.S. imports of products not processed according to GAPs. First, since the summer of 2005, four countries have stopped exporting to the United States because of USDA audits. Second, former President Bush chartered an interagency Working Group on Import Safety in July 2007. Represented by 12 departments and/or agencies in the federal government, the group developed a comprehensive plan not just for the safety of food imports but also for the safety of all imports. Because of this plan, we have seen a tremendous increase in the number of seized illegal products that would otherwise have entered the United States.

DeWaal agreed that the issue is an important one and one that deserves Congressional attention. She emphasized, again, the reality that the United States has two entirely different regulatory systems functioning (i.e., the recall system of the USDA and the enforcement system of the FDA) and different levels of confidence built into each (i.e., USDA checks and reviews all imports, whereas FDA relies on a take-all-commerce approach). She mentioned that CSPI started looking at this exact issue—the safety of incoming products—a number of years ago but did not come up with very satisfying answers. Part of the challenge stems from the reality that there is no border checkpoint that can be used to ensure the safety of imported products; which means that you must go to the country of origin and ensure that their food safety system is functioning. To that end, CSPI has worked with WHO, FAO, and about 25 other consumer organizations to develop guidelines for engaging NGOs in source countries in the effort to ensure food safety of those countries’ domestic programs. If their domestic programs are improved, DeWaal explained, their export programs would improve as well. While these efforts are not an entire solution, they are the beginning of “bootstrapping” food safety on a global level. DeWaal pointed to other efforts by the World Bank but noted that most of those efforts are “stove pipe” efforts that fail to benefit consumers of the source country. For example, if a country is having problems with its fish exports, the World Bank will build a fish export program to ensure safe products for export but without benefiting consumers in the source country. DeWaal also encouraged workshop participants to visit, where international outbreaks are tracked.

More Government Oversight on Farms?

Doyle asked the panel whether we need more government oversight on farms, given the importance of producing food under sanitary conditions and the problems that arise around that issue.

DeWaal responded first. “Absolutely,” she said. “We need more oversight on the farm.” As with all food safety programs, however, that oversight should begin with a written plan and should be within the farmer’s control. In fact, she argued, most farmers have effective food safety systems in place. Otherwise, more of what we are eating would be making us sick. Most of what we eat does not make us sick. So clearly something is working. However farmers should be able to document the effectiveness of their programs (e.g., their water quality, fertilizer inputs, and farm worker sanitation programs), and those programs should be capable of being audited. There should be government-set standards (i.e., FDA standards) regarding what the plans should entail. As with the processing and slaughter plants, food safety begins with industry and, as such, government oversight should be a function of audit, review, and discussion (e.g., informing farmers that their hazard analysis is incomplete or that their safety plan needs to be improved). Only in instances where the farmer does not implement their plan or otherwise meet federal requirements would governmental oversight become enforcement. The CSPI has petitioned the FDA for such a program. She urged the FDA to take action on this petition.

Doyle then offered his own opinion on the matter and said, “No, we don’t need more governmental oversight on the farm.” Having inspectors on farms is not really going to improve food safety. Rather, we need the private sector to be more diligent when building its food safety systems. This includes developing a mechanism for data sharing in the event that a regulatory agency needs that information. Doyle reiterated what Robach had argued both during his presentation and in response to the previous question about the global nature of the food supply chain: that we need to take a step back and re-evaluate our criteria for safe food production, distribution, and consumption and then make sure that those criteria are being employed. Verification and validation that those criteria are being met must be a transparent process and the information shared in a way that builds confidence in the safety of our food supply.

Do We Need a Single Unified Food Agency?: Contrasting Opinions

Elsewhere during the panel discussion, while urging FDA to “work harder” with respect to risk communication, DeWaal had referred to her dream of a single unified food agency.” This prompted Doug Podolsky of Consumer Reports to ask the panel to discuss this dream, or future solution.

Raymond was adamant that he did not share this vision. Rather, his “dream” was that policy makers meet with industry food-safety experts, consumers and scientists to identify the riskiest food products consumed in this country, where those products come from and discuss what should be done to increase their safety. He argued that a single food safety agency will not be able to do anything unless the laws change, as the USDA is bound by a set of very prescriptive laws, the FDA much less so. These legal problems are compounded by financial problems. If you were to combine the food safety components of the FDA and USDA into a single agency and provide them with the same amount of total funding they each have today, Raymond argued, the consequences would include no longer having continuous inspections of slaughter plants because some of the money currently being applied toward that aspect of USDA’s food safety program would be re-directed toward FDA activities. Right now, both the USDA and FDA do the best they can with the funds they have and the laws and statutes that mandate their actions. A single food agency cannot rectify these problems. Instead, we need to spend our time and energy on solving the legal and financial problems that already exist within each individual agency. For example, the law does not allow risk-based inspections: Not only does a ground beef plant receive the same level of inspection as a chicken noodle soup plant; tomato soup plants receive no inspections. This is despite the fact that not only is the risk much greater in the ground beef plant than it is in the chicken noodle soup plant, but there is probably very little difference in risk between a chicken noodle soup and a tomato soup plant. That is where we will be most productive, Raymond said, fixing that type of problem.

Sundlof agreed with Raymond. He said that the notion of a unified food safety system is a “noble idea” but that the amount of work that would be required to harmonize the two systems and re-write the laws and regulations of a new unified agency is daunting. If the United States were just at the beginning stages of developing a food safety system, then yes, it would be a valid idea. But not now.

DeWaal referred workshop participants to the IOM/NRC Ensuring Safe Food: From Production to Consumption report and remarked that, in fact, the notion of a single unified food safety agency is very much a topic of discussion in the 110th Congress. She mentioned efforts by the chairwoman of the Agricultural Appropriations Committee (Rosa DeLauro), who manages both the CFSAN and FDA budgets, and the efforts of Dick Durbin, Democratic Whip in the Senate and a key congressional legislator. There are many bills currently under consideration that would bring us closer to this vision of a unified agency, she said. It will be a step-wise effort, with most of the early efforts being directed to toward making improvements at the FDA because of the critical condition of that agency.

A member of the audience who identified himself as a local inspector who visits many different types of plants, including both FDA and USDA facilities, said that originally he was ambivalent about this notion of a single agency (i.e., nine years ago, when he started conducting these inspections). But now, after nine years of dealing with recalls, trace backs, detentions and other issues, he has realized that a single agency is in fact very important as a means of ensuring that the focus on food safety be retained and that decisions and actions around food safety not be impacted by political motivations.

Another audience member, Donna Rosenbaum, Executive Director of S.T.O.P. (Safe Tables Our Priority), commented on the fact that S.T.O.P. has been attending food safety meetings such as this one for a very long time and still finds many of the same issues (e.g., safety along the entire farm-to-fork continuum) discussed but unaddressed. She wondered if perhaps part of the inaction is because of agency focus and the fact that each agency is overseeing only its portion of the food safety system(s). No single agency is wrapping their arms around some of these very important issues, such as water quality and animal waste management. She asked, how would a single food safety agency be able to deal with these two issues in particular? Rather than addressing this specific question, the discussion turned toward the current regulatory and research status of these two issues.

Water Quality and Waste Management

Robach said that, from a processing standpoint, water quality is an extremely important issue. Water is obviously a key input into all food production processes, and Cargill treats water as a critical ingredient and evaluates it the same way it does any other product ingredient. Cargill assumes responsibility to demonstrate that the water it uses during food production meets drinking water standards, and USDA has access to those verification and validation records if necessary. Cargill also works closely with public water supply sources, as water is already fairly well-regulated at the state level. He argued that he is not sure how government would be able to contribute to that effort; likewise with animal waste management. Cargill works with state environmental protection agencies with respect to waste utilization, methane recycling, composting, and disposal issues, clean air standards, etc., all of which are again fairly well regulated (i.e., at the state, not federal level). There is already enough government oversight, he insisted.

Rosenbaum commented that state-level regulation of water is actually problematic since it contributes to varying water qualities. Sometimes the quality of the water used in irrigation is not the same as the quality of water being used to wash plants in the packing houses. In California, for example, tertiary water is sometimes used for plant irrigation. She asked, isn’t this a problem?

Sundlof agreed that, absolutely, the use of tertiary water is problematic. FDA’s GAP guidance deals specifically with the issue of water quality (e.g., that the water being used for irrigation or processing is not transmitting pathogens) as well as worker sanitation (e.g., that workers have access to port-a-potties, hand washing stations, and other sanitation tools and are not the cause of the problem). While that guidance is not law, FDA does have the statutory authority to require that food is produced under conditions that will render it not injurious to health, to identify problems with water quality and to take regulatory actions when those problems arise. As one example of FDA’s role in this area, after identifying a Salmonella problem in cantaloupe imported from Honduras, FDA conducted on-site inspections on the implicated cantaloupe farms and found many violations. There were problems with both water quality and biosecurity (e.g., birds roosting in areas where cantaloupe was processed). Because of these problems, Honduras is on import alert right now and will not be allowed to ship cantaloupe to the United States until they have improved their production system to a point at which we believe they can produce safe cantaloupe.

Bruhn noted that both issues (i.e., water quality and waste management) were the subjects of active areas of research. At the University of California, Davis, for example, a scientist in the School of Veterinary Medicine is funded by a major grant to examine and identify sources of contamination in aquifers. Other researchers are examining the various ways that composting can lead to sufficient heat generation (i.e., as a way to destroy microbes).

Food Safety Research

While on the subject of research, Bruhn also commented on the fact that most food safety research is conducted in the United States because of the constraints of the funding agency(-ies). This is true even though research conducted elsewhere might be pertinent in this country.

Bruhn also expressed frustration that a recent Request for Proposal (RFP) for research relating to consumer safe handling excluded consumer research (e.g., the type of research that Bruhn conducts) because it would involve human subject research. While Bruhn agrees that there is a need for laboratory research, which was the focus of this particular RFP, there is also a need to understand how people behave (e.g., that they eat raw dough) and what can be done to change that behavior.

Lessons Learned: Room for Improvement

Rosenbaum commented on the fact that upwards of 90 percent or more of food borne illnesses are sporadic cases, not outbreak situations, and by focusing on outbreaks we are failing to understand the greater majority of food borne disease. This was followed by a related comment by S.T.O.P. president and spokesperson Nancy Donley. Donley remarked that after closely following food safety issues for 15 years, ever since the Jack-in-the-Box epidemic really catapulted the issue of food safety into the public eye, despite new technologies and other innovations and new research, and new money allocated to agencies for food safety projects, food borne illnesses still occur and people still die from them. Why? Where have we gone wrong?

Raymond replied that improvements could be made in almost any area of food safety that has been discussed today. The complexity of the issue makes it very difficult to pinpoint any single most important problem area. It is an extremely difficult question to answer. Raymond agreed with Rosenbaum that we don’t understand most food borne illnesses and that, until we do, it is going to be difficult to make many dramatic changes. That said, there have been some significant improvements. For example, Salmonella numbers in poultry have gone down, although whether or not that translates into a decrease in Salmonella infections in people who eat poultry remains to be seen.

Raymond also noted that some of the trends, for example with E. coli O157, with the numbers getting worse, could be a result of improved surveillance and not an actual increased risk. With PFGE and other advanced technologies, we might be detecting outbreaks that may not have been identified as outbreaks in the past. Or, more people might be seeking medical care than in the past because they read about E. coli in the paper or otherwise are aware of the problem. So are we getting better at detecting contamination, or is the contamination getting worse? It could be either or both, he said. Raymond also noted that the USDA is currently considering several changes, which if approved, will hopefully lead to additional improvements in the next few years, such as the use of low-dose irradiation of beef carcasses to kill E. coli O157:H7.

Sundlof said that he too wished he had an answer. The FDA was encouraged by the drop in food borne outbreaks of Salmonella, E. coli O157:H7, Campylobacter, and other pathogens detected after initiating a tracking program in the mid-1990s (as part of the HHS Healthy People 2000 initiative). The agency set higher goals for Healthy People 2010. Some of the numbers have gone down, but the FDA has been very disappointed with the worsening trends for many foodborne pathogens. He said that the FDA asks itself this same question: Why? For example, are these microbes more capable of thriving in the open environment than in the past? Have they evolved the capacity to contaminate foods that they were not able to contaminate in the past? Or has our surveillance improved, and are we simply detecting outbreaks that we missed in the past?

Bruhn argued that perhaps part of the answer is to be open toward new technologies and to communicate the risks and benefits of these technologies to the public. For example, with respect to the issue of irradiation, Bruhn commented on how research in the past demonstrated that irradiating leafy greens wilted them. Now, with technological advances, irradiation can result in a “two to three log reduction, at least” and products that are indistinguishable from non-irradiated products both, in terms of nutritional value as well as texture. She noted a 2001 publication in Emerging Infectious Diseases by Robert Tauxe,8 where Tauxe argued that if half of all ground beef, poultry, and processed meats were irradiated, over 350 lives would be saved every year and more than 6,000 serious foodborne illnesses averted.

DeWaal remarked that FDA has approved irradiation of spinach and lettuce, which marks a step forward in improving food safety for leafy greens.

Groth agreed that FDA approval of irradiation for leafy greens is an important step forward but emphasized that irradiation is not a “silver bullet.” It is still unclear whether the technology will serve as a practical tool for solving various problems. Most of the work to date has been in the laboratory, leaving many unanswered questions about commercial production; and whether and how the technology can be fine-tuned to be effective against specific pathogens on specific substrates. Groth noted that it is unclear whether even a two- or three-log reduction is sufficient to meet the safety objectives of both producers and consumers given the way these products are handled after processing. There is still a lot more work to be done, and it will probably be a few more years before we see a market for irradiated leafy vegetables.


Caswell asked the panel about the quality of traceability in the United States. She said that she was really struck by press reports about traceability in tomatoes and stories about how tough it is to tell where tomatoes originate, as though it is impossible to do that. Likewise with ground beef. She asked, how much would traceability cost relative to the benefits? And is the United States falling behind other developed nations with respect to traceability? If so, how can we improve our traceability capacity?

Raymond commented that, obviously, the USDA does as much traceability as it can to determine contamination sources. This can be especially difficult with ground beef in particular, given that grinders mix and match products not only from different slaughterhouses but from different countries (i.e., trim from the United States is blended with trim from other countries in order to make a leaner burger). The Topps recall, for example, involved 12 sources, making traceback very difficult. Fortunately, in that outbreak, contaminated product was also identified in another plant involving only one supplier (i.e., one of Tops’s 12 suppliers), giving USDA the information they needed to trace the contamination back to that single slaughter plant. Otherwise, with multiple suppliers, it is a very difficult exercise. USDA is encouraging industry to hold all product that is tested before it is distributed. Most of the larger companies are now doing this, so there has been some progress. But it is more difficult for smaller companies. Another indication of progress, Raymond pointed out, is USDA’s “Steps Program,” whereby multiple suppliers to a single processing plant are entered into the program and, if identified twice within a certain time period as being a possible source of contamination, they are required to undergo testing at an increased frequency.

DeWaal answered, “With respect to animal identification, we are way behind.” She commented that, because of the BSE crisis, the European Union now has a very extensive animal ID system in place. In fact, most countries trading in the world market have extensive animal ID systems in place, since they must verify that animals entering EU markets have not been treated with hormones. With respect to tomatoes and other produce, DeWaal suggested that the stickers currently being used for check-out purposes at the supermarket could also be used as a way to track those food items back to the farm. In fact, the U.S. Congress passed a law in 1930, the Perishable Agricultural Commodities Act (PACA), requiring that farmers be able to trace their products from the farm to their first distribution point [one must have a PACA license in order to operate a produce business]. There is also the 2002 Bioterrorism Act, which requires that traceability subsequent to distribution. So a tomato distribution plant, for example, can identify where their tomatoes came from (as per PACA) and where they go (as per the Bioterrorism Act) but not necessarily what happens to those tomatoes while in the distribution plant. In other words, there is no internal traceability, and the identity of those tomatoes is lost. We need to examine both external and internal traceability, she argued. Again, in Europe, meat-processing plants have both good external and internal traceabilty, so it can be achieved. If you were to go to a veal plant in the Netherlands, for example, you would be able to identify not only where every calf was raised but also where every piece of meat from a single animal was shipped and what grain products were used to feed that animal.

Sundlof agreed that traceability is possible. But is it feasible? He mentioned that there are sophisticated technological solutions to some of these problems in the works but that tomatoes present a worst-case scenario. Tomatoes from multiple farms are combined and then sorted by size and color in the processing plant; so there is a lot of co-mingling of tomatoes from different farms. Then, that co-mingled group of tomatoes may be sent to another distributor and combined and then sorted again. Then, products co-mingle again during restaurant distribution, with distributors selling and buying back products from “jobbers,” etc. In order to sticker-mark a tomato for traceability purposes, he said, you would have to apply a sticker at ever step along the supply chain and could conceivably end up with a tomato completely covered by stickers. Again, there are some technological solutions being developed and tried, for example, the use of lasers to etch bar codes into produce, but it remains to be seen how well these will work.

Robach remarked that animal identification (ID)/traceability is not a major issue in the vertically integrated poultry and hog industries. It is, however, a problem with the beef industry, not just because of the way that beef animals are raised (e.g., being sold multiple times before reaching a feedlot) but also because there is no premise ID program in the United States.

Microwaveable Foods: Challenges and Changes

Doyle asked the panel about microwaveable foods and mentioned the recent E. coli O157:H7 outbreak associated with microwaveable pizza and the Salmonella outbreak traced back to microwaveable pot pies. He noted problems with the way microwave units are sold, such as the fact that they are not all the same wattage and do not cook at the same level, etc. Also, consumers often have different expectations about whether these products need to be actually cooked versus just heated. Do these types of foods need to be free of harmful bacteria, or should they be ready-to-eat with the onus on the consumer to make sure that he or she cooks the products sufficiently to kill E. coli and Salmonella?

Raymond responded by saying that all food should be free of pathogenic bacteria, even raw poultry. In reference to the Salmonella outbreak that was traced back to pot pies, he noted that in fact a lot of people did not cook the pies correctly. He noted the typically low wattage of microwaves in, for example, college dorm rooms and the reality that most people do not leave their microwaveable products in the microwave as per many product cooking instructions (e.g., cook on “high” for four minutes and then leave in the microwave for two minutes). He asked, why don’t companies instruct their consumers to simply cook it for the necessary time (i.e., without requiring that the cooked food sit before eating)? In fact, industry is working on this issue, trying to make their instructions easier to understand and less variable. Raymond also noted that some people like the raw dough of pot pies and will eat that dough and only microwave the rest of the contents. So there are a lot of consumer behavior issues that need to be considered. Finally, he noted that there were some problems with the epidemiology investigation surrounding that outbreak and that the initial assumption was that chicken pot pies were to blame; more specifically, that chicken was to blame, not the flour, even though flour could have been the problem. So the USDA received all of the initial blame, with FDA “off the hook.”

Sundlof noted that the FDA is currently dealing with similar issues: that products not labeled as ready-to-eat and with a cooking step are not always being appropriately prepared (cooked) by the consumer. For example, frozen peas and other vegetables meant to be cooked often end up in salads or salad bars without having had a kill step applied to them. This happens even though the labeling clearly states that the products should be cooked before consumption. Should we reclassify these foods as ready-to-eat foods and require a kill step during processing? The problem with that approach, he said, is that people like eating foods that they consider fresh.

Doyle added that raw products with grill marks (i.e., raw products that look like they have been cooked) can be problematic. He noted a recall of grill-marked raw chicken because of Salmonella contamination and the fact that the labeling did not caution consumers that the product was raw.

Gaining an International Perspective

Groth commented again on the lack of a developing country perspective in this workshop. He mentioned an upcoming food production meeting in Beijing, China, where food safety would be featured prominently. He also mentioned a meeting that was held earlier in the summer, in Europe, where about half of those in attendance were from exporting developing countries. He suggested that a similar meeting be held in the United States, where industry, government agencies, and civil society representatives from the United States and other importing countries meet with representatives from major exporting developing countries. Groth asked the panel if such a meeting would be useful and, if so, who would sponsor it.

Sundlof agreed that the European meeting was one of the most useful meetings he had ever attended. It was a forum where all importers from countries outside the EU attended to discuss requirements for meeting EU standards (i.e., not just HACCP requirements but also requirements demonstrating that foods are being produced under EU-mandated standards, etc.). Sundlof was amazed that what most people would consider the least developed countries in the world have been able to put systems in place to meet those very high European standards. There is a lot to be learned from that: It is possible to meet very high food safety standards when there is market incentive and with sufficient technical capacity or support.

Raymond agreed that any time experts with an interest in food safety meet the meeting is useful, no matter where the meeting is held. He mentioned that in fact the USDA already meets with some of those stakeholders regularly, for example through FAS and Codex (i.e., with respect to the food safety aspects of free trade agreements with Mexico and elsewhere) and as part of an organization known as QUAD (i.e., the Quadrilateral Group). QUAD is a collaborative group comprising food safety experts from Australia, Canada, New Zealand, and the United States. Australia, Canada, and New Zealand account for about 75 to 80 percent of imports into the United States. So there is a lot of this type of interaction and dialogue under way, he said, although nothing on the scale of the Beijing meeting. Raymond commented that he was unsure who would sponsor such a meeting.

Robach suggested that it might be possible to engage the United Nations (e.g., FAO), given its interest and work in food security, nutrition, and public health issues in the developing world. He mentioned that one of the most valuable meetings that he regularly attends is a meeting of Cargill food safety professionals from all 66 countries where Cargill operates. The meeting involves discussing key challenges faced over the past year and how those challenges were addressed, making for a tremendous learning opportunity.

DeWaal agreed that such a meeting would be “hugely beneficial” and noted that the CSPI has been trying to set one up with the NGO community but that the funding is unavailable. She mentioned that there is not much interest in standardizing and generating activity at that level.

Sustainability and Other Future Issues

Brackett asked a question about the issue of sustainability and whether the drive to use less energy, reuse water, etc., is sometimes in direct competition with the drive to achieve good food safety. How are the regulatory agencies going to deal with these competing interests? Sundlof replied that federal regulatory agencies are already dealing with some of these issues. As one example, he pointed to USDA’s “scorched earth policy” which involves eliminating wildlife within a certain area of agricultural land in Salinas Valley, California, and the problems created with respect to the loss of natural habitat for many species. Sundlof noted that sustainability in particular is not a major problem yet. When it does become a major problem, the agencies will deal with it in turn.

Robach commented that Cargill has had plentiful opportunities to deal with this sort of issue, for example in Brazil with respect to soybeans, Indonesia with respect to palm oil and in Ghana and the Ivory Coast with respect to cocoa beans. Robach also commented on the zoonotic origin of many food pathogens and the related issues around that. Cargill works with SSAFE (Safe Supply of Affordable Food Everywhere, Inc.) a public–private partnership dedicated to improving food production and control systems in developing countries while fostering sustainable agriculture and fisheries. Other SSAFE participants include, from the private sector, other large companies, like McDonalds, Nestle, and Pfizer, and from the NGO and public sectors, the Wildlife Conservation Society, the Center for Science in the Public Interest, CARE, the World Organization for Animal Health (OIE), and the Food and Agriculture Organization (FAO). Robach said that managing these issues is an ongoing process that requires both sides of each issue working together.

Raymond noted that there are lessons to be learned from other countries, for example Australia which has been struggling with a severe drought and has been forced to use reconditioned water in its slaughter facilities. In the United States, we use antimicrobial rinses as a way to save water. Chile, on the other hand, uses a tremendously high amount of water per bird in its poultry plants as a way to avoid the use of chlorine, since the EU doesn’t allow the import of poultry with chlorine.

Finally, DeWaal reminded the Forum and workshop participants of the transient nature of many of the specific issues addressed. Food scarcity, water use, climate change, and zoonotic issues, for example, are going to become much more important issues in the future. Some of what we may be dealing with today may only be a foreshadowing of what it is to come and we need to be thinking beyond the current problems.

Along this line, Doyle pointed to another future issue: the emergence of opportunistic pathogens that lead to serious illness only in certain populations (e.g., immunocomprised individuals) and which may be very difficult for the food industry to manage. He pointed to colitis-causing Clostridium difficile as an example. C. difficile often infects hospitalized patients being treated with antibiotics, although researchers are increasingly reporting its occurrence outside of the hospital environment as well—including in animal food products (e.g., it has been detected in braunschweiger and other meats). He asked, are we going to have to resort to making canned ham for all in order to protect the small percentage of people who are vulnerable to infection by C. difficile (or any other pathogen)? Ensuring safe food for all is going to become more difficult as these opportunistic pathogens continue to emerge.



Michael P. Doyle, PhD, is Regent Professor of Food Microbiology and Director of the Center for Food Safety at the University of Georgia, Athens, GA.


Stephen F. Sundlof, DVM, PhD, is Director of the FDA Center for Food Safety and Applied Nutrition, College Park, MD.


Ms. Caroline Smith DeWaal is Director of the Food Safety Program at the Center for Science in the Public Interest, Washington, DC.


Kessler was Commissioner from November 18, 1990, to February 28, 1997.


Mr. Michael C. Robach is Vice President of Corporate Food Safety and Regulatory Affairs for Cargill, Inc., Minneapolis, MN.


Richard Raymond, MD, was appointed Under Secretary for Food Safety at the USDA’s Food Safety and Inspection Service in 2005.


This section is a paraphrased summary of the lengthy discussion that followed the panelists’ comments.


Tauxe, R. V. 2001. Food safety and irradiation: Protecting the public from foodborne infections. Emerging Infectious Diseases 7(3 Suppl):516–521.

Copyright © 2009, National Academy of Sciences.
Bookshelf ID: NBK26434


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...