National Academies Press: OpenBook

Scientific Criteria to Ensure Safe Food (2003)

Chapter: 2 The Science of Public Health Surveillance

« Previous: 1 Historical Perspective on the Use of Food Safety Criteria and Performance Standards
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 28
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 29
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 30
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 31
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 32
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 33
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 34
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 35
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 36
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 37
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 38
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 39
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 40
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 41
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 42
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 43
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 44
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 45
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 46
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 47
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 48
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 49
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 50
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 51
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 52
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 53
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 54
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 55
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 56
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 57
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 58
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 59
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 60
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 61
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 62
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 63
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 64
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 65
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 66
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 67
Suggested Citation:"2 The Science of Public Health Surveillance." Institute of Medicine and National Research Council. 2003. Scientific Criteria to Ensure Safe Food. Washington, DC: The National Academies Press. doi: 10.17226/10690.
×
Page 68

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

The Science of Public Health Surveillance THE TOOLS OF PUBLIC HEALTH SURVEILLANCE Public health surveillance is the ongoing, systematic collection, analysis, interpretation, and dissemination of health outcome-specific data for use by the public health sector to reduce morbidity and mortality and to improve health (Thacker and Berkelman, 1988~. Surveillance of many infections and intoxica- tions, including those that are foodborne, has been a fundamental public health activity for many years. Human foodborne disease surveillance is conducted for three principal reasons: (1) to identify, control, and prevent outbreaks of food- borne disease, (2) to monitor trends and determine the targets and efficacy of control measures, and (3) to determine the burden of specific diseases on public health (Potter et al., 2000). By detecting outbreaks and their sources quickly, surveillance can lead to control of an acute health hazard, for example, by removing a contaminated product from the market or by temporarily closing a hazardous kitchen. Outbreak investigations can also identify critical gaps in knowledge, leading to applied research and ultimately to better long-term prevention as unsafe food handling processes are corrected or new food hazards are identified and controlled. The information gathered through surveillance and subsequent investiga- tions of outbreaks and of sporadic cases can reveal the magnitude and trends of foodborne diseases, which helps policy makers target prevention strategies. This information is also critical to the design and evaluation of risk assessments. Improved understanding of foodborne diseases, in turn, can help researchers recognize new problems, such as entirely new hazards (e.g., microbes or toxins) 28

29 Prevention Measures \ THE SCIENCE OF PUBLIC HEALTH SURVEILLANCE Surveillance Applied Targeted Research FIGURE 2.1 The cycle of public health prevention. - Epidemiologic Investigation / or known hazards that may appear in foods not previously associated with them. Most foodborne pathogens were discovered during outbreak investigations, and much of the knowledge we have about specific hazards and how they enter the food supply also was gained during the course of investigations. As new foodborne disease sources and agents emerge, the efforts to control them through application of the Hazard Analysis and Critical Control Point (HACCP) system and other control strategies must constantly evolve. Surveillance is a keystone in the effort to define, control, and prevent foodborne diseases (Figure 2.1~. In the United States, foodborne disease surveillance is primarily conducted by local and state public health agencies. In fact, local surveillance for diseases of public health concern has been conducted for centuries. In the nineteenth century, fear of cholera led to the establishment of permanent municipal health depart- ments and disease surveillance, even before the microbe that caused it was iden- tified (Rosenberg, 1987~. Reporting of typhoid fever cases and deaths drove many improvements in water and food safety at the beginning of the twentieth century. Increased concern following the large Escherichia cold 0157:H7 out- break in 1993 associated with consumption of undercooked ground beef (Bell et al., 1994) stimulated enhancements in surveillance for foodborne infections (FSIS, 1998c). Strategies in Public Health Surveillance There are specific strategies to collect information that may serve as a basis for making food safety policy decisions. The surveillance strategies for outbreaks and sporadic cases of diseases that are often foodborne are:

30 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD 1. Monitoring case reports of specific, notifiable infections 2. Investigating and reporting outbreaks of illnesses associated with events or establishments 3. Investigating and reporting unusual clusters of cases of specific infections 4. Vigilantly surveilling (termed sentinel site surveillance) for specific con- ditions that may or may not be notifiable 5. Laboratory subtyping of pathogens isolated from human infections 6. Surveying the population to measure trends in diarrhea! illness, consumer behavior, and food consumption One surveillance method may be more appropriate than another, and these methods may also be used alone or in combination, depending on the purpose. For example, subtyping of pathogens may be performed to confirm the source of an outbreak. The specific surveillance strategies are conducted either nationwide or in several sentinel sites that represent the whole population. Surveillance conducted to detect outbreaks and protect the public should cover the whole population, should include conditions most likely to appear in outbreak form, and in some instances, should focus on settings where outbreaks are likely to occur. Some outbreaks are not tightly clustered in time and space, and thus are not evident in surveillance conducted in one location. To detect dispersed outbreaks, it can be critical to compare specific markers of the infecting organisms, such as genetic "fingerprints," across many jurisdictions (Swaminathan et al., 2001~. Such com- parison of subtypes may reveal an unusual clustering of infections with a single strain of a pathogen that can then be further investigated. Public health laborato- ries use subtyping methods and are linked in a national network to permit rapid comparison of results and to provide warning of dispersed outbreaks. For example, the network of state public health laboratories detected a multistate cluster of Salmonella Newport infections that had the same pulsed-field gel electrophoresis profile. As a result of the investigation of genetic profiles, 78 infections in 13 states were linked to consumption of imported mangoes (Sivapalasingam et al., 2000~. If the purpose of surveillance is to measure the public health burden of disease or track long-term trends in the nation as a whole, more detailed data collected from a representative sample of sites around the country is likely to provide more accurate information (Angulo and Swerdlow, 1999~. This sentinel- site approach can provide data on important illnesses, such as Campylobacter or Vibrio infections, that are not well represented in national surveillance systems because they rarely appear in outbreak form and are not reportable in many jurisdictions. For the purpose of determining the food source of infections, surveillance based on outbreak investigations provides answers for those illnesses that fre- quently appear in outbreak form. For illnesses that rarely appear as outbreaks,

THE SCIENCE OF PUBLIC HEALTH SURVEILLANCE 31 case studies can give a general answer as to the source of illnesses that are strongly tied to specific sources, and case-control studies can provide information if the sources are complex. As described later, the committee feels that to con- struct a detailed quantification of the contribution of specific animal or food sources to foodborne diseases, systematic monitoring of pathogens in food and animal reservoirs using molecular subtyping and comparison of strains with iso- lates from human infections are urgently needed. The following sections describe how these strategies are utilized in both nationwide and sentinel site surveillance by public health agencies in the United States. Specialized surveys that relate the contribution of consumer behavior to the level of specific foodborne illness risk are described as well. Finally, several factors that limit the value of surveillance systems are discussed. Nationwide Surveillance of Notifiable Diseases Many counties and states have collected notifiable disease reports for more than a century, covering an ever-expanding list of illnesses. Since 1961, these reports have been voluntarily submitted to the Centers for Disease Control and Prevention (CDC), which publishes them as weekly and annual summaries (Thacker, 1994~. At its annual meetings, the Council of State and Territorial Epidemiologists decides which specific illnesses should be nationally notifiable. This general umbrella of reporting covers all areas of the United States; provides information useful to local, state, and national authorities; and is relatively inex- pensive. Most disease reporting is passive from the standpoint of the public health system, which means that clinicians and laboratories are asked to report cases on their own initiative. Basic case surveillance has been enhanced for some infections by further characterization of the infecting pathogen in public health laboratories. This voluntary case surveillance was first begun for Salmonella. Following large, multistate outbreaks of salmonellosis early in the 1960s, health department laboratories in states and large cities began to serotype strains of Salmonella isolated from humans; the results of this subtyping were shared with CDC as well in order to detect outbreaks affecting more than one state. Since 1962, national Salmonella surveillance has depended on this serotype-based reporting (Olsen et al., 2001~. These data have been critical to the detection of many outbreaks of salmonellosis each year. Since 1990, these data have been relayed electronically from states to CDC via the Public Health Laboratory Infor- mation System (Bean et al., 1992~. In addition, since 1995 these data have been routinely examined using an automated statistical outbreak detection algorithm that compares current reports with the preceding 5-year mean number of cases for the same geographic area and week of the year to look for unusual clusters of infection (Hutwagner et al., 1997~. The usefulness of the outbreak algorithm is limited by the timeliness of reporting and the high background rates of reporting for common serotypes such as S. Typhimurium and S. Enteritidis. The greatest

32 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD sensitivity for Salmonella serotyping to detect meaningful clusters is for the rare serotypes, whereas further differentiation is necessary for the most common sero- types. The utility of serotyping as an international designation for Salmonella sub- types has led to its widespread adoption. In a recent survey, 61 countries reported that they used Salmonella serotyping for public health surveillance (Herikstad et al., 2002a). A new World Health Organization (WHO) website (WHO, 2002) collects and presents the results of this serotyping. This website is a new mecha- nism for the global surveillance of foodborne diseases. Molecular subtyping is now expanding the power of surveillance to detect outbreaks that appear as sporadic cases and is improving the ability of public health authorities to investigate outbreaks by comparing the molecular "finger- print" of bacterial strains associated with sporadic cases of a foodborne disease. These new techniques can define subtypes within a single species or serotype and provide useful strain differentiation for a growing number of pathogens (Swaminathan et al.,2001~. State public health laboratories began using an assay standardized at CDC for E. cold 0157:H7 after it proved useful in the 1993 West Coast outbreak associated with the consumption of undercooked ground beef; they have now expanded the use of this technique to common serotypes of Salmonella such as Typhimurium and Enteritidis, and to Listeria monocytogenes (Swaminathan et al., 2001~. Developing this capacity at the state level also en- hanced rapid detection of multicounty clusters within the state (Bender et al., 1997, 2001~. Standardized subtyping protocols have now been developed for seven pathogens; next-generation, gene-based technologies are under develop- ment. Recently, PulseNet, a national network formed by linking all state public health laboratories via the Internet, with a national database maintained by CDC, made it possible to rapidly identify and investigate multistate clusters. Once a cluster of infections caused by strains with the same fingerprint is identified, rapid epidemiological investigation can determine whether the cluster is a true outbreak with a common source. Laboratories at the Food and Drug Administra- tion (FDA) and the U.S. Department of Agriculture (USDA) also participate in this network so that isolates from foods and animals can be compared within the system. It is noteworthy that Canada has already adopted a compatible system and that the European network for laboratory-based surveillance of foodborne infections, EnterNet, has similar plans. The participation of Canada, Europe, Asia, and other regions could make it possible to detect multiregional clusters of foodborne disease (Swaminathan et al., 2001~. Monitoring levels of antimicrobial resistance in foodborne pathogens is an- other form of subtype-based surveillance. Since 1996, the National Antimicrobial Resistance Monitoring System (NARMS) for enteric bacteria, a collaborative effort of CDC, USDA, and FDA, has been monitoring the prevalence of resis- tance in Salmonella, Campylobacter, and other foodborne bacterial pathogens

THE SCIENCE OF PUBLIC HEALTH SURVEILLANCE 33 isolated from humans, animals, and foods (Marano et al., 2000~. This type of surveillance provides information about the trends in microbial resistance to specific drugs, identifies the emergence of new resistance threats, and permits the comparison of strains identified in various locations. This information is useful to public health officials who are involved in controlling highly resistant strains, to clinicians making treatment decisions, and to regulators who can better evaluate the association between antibiotics used in animals or the environment and resis- tance developed in human pathogens. In summary, nationwide surveillance systems for cases of foodborne infec- tion are valuable tools for defining trends, identifying outbreaks, and evaluating food safety programs. In some situations, serotyping and subtyping of pathogens, coupled with nationwide surveillance, provide an ideal system to link a cluster of cases. Considering that state and local public health systems provide the only nation- wide population-based surveillance for foodborne diseases, and that outbreak investigations are critical to identify new pathogens and new food safety hazards, the committee recommends that foodborne outbreak investigation and reporting by state and local health departments be enhanced. Training and personnel and laboratory support should be provided to enable rapid, thorough, and accurate investigation and reporting of foodborne outbreaks by local and state health departments, with performance evaluated through systematic review of outbreak reports. In addition, timely analysis and dissemination of results to regulators, industry, and the public is essential. Time series analysis (as discussed in Chapter 3) would also be a valuable analysis technique in this area. Sentinel Site Surveillance In contrast to the national umbrella of routine notifiable disease surveillance supplemented with public health laboratory subtyping, a different strategy, senti- nel site surveillance, can provide more detailed information about specific ill- nesses that are likely to be foodborne. This strategy was first developed for monitoring cases of hepatitis, for which detailed laboratory and epidemiological data are crucial (Bell et al., 1998~. A more recent example of this type of surveillance is the Foodborne Disease Active Surveillance Network (FoodNet), a collaborative program of CDC, senti- nel sites (currently nine sites), USDA, and FDA under the aegis of CDC's Emerging Infections Program (Angulo and Swerdlow, 1999~. The establishment of FoodNet was stimulated by a request from USDA's Food Safety and Inspec- tion System (FSIS) for a system to ascertain the public health impact of USDA's Pathogen Reduction; Hazard Analysis and Critical Control Point Final Rule (PR/HACCP rule). FoodNet began with an initial five-site area in 1996 and expanded to nine sites by 2001. The current surveillance area covers 37.8 million persons, or approximately 13 percent of the U.S. population (CDC, 2002a).

34 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD FoodNet conducts active case ascertainment for foodborne diseases, accompa- nied by epidemiological studies designed to help better understand the epidemi- ology of foodborne diseases in the United States. Active case ascertainment means that public health authorities regularly contact clinical laboratories to obtain case reports of diagnosed illnesses; therefore, the results do not depend on which infections are locally notifiable or on local resources available for surveil- lance. Thus, because reporting is more uniform and complete, active case ascer- tainment yields better data than passive reporting systems. However, it is also more expensive and limited in geographic scope. In addition to case ascertain- ment, FoodNet surveys laboratory, physician, and patient practices that cause an individual case to be diagnosed. Also, FoodNet has been a platform for conduct- ing case-control studies of sporadic infections in order to identify general risk factors for infection that distinguish the persons who get ill from those who stay healthy. This information has been used to better define the burden of foodborne illness (Mead et al., 1999), to evaluate the risk factors for specific infections (e.g., in the Campylobacter case-control study [Friedman et al., 2000bj), and to track the trends in major foodborne infections (CDC, 2002a). To provide real-time tracking of human case surveillance, the committee recommends that the capacity of the sentinel sites of FoodNet to rapidly inter- view (i.e., as soon as possible after the case is diagnosed, as opposed to two to three weeks later when active surveillance contacts with the laboratory detect the case, a cluster is identified, or some other event shows the need for follow-up) individual illness cases that are potentially foodborne, to track real-time inter- views, and to collect and subtype Listeria, E. cold 0157:H7, and Salmonella isolates from human infections, be enhanced as soon as feasible. (Although sev- eral subtyping schemes exist for Campylobacter, none has yet been shown to be useful and practical in the public health setting for routine testing of all isolates.) All cases of infection from pathogens covered by FoodNet surveillance should be interviewed. In addition, the committee believes that international collaboration and the sharing of methods and microbiological and illness surveillance data between the United States and other surveillance systems such as WHO's Global SalmSurv (WHO, 2002) and Europe's EnterNet must be strongly supported. Foodborne Outbreak Reporting A foodborne outbreak is a cluster of two or more similar infections that are shown by investigation to result from ingestion of the same food (Olsen et al., 2000~. Local and state health departments conduct most foodborne outbreak investigations. Since 1967, CDC has collected reports of outbreaks of foodborne illnesses investigated by local, state, and national public health authorities (Olsen et al., 2000~. Reports of outbreaks include the nature of the pathogen or toxin, the type of food that caused the outbreak, and some information about factors that contributed to the outbreak. Before 1998, these reports were collected on paper

THE SCIENCE OF PUBLIC HEALTH SURVEILLANCE 35 and slowly reviewed and compiled. The system is now being overhauled with an improved form, the active solicitation of reports from states, the introduction of Internet-based reporting (Electronic Foodborne Outbreak Reporting System), and the more rapid analysis and dissemination of results (FDDB, 2002a). The foodborne outbreak surveillance system has provided useful informa- tion on long-term trends for many pathogens for which surveillance otherwise does not exist, as well as summaries of the outbreaks caused by a particular pathogen, hazard, or food (Bean and Griffin, 1990~. In the future, it may provide more systematic detection of clusters of outbreaks, based on both laboratory testing and epidemiological assessment of the outbreak presentation (Hall et al., 2001~. The committee considers the systematic analysis of information on out- breaks gathered through this system as an effective tool for allocating the burden of many infections and other hazards across broad food categories. Specialized Surveys of Behavior FoodNet and other surveillance efforts also provide systematic data on be- havior of the population and exposure to specific risks. Studies conducted through the CDC Behavioral Risk Factor Surveillance System (BRFSS) documented the high frequency of risky food behavior (Yang et al., 1998~. More recently, FoodNet population surveys have provided population-based data on the incidence of diarrhea! illness and the likelihood of seeking medical care for a diarrhea! illness; this information was critical to develop a general estimate of the burden of foodborne disease (Herikstad et al., 2002b; Mead et al., 1999~. The surveys also provided general population-based data on the frequency of exposure to a wide variety of foods and other potential sources of intestinal infection (Consumer Studies Branch, 2002; FDDB, 2002b). Another potential source of information is the complaint systems maintained by local and state health departments to which individuals can report illnesses or hazardous conditions they believe may be related to food (Samuel et al., 2001~. While such systems are far less specific than systems built on diagnosed cases of illness, they may provide an early warning of problems. Limitations of Surveillance One limitation inherent in all surveillance systems is that many cases go unrecognized for a variety of reasons. For example, cases may not be detected because people who are ill do not seek medical care, physicians and laboratories may not make a specific diagnosis, diagnosed cases may not be reported to authorities, and authorities with limited resources may not investigate or report cases. This last factor becomes especially significant if the surveillance program is voluntary, as is the case with outbreak reporting by local and state agencies.

36 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD Data collected in this voluntary manner do not correspond to a nationally representative sample of the population because reporting depends on other vari- ables, such as local resources or whether a particular disease is notifiable (CDC, 2001~. Even in active surveillance programs, such as FoodNet, the number of cases is underestimated because people do not seek medical care or because cases are reported only when they are confirmed by a laboratory. Therefore, the actual number of cases that occurs is likely to be substantially greater than the number of cases reported. For example, it has been estimated that 38 cases of salmonellosis occur for every 1 that is reported (Voetsch et al., 1998~. Many outbreaks are also likely to be unrecognized. A common-source outbreak in a restaurant may not be recognized because patrons were exposed in small groups that were unknown to each other. For some foodborne infections, the incubation period may be long enough to obscure the relationship with the meal unless persons attending a large gathering, such as a banquet or wedding reception, have some reason to compare their experiences afterwards. A second limitation is the difficulty in attributing a specific case to a specific food. Many infections can be transmitted by a variety of foods and by routes other than food. In the sporadic case of illness, the person may have consumed many foods and may have had other potentially risky exposures in the days preceding illness, making it difficult to determine the source of the illness. In an outbreak setting, where careful comparison of food consumption patterns of a group of ill persons with those of a group who remained well can identify the immediate food vehicle, it is still difficult sometimes to determine which of the various ingredients was the source of the illness. However, many outbreak inves- tigations are definitive, and comparison of patterns observed among groups of outbreaks can help define patterns. Finally, surveillance can only count what is measurable and known. Because diagnosis of Norwalk-like virus (recently designated "Noroviruses") infections is not routinely performed in clinical laboratories, for example, this extremely com- mon illness cannot be monitored with the same type of case-based surveillance that is conducted for infections caused by Salmonella or Campylobacter, for which routine diagnostic tests are available. The importance of Norwalk-like virus infections can be defined from outbreaks where the typical combination of signs, symptoms, incubation period, and duration of illness can be documented and where specimens reach specialized laboratories that can make the diagnosis (Bresee et al., 2002~. Similarly, enterotoxigenic E. colt, the cause of much travelers' diarrhea, is increasingly recognized as a cause of outbreaks in the United States, but may also be an unrecognized common cause of sporadic cases because the specialized tests to detect it are rarely applied (Dalton et al., 1999; Guerrant et al., 1990~. It is likely that there are many foodborne disease agents yet to be dis- covered which, consequently, are not currently tested for in any laboratory (Tauxe, 1997~.

THE SCIENCE OF PUBLIC HEALTH SURVEILLANCE 37 Similarly, behavioral risk-factor surveillance is subject to limitations. This type of surveillance depends on what people can and will report. People may overestimate how often they perform socially desirable behaviors such as hand washing. Questions about risk exposures also depend on what the consumer observes. People are not likely to know if the food they ate was cross-contaminated in the kitchen, even if they prepared it themselves. The observations individuals can make may be a less-than-perfect measure of risk. Although the FoodNet population survey used consumption of pink ground beef as an assessment of cooked meat doneness and safety, research has clearly demonstrated that cooked meat color is not an acceptable indicator for these parameters (Berry et al.,1998; FSIS, 1998a, 2000; Hunt et al., 1999~. Premature browning and a persistent pink color are two conditions that can occur in ground beef patties, influencing inter- nal beef patty color, whether or not a patty has been cooked to an internal temperature of 160°F (Hunt et al., 1999; Killinger et al., 2000~. In a nationwide evaluation, Berry and coworkers (1998) found 47.4 percent of hamburgers cooked to 160°F retained some pink color, and 15.8 percent still retained some pink color when cooked to 175°F. In addition, more than 25 percent of fresh-cooked ham- burgers (meat was never frozen) were brown or nearly brown internally although hamburgers were only cooked to 150°F. RESULTS FROM PUBLIC HEALTH SURVEILLANCE The Burden of Disease An estimation of the burden of disease is very useful when regulatory agen- cies make decisions about the focus and allocation of resources. The burden of disease attributable to foods has only been estimated in a general way; if the estimate of this burden was specific for particular foodborne diseases and food groups, more informed decisions could be made by regulatory agencies. Information from surveillance has recently been integrated into a general estimate of the overall burden of foodborne disease in the United States (Mead et al., 1999~. This estimate included the number of cases, hospitalizations, and deaths that were attributed to specific pathogens and to the large number of illnesses that remain unaccounted for. These pathogen-based point estimates can provide a benchmark for assessing the economic impact of foodborne diseases, such as the $6.9 billion estimated cost to society from the diseases caused by the major foodborne bacterial pathogens (Buzby and Roberts, 1996~. Some foodborne infections can also cause chronic complications in a small percentage of cases; for example, kidney failure related to E. cold 0157:H7 has been reported in 4 to 8 percent of cases (Griffin et al., 2002), and Guillain Barre syndrome paralysis may complicate 1 in 1,000 Campylobacter infections (Nachamkin et al., 2000~. There may be other complications of and sequelae from foodborne diseases. For ex- ample, it has recently been reported that people infected with multiresistant

38 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD salmonellae are more likely to die in the 6 months following the infection than uninfected individuals (Helms et al., 2002~. The full impact of illnesses includes acute morbidity and mortality, as well as the impact of subsequent complications and of long-term effects, such as life-long impairments from congenital toxoplas- mosis or early childhood diarrhea! illnesses in impoverished areas (Guerrant et al., 2002~. With more information about the frequency, duration, and disability caused by these complications, the burden of foodborne illness could be reesti- mated on a basis such as Disability Adjusted Life Years, a measure used to characterize the burden of many other public health problems (Murray and Lopez, 1997). Surveillance data can subdivide the burden of a specific infection. For ex- ample, the contribution of specific Salmonella serotypes to the overall burden of salmonellosis can be derived from their frequency. More specifically, the three most common serotypes of Salmonella, Typhimurium, Enteritidis, and Newport, together accounted for nearly half of all reported cases of salmonellosis in 2001, and thus of the burden of salmonellosis (Table 2.1~. The burden of reported foodborne outbreaks can also be measured. National foodborne outbreak reporting from 1998 through 2000 gave a combined annual incidence of 4.8 outbreaks per 1 million persons in the population (FDDB, 2002a). However, in addition to the limitations mentioned above, measuring the burden of disease due to outbreaks presents special challenges. For example, small out- breaks are particularly likely to go unrecognized and unreported, and it is likely that outbreak surveillance undercounts the true frequency of events for the rea- sons noted earlier. Moreover, a substantial fraction of outbreak investigations do not determine either the causative agent (the etiology) or the specific food that TABLE 2.1 The Top Ten Salmonella Serotypes Reported from Humans in 2001 Rank Serotype Number of Reported Cases Percentage of the Total 1 Typhimurium 6,999 22.1 2 Enteritidis 5,614 17.7 3 Newport 3,158 10.0 4 Heidelberg 1,884 5.9 5 Javiana 1,067 3.4 6 Montevideo 626 2.0 7 Oranienburg 595 1.9 8 Muenchen 5 8 3 1.8 9 Thompson 514 1.6 10 Saint Paul 469 1.5 Subtotal 21,509 67.9 Total 3 1,675 SOURCE: FDDB (2002c).

THE SCIENCE OF PUBLIC HEALTH SURVEILLANCE 39 was contaminated (the food vehicle), information that is critical in assigning the burden of disease and focusing resources on areas of most concern. Between 1993 and 1997, among the 2,751 foodborne outbreaks reported to CDC and included in a published summary, an etiology was reported for 838 (32 percent) and a food vehicle for 967 (35 percent) (Olsen et al., 2000~. Clinical and epide- m~ological profiling of outbreaks with unconfirmed etiology indicates that many of these can still be put into meaningful categories (Hall et al., 2001~. Among outbreaks investigated that affected at least ten persons in FoodNet sites in 1998 and 1999, 30 percent had a determined etiology and 57 percent had a reported vehicle (Jones et al., 2000~. One reason the etiology of many outbreaks goes unconfirmed is that appro- pnate clinical samples are not collected and tested (German et al., 2002~. Thus, only large outbreaks are likely to be charactenzed. The committee believes that allocation of more resources for diagnostic testing and investigation could in- crease the proportion of foodborne disease outbreaks that are charactenzed. Trends in Foodborne Disease Standard case surveillance data, such as that collected from the national Salmonella surveillance program, provide nationwide data on the prevalence and trends of specific serotypes of Salmonella. However, unreported cases due to not seeking medical attention or not performing the diagnostic occur. The re- sults over time show substantial variation in the incidence of specific serotypes as epidemics emerge and are controlled (Figure 2.2~. The national incidence of S. 14 o ~ 12 Q To A To To To Q C' C' 10 8 6 4 2 o li 1 ! ' ! i ,~/ `' ~ ~ n ~ Jon '*';~= ~~ ~ ,, by ~ ~ ~ it_ Enteritidis .+ -f a_ ~ ~ , ~~ AL —~ ~ - _ ~ is, ~_~ at=, ~ ~ Newport An, Heidelberg 1970 1974 1978 1982 1986 1990 1994 1998 Year FIGURE 2.2 Trends in incidence of the top four Salmonella serotypes 1970-2001. SOURCE: CDC (2002b).

40 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD Ententidis infections shows the progress of this primarily egg-associated epi- dem~c. The epidemic began in the 1980s and reached a peak in 1995. Although the incidence of this serotype has decreased by approximately 48 percent since 1995, it remains well above the pre-epidemic baseline of 1 per 100,000 popula- tion, at par with S. Typhimunum as the most prevalent salmonellae serotypes. The increase and subsequent return to baseline in S. Heidelberg, a serotype usu- ally associated with poultry, is also evident. In contrast, because significant vana- tion has occurred since 1995 in the number of reported cases of S. Newport, a serotype usually associated with cattle, the trend is not so clear, but there are indications that it is on the increase and it is currently the third most common serotype (FDDB, 2002c). Systematic review of the Salmonella surveillance data through 1997 indicates that there have been important declines in several sero- types associated with swine and with poultry, and increases in serotypes associ- ated with reptiles (such as pet turtles and snakes) (Olsen et al., 2001~. Another surveillance system that provides trends for the illnesses it tracks is FoodNet (Figure 2.3~. There have been sustained and important decreases in the reported incidence of Campylobacter, Yersinia, Listeria, and Salmonella infec- tions since 1996 (CDC, 2002a). These declines are not accounted for by changes in diagnostic procedures or in the surveillance system itself; the declines were significant even when the considerable regional variation in these infections was 1.4, ~ ~ - ~ 0.8 - . _ tar 0.6 - 0.4 - 0.2 - o Decrease since 1996 of: ,~, ,_ ~ '_ O \ 1996 1 997 1998 1999 2000 2001 Year ~ Salmonella ° Campylobacter - ~ Listeria 3< Yersinia 15% 27% 35% 49% FIGURE 2.3 Trends in relative incidence of selected foodborne infections, FoodNet 1996-2001. SOURCE: CDC (2002a).

THE SCIENCE OF PUBLIC HEALTH SURVEILLANCE 41 taken into account. Such declines coincided with the implementation of food safety assurance measures by USDA, including the PR/HACCP rule, in meat and poultry slaughter and processing plants. Additional interventions that have been introduced in the past several years include consumer safety warnings on raw meat and poultry, education efforts for the public, egg-quality assurance pro- grams for S. Enteritidis (see below), increased attention to fresh produce safety, implementation of HACCP in the seafood industry, application of HACCP to juice processing, and heightened awareness about the importance of food safety controls for imported foods. Changes in slaughter and processing procedures and sanitation are likely to have played an important role in reducing the incidence of four important food- borne diseases between 1996 and 2001. Y. enterocolitica infections, often associ- ated with pork (Lee et al., 1990; Tauxe et al., 1987), have declined the most: 49 percent (CDC, 2002a; FDDB, 2002a). This decline may have resulted in part from changes in pork carcass-dressing practices such as tying the bung (large intestine) early in the process. Because there have been no targeted public health control efforts for this infection in recent years, this decrease may also have been achieved partly as a result of basic food safety education and implementation of the PR/HACCP rule in pork processing. L. monocytogenes infections showed the second greatest decline: 35 percent. Outbreaks and sporadic cases of illness caused by this pathogen are most frequently associated with ready-to-eat and processed meats and raw-milk cheeses (Mead et al., 2002~. The recent decline in Listeria infections occurred as the ready-to-eat meat industry focused on improving fac- tory sanitation and implementation of HACCP programs in the wake of a large listeriosis outbreak in 1998 that was traced to hot dogs (CDC, 1998~. The 27 percent decline in Campylobacter infections, which are often associated with poultry, occurred alongside changes in poultry processing-plant operations that were introduced with the objective of reducing Salmonella contamination. These changes included the PR/HACCP rule implementation, as well as general food safety information dissemination efforts to increase public awareness (Shane, 2000~. The overall decline in Salmonella infections of 15 percent echoes the trends seen in national Salmonella surveillance. It includes declines in both S. Typhimurium (down 24 percent) and S. Enteritidis (down 22 percent), so it reflects more than the control of egg-associated S. Enteritidis infections. The overall decline in salmonellosis would be even greater except for the concurrent increase in infections due to S. Newport (up 32 percent; Figure 2.2~. The decline in the incidence of Salmonella infections in humans from 1996 to 2001 coincided with a decline in the prevalence of Salmonella isolated from FSIS-regulated products, according to comparisons of the baseline studies per- formed by USDA before (1994 to 1996) and after (2000) the PR/HACCP rule was implemented (Rose et al., 2002~. Similarly, the declines observed in the frequency of the four most common serotypes of Salmonella found in broiler chicken samples are matched by significant declines in the frequency of infec-

42 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD lions in humans with three of the four serotypes (S. Heidelberg, S. Typhimunum, and S. Hadar); the fourth serotype (S. Kentucky) was already rare among humans (RTI, 2002b). It is difficult to ascribe these trends to any one specific control measure because they are occurring in the setting of many simultaneous changes and improvements; nevertheless, the committee believes that these trends indi- cate that, collectively, the food safety efforts are making progress toward the national public health goals for 2010 (Table 2.2~. Infections with E. cold 0157:H7 do not show a sustained decline. Although their number decreased 21 percent in 2001 as compared with 1996, this decline is the result of a decrease only between 2000 and 2001 that does not imply a consistent trend; it may simply represent year-to-year variation and perhaps the effect of case-finding activities associated with specific outbreaks (Bender et al., in press). Trends in meat contamination from 2000 to 2002 indicate that the prevalence of this microorganism in ground beef has not changed. The trend, in percentage of positive samples, is flat at approximately 0.8 percent of tested samples (FSIS, 2003~. Among other pathogens tracked by FoodNet, Shigella, which has a human reservoir and is predominantly transmitted from person-to-person and only some- times via food, did not decrease significantly. Vibrio infections typically trans- m~tted via undercooked shellfish increased by 83 percent. This increase coin- cided with the recognition of a new epidemic strain of V. parahaemolyticus in 1997 (Daniels et al., 2000~. The incidence of parasitic infections with Cyclospora and Cryptosporidium, for which surveillance began in 1997, also decreased by 2001, although statistical trends were not calculated for Cyclospora because of the small number of cases and the shorter time of observation (CDC, 2002a). The committee recognizes that, ironically, because of some improvements in surveillance programs, the food safety problem in some cases may appear to have worsened. For example, the number of foodborne outbreaks reported to CDC increased sharply in 1998 from 400 to 500 per year (1990-1993) to 1,300 to TABLE 2.2 Incidence of Selected Foodborne Diseases in FoodNet, 2001, and the Healthy People 2010 Goals Pathogen Incidence, 2001a (per 100,000) National Goals 2010b (per 100,000) Listeria Campylobacter 13.8 Salmonella 15.1 1.6 0.3 30.8 Total 12.3 6.8 1.0 0.25 20.4 a Preliminary FoodNet data (CDC, 2002a). b Healthy People 2010 goals (HHS, 2000).

THE SCIENCE OF PUBLIC HEALTH SURVEILLANCE 43 1,400 per year (Figure 2.4~. As described earlier, this increase followed a revision of the reporting system and therefore largely reflects the impact of the new reporting procedures. In addition, better surveillance using PulseNet means that some outbreaks that were missed in the past are now being detected. For example, in the year following the introduction of PulseNet subtyping for E. cold 0157:H7 in Minnesota, four of ten common-source outbreaks caused by that pathogen were detected that would likely have been missed otherwise (Bender et al., 1997~.Si~l~ly, more Listeria outbreaks we being detected since the implementation of routine molecular subtyping; where these outbreaks used to be detected once every 5 years, they are being detected approximately twice a year (Mead et al., 2002~. Long-term trends can also be observed in reported foodborne outbreak in- vestigations. Since 1967, the number of outbreaks of staphylococcal and Clostridium perfringens food poisoning has decreased substantially (Bean and Gnffin, 1990~. Outbreaks of S. Ententidis infections increased in the 1980s to a peak in the m~d-1990s, but have since declined, as have the number of sporadic infections (FDDB, 2002c). Linking Pathogens to Specific Foods: Allocating the Burden of Disease Many foodborne pathogens are associated with a specific reservoir, either a food, an animal, or a human, and consequently the illnesses they cause are also 1~ Paper-based reporting 1 ,600 - 1 ,400 - ~ 1 200- Q 1,000- o o Q z 800 - 600 - 400 - 200 - O - ~ Web-based (EFORS) CDC initiates improved | reporting ,; 1990 1992 1994 1996 1998 2000 2002 Year FIGURE 2.4 Foodborne disease outbreaks reported to the Centers for Disease Control and Prevention, January 1, 1990, through March 15, 2002. SOURCE: FDDB (2002a).

44 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD often associated with a characteristic food group or reservoir. The current state of knowledge about the association of common food groups and common foodborne agents is summarized in Table 2.3. The data that link a pathogen to a specific reservoir often come from out- break investigations. For many pathogens, a series of investigated outbreaks provides the best information to define the association of the illness with specific foods. For example, the first investigation of E. cold 0157:H7 infections identi- fied the pathogen and linked the distinctive illness to eating undercooked ham- burgers (Riley et al., 1983~. Trace-back from an outbreak caused by ground beef and from sporadic cases caused by drinking raw milk led to identification of the bovine reservoir for E. cold 0157:H7; this finding is particularly noteworthy because infected cows are usually asymptomatic (Martin et al., 1986; Wells et al., 1991~. More recently, outbreaks of this infection have been associated with an expanding array of foods (Griffin et al., 2002~. Early investigations of Campylobacter outbreaks identified raw milk, undercooked poultry, and con- tam~nated water as common sources (Blaser et al., 1979; Demoing et al., 1987; Vogt et al., 1982). Pathogens that have human reservoirs can also be linked to specific foods, depending on the most characteristic mechanisms of contamination. In 1924, a large epidemic of typhoid fever was linked to raw oysters that were harvested and held near sewage sources (Lumsden et al., 1925~. More recently, outbreaks of Norwalk-like virus infection, which also has a human reservoir, have been linked to shellfish (and to direct contamination from ill fishermen) and to foods such as cold salads and sandwiches that are handled extensively in the kitchen (and to direct contamination from ill food handlers) (Kohn et al., 1995; Parashar and TABLE 2.3 Specific Association of Commodity Food Groups and Pathogens Food Group Pathogens Beef Salmonella, Escherichia cold 0157:H7 Poultry Campylobacter, Salmonella Pork Staphylococcus aureus, Yersinia enterocolitica, Salmonella, Toxoplasma, Trichinella Ready-to-eat meats Listeria monocytogenes Dairy L. monocytogenes, E. cold 0157:H7, Salmonella, Campylobacter Eggs Salmonella Fresh produce Norwalk-like virus, Salmonella, Shigella, E. cold 0157:H7, Hepatitis A, Cyclospora Finfish Histamine fish poisoning (scombroid), ciguatera poisoning, helminth parasites Shellfish Vibrio spp, Norwalk-like virus, Hepatitis A SOURCE: Doyle et al. (2001).

THE SCIENCE OF PUBLIC HEALTH SURVEILLANCE 45 Monroe, 2001~. For pathogens that rarely cause outbreaks, studies of sporadic cases and comparison with healthy controls can define associations with particu- lar foods. For example, V. vulnificus infection was definitively associated with consumption of raw oysters soon after it was first described (Blake et al., 1979~. Studies of E. cold 0157:H7 infections linked sporadic cases of infection with this pathogen to eating undercooked ground beef, thus supplementing the data from outbreaks (Kassenborg et al., 1998; Mead et al., 1997; Slutsker et al., 1998~. Studies of sporadic Campylobacter infection have linked it to eating poultry and other meats, as well as to drinking untreated water and to other sources. Around the world, poultry remains the dominant reservoir for this pathogen (Friedman et al., 2000a, 2000b; WHO, 2000~. Allocating the burden of infections quantitatively across specific food groups is a complex challenge that has been approached using several strategies. A main strategy draws from epidemiological and public health investigations. Data on outbreaks associated with foods, supplemented with data from sporadic cases, provide the most readily available public health information for allocating the burden of specific infections across food groups. For example, between 1993 and 1997, 1,152 foodborne disease outbreaks with a determined food vehicle, which involved 46,453 illnesses, were reported in the United States (updated from Olsen et al., 2000~. Among the 713 outbreaks for which the implicated food could be assigned to a single food group, 21 percent of the illnesses were associated with meat, l l percent with poultry, 28 percent with produce, 15 percent with seafood, and 26 percent with other foods. These findings indicate that food safety concerns exist for all major food groups. For those illnesses that rarely appear in outbreak form, data from individual case series or from case-control studies can be used to allocate the burden. Epidemiological investigations of outbreaks and cases can also provide im- portant insight into the precise mechanisms of exposure and the variations in human behavior that contribute to it. For example, illness in an outbreak was particularly associated with tasting raw ground beef in the process of seasoning and cooking it (Fontaine et al., 1978~. In an investigation of Campylobacter infections in Colorado, illness was associated particularly with handling and preparing chicken, rather than with eating it (Hopkins and Scott, 1983~. In an assessment of sporadic ground beef-associated E. cold 0157:H7 infections in New Jersey, ill persons were no less likely to have noticed the new meat handling recommendations on the meat wrapper than those who were well, but they were less likely to have washed their hands after handling raw beef (Mead et al., 1997~. Another strategy to help allocate the burden of foodborne disease relies on systematic sampling data from many foods. For example, the patterns of molecu- lar subtypes in strains of Salmonella isolated from people can be compared and matched to those of strains isolated from a variety of foods. This can help relate specific pathogenic subtypes and diseases to specific foods. To be successful, this strategy depends on extensive and systematic sampling of many foods and on the

46 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD use of standardized subtyping methods on a large number of strains. This strategy has been routinely applied in Denmark to track the burden of salmonellosis associated with various foods (Hald and Bronsted, 2002~. Finally, if data on pathogen prevalence are available for a large number of foods, a risk allocation can be constructed using the methods of risk analysis that have been used for L. monocytogenes (FSIS/CFSAN, 2001~. This approach depends on the assumptions that all strains of a pathogen are equally likely to cause disease, and that the distribution of the pathogen in foods can be reliably estimated from studies using a broad range of methods and conducted over a substantial time span. Once a food is implicated as a common source of a pathogen, a detailed review of its production process may reveal the likely points in the process where the food became contaminated. This is an important phase of intensive outbreak investigations that often involves tracing back along the production process from the implicated food the ill persons ate. Such a review may identify where the contamination was likely to have originated and where it may have been further amplified or controlled. This information, of particular interest to risk assessors, is only gathered in a minority of foodborne outbreak investigations and requires a multidisciplinary approach. Index of Consumer Behavior Surveys of consumer behavior can provide a useful index of behavior, sub- ject to the limitations associated with reporting by consumers. The 12-state BRFSS survey of 1995 to 1996 showed that in the preceding 12 months, 50 percent of those interviewed ate undercooked eggs, 20 percent ate pink ground beef, 19 percent did not wash their hands after handling raw meat or chicken, 8 percent ate raw oysters, and 1.4 percent consumed raw milk (Yang et al., 1998~. The frequency of consumption of pink hamburgers was higher in men, increased with education and salary, but decreased with age. As the correlation between hamburger color and degree of doneness is imperfect, these data do not mean that the persons interviewed necessarily ate undercooked hamburger (Berry et al., 1998; Hunt et al., 1999; Killinger et al., 2000~. In the most recent cycle of FoodNet population surveys, 27 percent of respondents reported that they ate a raw or runny egg dish in the preceding month, 26 percent ate pink ground beef, and 2.5 percent ate raw oysters. This survey also included questions about thermometer use in cooking (recommended by USDA to measure an internal temperature of 160°F as an indication of doneness); only 3 percent reported using a thermometer when cooking hamburgers (Yang et al., 1998~. Other food safety surveys were conducted by FDA in 1988, 1993,1998, and 2001 to gather data on consumer food-safety practices related to cross- contamination and consumption of potentially risky foods (Consumer Studies Branch, 2002~. The data showed large improvements consisting of the reduction

THE SCIENCE OF PUBLIC HEALTH SURVEILLANCE 47 of cross-contamination and the decreased consumption of potentially risky foods between 1993 and 1998 that were maintained at the time of the 2001 survey. However, notable exceptions to this trend were an increase in the consumption of raw clams, raw oysters, and raw fish from 1998 levels. The proportion of the population who reported not washing their hands after touching raw meat or after cracking eggs decreased from 29 percent and 66 percent, respectively, in 1993, to 15 percent and 55 percent in 2001. Although the proportion of the population who reported eating pink hamburger declined from 24 percent in 1993 to 16 percent in 2001, research (Berry et al., 1998; FSIS, 1998a, 2000; Hunt et al., 1999) has demonstrated that cooked ground-meat color is not an indication of safety. Based on this research, in the late 1990s USDA began recommending the use of a thermometer to check the internal temperature of cooked hamburgers. Analysis of existing surveys, focus groups, and observational data, con- ducted by the Research Triangle Institute for USDA, also indicated improved food safety knowledge and practices, as reported by consumers (RTI, 2002a). For example, this analysis indicated that the proportion of the population using ther- mometers when cooking hamburger doubled from 3 percent in 1998 to 6 percent in 2001. A certain degree of disparity between consumer-reported practices and observed behavior was also noted. RTI recommended additional educational efforts to encourage consumer changes in behavior concerning proper cleaning, heating, refrigeration, and use of thermometers. Overall, consumer behavior surveys indicate that although some changes in consumer behavior have occurred, consumer habits are still frequently less than optimal. The committee recommends periodic repetition of such surveys to help document behavioral changes concerning food safety in the population at large as a result of consumer education efforts, and to target food safety messages to subgroups of the population that engage in risky food-preparation and consump- tion behavior. MONITORING HAZARDS IN THE FOOD CHAIN Systematic Monitoring Routine systematic monitoring at various points of the food supply is the main form of surveillance for many toxic hazards for which the associated human illnesses are hard to diagnose and are persistent in nature. For example, FDA conducts a systematic pesticide residue monitoring program (CFSAN, 2002), shellfish beds are routinely monitored for evidence of fecal contamination, and imported shellfish are sampled for pathogens. As new foodborne hazards emerge, a system for rapid assessment of their prevalence at various points in the food supply is critical to developing prevention measures. For example, brains of cattle with evidence of neurological disease are tested for the presence of bovine

48 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD spongiform encephalopathy just after slaughter, providing an indication of the likely absence of the infectious prion in the food supply (APHIS, 2002~. For infectious pathogens, few systematic sampling programs exist in the public sector, although internal monitoring by industry is common. Although the PR/HACCP rule requires the monitoring of generic E. cold on carcasses in slaughter plants, these data are not publicly available and thus cannot be used to measure the overall effectiveness of the PR/HACCP rule or to compare contami- nation levels among individual producers or groups of producers. Because sys- tematic monitoring is a powerful tool for tracking specific microbial hazards, particularly if coupled with molecular subtyping, the committee recommends expansion of this type of monitoring to all high-risk food groups. For meat and poultry, although not designed to be an optimal surveillance system, product sampling as part of PR/HACCP verification provides some in- formation about the frequency of Salmonella in specific meat and poultry products, about the impact of plant size on contamination levels, and about trends in specific serotypes (Rose et al., 2002; RTI, 2002b). The committee believes that the value of this information would increase if such data were collected system- atically throughout the year, analyzed in ways that accounted for various process- ing plant characteristics, and used by the various plants to benchmark their per- formance compared with that of their peers. Further, anonymous linking of the library of subtype patterns thus generated for detected pathogens to public health subtyping systems could also provide valuable information regarding sources of contamination (e.g., to risk assessors). To this end, the committee suggests that a third-party repository be established for environmental and product testing data from industry, using subtyping methods comparable to those used in public health, and maintained in an anonymous fashion and with voluntary subscription. Disease-causing microorganisms and other hazards in the food chain can be tracked in targeted surveys of the environment, food animal reservoirs, and foods themselves. These surveys can be used to estimate risks associated with certain foods and to identify or design strategies to control or mitigate these risks. When systematically gathered, such information can also be used to monitor trends in contamination and to measure the impact of control strategies. In addition to the final public health surveillance outcome, this information can provide an indica- tion of the effectiveness of specific control measures. The committee concluded that systematic sampling of animals at the farm, and especially immediately before slaughter, may be particularly useful to mea- sure the frequency of the presence of important human pathogens such as S. Newport and E. cold 0157:H7 in animal populations. Given that the potential importance of pro- and postharvest infection of live animals needs to be assessed to obtain a clear understanding of contamination routes, the committee recommends that systematic sampling of animals for patho- gens at the point of slaughter be undertaken, analogous to the National Animal Health Monitoring System (NAHMS) surveys of producers conducted by

THE SCIENCE OF PUBLIC HEALTH SURVEILLANCE 49 USDA's Animal and Plant Health Inspection System (APHIS). The results of this sampling should be linked to those of other systematic sampling programs in existence, such as the NAHMS surveys and PR/HACCP monitoring. In addition, the sampling should be concurrent with an enhancement of the capacity of food and agriculture laboratories to rapidly subtype Salmonella and E. cold 0157:H7 isolates from the various systematic sampling programs. FSIS and FDA, in turn, should conduct periodic, systematic microbiological surveys of food-processing plants, with sampling at various points in the production chain. The results of these surveys should serve as the basis to revise baselines on the prevalence of pathogen and indicator microorganisms and to better characterize the structure of the industry and its pathogen reduction practices. The committee also recom- mends that, for all surveys, collection of samples for Salmonella be conducted within the same time frame each year, completed without interruptions or delays, and reported annually, in aggregate form, by size of establishment. In addition, the committee believes that further studies of farm, production, transport, and lairage-related risk factors for microbiological contamination of food animals are urgently needed to better define control points and strategies at these levels. Conducting additional studies on pathogen prevalence in animals arriving at processing plants would be a critical component for progress in foodborne disease prevention. The contamination is not likely to be random. By comparing sources, transport routes and conditions, and other characteristics of the incoming live animals, the factors that predict higher contamination levels could be defined. This information could target further research into how con- tamination occurs and how it may be prevented on the farm, in the feedlot, or during transportation. It could also be used to channel into special processing the animals most likely to be contaminated. For example, though unusual, it is standard practice for an individual animal on a slaughter line to be "passed for cooking" when a veterinary inspector iden- tifies a lesion that indicates localized tuberculosis (9 C.F.R. §311.2~. The carcass is removed from the main slaughter line and sent on a different path to receive a fully supervised cook. In recent years, egg farms that are known to have S. Enteritidis on the premises routinely send their eggs for pasteurization under voluntary Egg Quality Assurance Programs. In a new program, the Norwegian Agriculture Department is testing broiler flocks for Campylobacter and requiring positive flocks to be slaughtered after negative flocks to avoid cross-contamination at the plant; carcasses from positive flocks are then cooked or frozen under supervision (Norwegian Zoonosis Centre, 2002~. Therefore, in the future, groups of animals most likely to be contaminated may be designated for uses other than sale in raw form or may be processed in particular ways to minimize the contami- nation of raw final products. For produce, recent FDA surveys of imported and domestic items identified Salmonella or Shigella on 4.4 percent (44 out of 1,003 samples) of imported items (OPDFB, 2001) and on 1.6 percent (12 out of 767 samples) of domestic

so SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD produce items (CFSAN, 2001~. Among the latter, 0.8 percent (6 samples) were positive for Salmonella and an equal percentage were positive for Shigella. Considering the increasing importance of raw produce as a vehicle of food- borne infections in the United States, the committee recommends that high-risk (i.e., known to be frequently associated with foodborne infections) raw produce, both domestic and imported, be systematically monitored for such indicators of fecal contamination as generic E. coli, and for prevalence of such pathogens as Salmonella and Shigella. The results of this monitoring should be linked to studies of the specific determinants of such contamination and of the relationship between indicator organisms and pathogen prevalence. Periodic Monitoring In addition to systematic surveys, periodic surveys can also provide useful information. Following the release of a National Academies report (NRC, 1985), NAHMS began conducting surveys on food-animal production that provide snap- shots of the prevalence of animal and human pathogens and of management practices on farms (Wineland and Dargatz, 1998~. For example, a NAHMS sur- vey of layer-hen farms conducted in 1999 showed that 7.1 percent of farms had S. Enteritidis on their premises, that farms having high rodent populations were much more likely to be S. Enteritidis-positive, and that 56 percent of the farms participated in major egg quality assurance programs. Encouragingly, farms that practiced careful hen-house cleaning and disinfection between flocks did not have S. Enteritidis in their environments (APHIS, 2000~. Similar periodic, tar- geted surveys at other points in the food chain could provide important informa- tion. For example, although it is known that animal feeds may be contaminated with Salmonella, the source and frequency of contamination of specific feed ingredients remains undefined (Crump et al., 2002~. Pigs free of Salmonella at the farm were shown to acquire Salmonella infections in temporary holding pens just before slaughter, indicating that it was not the nonspecific stress of transport or holding, but specific exposure to Salmonella after the animals left the farm that was the most important determinant of carriage at slaughter (Hurd et al., 2001~. Similarly, a recent study of dairy animals at slaughter in the United Kingdom suggested that 75 percent of the E. cold 0157:H7 on their hides were the result of contamination that occurred after the animals had been transported to slaughter (Avery et al., 2002~. Standardization of Monitoring Methods Food hazard surveillance is usually a shared responsibility of the food indus- try and local, state, or federal regulatory agencies. Data generated are not stan- dardized and thus are difficult to compare. Lack of standardization of foodborne,

THE SCIENCE OF PUBLIC HEALTH SURVEILLANCE 51 microbial-hazard surveillance data hinders the development of nationwide hazard assessments and reduces the value of much of the surveillance. A consortium of state and federal regulatory agencies known as the National Food Safety System (NFSS) has begun to address the interrelated issues of labo- ratory accreditation, methods validation, and national data-sharing standards. Currently, clinical, environmental, and food microbiology laboratories are accredited by a variety of bodies, each with different standards and evaluation criteria. An NFSS workgroup is encouraging the accrediting bodies to accept the International Organization for Standardization 17025 standard, so that they abide by a single standard. To address methods validation, AOAC International is developing an electronic compilation of analytical methods (e-CAM) to serve as a repository of validated methods and is providing peer review for validating new methods (AOAC, 2002~. Finally, development of technical standards for the electronic exchange of data between food safety laboratories has begun with the electronic laboratory exchange network (eLEXNET), which was pilot tested in September 2000 and connected 38 laboratories in 26 states by 2002. The committee recommends that compatible subtype and antimicrobial re- sistance surveillance data from humans, animals, farms, and food products should be linked among such agencies and services as CDC, APHIS, FSIS, FDA (includ- ing its Center for Veterinary Medicine), and other state and federal laboratories. To facilitate these linkages, NFSS plans should be implemented to (1) provide for uniform accreditation of food safety laboratories, (2) promote the use of validated methods and the rapid validation of new methods, and (3) expand the scope of participation by food safety laboratories in eLEXNET. Association of Human Diseases with Specific Reservoirs Comparing information from monitoring and surveillance in animals, foods, and humans can document and even quantify the flow of specific pathogens from particular reservoirs to humans. For example, Denmark has established a compre- hensive surveillance system that includes extensive, systematic sampling of many foods and animal groups for Salmonella, and subtyping of Salmonella strains, which allows it to define the annual contribution of each of the animal reservoirs to human illness in that country (Hald, 2001; Hald and Bronsted, 2002~. These data provide a clear illustration of the link between the contamination of food and the resulting infections in humans and the effectiveness of targeted Salmonella control programs. In Denmark, these data drive the prevention strategies from farm to table. Hence, screening pork for antibodies to Salmonella on the farm has been used to identify pork herds that have a high prevalence of Salmonella. These animals are slaughtered separately from animals that come from herds with a low prevalence of Salmonella in order to avoid cross-contamination during slaughter and dressing; they are also used only in cooked products (Hald, 2001~.

52 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD In the United States, strains of Salmonella from NAHMS surveys, from HACCP monitoring, and from veterinary diagnostic laboratories are referred to USDA' s Agricultural Research Service for determination of antimicrobial resis- tance as part of the NARMS system (ARS, 2000~. Routine characterization of Salmonella from the NARMS system using molecular fingerprinting and comparison of these data to similar data on the human isolates from foodborne outbreaks would make it possible to connect human infections with specific subtypes of Salmonella to specific animal reser- voirs, similar to the Danish model. The committee recommends that Salmonella continue to be tracked in foods as an important foodborne pathogen. It is the only pathogen for which human surveillance systems are widely distributed. While food safety policy may be guided by monitoring hazard levels in animals or foods, and contaminated food certainly is associated with human illness, the relationships that link contamination levels in foods at processing with incidences of human illness is likely to be more complex than a simple one- to-one linear correspondence. Factors such as multiplication of microorganisms during distribution and preparation undoubtedly affect this relationship. Although careful cooking may eliminate many pathogens from the final food, cross- contamination in the kitchen may easily transfer microbes from raw products to other foods (Redmond et al., 2002~. Moreover, the state of the host may make exposure to a low dose of a pathogen highly problematic or inconsequential (see later section, "Pathogenesis"~. Risk assessment can attempt to model this com- plex series of relationships, but major uncertainties will still remain. Document- ing the level of a hazard in foods and comparing changes in that level with the final incidence of disease can empirically define the nature of the relationship. Thus, allocating the burden of illness to different foods and defining the points at which contamination occurs is a complex and imperfect process. It would be helpful to have a mathematical model that allocates hazards of food- borne illness across all food groups and allocates risks across all consumers, but the available data do not permit the development of such a complete and rigorous model. In the absence of such a model, the committee believes that monitoring microbiological contamination at various points in the food production and distri- bution chain can provide benchmarks to develop standards based on performance and current understanding of risk. These benchmarks and standards must be updated as new information emerges. The level of processing needed to make a food safe may depend on the likelihood that the product is contaminated. As mentioned before, microbiological methods are used to determine which flocks should send their eggs to pasteuriza- tion and whether to open or close shellfish beds to raw oyster harvest. In the future, scientific studies of sources and frequencies of contamination at several points in the process may differentiate various levels of contamination. This information could help identify sources with a higher risk of contamination, to

THE SCIENCE OF PUBLIC HEALTH SURVEILLANCE 53 target additional production and processing control steps, and to produce and evaluate a continually safer food supply (Guerrant and Theno, 1995~. PATHOGENESIS Whether a person is infected by a microorganism depends on a wide range of microbial and host factors. Numerous microbial virulence factors determine infectious doses and pathogenicity, while host susceptibility is determined by genetics, special conditions (e.g., pregnancy), immunity (e.g., vaccination, acquired immune deficiency syndrome), and behavior (e.g., hygiene, education, culture, food preparation methods). These factors and their estimates would con- tribute to the information required to make risk assessment modeling (as described later in Chapter 3) more complete and accurate. There is a dose-response relationship for many foodborne infections. The dose level at which 50 percent of exposed individuals will be infected will be much higher than the dose level at which only 5 percent of exposed individuals will be infected. In the context of an outbreak, dose-responses may correlate with attack rates. Thus, in many Salmonella outbreaks in which the food was contami- nated with only a few organisms, the attack rates were similarly small (Blaser and Newman, 1982~. However, even with a low attack rate, large-volume production can mean that the number of infected people, and therefore the outbreak itself, is very large. For example, in a large nationwide outbreak associated with ice cream, only 6 percent of persons who ate the ice cream became ill, perhaps because the ice cream was contaminated with only six or fewer Salmonella cells per serving (Hennessy et al., 1996~. Because of the large volume of production and its nationwide distribution, however, an estimated 224,000 cases occurred during this outbreak. Microbial-Related Factors Microbial threats to our food and water supplies range from toxins and viruses to bacteria, molds, and parasites. While many of these are easily inacti- vated or killed by sanitizers, heat, or radiation, or removed by filtration, others are resistant to these and other control measures. Unlike viruses and parasites that do not multiply outside their animal hosts, small numbers of bacteria typically multiply to large numbers when conditions permit. Infectious doses that cause disease in the majority of healthy hosts may range from over 1 million organisms for certain bacteria such as V. cholerae to as few as one to ten organisms for pathogens such as Cryptosporidium or Shigella (Guerrant and Steiner, 1999~. Many bacteria have the capacity to increase their resistance to acid, heat, drying, and peroxides through a range of inducible mecha- nisms. Bacteria stressed by one environmental challenge may become more resis- tant to a range of other environmental stresses and may become even more

54 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD invasive (Humphrey et al., 1996~. Furthermore, mobile genetic elements, some- times transmissible as plasmids, phages, or even as naked deoxyribonucleic acid, enable microorganisms to rapidly acquire new virulence and resistance properties. Host-Related Factors Host-related factors also influence microbial infectious doses. For example, neutralization of gastric acidity (e.g., higher stomach pH) reduces the infectious doses of Vibrio, Salmonella, and E. cold (Gitelson, 1971; Hornick et al., 1971~. This fact places the gastrectomized patient taking antacids at greatest risk when exposed to a potential pathogen (Baine et al., 1974~. Similarly, when critical defenses provided by normal bacterial flora are altered by antibiotics, a resistant pathogen may be favored and may complicate therapy for other infections (Barza and Travers, 2002~. For example, people taking antibiotics were at a sixfold higher risk than others of acquiring a resistant Salmonella infection in the 1985 Chicago outbreak of salmonellosis (Ryan et al., 1987~. An earlier report involv- ing Norwegian tourists visiting Spain in the 1960s showed that those who took prophylactic antibiotics were more likely to acquire salmonellosis than those who did not (Mentzing and Ringertz, 1968~. Furthermore, immunocompromised patients are not only at greater risk of acquiring enteric infections, but also of suffering from them more severely and experiencing difficulty in overcoming them; examples include salmonellosis and cryptosporidiosis in patients immunocompromised by age, chemotherapy, or immunodeficiency (Navin and Juranek, 1984; Sperber and Schleupner, 1987~. Finally, host educational, cul- tural, and behavioral factors also profoundly influence the risk of acquiring foodborne infections (Mead and Mintz, 1996~. Knowledge about food choices, cleanliness, storage, preparation, cooking, and serving practices can help reduce the risk posed to the host by microbial hazards in foods. USE OF PUBLIC HEALTH DATA TO IMPROVE FOOD SAFETY: SPECIFIC EXAMPLES Preventing foodborne disease is complex, requiring attention and interven- tion from farm or fishery to table (TOM/NRC, 1998~. There are no vaccines for the pathogens that are most commonly transmitted through foods, and while education of the consumer provides an important final safety barrier, it is not by itself sufficient. Making food safer before it reaches the consumer is critical to maintain confidence in the food supply. The consumer eats many foods without cooking them; prepares raw foods of animal origin with the same hands that prepare uncooked salads; is instructed by tradition and by cookery texts to pre- pare many meat, poultry, egg, and seafood dishes with more concern about over- cooking than undercooking; and is told routinely to season dishes "to taste" during the preparation process.

THE SCIENCE OF PUBLIC HEALTH SURVEILLANCE 55 When new foodborne hazards are identified, the knowledge base for defining effective preventions may be quite limited (Holmberg and Feldman, 1984~. Public health surveillance, with detailed investigations of outbreaks, can identify new and emerging hazards, can help define the likely points of control and the questions in need of further research, and can track the effectiveness of control measures. For some hazards, the control measures seem obvious and immediate. For example, requiring toilets with holding tanks on oyster boats made it less likely that oyster gatherers would contaminate the oyster beds with Norwalk-like virus (Kohn et al., 1995~. Similarly, providing appropriate toilet and hand washing facilities for field workers, and ensuring that such facilities are properly used, would likely reduce the incidence of workers contaminating produce with enteric pathogens. Providing restaurant kitchens with dedicated hand washing stations, in turn, would be expected to reduce the risk of microbiological cross-contamination of foods. For other situations, the relative merits of potential strategies to minimize or fully prevent microbial contamination of foods are less obvious at the outset, and development of controls must proceed by an iterative process. As more is learned about the settings of outbreaks, prevention strategies are progressively refined. Five examples are presented below to illustrate how this process can lead to improved prevention. Salmonella and Precooked Roast Beef From 1975 to 1977, surveillance detected repeated outbreaks of Salmonella infection associated with precooked deli roast beef (Parham,1984~. Evaluation of cooking temperatures revealed that they were sometimes insufficient to kill Salmonella present in raw beef, and consequently, an improved approach that used specific temperature requirements was applied as an emergency regulation in 1977. In 1981, outbreaks of salmonellosis were again traced to precooked roast beef prepared under these new regulations, showing that these measures were still insufficient (CDC, 1981~. In addition to time and temperature of cooking, further studies identified humidity inside the oven as a critical cofactor in deter- mining Salmonella survival (Parham, 1984~. Since further regulations have been promulgated, outbreaks traced to precooked roast beef have become extremely rare. Escherichia cold 0157:H7 and Apple Cider In 1992, investigation of an outbreak of E. cold 0157:H7 infections in Massachusetts linked this pathogen to apple cider for the first time (Besser et al., 1993~. This traditional beverage was often pressed from fallen apples, with mini- mal cleaning, but was long believed to be sufficiently acidic to be safe. However, assessment of survival of the microorganism in apple cider revealed that E. cold 0157:H7 was unusually acid tolerant and could easily survive in cider having a

56 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD pH lower than the value that was considered safe until then (pH 4.5) (Zhao et al., 1993~. Investigators of the Massachusetts outbreak thought that the apples were probably contaminated before they were pressed, possibly in the orchard, which was visited by deer. The first control measures adopted by the industry were simply to wash and brush the apples before pressing them. Yet recurrent out- breaks of E. cold 0157:H7 and Cryptosporidium infections occurred that were traced to cider made from apples that had been brushed and washed, which showed that even with cleansing of the apples, cider could be hazardous (CDC, 1997; Cody et al., 1999; Millard et al., 1994~. It was also shown that E. cold 0157:H7 could, under some circumstances, be internalized into apples and thus be protected from washing, brushing, or external disinfection (Buchanan et al., 1999; Burnett et al., 2000~. The occurrence of outbreaks of Salmonella infections also attributed to fruit juices, as well as recent related research, has led to the promulgation of juice regulations requiring a pathogen-reduction step such as pasteurization (FDA, 2001~. To date, no further commercial juice- or cider- associated outbreaks have been reported. Salmonella Enteritidis and Shell Eggs In the 1980s, dramatic outbreaks of S. Enteritidis infections were traced to Grade A shell eggs (St. Louis et al., 1988~. This was surprising, as the egg grading and disinfection process instituted in the 1960s (as a result of egg- associated salmonellosis related to contamination of the outside of the shell by Salmonella in chicken feces) had appeared to be effective. It was suggested that the new problem might reflect internal contamination of eggs, possibly as a result of infection of the hen's reproductive tissues. Sporadic cases of S. Enteritidis infections were also increasing, first in the Northeast, and later over most of the country (CDC, 1993~. It was possible to relate these cases to eggs and even to show a gradient of risk according to the degree of cooking, from hard-boiled and hard-cooked through over-easy, to soft-boiled and sunny-side-up (Hedberg et al., 1993; Passaro et al., 1996~. The Salmonella strains in the birds on farms that were the source of contaminated eggs were the same as the strains found in the affected humans, confirming that the source of contamination was the birds themselves (Altekruse et al., 1993; Mishu et al., 1991~. Experimental feeding of Salmonella to birds demonstrated that the birds developed silent ovarian infection and then laid normal-looking eggs that had contaminated contents (Gast, 1999~. In the early l990s, a pilot project to develop flock-based screening and control measures was begun: the Pennsylvania Egg Quality Assurance Program (Schlosser et al., 1999~. This project became the model for other states' egg quality assurance programs. The incidence of S. Enteritidis infections in mid- Atlantic states, for which Pennsylvania was the main egg source, began decreas- ing, followed later by decreases in other states (FDDB, 2000~. Microbiological screening of farms for S. Enteritidis is an integral part of an egg quality assurance

THE SCIENCE OF PUBLIC HEALTH SURVEILLANCE 57 program, with voluntary diversion of the eggs to liquid egg pasteurization if they are found positive. Thus, many potentially tainted eggs are sent for safe process- ing before they enter the shell egg market. As the epidemic among egg-laying flocks spread from the Northeast to virtually the entire country, outbreak investi- gations and the attendant trace-backs demonstrated the spread of this problem into new areas and stimulated local authorities to develop their own quality assurance programs for S. Enteritidis in eggs and egg products (Burr et al., 1999; CDC, 1993). A risk assessment was completed in 1998 (Baker et al., 1998~. The sustained epidemic prompted further measures, such as the refrigeration requirement for eggs in 1998 and 2000 (FDA, 2000; FSIS, 1998b) and the commercialization of a new in-shell pasteurization process. Current control policies of egg-associated S. Enteritidis appear to be having an impact. By 2000, the incidence of S. Enter- itidis had decreased to 2 per 100,000, down from the peak of nearly 4 per 100,000, although it remains above the pre-epidemic incidence of 1 per 100,000 (Figure 2.2~. However, egg-associated outbreaks continue to occur (CDC, 2003~. The surveil- lance data clearly show that progress is being made in slowing the S. Enteritidis nrnhlem in eggs, but further efforts are needed to completely control it. Salmonella, E. cold 0157:H7, and Alfalfa Sprouts Like S. Enteritidis in eggs, the new food safety problem with alfalfa sprouts is not an emerging pathogen, but rather the emergence of well-known pathogens in a different food. In 1995, shortly after the statistical outbreak detection algo- rithm was developed for the Salmonella surveillance system, a large, 22-state outbreak of infections caused by a rare serotype, S. Stanley, was detected in the United States (Matron et al., 1997~. Simultaneously, public health officials in Finland identified an outbreak caused by the same organism. Both outbreaks were linked to the consumption of alfalfa sprouts, sprouted from the same batch of seeds (Matron et al., 1997~. Research showed that the sprouting process could greatly amplify the number of salmonellae originally present in the seed and that the pathogen could be inside the sprout, where it might not be affected by wash- ing or disinfecting (Itoh et al., 1998; Jaquette et al., 1996~. The next three years witnessed at least seven outbreaks in the United States, caused by several sero- types of Salmonella and E. cold 0157:H7 in sprouts, often from contaminated seeds (Taormina et al., l999~. Japan experienced a devastating outbreak traced to radish sprouts that affected 6,000 school children (Michino et al., 1999; Watanabe et al., 1999~. Alfalfa and other seeds for sprouting are produced as raw agricul- tural commodities and may be easily contaminated in the field or warehouse, where they may be held for years before being sprouted (Breuer et al., 2001~. After researchers determined that disinfecting seeds with 20,000 ppm cal- cium hypochlorite could reduce contamination and preserve the ability of seeds to germinate, FDA promulgated guidelines on seed disinfection (FDA, 1999),

58 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD and the major seed distributors put these instructions on the seed packages. Since then, outbreaks of salmonellosis have been linked to a sprout producer that reported disinfecting the seeds following those guidelines (Proctor et al., 2000), as well as to a sprout producer using less chlorine than recommended (Winthrop et al., 2003~. Another recent outbreak involved a single lot of clover seed shipped to two sprout producers in Colorado (Brooks et al.,2001~. The first did not disinfect the seed before sprouting and caused 1.13 documented infections per 50 lb-bag of seed sprouted, whereas the second did disinfect the seeds and caused only 0.29 infections per bag of seed. These outbreaks show that the disinfection strategy works partially, but is by itself insufficient to completely protect the public. In addition to disinfection, FDA also recommended lot-by-lot testing of the irriga- tion water for Salmonella (FDA, l999~. One outbreak occurred that was linked to sprouts that had passed such a test, suggesting that false negative tests may occur (Winthrop et al., 2003~. Continued surveillance and investigation indicate that the challenge of preventing outbreaks of salmonellosis from sprouts has been par- tially met, but complete prevention has still not been achieved. Multidrug-Resistant Salmonella Newport and Foods of Bovine Origin One of the latest food hazards to emerge in the United States is a new and highly resistant strain of S. Newport (Zansky et al., 2002~. This strain was first identified through NARMS surveillance in 1998, and its detection increased rapidly in 1999 and 2000. The strain is resistant to at least nine antibiotics because it possesses a large plasmid bearing several resistance genes, including an unusual gene, the AmpC cmy2 gene, which confers resistance to most cephalosporins. In 2001, a retrospective study of these strains in Massachusetts identified the same strains in ill and dying dairy cattle, and showed that visiting or working on dairy farms was a risk factor for illness (Gupta et al.,2001~. Later that year, an outbreak in Connecticut was traced to traditional cheese made from insufficiently pasteur- ized milk from Massachusetts dairy farms (McCarthy et al., 2002~. In 2002, an investigation of a multistate cluster of cases in the Northeast linked the illness to eating ground beef traced to meat from a single slaughter plant (Zansky et al., 2002~. Surveillance of human infections indicates a sharp increase in S. Newport infections, which in 2001 represented 10 percent of human salmonellosis (FDDB, 2002c). Many of the S. Newport strains are multidrug resistant (CDC, 2002b). The same organism has been detected since 1998 among isolates from animals, including bovines (Fedorka-Cray et al., 2002~. Among S. Newport isolated from cattle in 2000,74 percent had the AmpC multidrug resistance profile (ARS, 2002~. The evidence to date indicates that this strain has spread in epidemic fashion among cattle herds and that it affects the animals themselves, persons in contact with the animals, and consumers of bovine products (including meat, cheese, and other foods). Once control measures begin, success can be measured by monitor- ing animals and meat for this strain, by trends in human illness, and by outbreak

THE SCIENCE OF PUBLIC HEALTH SURVEILLANCE 59 surveillance. Surveillance activities in animals, meat, and poultry can also pro- vide early warning of the spread of this strain or its plasmid to other food-animal populations. ANTICIPATING THE FUTURE In the future, it can be expected that new pathogens and new foodborne modes for transmission of such pathogens will continue to be recognized. New diagnostic strategies will identify some pathogens that currently are often or completely missed. Globalization of the food supply and concentration of food production, in turn, will create new challenges for detection, investigation, con- trol, and prevention of microbial foodborne hazards. The committee concludes that enhanced public health surveillance for human foodborne illnesses will be vital to identify and investigate these new challenges. In addition, it believes that a flexible monitoring system is needed that permits comparison of information from multiple points in the food supply. Just as moni- toring individual cattle at slaughter is an important strategy for documenting the continuing absence of bovine spongiform encephalopathy, a system for docu- menting the frequency of microbial or other foodborne hazards at the point of slaughter or processing could be critical to assessing and controlling these haz- ards in the future. Systematic surveys of potential hazards, such as the appearance of antibiotic resistant microbial strains in live animals in production, already provide information useful to industry, regulators, and the public health sector. In the future, similar systematic surveys of microbial contamination in various cat- egories of processing plants and at various points along processing lines could be equally useful for risk assessors. Preventing or minimizing contamination early in the chain, as well as identifying foods at higher risk of being contaminated so that they can be diverted out of the raw product market and into safer processing, may become the norm. For some foods, irradiation and other terminal microbial decontamination steps hold great potential (Tauxe, 2001~. High-pressure process- ing, for example, is already commercially available. Preventing foodborne dis- ease means preventing contamination before food reaches the consumer. Risk- management policies applied throughout the food system on farms, fisheries, and orchards; in slaughter facilities and processing plants; during transportation and storage; and in retail food stores, food service establishments, and homes- are all key parts of food safety. For certain products, it may be possible to define varying levels of process- ing depending on microbiological and other markers of the risk that they are contaminated. Already, eggs that are cracked or that come from farms contami- nated with S. Enteritidis are routinely approved for marketing after pasteuriza- tion; milk for manufacturing purposes meets standards that are different from Grade A milk; and an occasional carcass is passed for cooking rather than being

60 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD allowed to go through standard slaughter. In the future, such treatments of higher- nsk food may be a useful tool for achieving pathogen reduction in other foods. REFERENCES Altekruse S. Koehler J. Hickman-Brenner F. Tauxe RV, Ferris K. 1993. A comparison of Salmonella Enteritidis phage types from egg-associated outbreaks and implicated laying flocks. Epidemiol Infect 110:17-22. Angulo FJ, Swerdlow DL. 1999. Epidemiology of human Salmonella enterica serovar Enteritidis infections in the United States. In: Saeed AM, Gast RK, Potter ME, Wall PG, eds. Salmonella enterica Serovar Enteritidis in Humans and Animals: The Epidemiology, Pathogenesis and Control. Ames, IA: Iowa State University Press. Pp. 33-41. AOAC (Association of Analytical Communities). 2002. A vision: The National Food Safety System and the three-legged stool. Inside Laboratory Management May/June:e7-el2. Online. Avail- able at http://www.aoac.org/AOAC_e-CAM.pdf. Accessed January 19, 2003. APHIS (Animal and Plant Health Inspection Service). 2000. Highlights of Layers '99 Study Results: Salmonella enterica serotype Enteritidis. Online. U.S. Department of Agriculture (USDA). Available at http://www.aphis.usda.gov/vs/ceah/cahm/Poultry/lay99SEhi.htm. Accessed Sep- tember 11, 2002. APHIS. 2002. Bovine Spongiform Encephalopathy. Online. USDA. Available at http:// www.aphis.usda.gov/lpa.pubs/fsheet_faq_notice/fs_ahbse.html. Accessed December 5, 2002. ARS (Agriculture Research Service). 2000. National Antimicrobial Resistance Monitoring System (NARMS)—Animal Veterinary Isolate Data. Online. USDA. Available at http:// www.arru.saa.ars.usda.gov/narms/narms.htm. Accessed September 5, 2002. ARS. 2002. NARMS—EB 2000; Veterinary isolates. Figure 18. Resistance among Salmonella sero- types for isolates from cattle. Online. USDA. Available at http://www.ars-grin.gov/ars/ SoAtlantic/Athens/arru/narms_2000/2000_figure_18.pdf. Accessed September 11, 2002. Avery SM, Small A, Reid CA, Buncic S. 2002. Pulsed-field gel electrophoresis characterization of Shigatoxin-producing Escherichia cold 0157 from hides of cattle at slaughter. J Food Prot 65:1172-1176. Baine WB, Zampieri A, Mazzotti M, Angioni G. Greco D, Di Gioia M, Izzo E, Gangarosa E, Pocchiari F. 1974. Epidemiology of cholera in Italy in 1973. Lancet 2:1370-1381. Baker AR, Ebel ED, Hogue AT, McDowell RM, Morales RA, Schlosser WD, Whiting R. 1998. Salmonella Enteritidis Risk Assessment: Shell Eggs and Egg Products. Washington DC: Food Safety and Inspection Service, U.S. Department of Agriculture. Barza M, Travers K. 2002. Excess infections due to antimicrobial resistance: The "attributable frac- tion." Clin Infect Dis 34: S 126-S 130. Bean N. Griffin P. 1990. Foodborne disease outbreaks in the United States, 1973-1987: Pathogens, vehicles, and trends. J Food Prot 53:804-8 17. Bean N. Martin S. Bradford H. 1992. PHLIS: An electronic system for reporting public health data from remote sites. Am J Public Health 82:1273-1276. Bell BP, Goldoft M, Griffin PM, Davis MA, Gordon DC, Tarr PI, Bartleson CA, Lewis JH, Barrett TJ, Wells JG, Baron R. Kobayashi J. 1994. A multistate outbreak of Escherichia cold 0157:H7 Associated bloody diarrhea and hemolytic uremic syndrome from hamburgers. The Washington experience. JAMA 1349-1353. Bell BP, Shapiro CN, Alter MJ, Moyer LA, Judson FN, Mottram K, Fleenor M, Ryder PL, Margolis HS. 1998. The diverse patterns of hepatitis A epidemiology in the United States Implications for vaccination strategies. J Infect Dis 178:1579-1584.

THE SCIENCE OF PUBLIC HEALTH SURVEILLANCE 61 Bender JB, Hedberg CW, Besser JM, Boxrud DJ, MacDonald KL, Osterholm MT. 1997. Surveil- lance for Escherichia cold 0157:H7 infections in Minnesota by molecular subtyping. N Engl J Med 337:388-394. Bender JB, Hedberg CW, Boxrud DJ, Besser JM, Wicklund JH, Smith KE, Osterholm MT. 2001. Use of molecular subtyping in surveillance for Salmonella enterica serotype Typhimurium. N Engl J. Med 344:189-195. Bender J. Smith K, McNees A, Rabafsky-Ehr T. Segler S. Hawkins M, Spina N. Keene W. VanGilder T. Kennedy M, Hedberg C. In press. Surveillance for E. cold 0157:H7 infection in FoodNet sites, 1996-1999. Clin Infect Dis. Berry BW, Lyon BG, Soderberg DL, Philipo T. Bakowski R. Henry C, Dreas J. Powell S. Agmata E, Yee C, Clinch N. Vanlonkhuyzen D. 1998. Visual and Instrumental Evaluations of Color in Beef Patties Obtained Nationwide and Cooked to Four Internal Temperatures. Presented at the Reciprocal Meats Conference, American Meat Science Association, Indianapolis, IN, July 24-28. Besser RE, Lett SM, Weber JT, Doyle MP, Barrett TJ, Wells JG, Griffin PM. 1993. An outbreak of diarrhea and hemolytic uremic syndrome from Escherichia cold 0157:H7 in fresh-pressed apple cider. JAMA 269:2217-2220. Blake PA, Merson MH, Weaver RE, Hollis DG, Heublein PC. 1979. Disease caused by a marine vibrio. Clinical characteristics and epidemiology. N Engl J Med 300:1-5. Blaser MJ, Newman LS. 1982. A review of human salmonellosis. I. Infective dose. Rev Infect Dis 4: 1096-1 106. Blaser MJ, Cravens J. Powers BW, Laforce FM, Wang WL. 1979. Campylobacter enteritis associ- ated with unpasteurized milk. Am JMed 67:715-718. Bresee JB, Widdowson M-A, Monroe SS, Glass RI. 2002. Foodborne viral gastroenteritis: Chal- lenges and opportunities. Clin Infect Dis 35:748-753. Breuer T. Benkel DH, Shapiro RR, Hall WN, Winnett MM, Linn MJ, Neimann J. Barrett TJ, Dietrich S. Downes FP, Toney DM, Pearson JL, Rolkalt H. Slutsker L, Griffin PM. 2001. A multistate outbreak of Escherichia cold 0157:H7 infections linked to alfalfa sprouts grown from contami- nated seeds. Emerg Infect Dis 7:977-982. Brooks JT, Rowe SV, Shillam P. Heltzel DM, Hunter SB, Slutsker L, Hoekstra RM, Luby SP. 2001. Salmonella Typhimurium infections transmitted by chlorine-pretreated clover sprout seeds. Am JEpidemiol 154:1020-1028. Buchanan RL, Edelson SG, Miller RL, Sapers GM. 1999. Contamination of intact apples after immersion in an aqueous environment containing Escherichia cold 0157:H7. J Food Prot 62:444-450. Burnett SL, Chen J. Beuchat LR. 2000. Attachment of Escherichia cold 0157:H7 to the surfaces and internal structures of apples as detected by confocal scanning laser microscopy. Appl Environ Microbiol 66:4679-4687. Burr R. Effler P. Kahana J. Kanenaka R. Balmer M, Angulo F. Barrett T. Hayes P. Nekomoto T. Nakata M, Sasaki D, Higa H. Sawa T. Choy B. 1999. Emergence of Salmonella Serotype Enteritidis Phage Type 4 in Hawaii Caused by Locally Produced Contaminated Eggs. Abstract presented at the 48th Annual Epidemic Intelligence Service Conference, Centers for Disease Control and Prevention (CDC), Atlanta, GA, April 19-23. Buzby JC, Roberts T. 1996. ERS updates U.S. foodborne disease costs for seven pathogens. Food Review 19:20-25 CDC (Centers for Disease Control and Prevention). 1981. Multiple outbreaks of salmonellosis asso- ciated with precooked roast beef Pennsylvania, New York, Vermont. Morb Mortal Wkly Rep 30:569-570. CDC. 1993. Outbreaks of Salmonella Enteritidis gastroenteritis California, 1993. Morb Mortal Wkly Rep 42:793-797.

62 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD CDC. 1997. Outbreaks of Escherichia cold 0157:H7 and cryptosporidiosis associated with drinking unpasteurized apple cider Connecticut and New York, October 1996. Morb Mortal Wkly Rep 46:4-8. CDC. 1998. Update: Multistate outbreak of listeriosis United States, 1998. Morb Mortal Wkly Rep 47: 1085-1086. CDC. 2001. Updated guidelines for evaluating public health surveillance systems: Recommenda- tions from the Guidelines Working Group. Morb Mortal Wkly Rep 50:2-35. CDC. 2002a. Preliminary FoodNet data on the incidence of foodborne illnesses Selected sites, United States, 2001. Morb Mortal Wkly Rep 51:325-329. CDC. 2002b. 2000 Annual Report. HARMS. National Antimicrobial Resistance Monitoring System: Enteric Bacteria. Online. Available at http://www.cdc.gov/narms/annual/2000/ NARMS_final_report_2000.pdf. Accessed April 15, 2003. CDC. 2003. Outbreaks of Salmonella serotype Enteritidis infection associated with eating shell eggs United States, 1999-2000. Morb Mortal Wkly Rep 51:1149-1152. CFSAN (Center for Food Safety and Applied Nutrition). 2001. Survey of Domestic Fresh Produce: Interim Results (July 31, 2001). Online. Food and Drug Administration (FDA). Available at http://www.cfsan.fda.gov/~dms/prodsur9.html. Accessed September 3, 2002. CFSAN. 2002. FDA Pesticide Program: Residue Monitoring, 1993-2001. Online. FDA. Available at http://www.cfsan.fda.gov/~dms/pesrpts.html. Accessed January 22, 2003. Cody SH, Glynn KM, Farrar JA, Cairns KL, Griffin PM, Kobayashi J. Fyfe M, Hoffman R. King AS, Lewis JH, Swaminathan B. Bryant RG, Vugia DJ. 1999. An outbreak of Escherichia cold 0157:H7 infection from unpasteurized commercial apple juice. Ann Intern Med 130:202-209. Consumer Studies Branch. 2002. Food Safety Survey: Summary of Major Trends in Food Handling Practices and Consumption of Potentially Risky Foods. Online. CFSAN, FDA. Available at http://www.cfsan.fda.gov/~dms/fssurvey.html. Accessed September 23, 2002. Crump JA, Griffin PM, Angulo FJ. 2002. Bacterial contamination of animal feed and its relationship to human foodborne illness. Clin Infect Dis 35:859-65. Dalton C, Mintz E, Wells J. Bopp C, Tauxe R. 1999. Outbreaks of enterotoxigenic Escherichia cold infection in American adults: A clinical and epidemiologic profile. Epidemiol Infect 123:9-16. Daniels NA, MacKinnon L, Bishop R. Altekruse S. Ray B. Hammond RM, Thompson S. Wilson S. Bean NH, Griffin PM, Slutsker L. 2000. Vibrio parahaemolyticus infections in the United States, 1973-1998. J Infect Dis 181: 1661-1666. Deming MS, Tauxe RV, Blake PA, Patton C, Dixon SE, Fowler BS, Jones TS, Lockamy EA, Patton CM, Sikes RO. 1987. Campylobacter enteritis at a university: Transmission from eating chicken and from cats. Am J Epidemiol 126:526-534. Doyle MP, Beuchat LR, Montville TJ. 2001. Food Microbiology: Fundamentals and Frontiers, 2nd ed. Washington, DC: American Society for Microbiology Press. FDA (Food and Drug Administration). 1999. Guidance for industry: Reducing microbial food safety hazards for sprouted seeds and Guidance for industry: Sampling and microbial testing of spent irrigation water during sprout production. Fed Reg 64:57893-57902. FDA. 2000. Food labeling, safe handling statements, labeling of shell eggs; Refrigeration of shell eggs held for retail distribution; Final rule. Fed Reg 65:76091-76114. FDA. 2001. FDA Publishes Final Rule to Increase Safety of Fruit and Vegetable Juices. Online. Available at http://www.fda.gov/bbs/topics/NEWS/2001/NEW00749.html. Accessed Septem- ber 9, 2002. FDDB (Foodborne and Diarrheal Diseases Branch). 2000. FoodNet Surveillance Report for 1999 (Final Report). Online. Division of Bacterial and Mycotic Diseases (DBMD), CDC. Available at http://www.cdc.gov/foodnet/annual/1999/pdf/FoodNet_l999_Annual_Report.pdf. Accessed September 11, 2002.

THE SCIENCE OF PUBLIC HEALTH SURVEILLANCE 63 FDDB. 2002a. Foodborne-Disease Outbreaks Reported to CDC January 1, 1990 through March 15, 2002. Online. DBMD, CDC. Available at http://www.cdc.gov/foodborneoutbreaks/ outbreak_satistics/cdc_reported.htm. Accessed October 20, 2002. FDDB. 2002b. FoodNet Data Sources. Population Survey. Online. DBMD, CDC. Available at http:/ /www.cdc.gov/foodnet/surveys/pop_cov.htm. Accessed September 11, 2002. FDDB. 2002c. Salmonella: Annual Summary, 2001. Online. DBMD, CDC. Available at http:// www.cdc.gov/ncidod/dbmd/phlisdata/salmtab/2001/SalmonellaAnnualSummary2001.pdf. Ac- cessed September 11, 2002. Fedorka-Cray PWF, Gray T. Hermosilla J. Ball T. Anandaraman M, Salamone B. Rose B. Dargatz D, Headrick M. 2002. Prevalence and resistance of Salmonella enterica serotype Newport 1997-2000: A NARMS Report. In: Proceedings of the National Conference on Antimicrobial Resistance. Bethesda, MD: National Foundation for Infectious Disease. Fontaine RE, Arnon S. Martin WT, Vernon TM Jr, Gangarosa EJ, Farmer JJ 3rd, Moran JB, Silliker JH, Decker DL. 1978. Raw hamburger: An interstate common source of human salmonellosis. Am JEpidemiol 107:36-45. Friedman CR, Neimann J. Wegener HC, Tauxe RV. 2000a. Epidemiology of Campylobacter jejuni infections in the United States and other industrialized nations. In: Nachamkin IJ, Blaser MJ, eds. Campylobacter, 2nd ed. Washington, DC: American Society for Microbiology. Pp. 121- 138. Friedman C, Reddy S. Samual M, Marcaus R. Bender J. Desai S. Shiferaw B. Helfrick D, Carter M, Anderson B. Hoekstra M, The EIP FoodNet Working Group. 2000b. Risk Factors for Sporadic Campylobacter Infections in the United States: A Case-Control Study on FoodNet Sites. Pre- sented at the International Conference on Emerging Infectious Diseases, CDC, Atlanta, GA, July 16-19. FSIS (Food Safety and Inspection Service). 1998a. An FSIS/ARS Study: Premature Browning of Cooked Ground Beef. Online. USDA. Available at http://www.fsis.usda.gov/OPHS/ prebrown.htm. Accessed September 11, 2002. FSIS. 1998b. Refrigeration and labeling requirements for shell eggs. Final rule. Fed Reg 63:45663- 45675. FSIS. 1998c. Report to Congress. FoodNet: An Active Surveillance System for Bacterial Foodborne Diseases in the United States. Online. USDA. Available at http://www.fsis.usda.gov/OPHS/ rpcong98/rpcong98.htm. Accessed April 10, 2003. FSIS. 2000. Color of Cooked Ground Beef as it Relates to Doneness. Online. USDA. Available at http://www.fsis.usda.gov/OA/pubs/colortech.htm. Accessed September 10, 2002. FSIS. 2003. Microbiological Results of Raw Ground Beef Products Analyzed for Escherichia cold 0157:H7, Calendar Year 2003. Online. USDA. Available at http://www.fsis.usda.gov/OPHS/ ecoltest/ecpositives.htm. Accessed March 7, 2003. FSIS/CFSAN. 2001. Draft Assessment of the Relative Risk to Public Health from Foodborne Listeria monocytogenes Among Selected Categories of Ready-to-Eat Foods. Online. USDA, FDA. Available at http://www.foodsafety.gov/~dms/lmrisk.html. Accessed September 9, 2002. Garman R. Jones T. Bulens S. Tong S. Myers R. Gettner S. Mead P. Parashar U. EIP FoodNet Working Group. 2002. A Pilot Study in FoodNet of the use of Stool Collection Kits Delivered to the Home to Improve Confirmation of Etiology in Gastroenteritis Outbreak Investigations. Presented at the International Conference on Emerging Infectious Diseases, CDC, Atlanta, GA, March 24-27. Gast R. 1999. Applying experimental infection models to understand the pathogenesis, detection and control of Salmonella enterica serovar Enteritidis in poultry. In: Saeed AM, Gast RK, Potter ME, Wall PG, eds. Salmonella enterica Serovar Enteritidis in Humans and Animals: The Epi- demiology, Pathogenesis and Control. Ames, IA: Iowa State University Press. Pp. 233-243. Gitelson S. 1971. Gastrectomy, achlorhydria, and cholera. Isr JMed Sci 7:663-667.

64 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD Griffin P. Mead P. Sivapalasingam S. 2002. Escherichia cold 0157:H7 and other enterohemorrhagic E. colt. In: Blaser MJ, Smith PD, Ravdin JI, Greenberg H. Guerrant RL, eds. Infections of the Gastrointestinal Tract, 2nd ed. Philadelphia: Lippincott Williams and Wilkins. Pp. 627-642. Guerrant RL, Steiner TS. 1999. Principles and syndromes of enteric infection. In: Mardell G. Bennet J. Dolin R. eds. Principles and Practice of Infection Diseases, 4th ed. Kent, UK: Churchill, Livingstone. Pp. 1076-1093. Guerrant RL, Theno DM. 1995. Escherichia cold 0157:H7. NEngl JMed 333:1711-1712. Guerrant RL, Hughes JM, Lima NL, Crane J. 1990. Diarrhea in developed and developing countries: Magnitude, special settings, and etiologies. Rev Infect Dis 12: S41-S50. Guerrant RL, Kosek M, Lima AA, Lorntz B. Guyatt HL. 2002. Updating the DALYs for diarrhoeal disease. Trends Parasitol 18: 191-193. Gupta A, Crowe C, Bolstorff B. Fontana J. Stout A, Montgomery S. McGuill M, Matayas B. Schoenfeld S. Angulo F. 2001. Multistate investigation of multi-drug resistant Salmonella sero- type Newport infections in the Northeastern U.S., 2000: Human infections associated with dairy farms. In: Proceedings of the Annual Meeting of the Infectious Disease Society of America. Alexandria, VA: Infectious Disease Society of America. P. 36. Hald T. 2001. Salmonella in Pork—Epidemiology, Control and the Public Health Impact. Manu- script 3: Quantifying the Contribution of Animal-Food Sources to Human Salmonellosis in Denmark in 1999. PhD Thesis. Copenhagen, Denmark: Royal Veterinary and Agricultural University. Hald T. Bronsted T. 2002. Annual Report on Zoonoses in Denmark, 2001. Copenhagen: Ministry of Food, Agriculture and Fisheries. Hall JA, Goulding JS, Bean NH, Tauxe RV, Hedberg CW. 2001. Epidemiologic profiling: Evaluat- ing foodborne outbreaks for which no pathogen was isolated by routine laboratory testing: United States, 1982-9. Epidemiol Infect 127:381-387. Hedberg CW, David MJ, White KE, MacDonald KL, Osterholm MT. 1993. Role of egg consump- tion in sporadic Salmonella Enteritidis and Salmonella Typhimurium infections in Minnesota. J Infect Dis 167: 107-111. Helms M, Vastrup P. Gerner-Smidt P. Molbak K. 2002. Excess mortality associated with antimicro- bial drug-resistant Salmonella Typhimurium. Emerg Infect Dis 8:490-495. Hennessy TW, Hedberg CW, Slutsker L, White KE, Besser-Wiek JM, Moen ME, Feldman J. Coleman WW, Edmonson LM, MacDonald KL, Osterholm MT. 1996. A national outbreak of Salmonella Enteritidis infections from ice cream. N Engl J Med 334: 1281-1286. Herikstad H. Motarjemi J. Tauxe RV. 2002a. Salmonella surveillance: A global survey of public health serotyping. Epidemiol Infect 129: 1-8. Herikstad H. Yang S. Van Gilder TJ, Vugia D, Hadler J. Blake P. Deneen V, Shiferaw B. Angulo FJ. 2002b. A population-based estimate of the burden of diarrhoeal illness in the United States: FoodNet, 1996-1997. Epidemiol Infect 129:9-17. HHS (U.S. Department of Health and Human Services). 2000. Healthy People 2010. Vol.1. 2nd Ed. Washington, DC: U.S. Government Printing Office. Holmberg SD, Feldman RA. 1984. New and newer enteric pathogens: Stages in our knowledge. Am J Public Health 74:205-207. Hopkins RS, Scott AB. 1983. Handling raw chickens as a source for sporadic Campylobacter jejuni infections. J Infect Dis 148:770. Hornick RB, Music SI, Wenzel R. Cash R. Libonati JP, Snyder MJ, Woodward TE. 1971. The Broad Street pump revisited: Response of volunteers to ingested cholera vibrios. Bull NY Acad Med 47:1181-1191. Humphrey TJ, Williams A, McAlpine K, Lever MS, Guard-Petter J. Cox JM. 1996. Isolates of Salmonella enterica Enteritidis PT4 with enhanced heat and acid tolerance are more virulent in mice and more invasive in chickens. Epidemiol Infect 117:78-88.

THE SCIENCE OF PUBLIC HEALTH SURVEILLANCE 65 Hunt MC, Soerheim O. Slinde E. 1999. Color and heat denaturation of myoglobin forms in ground beef. J Food Sci 64:847-851. Hurd HS, McKean ID, Wesley IV, Karriker LA. 2001. The effect of lairage on Salmonella isolation from market swine. J. Food Prot 64:939-944. Hutwagner LC, Maloney EK, Bean NH, Slutsker L, Martin SM. 1997. Using laboratory-based sur- veillance data for prevention: An algorithm for detecting Salmonella outbreaks. Emerg Infect Dis 3:395-400. IOM/NRC (Institute of Medicine/National Research Council). 1998. Ensuring Safe Foodfrom Pro- duction to Consumption. Washington, DC: National Academy Press. Itoh Y. Sugita-Konishi Y. Kusuga F. Iwaki M, Hara-Kudo Y. Saito N. Noguchi Y. Konuma H. Kumagai S. 1998. Enterohemorrhagic Escherichia cold 0157:H7 present in radish sprouts. Appl Environ Microbiol 4:1532-1535. Jaquette C, Beuchat L, Mahon BE. 1996. Efficacy of chlorine and heat treatment in killing Salmo- nella Stanley inoculated onto alfalfa seeds and growth and survival of the pathogen during sprouting and storage. Appl Environ Microbiol 62:2212-2215. Jones T. Imhoff B. Deneen V, Samuel M, Mshar R. Gibbs-McCombs K, Kamer T. Cambridge M, Slutsker L, EIP FoodNet Working Group. 2000. Limitations of Field Investigations: Etiology and Food Vehicle not Determined for a Substantial Proportion of Foodborne Outbreaks in FoodNet, 1998-1999. Presented at the 2nd International Conference on Emerging Infectious Diseases, CDC, Atlanta, GA, July 16-19. Kassenborg H. Hedberg C, Evans M, Chin G. Fiorentino T. Vugia D, Bardsley M, Slutsker L, Griffin P. 1998. Case-control study of sporadic Escherichia cold 0157:H7 infections in 5 FoodNet sites. In: Proceedings of the International Conference on Emerging Infectious Diseases. At- lanta, GA: CDC. P.50. Killinger KM, Hunt MC, Campbell RE, Kropf DH. 2000. Factors affecting premature browning during cooking of store-purchased ground beef. J Food Sci 65:585-587. Kohn MA, Farley TA, Ando T. Curtis M, Wilson SA, Jin Q. Monroe SS, Baron RC, McFarland LM, Glass RI. 1995. An outbreak of Norwalk virus gastroenteritis associated with eating raw oys- ters: Implications for maintaining safe oyster beds. JAMA 273:466-471 Lee LA, Gerber AR, Longsway DR, Smith JD, Carter GP, Puhr ND, Parrish CM, Sikes RK, Finton RJ, Tauxe RV. 1990. Yersinia enterocolitica 0:3 infections in infants and children, associated with the household preparation of chitterlings. N Engl J Med 322:984-987. Lumsden L, Hasseltine H. JP L, Veldee M. 1925. A typhoid fever epidemic caused by oyster-borne infections (1924-1925). Public Health Rep 50:1-102. Mahon BE, Ponka A, Hall WN, Komatsu K, Dietrich SE, Siitonen A, Cage G. Hayes PS, Lambert- Fair MA, Bean NH, Griffin PM, Slutsker L. 1997. An international outbreak of Salmonella infections caused by alfalfa sprouts grown from contaminated seeds. J Infect Dis 175:876-882. Marano NN, Rossiter S. Stamey K, Joyce K, Barrett TJ, Tollefson LK, Angulo FJ. 2000. The National Antimicrobial Resistance Monitoring System (NARMS) for enteric bacteria, 1996- 1999: Surveillance for action. JAm Vet Med Assoc 217:1829-1830. Martin ML, Shipman LD, Wells JG, Potter ME, Hedberg K, Wachsmuth IK, Tauxe RV, Davis JP, Arnoldi J. Tilleli J. 1986. Isolation of Escherichia cold 0157:H7 from dairy cattle associated with two cases of haemolytic uraemic syndrome. Lancet 2:1043. McCarthy T. Phan Q. Mshar P. Mshar R. Howard R. Hadler J. 2002. Outbreak of Multi-Drug- Resistant Salmonella Newport Associated with Consumption of Italian-Style Soft Cheese, Con- necticut. Presented at the International Conference on Emerging Infectious Diseases, CDC, Atlanta, GA, May 24-27. Mead PS, Mintz ED. 1996. Ethnic eating: Foodborne disease in the global village. Infect Dis Clin Pract 5:3 19-323.

66 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD Mead PS, Finelli L, Lambert-Fair MA, Champ D, Townes J. Hutwagner L, Barrett T. Spitalny K, Mintz E. 1997. Risk factors for sporadic infection with Escherichia cold 0157:H7. Arch Intern Med 157:204-208. Mead PS, Slutsker L, Dietz V, McCaig LF, Bresee IS, Shapiro C, Griffin PM, Tauxe RV. 1999. Food-related illness and death in the United States. Emerg Infect Dis 5:607-625. Mead P. Slutsker L, Gellin B. 2002. Listeriosis. In: Blaser M, Smith P. Ravdin J. Greenberg H. Guerrant R. eds. Infections of the Gastrointestinal Tract, 2nd ed. Philadelphia: Lippincott Williams & Wilkins. Pp. 802-811. Mentzing LO, Ringertz O. 1968. Salmonella infection in tourists. 2. Prophylaxis against salmonellosis. Acta Pathol Microbiol Scand 74:405-413. Michino H. Araki K, Minami S. Takaya S. Sakai N. Miyazaki M, Ono A, Yanagawa H. 1999. Massive outbreak of Escherichia cold 0157:H7 infection in schoolchildren in Sakai City, Ja- pan, associated with consumption of white radish sprouts. Am J. Epidemiol 150:787-796. Millard PS, Gensheimer KF, Addiss DG, Sosin DM, Beckett GA, Houck-Jankoski A, Hudson A. 1994. An outbreak of cryptosporidiosis from fresh-pressed apple cider. JAMA 272: 1592-1596. Mishu B. Griffin PM, Tauxe RV, Cameron DN, Hutcheson RH, Schaffner W. 1991. Salmonella Enteritidis gastroenteritis transmitted by intact chicken eggs. Ann Intern Med 115: 190-194. Murray C, Lopez A. 1997. The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases, Injuries, and Risk Factors in 1900 and Projected to 2020. Cam- bridge, MA: Harvard University Press. Nachamkin I, Allos B. Ho T. 2000. Campylobacter jejuni infection and the association with Guillain- Barre syndrome. In: Nachamkin I, Blaser M, eds. Campylobacter, 2nd ed. Washington, DC: ASM Press. Pp. 155-175. NRC (National Research Council). 1985. Meat and Poultry Inspection: The Scientific Basis of the Nation's Program. Washington, DC: National Academy Press. Navin T. Juranek D. 1984. Cryptosporidiosis: Clinical, epidemiologic, and parasitologic review. Rev Infect Dis 6:313-327. Norwegian Zoonosis Centre. 2002. Action plan—Campylobacter spp. in Norwegian broilers. Summary of results from the surveillance of broiler flocks 2001; Survey of broiler meat prod- ucts 2001. Available at http ://www.vetinst.no/Zoonosesenteret/Camp_handlingsplan2002.pdf. Accessed September 11, 2002. Olsen SJ, MacKinnon LC, Goulding JJ, Bean NH, Slutsker L. 2000. Surveillance for foodborne disease outbreaks United States, 1993-1997. Morb Mortal Wkly Rep 49:1-62. Olsen SJ, Bishop R. Brenner FW, Roels TH, Bean N. Tauxe RV, Slutsker L. 2001. The changing epidemiology of Salmonella: Trends in serotypes isolated from humans in the United States, 1987-1997. J Infect Dis 183:753-761. OPDFB (Office of Plant and Dairy Foods and Beverages). 2001. FDA Survey of Imported Fresh Produce: FY 1999 Field Assignment. Online. CFSAN, FDA. Available at http:// www.cfsan.fda.gov/~dms/prodsur6.html. Accessed September 3, 2002. Parashar UD, Monroe SS. 2001. "Norwalk-like viruses" as a cause of foodborne disease outbreaks. Rev Med Virol 11 :243-252. Parham G. 1984. Salmonellae in cooked meat products. In: Snoeyenbos G. ed. Proceedings of an International Symposium on Salmonella. Kennett Square, PA: University of Pennsylvania. Pp. 275-280. Passaro DJ, Reporter R. Mascola L, Kilmon L, Malcolm GB, Rolka H. Werner SB, Vugia DJ. 1996. Epidemic Salmonella Enteritidis infections in Los Angeles County, California. The predomi- nance of phage type 4. West JMed 165:126-130. Potter M, Archer D, Benson A, Busta F. Dickson J. Doyle M, Farber J. Finlay BB, Goldblatt M, Hedberg C, Hoober D, Jahncke M, Jaykus L-A, Kaspar C, Liang AP, Lindsay J. Pestka J. Pierson M, Slade P. Tompkin RB, Tortorello ML. 2000. Emerging Microbiological Food Safety Issues: Implications for Control in the 21st Century. Chicago: Institute of Food Technology.

THE SCIENCE OF PUBLIC HEALTH SURVEILLANCE 67 Proctor M, Hamacher M, Tortorello M, Archer J. Barnett M, Ying M, Slutsker L, Davis J. 2000. Multistate outbreak of Salmonella Muenchen infections associated with alfalfa sprouts grown from seeds pretreated with calcium hypochlorite. In: Proceedings of the 2nd International Conference on Emerging Infectious Diseases. Atlanta, GA: CDC. P. 98. Redmond E, Griffith C, Slader J. Humphrey T. 2002. Assessment of risks associated with consumer food handling practices using real-time microbiological analysis. In: Proceedings of the Annual Conference of the International Association for Food Protection. Des Moines, IA: International Association for Food Protection. P. 67. Riley LW, Remis RS, Helgerson SD, McGee HE, Wells JG, Davis BR, Hebert RJ, Olcott ES, Johnson LM, Hargrett NT, Blake PA, Cohen ML. 1983. Hemorrhagic colitis associated with a rare Escherichia cold serotype. NEngl JMed 308:681-685. Rose BE, Hill WE, Umholtz R. Ransom GM, James WO. 2002. Testing for Salmonella in raw meat and poultry products collected at federally inspected establishments in the United States, 1998 through 2000. J Food Prot 65:937-947. Rosenberg C. 1987. The Cholera Years: The United States in 1832, 1849, and 1866. Chicago: University of Chicago Press. RTI (Research Triangle Institute). 2002a. PR/HACCP Rule Evaluation Report. Changes in Con- sumer Knowledge, Behavior and Confidence since the 1996 PR/HACCP Final Rule. Final Report. Online. FSIS, USDA. Available at http://www.fsis.usda.gov/oa/research/ HACCPimpacts.htm. Accessed September 23, 2002. RTI. 2002b. Revised Final Draft: PR/HACCP Rule Evaluation Report. Effect of the 1996 PR/HACCP Rule on Foodborne Illness: Young Chicken Salmonella Study. Research Triangle Park, NC: RTI. Ryan CA, Nickels MK, Hargrett-Bean NT, Potter ME, Endo T. Mayer L, Langkop CW, Gibson C, McDonald RC, Kenney RT, Purh ND, McDonnell PJ, Martin RJ, Cohen ML, Blake PA. 1987. Massive outbreak of antimicrobial-resistant salmonellosis traced to pasteurized milk. JAMA 258:3269-3274. Samuel MC, Portnoy D, Tauxe RV, Angulo FJ, Vugia DJ. 2001. Complaints of foodborne illness in San Francisco, California, 1998. J Food Prot 64:1261-1264. Schlosser W. Henzler D, Mason J. Kradel D, Shipman L, Trock S. Hurd S. Hogue A, Sischo W. Ebel E. 1999. The Salmonella enterica serovar Enteritidis pilot project. In: Saeed AM, Gast RK, Potter ME, Wall PG, eds. Salmonella enterica Serovar Enteritidis in Humans and Animals: The Epidemiology, Pathogenesis and Control. Ames, IA: Iowa State University Press. Pp. 353-365. Shane SM. 2000. Campylobacter infection of commercial poultry. Rev Sci Tech Off Int Epizoot 19:376-295. Sivapalasingam S. Kimura A, Ying M, Frisch A, Barrett E, Phan Q. Shillam P. Reddy S. Breslowsky T. Gould E, Van Duyne MS, Slutsker L. 2000. A Multistate Outbreak of Salmonella Newport Infections Linked to Mango Consumption, November-December 1999. Abstract presented at the 38th Annual Meeting of the Infectious Diseases Society of America, New Orleans, LA, September 7-10. Slutsker L, Ries AA, Maloney K, Wells JG, Greene KD, Griffin PM. 1998. A nationwide case- control study of Escherichia cold 0157:H7 infection in the United States. JInfectDis 177:962- 966. Sperber SJ, Schleupner CJ. 1987. Salmonellosis during infection with human immunodeficiency virus. Rev Infect Dis 9:925-934. St. Louis ME, Morse DL, Potter ME, DeMelfi TM, Guzewich JJ, Tauxe RV, Blake PA. 1988. The emergence of grade A eggs as a major source of Salmonella Enteritidis infections: Implications for the control of salmonellosis. JAMA 259:2103-2107. Swaminathan B. Barrett TJ, Hunter SB, Tauxe RV. 2001. PulseNet: The molecular subtyping net- work for foodborne bacterial disease surveillance, United States. Emerg Infect Dis 7:382-389.

68 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD Taormina PI, Beuchat LR, Slutsker L. 1999. Infections associated with eating seed sprouts: An international concern. Emerg Infect Dis 5:626-634. Tauxe RV. 1997. Emerging foodborne diseases: An evolving public health challenge. Emerg Infect Dis 3:425-434. Tauxe RV. 2001. Food safety and irradiation: Protecting the public health from foodborne infections. Emerg Infect Dis 7:516-521. Tauxe RV, Vandepitte J. Wauters G. Martin SM, Goossens V, De Mol P. Van Noyen R. Thiers G. 1987. Yersinia enterocolitica infections and pork: The missing link. Lancet 1:1129-1132. Thacker SB. 1994. Historical development. In: Teutsch S. Churchill R. eds. Principles and Practice of Public Health Surveillance. New York: Oxford University Press. Pp. 3-17. Thacker SB, Berkelman RL. 1988. Public health surveillance in the United States. Epidemiol Rev 10: 164-190. Voetsch D, Angulo F. Shallow S. Marcus R. Farley M, Hedberg C, Cieslak P. Tauxe R. FoodNet Working Group. 1998. Estimate of the Burden of Illness Caused by Nontyphoidal Salmonella Infections in the United States from FoodNet. Presented at the American Public Health Asso- ciation 126th Annual Meeting Exposition, Washington DC, November 15-19. Vogt RL, Sours HE, Barrett T. Feldman RA, Dickinson RJ, Witherell L. 1982. Campylobacter enteritis associated with contaminated water. Ann Intern Med 96:292-296. Watanabe Y. Ozasa K, Mermin JH, Griffin PM, Masuda K, Imashuku S. Sawada T. 1999. Factory outbreak of Escherichia cold 0157:H7 infection in Japan. Emerg Infect Dis 5:424-428. Wells JG, Shipman LD, Greene KD, Sowers KG, Green JH, Cameron DN, Downes FP, Martin ML, Griffin PM, Ostroff SM, Potter ME, Tauxe RV, Wachsmuth IK. 1991. Isolation of Escherichia cold serotype 0157:H7 and other Shiga-like-toxin-producing E. cold from dairy cattle. J Clin Microbiol 29:985-989. WHO (World Health Organization). 2000. The Increasing Incidence of Human Campylobacteriosis. Report and Proceedings of a WHO Consultation of Experts. Online. Available at: http:// www.who.int/emc-documents/zoonoses/docs/whocdscsraph20017.pdf. Accessed April 11, 2003. WHO. 2002. Global Salm-Surv. Online. Available at http://www.who.int/emc/diseases/zoo/SALM- SURV/index.html. Accessed September 9, 2002. Wineland NE, Dargatz DA. 1998. The national animal health monitoring system: A source of on- farm information. Vet Clin North Am Food Anim Pract 14:127-139. Winthrop KL, Palumbo MS, Farrar JA, Mohle-Boetani JC, Abbott S. Beatty ME, Inami G. Werner SB. 2003. P101 Alfalfa sprouts and Salmonella Kottbus infection: A multistate outbreak fol- lowing inadequate seed disinfection with heat and chlorine. J Food Prot 66:13-17. Yang S. Leff MG, McTague D, Horvath KA, Jackson-Thompson J. Murayi T. Boeselager GK, Melnik TA, Gildemaster MC, Ridings DL, Altekruse SF, Angulo FJ. 1998. Multistate surveil- lance for food-handling, preparation, and consumption behaviors associated with foodborne diseases: 1995 and 1996 BRFSS food-safety questions. Morb Mortal Wkly Rep 47:33-57. Zansky S. Wallace B. Schoonmaker-Bopp D, Smith P. Ramsey F. Painter J. Gupta A, Kalluri P. Noviello S. 2002. Outbreak of multidrug-resistant Salmonella Newport United States, January-April 2002. Morb Mortal Wkly Rep 51:545-548. Zhao T. Doyle MP, Besser RE. 1993. Fate of enterohemorrhagic Escherichia cold 0157:H7 in apple cider with and without preservatives. Appl Environ Microbiol 59:2526-2530.

Next: 3 Food Safety Tools »
Scientific Criteria to Ensure Safe Food Get This Book
×
Buy Paperback | $68.00 Buy Ebook | $54.99
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

Food safety regulators face a daunting task: crafting food safety performance standards and systems that continue in the tradition of using the best available science to protect the health of the American public, while working within an increasingly antiquated and fragmented regulatory framework. Current food safety standards have been set over a period of years and under diverse circumstances, based on a host of scientific, legal, and practical constraints.

Scientific Criteria to Ensure Safe Food lays the groundwork for creating new regulations that are consistent, reliable, and ensure the best protection for the health of American consumers. This book addresses the biggest concerns in food safety—including microbial disease surveillance plans, tools for establishing food safety criteria, and issues specific to meat, dairy, poultry, seafood, and produce. It provides a candid analysis of the problems with the current system, and outlines the major components of the task at hand: creating workable, streamlined food safety standards and practices.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!