THE NATIONAL ACADEMIES

Advisers to the Nation on Science, Engineering, and Medicine

BOARD ON AGRICULTURE AND NATURAL RESOURCES

500 Fifth Street, NW Washington, DC 20001 Phone: 202 334 3062 Fax: 202 334 1978 E-mail: banr@nas.edu www.dels.nas.edu/banr

March 11, 2009

Mr. Donald W. Anderson

Deputy Director

Program Evaluation and Improvement Staff

USDA/FSIS/OPEER

1400 Independence Ave., SW Room 3833 South Building

United States Department of Agriculture

Washington, DC 20250-3700

Dear Mr. Anderson:

At the request of the Food Safety and Inspection Service (FSIS), the National Academies’ Division on Earth and Life Studies established the ad hoc Committee for the Review of the Methodology Proposed by the Food Safety and Inspection Service for Risk-Based Regulation of In-Commerce Activities. The committee’s charge was to comment on the new methods proposed by FSIS to organize its regulation and inspection of in-commerce businesses.


The committee held two in-person meetings. At the first meeting, FSIS staff presented their proposed approach and answered questions raised by committee members. The second meeting was spent writing and editing the report. Staff members from FSIS were available throughout the process to answer any additional questions.


This letter report contains the committee’s responses. The committee commends FSIS for its efforts in developing a useful first approach to risk-based surveillance activities for in-commerce establishments. The report contains several recommendations for consideration by the agency as it continues to develop in-commerce surveillance activities to protect public health.

Sincerely,

John N. Sofos,

Chair

Committee for the Review of the Methodology Proposed by the Food Safety and Inspection Service (FSIS) for Risk-Based Regulation of In-Commerce Activities



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 1
BOARD ON AGRICULTURE AND NATURAL RESOURCES 500 Fifth Street, NW Washington, DC 20001 Phone: 202 334 3062 Fax: 202 334 1978 E-mail: banr@nas.edu www.dels.nas.edu/banr March 11, 2009 Mr. Donald W. Anderson Deputy Director Program Evaluation and Improvement Staff USDA/FSIS/OPEER 1400 Independence Ave., SW Room 3833 South Building United States Department of Agriculture Washington, DC 20250-3700 Dear Mr. Anderson: At the request of the Food Safety and Inspection Service (FSIS), the National Academies’ Division on Earth and Life Studies established the ad hoc Committee for the Review of the Methodology Proposed by the Food Safety and Inspection Service for Risk-Based Regulation of In-Commerce Activities. The committee’s charge was to comment on the new methods proposed by FSIS to organize its regulation and inspection of in-commerce businesses. The committee held two in-person meetings. At the first meeting, FSIS staff presented their proposed approach and answered questions raised by committee members. The second meeting was spent writing and editing the report. Staff members from FSIS were available throughout the process to answer any additional questions. This letter report contains the committee’s responses. The committee commends FSIS for its efforts in developing a useful first approach to risk-based surveillance activities for in- commerce establishments. The report contains several recommendations for consideration by the agency as it continues to develop in-commerce surveillance activities to protect public health. Sincerely, John N. Sofos, Chair Committee for the Review of the Methodology Proposed by the Food Safety and Inspection Service (FSIS) for Risk-Based Regulation of In-Commerce Activities ix

OCR for page 1

OCR for page 1
SUMMARY The committee believes that a risk-based approach for surveillance activities of in- commerce establishments is appropriate. It commends the Food Safety Inspection Service (FSIS) for its efforts. The 13 business types identified by FSIS adequately cover the array of in-commerce facilities that handle meat, poultry, and egg products. However, many of the business types are very diverse and the committee recommends dividing some of the 13 types into more precise categories. The committee reviewed the five risk considerations. Although all of them are considered important, several modifications are recommended. It was difficult for the committee to evaluate fully the concept and use of the “inherent hazard” risk consideration because of insufficient description and lack of clear direction for its use in categorization of facilities into risk groups. There is little discrimination among a majority of the establishment types based on inherent hazard as a risk factor; this is evident by the fact that most were rated by FSIS as “medium.” The committee agrees with most of the food defense vulnerability rankings, but it believes rendering facilities could also qualify for the “high” risk category. Product volume is an important risk consideration and the committee found that the risk rankings based on product volume are appropriate. Additionally, however, the committee determined that FSIS should seek adequate and reliable data on product volume handled by various facilities. The committee found that the consumer susceptibility risk consideration is not equally applicable to all institutions. This limits the value of this risk consideration in prioritizing surveillance frequency for all types of institutions, some of which serve less susceptible populations. The committee concluded that surveillance by other authorities may be the primary risk consideration as a selection criterion for surveillance prioritization. However, because surveillance activities by other jurisdictions vary by region and because there is considerable diversity within and among various business types, the four other risk considerations may be used to enable secondary decisions about surveillance prioritization. In addition, the committee found that FSIS lacks adequate data on inspection activities by other jurisdictions. Availability of such information is critical in establishing surveillance objectives, approaches, and priorities. The FSIS should consider partnering with local and state regulatory agencies to increase its knowledge of existing meat, poultry, and egg product surveillance activities and use the information to improve priority-setting of its own surveillance activities. In addition, FSIS should consider serving as a resource for proper meat and poultry inspection or signing memoranda of understanding with other jurisdictions in order to avoid duplication of effort and to achieve optimal inspection of all facilities in need. Finally, the committee recommends use of a stepwise decision process to set priorities for surveillance of establishments rather than the proposed three-tier system. It suggests that FSIS creates a single category of high-risk business types and areas where it has sole jurisdiction and that most (potentially 90%) of available resources be devoted to surveillance of such facilities. 1

OCR for page 1
2 REVIEW OF IN-COMMERCE SURVEILLANCE INTRODUCTION The Food Safety and Inspection Service (FSIS) of the U.S. Department of Agriculture (USDA) is the federal agency responsible for enforcing the regulatory statutes applicable to meat, poultry, and egg products. The goal of the regulations is to ensure that the products are wholesome and safe for human consumption. To enforce its regulatory mandate even outside processing plants (receiving continuous inspection), FSIS also conducts surveillance of businesses that are engaged in the transport, storage, distribution, and sale of meat, poultry, and egg products. The subject of this report is the review of plans by FSIS to use a risk-based approach in enforcing regulations for business types involved in so-called in-commerce activities for products under its authority (meat, poultry, and egg products). Throughout the report the term surveillance is used to describe the oversight of in-commerce businesses because this is the term preferred by FSIS. As part of its effort to move toward a more risk-based regulatory enforcement approach and food-safety assurance, FSIS has proposed new methods to organize its surveillance of in- commerce businesses. The agency has asked the National Academies to review and comment on its proposals. The Committee for the Review of the Methodology Proposed by the Food Safety and Inspection Service (FSIS) for Risk-Based Regulation of In-Commerce Activities (hereinafter referred to as “the committee”) was appointed for that purpose. The proposals were prepared by an FSIS In-Commerce System Public Health Risk-Based Surveillance Workgroup (the ICS Workgroup) and were described in an FSIS summary paper1 and in material presented to the committee at public sessions on November 6 and 7, 2008 (Appendix A). The specific task assigned to the committee is presented in Box 1. This report contains the committee’s findings and recommendations; it was reviewed by the persons acknowledged in the front matter. Box 1 Statement of Task An ad hoc committee will examine the methods proposed by the Food Safety and Inspection Service (FSIS) to inspect and regulate in-commerce activities using a risk-based approach. The committee will review and comment on the data, assumptions, risk factors, and methodology FSIS will use to rank the relative public health risks posed by the various types of in-commerce businesses that handle meat and poultry products. The committee will also review and comment on the agency's preliminary risk rankings. A letter report will be issued. According to FSIS, its Office of Program Evaluation, Enforcement, and Review has historically conducted “surveillance reviews” at in-commerce facilities to help to ensure that meat, poultry, and egg products in commerce are wholesome; unadulterated; appropriately marked, labeled, and packaged; and secure from intentional acts of contamination. However, because the number of in-commerce facilities is large (over 700,000; the number was determined by FSIS from information provided to the agency by a vendor on the basis of the Standard 1 Document provided to the committee by FSIS titled Public Health Risk-Based “In-Commerce” Surveillance, October 25, 2008.

OCR for page 1
LETTER REPORT 3 Industrial Classification Codes of the U.S. Department of Labor) and the number of Compliance Investigations Division (CID) investigators is small (approximately 108), CID managers informally set priorities among facilities for surveillance reviews. In 2008, an electronic In- Commerce System (ICS) was implemented to improve the use of FSIS resources by capturing the results of surveillance reviews and enabling more formal priority-setting. According to testimony of the then USDA Undersecretary for Food Safety Richard Raymond2 and FSIS Directive 8010.1,3 the ICS was intended to complement inspection activities performed in plants and surveillance activities of in-commerce business types were to be prioritized to protect public health. An important element of the plan is that the priority-setting process targets facilities with no prior surveillance history. It should be noted, however, that the majority of CID investigators’ time is allocated to emergency response, outbreak investigations, and product recall and withdrawal activities. Those activities merit a higher priority because of their immediate and critical nature. However, many other in-commerce facilities have not been subject to surveillance activity, and these are the facilities the CID wishes to prioritize to ensure that surveillance activities maximize the public health benefit. RESPONSE TO FSIS QUESTIONS Adequacy of the 13 Business Types Thirteen types of businesses—distributors, warehouses, transporters, restaurants, institutions, retail salvage, food banks, retailers, 4-D (dead, dying, disabled, or diseased) facilities, renderers, animal feed, custom exempt, and exempt poultry—were identified by FSIS for in-commerce risk-based surveillance, and the committee was asked to comment on whether the businesses were identified and categorized correctly and whether any additional businesses should be considered. The committee found the identification and categorization of business useful for priority-setting for CID surveillance. Although the committee agreed that all relevant types of businesses were identified by the FSIS ICS Workgroup, it raised several questions about the categorization. The intent of the FSIS ICS Workgroup in identifying the 13 “significant” in-commerce business types was to place each type into one of three tiers according to the need for CID surveillance (see section “Three Surveillance Tiers”). The committee’s consensus on the business types is presented below. 1. Distributors—Distributors purchase meat, poultry, egg, and other food products from multiple domestic and overseas suppliers, store them (refrigerated, frozen, or at ambient temperature), and sell them to multiple customers, including retailers, hotels, restaurants, and institutions. The FSIS indicated that there are 8,262 distributors in its database.4 Until 2 http://www.fsis.usda.gov/PDF/Testimony_Raymond_030806.pdf. Accessed January 26, 2009. 3 http://www.fsis.usda.gov/OPPDE/rdad/FSISDirectives/8010.1Rev1.pdf. Accessed January 26, 2009. 4 The numbers of businesses in the FSIS database were provided to the committee. The committee did not seek independent confirmation of the numbers of businesses in each business type.

OCR for page 1
4 REVIEW OF IN-COMMERCE SURVEILLANCE an investigator visits an establishment, the amounts or kinds of meat, poultry, and egg products present, which may vary from time to time, are unknown to the agency. Little or no processing occurs at distributors, and in most situations they do not even “break bulk” (i.e., remove packages from cases or pallets of product). However, temperature control and maintenance of package integrity are important food-safety controls for meat, poultry, and egg products in these and other facilities because of the potential for cross-contamination, recontamination, or microbial growth in potentially large amounts of products. Some distributors may export products and may perform identification service under the supervision of an inspector. Additional surveillance by a CID investigator seems redundant in such situations. Furthermore, because states commonly have licensing and inspection programs for food-distribution centers, it may be useful for the CID to obtain information from state agencies on surveillance history, products handled, amounts of products, and other relevant information. The Food and Drug Administration (FDA) may also be consulted as a source of input inasmuch as it may conduct surveillance in distribution centers that handle FDA-regulated products. 2. Warehouses—Warehouses are similar to distributors in that they receive meat, poultry, egg, and other food products from domestic and overseas suppliers and store them under refrigerated, frozen, or ambient conditions. The primary difference is that warehouses typically do not take legal ownership of the products. The FSIS indicated that there are 23,625 warehouse businesses in its database. Until an investigator visits an establishment, the amounts or kinds of meat, poultry, and egg products that are present and handled are unknown to the agency. As with distributors, temperature control and package integrity are important for product safety. Some warehouses may operate under a voluntary grant of inspection. In such cases, additional CID surveillance may not be warranted. States commonly have licensing and inspection programs for food warehouses, which may be useful to the CID for information on surveillance history, products handled, volume, and other relevant information. The FDA may also conduct surveillance in warehouses that handle FDA-regulated products and could provide the CID with useful information. 3. Transporters—Transporters move products from one facility to another for a fee. They do not buy, sell, or process products. The FSIS indicated that there are 16,754 transporters in its database. Until an investigator is involved, the amounts or kinds of meat, poultry, and egg products that may be transported at any given time are unknown to the agency. This is a large category and some types of transporters (e.g., those involving importation of product) may have greater relevance for surveillance activity. Systematic surveillance of transporters, other than those involved in product importation, may not be the best use of the limited resources available, because customer–supplier relationships often require transport-vehicle inspection, temperature monitoring, and similar practices to protect the safety and quality of products. 4. Restaurants—Restaurants prepare meals for on-site or off-site consumption by the general public. The FSIS indicated that there are 500,000 restaurants in its database. Until an investigator visits an establishment, the amounts or kinds of meat, poultry, and egg products that are present are unknown to the agency. The FSIS ICS Workgroup did not distinguish among types of restaurants—quick- serve or full-serve, large or small, whether they are parts of a chain or are independently

OCR for page 1
LETTER REPORT 5 operated, and other factors that could influence the level of surveillance needed for control of public-health risk. Such considerations can be important in setting priorities for surveillance or investigation. Therefore, FSIS should reexamine whether restaurants should remain as one large group, since this would complicate assignment of risk factors among diverse establishments within this category. Because restaurants typically handle a wide variety of food products, surveillance activities focused solely on meat, poultry, and egg products may overlook other important food-safety risks. For example, Centers for Disease Control and Prevention (CDC) surveillance data5 suggest that meat and poultry products pose a substantial risk as sources of contamination of other products in restaurants and other settings, and this is probably not a focus for CID surveillance activities according to procedures in FSIS Directive 8010.1. In general, restaurants are under the regulatory jurisdiction of state and local agencies, which should conduct comprehensive inspections of all aspects of food preparation for all food categories. The intensity and frequency of inspection activities by such jurisdictions may vary, however, among states and localities. Considering the limited resources available, it seems that public health could be served better if FSIS acted as a resource for state and local inspectors in providing guidance on inspection parameters specific to meat, poultry, and egg products, and ensuring that adequate inspection was applied in all situations needed. That would remove many of in-commerce establishments from the list of those potentially in need of surveillance by FSIS. If, however, FSIS investigators receive information regarding specific issues related to adulteration of meat, poultry, or egg products in a facility and the problem is not being handled by a local authority, investigation would be warranted. 5. Institutions—Like restaurants, institutions use a wide variety of ingredients (in addition to meat, poultry, and egg products) to prepare meals for consumption. However, institutions prepare meals for resident populations in such facilities as hospitals, nursing homes, schools, military bases, and prisons. As with restaurants, the public-health risk associated with those settings involves the preparation and handling practices of all ingredients and foods, not just meat, poultry, and egg products. The FSIS indicated that there are 54,621 institutions in its database. Until an investigator visits an establishment, the amounts or kinds of meat, poultry, and egg products that are present are unknown to the agency. As with restaurants, the impact on public-health protection would be greater if FSIS served as a resource for state and local inspectors in providing guidance on inspection parameters specific to meat, poultry, and egg products rather than including these types of establishments in their surveillance efforts. If the agency continues to include institutions on its list for surveillance, the committee suggests dividing the institution category according to the populations served (see section “Adequacy of the Five Risk Factors”). Hospitals, nursing homes, and child- care facilities serve food to populations that include persons very vulnerable to foodborne illness, and it is reasonable to conclude that, as a whole, their clients are likely to be more susceptible than the general public to foodborne illness. However, the populations in other institutions such as military bases, middle and high schools, and colleges may, in 5 http://www.cdc.gov/mmwr/preview/mmwrhtml/ss5510a1.htm. Accessed January 26, 2009.

OCR for page 1
6 REVIEW OF IN-COMMERCE SURVEILLANCE general, be less susceptible than the general public. Therefore, maintaining institutions as one large group complicates assignment of risk factors. Like restaurants, most institutions are under state and local jurisdiction for inspection, while military bases have their own food-safety system and personnel. Coordination of efforts with the other jurisdictions could help to direct CID resources to areas in greater need and consequently help to maximize public-health benefits. 6. Retail Salvage—Retail salvage businesses purchase “distressed” products that other businesses may not be able to sell. They include products that may not meet quality standards, products that were frozen and may have undergone temperature abuse that has not affected safety, and products near the end of their shelf-life. The FSIS indicated that there are 162 retail salvage businesses in their database. Until an investigator visits an establishment, the amounts or kinds of meat, poultry, and egg products that are present are unknown to the agency. The committee does not know how often retail salvage businesses are inspected by officials of other jurisdictions, but it assumes that the frequency varies widely depending on facility location. Because their products are handled outside “normal” channels, surveillance of the facilities may be warranted to ensure that proper controls are in place to maintain the safety and wholesomeness of meat, poultry, and egg products. The products might also become part of food-bank inventories (see next section). 7. Food Banks—Food banks receive product from retail salvage, manufacturers, wholesalers, retailers, and government agencies for storage and redistribution to not-for- profit emergency and community food programs. The FSIS indicated that there are 642 food banks in their database. Until an investigator visits an establishment, the amounts or kinds of meat, poultry, and egg products that are present are unknown to the agency; however, because food banks attempt to provide a balanced diet for the populations they serve, it is expected that meat, poultry, and egg products are present. Food banks handle foods in many formats, including fresh, frozen, and shelf-stable products. Large food banks should typically undergo some type of third-party audit to ensure the safety and wholesomeness of the products that they provide. Some states also provide guidance and oversight regarding food banks and their activities. For example, in late 2008, Minnesota detected lead at high concentrations in venison donated to food banks and changed registration and metal-detection requirements for establishments that provided this meat to food banks. Ensuring the safety of food in food banks is important because people consuming it may be under considerable stress, which may increase their susceptibility to foodborne illness. In order to help with prioritization of visits, it is important for FSIS to determine the existence and extent of state and local inspection or other types of surveillance of specific food banks of all sizes. 8. Retailers—Retailers sell products directly to consumers for consumption off the premises and include grocery stores, supermarkets, convenience stores, street food vendors, and similar businesses. The FSIS indicated that there are 120,249 retailers in their database. Until an investigator visits an establishment, the amounts or kinds of meat, poultry, and egg products that are present are unknown to the agency. There is substantial variation in how food is handled in retail establishments. Some, such as many convenience stores, offer only manufacturer-packaged products, whereas others, such as delicatessens and street food vendors, handle open food products.

OCR for page 1
LETTER REPORT 7 Some retailers provide elaborate “ready-to-eat” meals, which may require extensive preparation, handling, and potentially transportation for catered meals. A serious concern for FSIS is the safety of luncheon meats, which are handled in retail or food service delicatessens. State and local inspection should occur in these establishments, and should address issues related to open packages and compliance with the FDA Food Code.6 As mentioned in the discussion of restaurants, food-safety issues in retail establishments require evaluation of handling and preparation practices for a wide variety of foods. Protection of public health in connection with retail establishments requires an integrated approach, rather than a focus solely on meat, poultry, and egg items. Except for retail establishments that focus primarily on meat, poultry, and egg products and have substantial handling of products on site, FSIS resources would be used better if it partnered with state and local inspection agencies to provide guidance for those jurisdictions in evaluating meat and poultry activities in retail establishments. 9. 4-D Facilities—4-D facilities handle dead, dying, disabled, and diseased animals, and they cannot legally put products into commerce for human consumption. The FSIS indicated that there are 64 4-D facilities in its database. The committee is unaware of any entity other than FSIS that would inspect these facilities frequently. To ensure, through record verification, that products of 4-D facilities are disposed properly and that they do not enter the human food chain, surveillance of the facilities seems appropriate for protecting public health. For example, concerns over bovine spongiform encephalopathy (BSE) justify surveillance of 4-D facilities to ensure that products are properly marked and to prevent them from becoming, unintentionally or intentionally, part of the human food supply. 10. Renderers—Renderers handle meat and poultry carcasses and their parts that are distressed or otherwise unsuitable for direct human use. The FSIS indicated that there are 383 renderers in its database. Like surveillance of 4-D facilities, surveillance of renderers seems appropriate for the protection of public health, if entities other than FSIS are not involved. The objective is to ensure, through record verification, the proper use or disposition of products that should not enter into the human or animal food chain. For example, alteration of rendering practices is thought to have contributed to the incidence of BSE in Europe. 11. Animal Food—Animal food businesses manufacture or purchase meat and poultry and use them in the manufacture of pet foods or to feed animals in zoos, mink farms, and similar facilities; they do not include manufacturers of feed for livestock species, such as poultry, swine, and cattle. It is not legal for meat and poultry handled in these facilities to be put into commerce for human consumption. The FSIS indicated that there are 2,754 animal food establishments in its database. Meat and poultry products are handled in diverse ways in the establishments. For establishments like zoos and animal parks that purchase meat to be fed to animals in their care, the committee concludes that there is minimal potential for the meat to enter the human supply chain. However, surveillance of facilities that manufacture pet food products seems appropriate, if entities other than FSIS are not involved, to ensure that procedures are in place for pathogen control and to keep the products from entering the human food supply. 6 http://www.cfsan.fda.gov/~dms/fc05-toc.html. Accessed January 26, 2009.

OCR for page 1
8 REVIEW OF IN-COMMERCE SURVEILLANCE 12. Custom Exempt—Custom exempt businesses do not sell meat but rather offer slaughtering and processing services. They slaughter animals or process carcasses brought to them by the owners. The FSIS indicated that there are 170 custom slaughter establishments in its database. Some states inspect such facilities, others do not. Because the products are similar to those under routine FSIS in-plant inspection, surveillance of the facilities seems appropriate if states do not duplicate the effort. However, the amount of products involved in the facilities is relatively small, and the products are used by a relatively small number of people, so the overall public health-impact of FSIS inspection is likely to be low, especially if the facilities are inspected by states. In states in which such facilities are not inspected, periodic surveillance by FSIS is warranted. 13. Exempt Poultry—Exempt poultry businesses slaughter and/or process small volumes of poultry (less than a total of 20,000 birds/year) and are exempt from continuous FSIS inspection, although they sell products to consumers and other establishments. The FSIS indicated that there are 100 exempt poultry establishments in its database. Some states may inspect the facilities, other states may not. Because the products are similar to those under routine FSIS inspection, surveillance of the facilities seems appropriate in states that do not have an inspection program. This may increase in importance in that interest in consumption of locally produced foods is increasing. Because the products may be sold to the general public, the committee concludes that surveillance of this type of business warrants higher priority than surveillance of custom exempt facilities. Finding: The committee finds that the 13 business types identified by FSIS are adequate to cover the array of in-commerce facilities that handle meat, poultry, and egg products. However, the diversity of several of the 13 business categories complicates effective priority-setting for surveillance based on public-health risk. Furthermore, because several types of businesses and some businesses within types are likely to be under the jurisdiction of other regulatory agencies, FSIS surveillance of them may result in duplication of effort. The amounts of meat, poultry, and egg products relative to other foods handled in some facilities are unknown, whereas they are the only products handled in other facilities. Recommendation: To improve priority-setting for risk-based surveillance of in- commerce facilities, FSIS should consider dividing business types that are very diverse—such as institutions, restaurants, and retailers—into more precise categories. Adequacy of the Five Risk Factors The FSIS ICS Workgroup used five public-health risk factors in categorizing the risk posed by each of the 13 business types discussed above. • Inherent hazards—food-safety hazards that are inherent in activities typically conducted or products typically handled by a business.

OCR for page 1
LETTER REPORT 9 • Food defense vulnerability—“weakness” in operations that can potentially be exploited by adversaries and lead to public-health harm. • Product volume—average volume of meat, poultry, and egg products handled by a business. • Consumer susceptibility—susceptibility of people who typically purchase or receive products from a business. • Surveillance by other authorities—surveillance by federal, state, local, and other jurisdictions. The committee was asked to comment on the adequacy of those risk factors for FSIS use in a risk-based approach to in-commerce surveillance activities. In general, a risk-based approach is appropriate for this activity; however, the term risk factors may cause some confusion for businesses that operate in accordance with the FDA Food Code, such as restaurants, retail establishments, and institutions. The FDA Food Code identified five major “risk factors” for foodborne illness in retail and food-service establishments: improper holding temperatures, inadequate cooking, contaminated equipment, food from unsafe sources, and poor personal hygiene. The committee proposes the use of another term, such as risk considerations, to avoid confusion and uses risk considerations, instead of risk factors, throughout the remainder of this report. Inherent Hazards As addressed in FSIS Directive 8010.1 (6/25/08), the agency’s main concern focuses on the potential to adulterate meat, poultry, or egg products in in-commerce business facilities. The FSIS report Public Health Risk-Based “In-Commerce” Surveillance, dated October 25, 2008, stated that “the inherent hazards in the different types of facilities depend mainly on the characteristics of the products and how the facilities handle them.” The report further stated that “facilities that handle products with inherently lower risk, and/or perform operations that are inherently less hazardous, pose less public health risk.” On the basis of the information provided, committee members had various interpretations of the meaning of inherent hazard; therefore, more explicit criteria are needed to ensure that hazard assessment is applied consistently. In addition, however, raw meat, poultry, or egg products can serve as sources of contamination of other foods that are handled in the same facilities, and might increase the inherent hazard. The focus on meat and poultry products alone may therefore not maximize the public-health benefit of surveillance. For example, CDC surveillance data referred to in footnote 5 in the section “Adequacy of the 13 Business Types” above indicate that 22% of outbreaks in which investigators identified contributing factors were associated with inadequate cleaning of processing or preparation equipment and utensils, 12% with cross-contamination from raw ingredients of animal origin, and 11% with raw produce or ingredients contaminated by pathogens from animals or the environment (see Appendix B). In determining the significance of inherent hazard, it is important to consider that a large portion of foodborne illnesses are sporadic or not part of recognized outbreaks, the majority of outbreaks are of unknown source, and the vast majority of illnesses remains unreported. Therefore, caution must be used in making generalizations based on outbreak data. It is possible that sporadic cases are very different from those associated with identified outbreaks (i.e., transmission, pathogen, etc).

OCR for page 1
LETTER REPORT 11 • Animal food and 4-D facilities are similar in that their products are not intended to enter the human food chain. The agency’s concern is that the products could enter human food illegally. However, in the absence of additional information, the committee concluded that the inherent hazard in both business types is low. Additional information, such as suspicion that the products are entering the human food supply, may change that ranking. Finding: The insufficient description of “inherent hazard” by FSIS did not provide clear direction for categorization of facilities into risk groups. Significant diversity exists among some business types (e.g., restaurants), which prevents useful categorization and risk assignment. There is little discrimination among most business types on the basis of inherent hazard, inasmuch as most were rated “medium.” Public-health data suggest that the risk of cross-contamination between raw meat and poultry and other products is important, but CID surveillance procedures do not seem to address this in establishments that handle a wide variety of food products. Recommendation 1: “Inherent-hazard” criteria should be more explicitly defined. If this risk consideration is to be used successfully, criteria need to be defined in more detail to provide better discrimination for risk ranking. Recommendation 2: FSIS should consider partnering with other agencies to ensure that facilities with high inherent hazard receive adequate surveillance. FSIS should partner with agencies that take a comprehensive approach to inspection of facilities that handle a wide variety of food products because of the potential for meat and poultry products to serve as sources of contamination for other foods. The FSIS should collect data to define better the nature and frequency of hazardous “events” in categories and subcategories of facilities it may inspect. Food Defense Vulnerability The committee found that it could be helpful to consider the proposed risk considerations and surveillance tiers in the context of the complete food supply chain, which consists of three major components—production agriculture, food-processing operations, and in-commerce activities. Many potential food-safety and food-defense risks inherent in the food supply chain can be controlled by measures taken in production agriculture and food-processing operations. Thus, it is important that the in-commerce surveillance activities of the CID are undertaken with an awareness of such protective measures taken earlier in the food supply chain. The committee acknowledged that foods stored or handled in some types of in-commerce businesses may be more vulnerable than foods in others to deliberate contamination or adulteration with hazardous agents, as well as economic fraud, that may be purposefully introduced by political adversaries, deranged or criminal individuals, or disgruntled employees. These may be rare but are potentially very disruptive to both the economy and to the public’s trust in the food industry and the government oversight system. A case in point is specified risk material linked to the potential for survival of prions associated with BSE that have in the past contaminated beef destined for consumption within a country and for export to other countries. The degree of vulnerability of a particular food or food ingredient is related to its exposure and access during processing, transportation, and storage. Although all documented attacks on the

OCR for page 1
12 REVIEW OF IN-COMMERCE SURVEILLANCE U.S. food supply have been within the United States, vulnerable nodes for deliberate contamination may also exist during handling of food ingredients at their points of origin in exporting countries, during processing or final assembly of a food product in the United States, and during final food handling before consumption. Thus, although the responsibility for preventing adulteration of meat, poultry, or egg ingredients or products lies with the U.S. manufacturer or importer, the FSIS need to be assured that in-commerce businesses have systems in place to limit any illegal penetration of the food chain. Some food-defense activities can be based on established food-safety measures. Food- safety programs, such as hazard analysis and critical control point (HACCP) and good manufacturing practices (GMP), are widely used by the food industry to control naturally occurring contamination with microbial pathogens or chemical toxicants through a system of processing controls, robust supervision, effective training, and adequate management. Those programs can also provide protection against the deliberate contamination of food. The same rigorous procedures that are applied for food-safety assurance (e.g., process controls, supervision, and training) may often be applied for food-defense assurance. For that reason, food protection is used increasingly to include both food-safety and food-defense activities. Facility and transportation security and controlled access to exposed product areas are examples of protective features that are important for both food safety and food defense. To the extent practical, the “high-risk” designation may be determined by using such food defense tools as CARVER + Shock.8 The committee determined that the “low” risk ranking assigned to renderers should be reconsidered because of the potential for deliberate contamination with persistent organic pollutants9 and accidental contamination with BSE specified risk materials. Finding: The committee is in general agreement with the CID proposed “high” risk ranking for transporters, institutions, retailers, restaurants, and food banks, in terms of food defense vulnerability, but it believes that rendering facilities should also be included in the “high” risk category. Recommendation: The committee recommends that the “low” risk ranking for renderers be changed to “high” because of the lack of extensive oversight and the potential for deliberate contamination with persistent organic pollutants and accidental contamination with BSE specified risk materials. Product Volume The committee agrees that the volume (or amount) of a particular food product handled at various points in the food-distribution chain is an important consideration for public-health protection. Larger volumes of food contaminated, either intentionally or unintentionally, with infectious or lethal doses of an agent could be consumed by large numbers of people before the source of contamination is identified and controlled. Therefore, it is reasonable to assume that the volume of a product may be directly correlated with the number of consumers potentially at 8 http://www.cfsan.fda.gov/~dms/vltcarv.html. Accessed January 26, 2009. 9 Larebeke, van N., L. Hens, P. Shepens, A. Covaci, J. Baeyens, K.Everaert, J. L. Bernheim, R. Vlietinck, and G. De Poorter. 2001. The Belgian PCB and dioxin incident of January–-June, 1999: Exposure data and potential impact on health. Environ. Health Perspect. 109:265–273.

OCR for page 1
LETTER REPORT 13 risk. The “high” risk rankings assigned to transporters, distributors, and warehouses are appropriate. Additional consideration could be given to larger retailers and food banks. However, although individual retail stores, restaurants, or institutions may handle relatively small volumes of food, a single distributor, transporter, or warehouse may supply numerous sites. Accordingly, the detection of contaminated food at a “small-volume” retail store, restaurant, or institution may signal that a much larger volume of food has been contaminated at an earlier point in the food- supply chain. Finding: The committee agrees that product volume is an important risk consideration and that the CID risk rankings for volume are appropriate. It is noted, however, that FSIS is in need of adequate and reliable information on volume in various facilities as well as other pertinent and necessary data. Consumer Susceptibility The committee agrees that consumer susceptibility is an important risk consideration. The probability of infection and subsequent illness is a function of (1) the vulnerability of the host (e.g., age and immune resistance), (2) the number of units of the infectious agent ingested with food (e.g., viral particles, bacterial cells, and parasitic cysts), and (3) the virulence or pathogenicity of the agent. The first is related to consumer susceptibility and the second and third to the inherent hazard. According to the CDC, highly susceptible populations are more likely to experience foodborne disease than the general population because they include the immunocompromised, the ill, preschool-age children, and older adults. Establishments that serve such populations include child-care centers, kidney dialysis centers, hospitals, nursing homes (e.g., custodial-care, health-care, and assisted-living facilities), and nutritional or socialization services (e.g., senior centers). The committee notes that the estimate of 30 million people as likely to be particularly susceptible to foodborne disease in the United States (Healthy People 201010), which was based on a 1994 report, must be far greater at this time. By 2025, people 60 years old and older are predicted to make up 25% of the total population; FSIS will need to consider adjusting its priorities to respond to such projected demographic changes. Comparisons of establishments implicated in foodborne illness outbreaks should be considered carefully if they are to be used for surveillance priority-setting. Illnesses in schools, child-care centers, and residential-care facilities are usually caused by highly infectious viruses, such as norovirus and rotavirus, which are transmitted predominantly through person-to-person contact and less frequently through consumption of meat and poultry products (Appendix C). In hospitals, nursing homes, long-term–care facilities, schools, colleges, universities, and child-care centers, the contaminated vehicles involved in outbreaks are mainly ready-to-eat sandwiches, salads, and fruit and less likely to be meat, poultry, or egg products (see footnote 5 in the section “Adequacy of the 13 Business Types”). Although this observation uses outbreak data, which may be different than the contributing factors for sporadic cases, these are the best available data because sporadic cases are difficult to investigate. 10 http://www.healthypeople.gov/Document/pdf/Volume1/10Food.pdf. Accessed January 26, 2009.

OCR for page 1
14 REVIEW OF IN-COMMERCE SURVEILLANCE Thus, priority-setting for surveillance of institutions on the basis of food provided to several groups of susceptible populations is not clear, and the CID should consider hospitals and nursing homes, especially those catering to cancer and organ-transplantation patients, to be at higher risk with regard to susceptibility. In addition, commissaries that provide ready-to-eat food with meat, poultry, and egg ingredients to those types of “high-risk” institutions should be considered to be in the high-risk category for susceptibility. Food banks that distribute meat, poultry, and egg products may also be considered to be at high risk because many of the recipients are in the susceptible population group (e.g., the ill, the malnourished, and the elderly). However, in light of the limitations of FSIS resources, it is suggested that other facilities—such as middle schools and high schools, colleges/universities, and military bases, which are less often documented in foodborne-disease summary reports—should be considered to be at lower risk with regard to susceptibility. One approach that FSIS could consider is to separate institutions into only high and low risk for populations that can be considered susceptible. If that is done, it would lower the priority for CID surveillance of many institutional facilities. Also, as noted in Appendix C, the percentage of illnesses associated with institutions is relatively low compared with the total population of restaurants, caterers, and other food-service facilities. Finding: The committee finds that the consumer-susceptibility risk consideration is not equally applicable to all business types. It is apparent from the literature that although institutions with highly susceptible populations do have outbreaks, they are not often related to FSIS-regulated food: this, limits the value of this risk consideration in setting priorities for surveillance frequency. Recommendation: The FSIS CID should consider using the consumer-susceptibility risk consideration only in connection with institutions that primarily serve susceptible populations, which should be considered as at “high” risk; these institutions include child-care centers, primary schools, nursing homes, hospitals, and commissaries that cater to these institutions. Surveillance by Other Authorities The stated goal of the FSIS CID is to “prioritize surveillance activities based on public health risk.” The total available personnel resources for surveillance of in-commerce facilities is approximately 10.8 full-time equivalents (10% of the time of 108 inspectors) per year for coverage of approximately 730,000 establishments. With such limited resources, it is important for FSIS to identify whether and which types of facilities have undergone surveillance or are inspected by other authorities, for example, state or local regulatory agencies or other federal agencies such as the FDA. In addition, the FDA offers training and assistance to the approximately 75 state and territorial agencies and more than 3,000 local departments responsible for inspection of various establishments within the restaurant and retail segment of the food industry. Furthermore, FSIS and FDA sometimes enter into memoranda of understanding (MOU) for certain activities, which is a valuable approach in avoiding inspection gaps, while optimizing use of resources. Because minimizing gaps in food inspection is imperative, it is important to ensure that all 13 business types are adequately inspected and

OCR for page 1
LETTER REPORT 15 consumers are protected. This is more easily accomplished when each relevant agency clearly understands its responsibilities and agencies are communicating and coordinating activities. It is important to note that the level of surveillance provided is often variable both between and within states and local jurisdictions. In addition, state and local resources for food- safety inspections and compliance activities are commonly subject to budgetary cuts in the state or locality,11 which affect their ability to conduct food-safety activities. Furthermore, some state and local jurisdictions have been unable to perform required food inspections.12 Therefore, the CID must be cognizant of those realities when incorporating this risk consideration into its priority-setting activities. To use the risk consideration effectively, increased communication and data-sharing between the CID and state and local regulatory agencies is needed. When the committee notes that other jurisdictions are able to perform some surveillance activities better, it assumes that the activities are being performed adequately. Inspections and audits by third-party entities that examine establishments’ overall approach to food safety should also be considered. The frequency of surveillance by such entities varies and depends on contracts, inspection results, or the complexity of the operation. Some third-party programs offer a certification based on an establishment’s food-safety management programs, including food storage, cooking, serving, and sanitation practices. If third-party audits are accessible, FSIS should use the information in setting priorities for surveillance. • Distributors, warehouses, and transporters—Surveillance of these business types by other authorities varies with the location and the type of business. Although the committee does not have specific information on inspection frequencies, FDA performs inspections in these types of establishments. However, the state health or designated agency for food safety and/or the state department of agriculture may also conduct inspections, which vary by state and jurisdiction; the frequency is unknown. • Restaurants, institutions, and retailers—Of the business types listed in the report, these are the most frequently inspected by state or local regulatory agencies. They operate under the state or local food code, and inspection visits depend on results of previous inspections, consumer complaints, the primary population served, and other programs or plans at the operation. The FDA Food Code states that these operations should be inspected at least every 6 months (Food Code 8-401.10). However, inspection frequencies may vary depending on jurisdiction, available resources, and other factors. • Retail salvage and food banks—In general, these business types receive an unknown level of inspection. For example, there are many reasons why foods are diverted to salvage retailers, and it should not be assumed that the manufacturer maintains oversight of such products. Since FSIS lacks information about the risks associated with products in these outlets, the CID could conduct a survey of a representative sample of the 162 outlets to understand the extent of inspection and enforcement that may be needed to ensure consumer protection. Food banks should have written policies that define the condition of the foods they will or will not accept. In addition, their policies likely specify the conditions of storage and when products in their possession become unacceptable. The existence of such policies could facilitate the surveillance process, if they are reviewed by CID investigators to ensure they address food-safety concerns and whether the food bank is following them. 11 http://www.tribune-democrat.com/local/local_story_025232154.html. Accessed January 29, 2009. 12 http://www.oppaga.state.fl.us/reports/pdf/0867rpt.pdf. Accessed January 29, 2009.

OCR for page 1
16 REVIEW OF IN-COMMERCE SURVEILLANCE • 4-D facilities, renderers, custom exempt, and exempt poultry—In general, more information should be gathered by FSIS to understand the extent to which other regulatory agencies inspect the 4-D, renderer, custom exempt, and exempt poultry business types. Compliance in some of these facilities may rely solely on random visits by FSIS investigators. • Animal food—This category was ranked “low” by FSIS. The committee questions whether certain animal-food manufacturers (e.g., those serving zoos, mink farms, and animal parks) receive any other inspection, while pet-food manufacturers may be under the surveillance of FDA. However, these types of facilities may warrant surveillance to ensure products do not enter the human food chain. In addition, pet food, such as dog treats made with pork skin or ears, have been found to be sources of Salmonella. The concern is that when such contaminated food is given to pets there may be transmission of this or other pathogens to humans and especially to sensitive household members such as the young and elderly. This is an example where specific products or processes within a business type deserve greater inspection. The FSIS is well-suited to inspecting facilities making such products, and it may be appropriate for FSIS to collaborate with FDA or other agencies and to consider a requirement for HACCP implementation by pet-food manufacturing facilities. Finding: Other authorities may be conducting surveillance of many business types, but it often varies among states and localities. Such information seems to be underutilized by the CID. In contrast with the FSIS report, a review of surveillance of distributors and warehouses revealed that state and local regulatory agencies conduct periodic inspections of these business types. Recommendation 1: The FSIS CID should consider absence of surveillance by other jurisdictions to be more important in setting priorities for surveillance than the other risk considerations. For this to succeed, FSIS should gather more information in order to understand the extent (frequency and adequacy) to which other regulatory agencies inspect businesses, especially those such as retail salvage, food banks, 4-D, and renderers. Recommendation 2: The FSIS CID should consider the wide variation in surveillance activities in each state. In addition, FSIS CID should consider partnering with local and state regulatory agencies to increase its knowledge of existing meat, poultry, and egg product surveillance activities, and use the information to improve priority-setting of its own surveillance activities. To avoid or minimize gaps in inspection or surveillance activities, it may be appropriate for relevant agencies to enter into MOU in which responsibilities for inspection and enforcement are clearly defined. This approach may be appropriate for all business types.

OCR for page 1
LETTER REPORT 17 Three Surveillance Tiers The FSIS, on the basis of FSIS expert opinion, used a three-tier risk-ranking approach to set priorities for surveillance of the identified business types. The three tiers proposed are based on three levels of risk within each of the five risk considerations. However, as discussed above, the committee determined that the three-tier system neither takes into account differences in state, local, and other surveillance activities nor captures the diversity within various business types. In particular, the committee thought that a three-tier approach that results in broad risk categories that would be applied in the same manner throughout the country would not be particularly useful to CID investigators. The committee believes that the three risk rankings are not sufficiently discriminatory for adequate priority-setting for surveillance activities. A comparison of the FSIS ranking of risk consideration with the committee’s recommendations is presented in Appendix D. The committee suggests that a stepwise decision process approach would be more appropriate than the three-tier ranking system. In addition to the above considerations, it could assist investigators in selecting specific businesses and facilities for surveillance in the field on the basis of regional or other variations. However, if the CID prefers to use the three-tier approach, the committee recommends consideration of the suggested changes in each of the risk considerations and consequent changes in the three tiers. It should be noted that ranking by tiers and ranking in a stepwise process both involve the interaction of the risk considerations and the characteristics of the business types. An example of a stepwise decision process that could be considered for this purpose is presented in Appendix E; variations to this approach may also be considered. The five FSIS risk considerations would still be used to determine surveillance activities by CID investigators but with the following modifications. The committee believes that “jurisdictional risk” (i.e., surveillance by other authorities) should out-rank all other risk considerations as a priority- setting criterion, so it is presented as the first risk consideration in the stepwise decision process. Specifically, if risks posed to the public by particular business types or facilities are addressed adequately by other jurisdictions, those businesses or facilities would be assigned low priority for CID surveillance and FSIS would place the highest priority for CID surveillance on businesses that pose public-health risks but are not addressed by other jurisdictions. Surveillance by other authorities would be considered first in deciding on routine surveillance activities. It is important that FSIS has confidence that the frequency and quality of surveillance by other jurisdictions is sufficient. The remaining, and secondary, risk considerations would be considered in order of decreasing order of importance (as depicted by their prominence in the chart) product volume, inherent hazards, consumer susceptibility, and food defense vulnerabilities. In addition to the decreasing order of importance for surveillance presented in the chart, priority would be determined on the basis of increasing number of risk considerations present and as available CID resources would permit. It should be stressed that even if a stepwise decision process is adopted, emergencies, alerts, special incidents, and information about potential problems would remain the highest priority for CID investigation activities. With respect to a stepwise decision process, the committee refers FSIS to an article by Cox and coauthors13 that addresses problems that may arise when qualitative systems make severe rating errors. The authors suggest that the value of information produced by such systems can approach zero if most risks are small and a few are large. That happens if a qualitative 13 Cox, L. A., D. Babayev, and W. Huber. 2005. Some limitations of qualitative risk rating systems. Risk Anal. 25:651–662.

OCR for page 1
18 REVIEW OF IN-COMMERCE SURVEILLANCE system is unable to distinguish between large and very small risks. Another source of information to consider, in efforts to set priorities for the use of resources, is the book Worst Things First,14 which describes various ways to approach the problem of setting priorities. The book discusses many kinds of decision models. One aspect that is often neglected in a risk-only model is the cost and effect of different kinds of actions. For example, some establishments may be more expensive to inspect and have a lower probability of affecting public health. The effectiveness of any of these approaches depends on information available; more sophisticated models require more extensive and quantitative data; something that at this stage is lacking for FSIS to better prioritize its surveillance efforts. Ultimately, FSIS may consider a Bayesian “learning” model that begins with a determination, such as information from prior surveillance activities, and makes adjustments on the basis of new surveillance and information from other inspections. As stressed throughout this report, adequate and appropriate data are needed for such an approach to become feasible. The committee also considered deterrence as a potential reason for making FSIS activities more broad-based than if it focused on businesses that lack other regulatory oversight; this consideration is not discussed in the documents provided by FSIS. The committee concluded that proclaiming the intention to inspect broad categories of businesses may act as a deterrent even though a very small proportion of the individual businesses will be inspected. However, such a deterrent effect maybe negligible for businesses that are routinely inspected by other authorities because they would not perceive that different types of inspections would have different outcomes. In addition, for the deterrence to be credible, businesses must be aware of the potential for frequent FSIS surveillance and must believe that failing FSIS surveillance would have severe consequences. The FSIS provided no rationale for its 50–30–20 allocation of enforcement effort among the three proposed tiers. It appears that, although the system proposed by FSIS uses risk-based criteria, the proposed distribution of resources is not data driven. Also, it seems unlikely that the limited resources available from the CID would make additional contributions to protecting public health in business types and geographic areas where other agencies already make inspections. It seems likely that public health would be served better if the CID is able to focus its surveillance efforts on business types and facilities that handle only products that are solely under FSIS jurisdiction (meat, poultry, and egg products) and assign specific efforts based on pertinent data. However, as data available to FSIS are limited at this time, it is important that data and information collection become an important priority for the agency. Based on this, the committee concluded that because using a risk-based system is new to the CID, it may be premature to arbitrarily assign percentages of its limited resources. Rather, it may be more informative to benchmark where surveillance should take place for a few years and then allocate surveillance time on the basis of data collected, where problems have been identified, and estimation of where the greatest benefit to public health could be achieved. Although the committee agreed that the tactic of assigning different levels of overall public-health risk to categories of businesses may be reasonable, it was not convinced that three tiers are necessary. As available resources are always limited, it may be more useful to classify facilities into those that are in very high need for surveillance and those that, for various reasons, do not need frequent FSIS attention. The facilities in need of CID surveillance would then 14 Finkel, A. M., and D. Golding (eds). 1996. Worst Things First: The Debate over Risk-Based National Environmental Priorities. Washington, D.C.: Resources for the Future Press.

OCR for page 1
LETTER REPORT 19 receive the vast majority (possibly at least 90%) of the surveillance resources available; facilities in the other category may be investigated only when there is reason to do so. As indicated, specific time allocation percentages may be easier to assign after a few years of experience and data collection. Finding: The committee concludes that surveillance by other authorities may be the primary risk consideration as a selection criterion for surveillance priority-setting. However, because surveillance activities by other jurisdictions vary by region and because there is considerable diversity within business types, the four other risk considerations may be used to enable secondary decisions about surveillance priorities. In addition, FSIS lacks adequate pertinent data on various business types and on related inspection activities by other jurisdictions. Availability of such information is critical in establishing surveillance objectives, approaches, and priorities. Recommendation: The committee recommends that the FSIS CID creates a single category of high-risk business types and areas where it has sole jurisdiction, that most (potentially 90%) of available resources be devoted to surveillance of such facilities, and that a stepwise decision process to set priorities for establishments for surveillance be considered. SUMMARY OF FINDINGS A summary of the committee findings follows: • The committee determined that the 13 business types identified by FSIS adequately cover the array of in-commerce facilities that handle meat, poultry, and egg products. In addition, however, it was determined that several of the 13 business types are very diverse in activities and this complicates prioritization of surveillance based on public- health risk. • The committee found that there may be a duplication of effort if FSIS conducts surveillance in certain types of businesses or businesses within types that are also under the authority of other jurisdictions. The committee acknowledges, however, that inspection activity by other jurisdictions may vary by state, locality, and business type. • Another finding that may influence allocation of surveillance resources is that the amounts of meat, poultry, and egg products handled in some facilities are unknown, whereas meat and poultry items are the sole products handled in other facilities; the latter facilities may not receive inspection by other authorities. Another related finding, as suggested by public-health data, is that the risk of transfer of contamination from raw meat and poultry to other products, although significant, did not seem to be addressed by proposed CID surveillance procedures. • It was difficult for the committee to evaluate fully the concept and use of the “inherent hazard” risk consideration because of insufficient description and lack of clear direction for its use in categorization of facilities into risk groups. The issue becomes more

OCR for page 1
20 REVIEW OF IN-COMMERCE SURVEILLANCE complicated because there is significant diversity among and within some business types (e.g., restaurants), which prevents meaningful categorization and risk assignment. According to the committee’s findings, there is little discrimination among a majority of the establishment types based on inherent hazard as a risk factor; this is evident by the fact that most were rated by FSIS as “medium.” • The committee agrees with the CID proposed “high” risk ranking for transporters, institutions, retailers, restaurants, and food banks, in terms of food defense vulnerability; however, rendering facilities could also qualify for the “high” risk category. • The committee found that product volume is an important risk consideration and that the CID risk rankings based on product volume are appropriate. Additionally, however, the committee determined that FSIS needs adequate and reliable data on product volume handled by various facilities as well as other pertinent information. • The committee found that the consumer susceptibility risk consideration is not equally applicable to all business types. This limits the value of this risk consideration in prioritizing surveillance frequency for all types of institutions, some of which serve less susceptible populations. • The committee determined that surveillance of certain business types should be ongoing by other authorities but it is often variable among states and localities. • The committee concluded that surveillance by other authorities may be the primary risk consideration as a selection criterion for surveillance prioritization. However, because surveillance activities by other jurisdictions vary by region and because there is considerable diversity within and among various business types, the four other risk considerations may be used to enable secondary decisions about surveillance prioritization. • In addition, the committee found that FSIS lacks adequate pertinent data on various business types and on related inspection activities by other jurisdictions. Availability of such information is critical in establishing surveillance objectives, approaches, and priorities. GENERAL RECOMMENDATIONS The committee commends FSIS for its efforts in developing a useful first approach to risk-based surveillance activities for in-commerce establishments. The committee offers the following recommendations for consideration by FSIS as it continues to develop its approach for in-commerce surveillance activities to contribute to public health protection. • The committee recommends use of a stepwise decision process that sets priorities for surveillance of establishments in each investigator’s jurisdiction rather than the proposed three-tier system. An example is given in Appendix E; other similar approaches may also be considered.

OCR for page 1
LETTER REPORT 21 • The committee recommends that FSIS partners with other agencies. • In light of the wide variation in surveillance activities in each state and locality, FSIS and CID investigators should partner with state and local regulatory agencies to collect data and better understand existing inspection activity related to meat, poultry, and egg products, and should consider the resulting information in setting priorities for surveillance activities. Specifically this approach should include developing a process or system to collect additional pertinent data and information on surveillance by federal, state, and local jurisdictions; could serve as a resource for those jurisdictions; and would avoid duplication of surveillance efforts. • FSIS should partner with agencies that undertake a comprehensive approach to inspection of facilities that handle a wide variety of food products, because of the potential for meat, poultry, and egg products to serve as sources of contamination of other ready-to-eat foods. • If the three-tier approach is retained, FSIS should focus its limited CID resources on facilities and localities that are not inspected by other jurisdictions, such as renderers, 4-D, and animal-food businesses to provide assurance that their products do not enter the human food supply. If other jurisdictions do not inspect custom exempt and exempt poultry businesses, these could be included in FSIS CID activities. • Information on the volume of meat, poultry, and egg products should be available to the CID investigators before surveillance is undertaken. Such data will assist in directing CID resources toward facilities that handle the highest amounts of products, which represent the greatest potential exposure of the public. The system or process developed by FSIS for data and information collection should capture volume information from CID ongoing surveillance activities and existing information from federal, state, local, and private sources. • The committee recommends using an initial period of activity of at least a year to benchmark where surveillance activities should take place before establishing a specific allocation of time for these activities. Risk ranking may be modified appropriately as specific knowledge and data become available.