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.
7 Scientific Criteria and Performance Standards to Control Hazards in Dairy Products High morbidity and mortality rates associated with diseases such as typhoid fever and infantile diarrhea, which may be contracted through consumption of microbiologically contaminated foods, led to initiation of food- and water-borne disease reporting in the United States more than 75 years ago (Olsen et al., 2000~. Anecdotal observations that linked consumption of milk with the spread of dis- ease spurred various scientists and physicians in the United States and around the world to undertake public health research to investigate the role of milk con- sumption in foodborne disease as early as the turn of the twentieth century. As a result of these investigations, consumption of unpasteurized milk was found to be associated with many serious diseases, including diphtheria, typhoid, tuberculo- sis, and brucellosis (Johnson et al., 1990~. The first reports of gastrointestinal disease outbreaks attributed to milk con- sumption were published by the Public Health Service (PHS) in 1925. These early reports provided evidence suggesting that to control milk-borne diseases, sanitation measures would need to be applied at all points in the food system, from the farm to the consumer (CFSAN,2002~. Further, these observations high- lighted the need for technical research that would determine the bacterial destruc- tion characteristics of food-processing treatments for pathogenic microbes likely to be present in raw milk (Enright et al., 1957; Gilman et al., 1946~. The results of these studies led to the development of specific recommendations for pasteuriza- tion and other intervention strategies (described below) that were designed to protect the public from exposure to hazardous microorganisms that may be present in raw milk. In the case of cheese, however, investigations were initiated not because of the association between illness and cheeses made from unpasteurized 225
226 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD TABLE 7.1 Equivalent Temperature and Time Combinations for Milk Pasteurization According to U.S. Regulations Temperature Time Temperature Time 63°C (145°F)a 30 men 94°C (201 °F) 0.1 see 72°C (161°F)a 15 sec 96°C (204°F) 0.05 sec 89°C (191°F) 1.0 see 100°C (212°F) 0.01 see 90°C (194°F) 0.5 see a If the fat content of the milk is 10% or more, or if it contains added sweeteners, the required minimum temperature must be increased by at least 3°C (5°F). SOURCE: CFSAN (2002). milk, but to assess the survival of Brucella abortus in the product (Gilman et al., 1946~. In the past few decades, foodborne disease outbreaks have been linked to various cheeses and, therefore, the need to evaluate the survival of human patho- gens during cheese manufacturing and aging has been revisited. In light of data indicating that certain pathogens (Listeria monocytogenes and Escherichia cold 0157:H7) may survive the 60-day aging period in cheese, research is currently being conducted to determine if this process criterion is adequate to protect public health. For the purpose of this report, "raw milk" is defined as milk, harvested from an animal, that may have been cooled to refrigeration temperatures or below, but that has not been subjected to processing with the objective of eliminating patho- genic bacteria that may be present. "Unpasteurized milk" is milk that may have been cooled or heated, but that has not been subjected to the minimal pasteuriza- tion processing conditions described in Table 7.1. While these terms are typically used interchangeably, unpasteurized milk is a broader term than raw milk as, for example, milk that can be processed into some types of cheeses may be subjected to heat treatments below minimum pasteurization conditions. Milk treated in this manner would be considered unpasteurized but not raw. (For a full discussion of the use of unpasteurized milk in dairy-product manufacturing, see later section, "Cheese and Other Dairy Food Products.") MILK Current Criteria and Standards PHS implemented the Standard Milk Ordinance in 1924 to assist states and cities in the voluntary adoption of programs designed to control milk-borne dis- ease. In 1950, the U.S. Surgeon General invited state milk-sanitation regulatory agencies to establish procedures for a voluntary Interstate Milk Shipper Certifica-
CONTROLS FOR HAZARDS IN DAIRY PRODUCTS 227 tion Program, which resulted in the formation of the National Conference on Interstate Milk Shipments (NCIMS) and the Cooperative State-Public Health Service Program for the certification of interstate milk shippers (CFSAN, 2000~. Responsibilities under this program were divided between state agencies and PHS. In 1969, the PHS responsibilities were transferred to the Food and Drug Administration (FDA). Currently, all states, the District of Columbia, and the United States trust territories participate in NCIMS. PHS, and later FDA, recommended the application of the current Grade A Pasteurized Milk Ordinance, commonly referred to as the PMO (CFSAN, 2002), to provide national uniformity for milk sanitation standards. Milk products covered by the PMO include products such as creams, concentrated milks, yogurts, and low-fat and skim milks (CFSAN,2002~. FDA's Division of Dairy and Egg Safety, Office of Plant and Dairy Foods and Beverages, is responsible for the develop- ment of additional regulations to protect the safety of cheese and other dairy foods (infant formula, dried milk products, ice cream or other frozen desserts, butter, and cheese) that enter interstate commerce, but that are not specifically covered by the PMO. The PMO covers production, transportation, processing, handling, sampling, examination, labeling, and sale of milk and milk products; the inspection of dairy farms and milk plants; the issuing and revocation of permits to milk producers, haulers, and distributors; and the fixing of penalties (CFSAN,2002~. The PMO is considered the reference for federal specifications for the procurement of milk and dairy products and as the sanitary regulation for dairy products served during interstate travel. It is also recognized by public health agencies and the dairy industry as the national standard for milk sanitation. As knowledge and experi- ence is gained, modifications to the PMO are recommended during biennial NCIMS meetings, which then must be approved by FDA before incorporation into the PMO. Since 1924, the PMO has evolved with input from many sources, including federal, state, and local government health and agriculture departments, the dairy industry (from producers to associations), academic organizations, and individuals. Hence, the PMO is derived from broad-based consensus of current knowledge and experience with milk sanitation standards in the United States. The implementation and enforcement of the PMO is another key element to protect the public from milk-borne illness. In this regard, FDA has no legal jurisdiction to enforce milk sanitation standards, except for interstate carriers and for products in interstate commerce. In general, although state and local agencies bear the majority of enforce- ment responsibilities for dairy regulatory programs, they still commonly use the PMO as the basis for developing their programs. Since 1924, government, academia, and industry have worked together to address targeted research needs as new pathogens have been identified and to modify regulations when science- based research has revealed appropriate measures for destruction and control of microbiological hazards.
228 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD The development, implementation, and enforcement of the PMO provide a good model for an integrated "cow-to-cup" strategy for product safety assurance. In addition, this model also provides a specific structure and mechanism for biennial review of existing regulations directed toward the fluid milk industry. Although FDA has the authority to enforce the implementation of the PMO standards in milk for interstate commerce, milk for local consumption is not subject to FDA oversight. Therefore, consumption of unpasteurized or raw milk continues to be an issue of concern, since it has been clearly established as a high- risk behavior for contracting foodborne illness. The committee concludes that targeted educational programs that illustrate the hazards of raw milk and raw milk-product consumption for milk producers and for the general public are warranted. The Public Health Objective of Fluid Milk Processing The public health objective for milk pasteurization, as defined in the PMO, is to eliminate all nonspore-forming pathogens commonly associated with milk; nevertheless, the guidance document cautions that pasteurization may not destroy preformed toxins (CFSAN, 2002~. According to the PMO, an analysis of milk-borne outbreak data over the years indicates that the risk of contracting disease is about 50 times less when consuming pasteurized versus unpasteurized milk. Pasteurization, as first adopted in the United States, was defined in the 1939 Milk Ordinance and Code as "the process of heating every particle of milk to at least 143°F (61.7°C) and holding at such temperature for at least 30 minutes, or to at least 160°F (71.1 °C) and holding at such temperature for at least 15 seconds, in approved and properly operated equipment" (PHS, 1940~. These heat treatments were referred to as the "holding method" or vat/batch pasteurization, and the "flash method" or high-temperature, short-time pasteurization, respectively. Table 7.1 contains these and other equiva- lent temperature/time combinations allowed by U.S. regulations. To address recognized scientific gaps regarding knowledge of the microbes associated with milk-borne disease, extensive research was conducted to deter- mine the heat treatment required to kill Mycobacterium tuberculosis which, at the time, was considered to be the most heat-resistant pathogen associated with milk (Hammer, 1948~. This work led to the widespread recognition of the public health significance of thermal milk processing and formed the basis for modern pasteurization processes (Hammer, 1948~. In 1956, minimal pasteurization tem- peratures were slightly increased to those listed in Table 7.1 to assure destruction of Coxiella burnetti, the organism associated with Q fever, which was found to be more heat resistant than M. tuberculosis (Enright et al., 1957~. As described above, the PMO prescribes highly specific pasteurization con- ditions (i.e., time and temperature combinations), equivalent to process standards, to ensure the safety of dairy products.
CONTROLS FOR HAZARDS IN DAIRY PRODUCTS The Scientific Basis for Current Pasteurization Requirements 229 The observation of a significant number of cases of Q fever attributed to the consumption of raw milk in the United States in the 1940s and 1950s inspired targeted research to precisely define conditions required for thermal destruction of C. burnetti (Enright et al., 1957~. Q fever, which was first described in the mid- 1930s, is a rickettsial disease characterized by chills, fever, weakness, and head- ache, with endocarditis as a possible complication in immunocompromised patients. C. burnetti is an obligate intracellular parasite that cannot multiply outside of living host cells; therefore, it cannot be cultured in laboratory media. While new detection strategies (e.g., polymerase chain reaction-based methods) are under development, current diagnostic strategies for Q fever are still based on the measurement of antibody titers for C. burnetti in blood samples taken from patients. High numbers of C. burnetti-specific antibodies in a patient's blood are considered to be indicative of exposure to this organism. Experiments to ascertain the thermal destruction of C. burnetti are techni- cally challenging because the presence of this organism in a heat-treated milk sample can only be measured indirectly by assessing the presence and concentra- tion of antibodies in a host animal that has been inoculated with a sample of the milk. Current milk processing strategies, which are designed to destroy C. burnetti in raw milk, are the outcome of a collaborative project between PHS and the University of California in the mid-1950s. The objectives of this study were to determine the maximum number of C. burnetti that might be found in the milk of an infected cow, to develop a sensitive method for determining the presence of small numbers of C. burnetti in pasteurized milk, and to ascertain the thermal resistance of C. burnetti in whole raw milk to ensure the absence of viable organisms in processed milk products (Enright et al., 1957~. The guinea pig was the host animal selected for monitoring residual levels of C. burnetti in heat- treated milk samples. Numbers of C. burnetti present in the milk were referred to as "infective guinea pig doses" because they were assessed through a determina- tion of the highest tenfold milk dilution that caused an intraperitoneally inocu- lated guinea pig to have a significant rise (at least fourfold) in antibody titer to C. burnetti (Enright et al., 1957~. The highest level of C. burnetti in milk of infected cows from samples collected around the state of California was deter- mined to be 10,000 infective guinea pig doses. Therefore, to provide an additional margin of safety, the authors selected thermal destruction of 100,000 infective guinea pig doses as the goal for minimal pasteurization conditions (Enright et al., 1957~. For the purpose of contrasting this thermal processing goal with other microbial destruction strategies described in this report, destruction of 100,000 infective guinea pig doses would be equivalent to a 5-D reduction in infective capacity for a given volume of milk. The pasteurization conditions described in Table 7.1 were found to result in destruction of 100,000 infective guinea pig doses of C. burnetti. Therefore, on July 16, 1956, the U.S. Assistant Surgeon
230 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD General released a recommendation for a minimum raw milk heat treatment of 145°F for 30 min or 161°F for 15 sec to ensure protection of the public from exposure to C. burnetti through consumption of milk. While the results reported by Enright and colleagues (1957) still serve as the scientific basis for current milk pasteurization practices, many processors apply time and temperature combina- tions that are above the minimum conditions (Douglas et al., 2000~. As discussed in Chapter 3 and mentioned in Chapters 4 and 5, the committee reiterates its belief that the implementation of performance standards that specify the reduction in numbers required for a targeted organism (e.g., a 5-D reduction for infective guinea pig doses for C. burnetti) in a food product (milk in this case), rather than specifying the precise conditions (i.e., process standards) for achieving that end, as currently practiced, could enable greater flexibility and innovation in the dairy industry, perhaps enabling the adoption of effective new processing technologies. Emerging Food Safety Concerns That May Justify a Reexamination of Current Milk Pasteurization Conditions As mentioned previously, currently applied thermal processing conditions for Grade A raw milk are designed to destroy the most heat-resistant of currently recognized nonspore-forming human pathogens, namely C. burnetti. However, some microbes that may be present in raw milk can survive pasteurization (Hammer et al., 1995~. Spore-forming bacteria, including those of the Bacillus and Clostridium genera, are among the heat resistant organisms that can be iso- lated from pasteurized milk. While the public health risk associated with the presence of these organisms in processed milk products is considered insignifi- cant under the current PMO, it is very important to recognize the fact that the pasteurization process is not intended to sterilize raw milk. In addition to incomplete destruction of spore-forming bacteria, the efficacy of milk pasteurization in killing M. avium subspp. paratuberculosis, a bacterium that causes Johne's disease in cattle but that has not been proven to cause human disease is uncertain (Klijn et al., 2001; Mechor, 1997~. Furthermore, although no evidence exists linking development of encephalopathy to con- sumption of milk from cows infected with bovine spongiform encephalopathy (commonly referred to as mad cow disease), current pasteurization conditions do not inactivate the causative prion. This prion, an infectious protein, shows little loss of infectivity even after prolonged exposure to temperatures up to 176°F (80°C) (Asher et al., 1986~. Although mice injected with milk from bovine spongiform encephalopathy-infected cattle did not develop the disease nor have epidemiological analyses suggested transmission of the disease to calves via milk (Hillerton, 1997), the possibility of such a transmission route should not be totally ruled out.
CONTROLS FOR HAZARDS IN DAIRY PRODUCTS 231 A more recently emerging food pathogen is Enterobacter sakazakii. On April 12, 2002, FDA alerted health care professionals about this pathogen, which has been associated with consumption of milk-based infant formulas. E. sakazakii can cause sepsis, meningitis, or necrotizing enterocolitis in newborn infants, particularly in premature or immunodeficient infants. Investigations of multiple outbreaks of E. sakazakii infection occurring in neonatal intensive care units worldwide over the past several years have associated illnesses with milk-based powdered infant formulas. To date, FDA is not aware of E. sakazakii infections among healthy, full-term infants in home settings, nor have illnesses been associ- ated with liquid infant formulas (FDA, 2002~. The emergence of new or newly recognized human pathogens that may be transmitted through milk or through consumption of other animal products highlights the need for food safety regula- tions that can be changed in a timely and responsive fashion as new hazards are identified and characterized. Other Fluid Milk Standards Although adequate refrigeration, aseptic processing, and a specified sub- pasteurization heat treatment to separate cream prior to bulk shipment of milk are processes included in the PMO, only pasteurization and ultrapasteurization (defined in Table 7.2) are recognized by the PMO as acceptable processes for removing or deactivating microorganisms in milk (CFSAN, 2002~. In addition to specific recommendations for pasteurization conditions, chemi- cal, bacteriological, and temperature standards have been established for grade A raw milk products intended for pasteurization, as well as for grade A pasteurized and bulk-shipped, heat-treated milk products (CFSAN, 2002~. For these products, milk must be cooled to 7°C or less within two hours after milking. Further, the TABLE 7.2 Minimum Temperature and Times for Fluid Milk Heat Treatments High-Temperature, Ultrahigh- Short-Time Temperature Country Pasteurization Ultrapasteurization Processing United Statesa 72°C/15 sec 138°C/2 sec Not definedd European Economic Communityb 71.7°C/15 sec Not defined 135°C/1 sec Australia/New ZealandC 72°C/15 sec 132°C/1 sec 132°C/1 sec a CFSAN (2002). b EEC (1994). c ANZFA (2000). d 21 C.F.R. part 113: "The thermal process and procedures for manufacturing UHT aseptically processed milk and milk products must comply with U.S. Food and Drug Administration require- ments for sterilizing low acid foods."
232 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD blend temperature after the first and subsequent milkings cannot exceed 10°C. Pasteurized products must not exceed 7°C throughout distribution. Raw milk and pasteurized products cannot test positive for drug residues as specified in sec- tion 6 of the PMO (CFSAN, 2002), a chemical performance standard related to good on-farm practices. Residual phosphatase activity may be measured in pasteurized products to reflect pasteurization efficacy. Pasteurized products must have less than 350 munits/L phosphatase activity for fluid products and less than 500 munitslL for other milk products. Table 7.2 provides current standards from the United States, the European Economic Community (now the European Union), and Australia and New Zealand for minimum heat treatments for milk products. Table 7.3 provides microbial and somatic cells limits for raw milk intended for pasteurized products, and Table 7.4 provides microbial standards for pasteurized fluid milk products. Somatic cell count limits for raw milk intended for pasteurized products are arguably a safety standard, as exceeding these limits may prevent effective appli- cation of a pasteurizing process. Similarly, the microbial standards for pasteurized fluid milk products (total bacteria and coliform bacteria) were not implemented on the basis of food safety per se; instead, the rationale behind these standards is that keeping total bacteria and coliform cell numbers within the specified limits reflects good management practices such as equipment cleanliness and sanitation or refrigeration control, which are essential elements of a food safety program. TABLE 7.3 Microbial and Somatic Cell Count (SCC) Standards for Raw Milk Intended for Pasteurized Milk Products Country Producer Raw Milka Plant Raw Milkb United StatesC 100,000 cfu/mLd 300,000 cfu/mL 750,000 SCCe Canadaf 50,000 cfu/mL 50,000 cfu/mL 500,000 SCC European Economic Communityg 100,000 cfu/mL 300,000 cfu/mL 400,000 SCC Australia/New Zealandh 150,000 cfu/mL 150,000 cfu/mL a Unpasteurized milk before it has left the holding tank on the farm. b Unpasteurized milk after it has left the farm holding tank. c CFSAN (2002). d cfu/mL were measured by aerobic plate count. e SCC must not exceed 1,000,000 in goat milk. fCFIS (1997). g EEC (1994). h ANZFA (2000).
CONTROLS FOR HAZARDS IN DAIRY PRODUCTS TABLE 7.4 Microbial Standards (per mL) for Pasteurized Milk Products 233 Country Total Bactenaa Coliform Bactenaa United Statesb 20,000C 10, except in heat-treated, bulk milk transport tank shipments, which may not exceed 100 Canadad m = 10,000 m = 1 M= 25,000 M= 10 n=5 n=5 c=2 c=2 European Economic Community After 5 d at 6°C (EEC)e m = 50,000 m = 0 M = 500,000 M = 5 n=5 n=5 c=1 c=1 Australia/New Zealandf m = 50,000 m = 1 M = 100,000 M = 10 n=5 n=5 c=1 c=1 a Total bacteria (as measured by aerobic plate count) and coliform bacteria counts given as the upper limit of cfu/mL for the United States. For Canada, EEC, arid Australia/New Zealand, two-tiered limits are given, with allowable results based on n number of samples, where n = number of sample units (subsamples) to be examined per lot, m = maximum number of bacteria per g or mL of product that is of no concern (acceptable level of contamination), M = maximum number of bacteria per g or mL of product, that if exceeded by any one sample unit (subsamples) renders the lot in violation of the regulations, c = maximum number of sample units (subsamples) per lot that may have a bacterial concentration higher than the value for m but less than value for M without violation of the regulations. b PHS (1999). c Not applicable in cultured dairy products. d CFIS (1997). e EEC (1994). f ANZFA (2000). CHEESE AND OTHER DAIRY FOOD PRODUCTS As with milk, FDA's Division of Dairy and Egg Safety, Office of Plant and Dairy Foods and Beverages is also responsible for the development and imple- mentation of regulations to protect the safety of cheese and other dairy foods that enter interstate commerce. According to 21 C.F.R. §1240.61, no milk or milk products in final package form intended for direct human consumption shall enter interstate commerce unless they are manufactured from pasteurized milk or pas- teurized milk ingredients, except where alternative procedures are provided for by regulation, such as in 21 C.F.R. part 133, which contains regulations for cheeses and related cheese products.
234 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD Standards of identity have been established for most natural cheeses, process cheeses, cheese foods, and cheese spreads (21 C.F.R. part 133~. All cheeses belonging to a given variety must comply with the published standard and must be labeled with the name prescribed in the standard. In general, identity standards specify a maximum permissible moisture content and minimum milk fat content. A few natural cheeses are required to be made from pasteurized milk (e.g., Monterey Jack, cream cheese, mozzarella cheese); however, most, including many soft ripened cheeses (21 C.F.R. § 133.182) and semi-soft cheeses (21 C.F.R. §133.187), may be made from either raw or pasteurized milk. The regulation states that "if cheese is labeled as 'heat treated,' 'unpasteurized,' 'raw milk,' or 'for manufacturing,' the milk may be raw or heated at temperatures below pas- teurization. Cheese made from unpasteurized milk shall be cured for a period of 60 days at a temperature not less than 35°F. If the milk is held more than 2 hours between time of receipt or heat treatment and setting, it shall be cooled to 45°F or lower until time of setting" (7 C.F.R. §58.439~. Standards of identity may stipu- late a holding period longer than 60 days if further aging is required to develop the characteristics of the cheese variety. The Scientific Basis for the 60-Day Aging Period for Cheeses Made with Unpasteurized Milk Origins Although not explicitly stated in the regulations, the 60-day holding period recommendation is intended to provide a measure of pathogen reduction in cheeses manufactured from milk that has not been pasteurized. This recommen- dation, which was first published in 1950 (15 C.F.R. §5653), was established by expert testimony provided during hearings that were conducted during the devel- opment of the current cheese standards of identity (Personal communication, J. Mowbray, FDA, September 25, 2002~. The scientific underpinnings of this recommendation are obscure, but appear to be derived at least partially from a study that investigated survival of B. abortus in Cheddar cheese (Gilman et al., 1946~. This study reported that B. abortus survived for up to 6 months in cheeses that had been artificially inoculated at levels of approximately 1,000 cfu/mL and held at 4.4°C. Bacterial survival was monitored directly by culturing viable B. abortus, and indirectly by guinea pig infection. In these initial experiments, 6-month-old cheeses were reported as positive for B. abortus, but no numbers were given; guinea pig lesions were described as slight. When these cheeses had been held for 1 year, inoculated guinea pigs showed no sign of B. abortus infection (i.e., no blood agglutination reactions, no characteristic lesions, and no B. abortus recovery from the spleen). No B. abortus was recovered from commercial Limburger cheeses that had been held for 57 days (no temperature or other conditions were described), despite the
CONTROLS FOR HAZARDS IN DAIRY PRODUCTS 235 fact that the milk used to manufacture two of the cheeses had tested positive for both viable B. abortus (no numbers given) and for guinea pig infection. Cheddar cheese made from milk that was naturally contaminated at levels of 700 to 800 cfu/mL was positive for culturable B. abortus (no numbers given) for 3 months. Viable B. abortus were recovered from some, but not all, of these test cheeses after 6 months; after 1 year, all guinea pigs were negative for signs of B. abortus infection. Cheeses made from milk collected from herds positive for B. abortus (no numbers given for initial levels of B. abortus in cheese milk) were negative (apparently for the presence of viable B. abortus, but the authors did not distin- guish between this possibility or whether these negative results reflected guinea pig inoculation experiments) after storage for at least 41 days at temperatures ranging from 1.1°C to 2.7°C. Unfortunately, many of the cheeses that were intended for examination in this part of the study were not tested for the presence of B. abortus, as samples were lost. Further, initial cheese storage period lengths were not standardized, but rather ranged from 41 to 84 days, making it very difficult to compare results among the cheeses. As part of the manuscript discus- sion, the authors claimed that Cheddar cheese had not been proven as a vector for human brucellosis (undulant fever), and that typhoid fever epidemics had not been attributed to cheeses cured for more than 63 days. Therefore, despite their own laboratory results, they believed that the epidemiological evidence suggest- ing a lack of association between cheese consumption and disease provided strong support for an aging period of approximately 2 months for commercial cheeses. The final stated conclusion was that "an aging period of 60 days is reasonable assurance against the presence of viable B. abortus organisms in Cheddar cheese" (Gilman et al., 1946~. Emerging Food Safety Concerns Recent evidence of the ability of bacterial pathogens to survive throughout a 60-day holding period has arisen from investigations of outbreaks of foodborne illnesses that have been traced back to aged cheeses, as well as from additional scientific research. Specifically, three outbreaks of salmonellosis following con- sumption of Cheddar cheese, two in Canada and one in the United States, suggest that various Salmonella strains can survive for extended periods in cheese prod- ucts, as described below. In the first outbreak, which was traced to Cheddar cheese manufactured in Kansas in 1976, raw milk had been held without refrigeration in the processing plant for 1 to 3 days prior to pasteurization and cheese manufacture. While it is not known for certain, total bacterial numbers in the prepasteurized, raw milk could have exceeded the thermal destruction capacity of the pasteurizing process. Microbiological analyses revealed the presence of Salmonella Heidelberg at very low levels (0.36-1.8 cfu/100 g of cheese) in the aged cheeses. The average pH of cheese batches bearing Salmonella was 5.6 vs. 5.4 for uncontaminated product;
236 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD thus, it is possible that slow acid production by starter cultures could have contributed to Salmonella survival. This outbreak resulted from numerous lapses in good manufacturing practices, and cannot be attributed solely to inadequacy of a 60-day holding period for pathogen reduction (Johnson et al., l990~. The second incident was comprised of a series of Salmonella outbreaks that occurred in Ontario, Canada, from 1980 to 1982. In these cases, Salmonella Muenster was isolated from raw-milk Cheddar cheese even after 125 days of curing at 41°F. In the third outbreak, which affected over 2,700 people in Canada in 1984, S. Typhimurium was isolated at very low levels from Cheddar cheese (0.39-9.3 cfu/100 g of cheese) that may have been prepared from a mix of raw and pasteur- ized milk. S. Typhimurium was found to persist in this cheese for 8 months at 41°F (Johnson et al., 1990~. In addition to the epidemiological evidence, research by Ryser and Marth (1987) and by Reitsma and Henning (1996) demonstrated the survival of L. monocytogenes and E. cold 0157:H7 for more than 60 days in Cheddar cheese. Ryser and Marth (1987) showed that L. monocytogenes could persist for up to 434 days postprocessing in artificially contaminated Cheddar cheese. Together with the outbreak information, these laboratory findings suggest the possibility that various foodborne pathogens may be capable of surviving current raw-milk Cheddar cheese manufacturing practices. These data suggest the need for addi- tional research on the persistence of pathogens during cheese manufacture and ripening, with a particular need to focus on survival of pathogens recognized as human hazards since 1946. As a consequence of the outbreaks described above, and due to reports in the scientific literature regarding pathogen persistence beyond 60 days, FDA is currently reviewing the 60-day aging policy. This policy review includes an examination of the literature to identify data gaps, research to confirm some findings and to fill identified data gaps, and input from stake- holders (Personal communication, J. Mowbray, FDA, September 25, 2002~. The committee recommends that for finished cheese products, a scientifi- cally appropriate performance standard for the reduction of targeted pathogens that result from the processing strategies or aging periods be developed and implemented. The committee recommends that the cheese industry, FDA, and state authorities work together to conduct and/or sponsor research to assess patho- gen reduction efficacies of cheese manufacturing conditions. The use of pasteur- ized milk in cheese manufacturing may provide an appropriate safe harbor for the manufacture of products for which adequate pathogen reduction may not occur during manufacture or during a holding period without an additional intervention. In the meantime, to enable consumers to make informed decisions regarding consumption of unpasteurized milk products, the committee recommends that FDA and state authorities require cheeses manufactured from subpasteurized milk to be clearly and prominently labeled as such at the point of purchase.
CONTROLS FOR HAZARDS IN DAIRY PRODUCTS Food Safety Policy for Imported Cheeses 237 FDA is charged with enforcing the Federal Food, Drug and Cosmetic Act, along with other laws that are designed to protect the health of consumers. These laws apply equally to domestic and imported products. Therefore, as with domes- tic products, imported foods must be pure, wholesome, safe to eat, and produced under sanitary conditions. All products must contain truthful and informative labeling in English. Under some circumstances, based on past history of a product or on other information indicating that a product may be violative, imported products may be detained upon arrival into the United States. FDA can identify and detain products from an entire country or geographic region if violative conditions appear to be widespread (this procedure is called "detention without physical examinational. Cheeses and other dairy foods have occasionally been subjected to detention. For example, due to widespread contamination with L. monocytogenes, French cheese was ordered to be detained in mid-1986. This action occurred despite a French program that already had been implemented in 1974, which allowed only plants that were certified by the French government to be following good manufacturing practices to export soft-ripened cheese to the United States. In January 1987, this certification program was expanded to include a requirement for the use of pasteurized milk in the manufacture of soft-ripened cheeses, as well as for Listeria testing of those products intended for export to the United States. Currently, only French processing plants that are certified by the French Ministry of Agriculture to export soft-ripened cheese manufactured from pasteurized milk can legally market their products in the United States. The FDA's Office of Regulatory Affairs maintains a listing of products that are cur- rently subject to import action. The Food and Drug Administration Food Compliance Program for Domestic and Imported Cheese and Cheese Products In response to a stated increase in the association of cheese and cheese products with outbreaks of human illness, in 1998 FDA issued a Food Compliance Program document that detailed plans for inspecting domestic cheese firms; examining domestic and imported cheeses for microbiological contamination, phosphatase, and filth; and taking action on cheese lots when violations are detected (CFSAN, 1998~. Sampling priorities were established in the following order: soft cheeses, hard cheeses, and cheese products. When cheese samples are taken as part of this program, mandated analyses include testing for L. monocytogenes, Salmonella, E. colt, enterotoxigenic E. cold (enterotoxigenic E. cold analyses are performed only when E. cold are present at levels of 104/g), E. cold 0157:H7, Staphylococcus aureus, and phosphatase. The testing is per- formed as a result of a public health concern and with the objective of identifying contaminated product and keeping it off the market; therefore, the sampling is not
238 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD designed for batch-to-batch or process verification purposes. Samples are col- lected during scheduled inspections when either of the following criteria are met: (1) the firm's products have a previous history of microbiological contamination (e.g., as a follow-up to a complaint or illness), or (2) sampling is conducted for some specific reason (e.g., observations during inspection indicate that sampling is warranted). This program is an example of a finished-product testing strategy initiated in response to illnesses associated with specific foods. THE ROLE OF THE U.S. DEPARTMENT OF AGRICULTURE IN DAIRY PRODUCT QUALITY AND WHOLESOMENESS Dairy Products Grading and Inspection Program In addition to FDA oversight of dairy product safety, many U.S. dairy plants participate in a voluntary grading and inspection program offered by the U.S. Department of Agriculture (USDA) through its Agricultural Marketing Service (AMS). USDA inspection and grading services are performed under the regula- tions in 7 C.F.R. part 58. The overarching goal of the AMS inspection and grading program is "to aid in the marketing of milk and dairy products by provid- ing a common language of trade through the development, improvement, and interpretation of standards, specifications, and quality improvement programs" (AMS, 2002~. The specific objectives of the program are to develop, maintain, and disseminate (1) sanitary requirements and model regulations to enhance the availability of safe, wholesome, high-quality dairy products, (2) definitions for product quality and wholesomeness, (3) requirements for participation in the USDA-Approved Dairy Plant Program, and (4) model state requirements for sanitary production and processing of manufacturing grade milk and milk products. The USDA grading program was initiated in the early 1900s as a conse- quence of a recognized need for a common language for dairy product character- istics. The Office of Markets, which predates the AMS, was established in 1913 to lay the groundwork for dairy market news and product standardization and grading. The Dairy Grading Branch of AMS currently administers this program. Plants participating in this program are inspected at least twice yearly. Plant inspections are unannounced and cover more than 100 items, including milk supply, plant facilities, equipment condition, sanitary practices, and processing procedures. AMS publishes specifications to guide dairy plants toward meeting approval requirements (Dairy Division, 2002~. In some cases, buyers may require that products meet specifications or grade standards. Therefore, despite a fee required to participate, this voluntary program is widely used by the dairy indus- try (AMS, 2002) because it provides guidance regarding how to achieve these quality standards. Although almost all dairy products can be inspected or graded, the products most commonly inspected and graded are butter, Cheddar cheese, and instant and
CONTROLS FOR HAZARDS IN DAIRY PRODUCTS 239 regular nonfat dry milk (AMS, 2002~. Official USDA grades (e.g., U.S. Grade AA for butter and Cheddar cheese and U.S. Extra Grade for nonfat dry milk) are derived from uniform standards of quality developed by the Standardization Branch of USDA. An official USDA grade indicates the product's quality by use of designated letters such as "AA" or words such as "extra" (AMS,2002~. Product specifications reflect minimum acceptable requirements for dairy products for which official grade standards have not been determined. The official USDA quality approved shield can be applied to products that meet the requirements of a specification. USDA standards and specifications are designed to ensure that products are free from defects that affect usability, which include, but are not limited to, the state of preservation of the product, cleanliness, wholesomeness, and fitness for human food (7 C.F.R. part 58~. Development of USDA product standards and specifications is usually initi- ated by requests from outside USDA, often as a consequence of the development of a new product or a change in processing technology (AMS, 2002~. Many requests are industry-driven, but other groups may initiate the process as well. Standard and specification development includes four elements: (1) research to determine quality factors and the range of quality encountered for the product, (2) investigation of production practices, including types of processing operations, packing, and equipment used, and consumer buying practices, (3) a statistical plan for product sampling, and (4) interviews with producers, packers, processors, shippers, receivers, consumers, and scientists. The standards and specifications are field-tested after the research is completed. At the end of this process, the standard or specification is published in the Federal Register. Standards and specifications increasingly rely upon scientific measurements, microscopic exami- nations, and written descriptions of quality aspects, but the process is still largely subjective (AMS, 2002~. Conformance to standards and grades is largely based on the grader' s perception of product taste, smell, appearance, and feel. Standards and specifications are reviewed and updated periodically to reflect changes in technology and milk quality (AMS, 2002~. Milk for Manufacturing Purposes Milk for manufacturing purposes includes "milk produced for processing and manufacturing into products for human consumption but not subject to Grade A or comparable requirements" (AMS, 2002~. USDA has established bacterial standards for milk to be used for manufacturing purposes. The goal of these requirements is to promote uniformity in state dairy regulations and laws, which should promote national uniformity in the sanitary processing of milk for manufacturing purposes. Enforcement of manufacturing milk regulations lies solely with the states. Lists of recommended microbiological standards for raw milk intended for manufacturing purposes and of AMS dairy product grade standards are presented in the tables in Appendixes F and G.
240 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD THE USE OF CURRENT STANDARDS AND CRITERIA UNDER HACCP As described earlier, through evolution of the PMO and other dairy stan- dards, the dairy industry has a long history of application of regulations to ensure the safety of its products intended for interstate commerce. Nevertheless, NCIMS has proposed testing the Hazard Analysis and Critical Control Point (HACCP) system under NCIMS as an alternative to the traditional dairy inspection/rating/ check system. In 1997, NCIMS conference delegates voted to evaluate the possi- bility of implementing HACCP systems in the dairy industry, and in 1999 they voted to implement a voluntary HACCP pilot program. The NCIMS HACCP committee has had oversight responsibilities for implementation of this pilot program since 1999. One of the greatest challenges facing the dairy industry has been the incorporation of HACCP into the regulatory format already in place. The NCIMS proposal has been developed in ways that harmonize HACCP with traditional NCIMS requirements, in terms of regulatory reciprocity and over- sight. For example, NCIMS proposes that the role of FDA in dairy HACCP could be similar to its current oversight and technical assistance role in the NCIMS system. The current regulatory authority is envisioned to perform the HACCP auditing function to verify that HACCP plans are effective. Implementation of HACCP requires establishing prerequisite programs such as Good Manufacturing Practices and Standard Sanitary Operating Procedures. Various aspects of these programs (e.g., safety of process water, condition and cleanliness of food contact surfaces, prevention of cross-contamination, control of employee health conditions and personal hygiene facilities, proper labeling and storage of toxic compounds, and pest exclusion) are already addressed in various sections of the existing PMO. Hence, the dairy industry already has in place the background Good Manufacturing Practices and Standard Sanitary Operating Procedures to reduce the potential occurrence of food safety hazards. The most likely critical control points for dairy processing operations will be pasteurization time and temperature conditions and control of raw and processed product storage temperatures. Microbial specifications and standards have been and will continue to be used for regulatory purposes in the dairy industry; how- ever, microbiological CCPs are unlikely to be adopted in the dairy industry. In July 1999, applications to participate in the dairy HACCP pilot program were sent to all 50 states by the NCIMS HACCP committee. Of 16 dairy industry applicants, 6 plants representing 6 states were chosen to participate. To provide essential ongoing technical support for the participating plants and state regula- tors, NCIMS and FDA's State Training Branch have held HACCP training work- shops for program participants. Further, the NCIMS HACCP established a Tech- nical Resource Team comprised of FDA, state, and industry representatives. Questions are generally submitted by e-mail, and responses are posted on the NCIMS HACCP website (CFSAN, 2003~. In May 2001, the NCIMS Conference
CONTROLS FOR HAZARDS IN DAIRY PRODUCTS 241 extended the pilot program to 2003 and expanded the program to invite all Grade A plants to participate. The pilot program now includes 15 plants in 10 states. The dairy processing industry's continued participation in this program will help to promote the continued availability of the NCIMS HACCP program as a voluntary alternative to the more prescriptive PMO program. The committee commends the dairy industry for voluntarily implementing a HACCP pilot program and strongly encourages timely adoption of HACCP systems throughout various sectors of the dairy processing industry. Adoption of performance standards for pathogen reduction, such as that proposed for cheese manufacturing, would more appropriately fit into a HACCP framework than in the dairy industry's current regulatory system. ARE THE STANDARDS AND SCIENTIFIC CRITERIA FOR MILK AND DAIRY PRODUCTS ACHIEVING THEIR GOAL? The committee recognizes that the application of regulations within the evolving PMO has been directly credited with reducing the incidence of milk- borne disease (Olsen et al., 2000~. To illustrate this point, the 1999 revision of the PMO stated that 25 percent of all disease outbreaks due to contaminated food and water were a consequence of consumption of milk products in 1938, but that, more recently, the prevalence of milk-borne disease has dropped to less than 1 percent of reported outbreaks. While dairy foods appear to be responsible for a relatively small proportion of U.S. foodborne-illness outbreaks that currently are successfully tracked to their source, occasional outbreaks of illness from consumption of contaminated dairy products do occur. The outbreaks listed in Table 7.5 do not provide a comprehensive listing of dairy food-associated illnesses since 1985, but rather provide a description of a selection of outbreaks associated with an international variety of dairy products, a variety of foodborne pathogens, and a variety of routes of product contamination. The goal of the table is to illustrate routes of entry for foodborne pathogens in dairy products. Determination of patterns among outbreak incidents may assist in identifying the most effective interventions and allocation of resources to further reduce dairy food-associated illnesses. Of the 20 outbreaks listed in Table 7.5, 11 are associated with consumption of raw milk products or of products contaminated by raw milk or by close contact with farm animals. These outbreaks further illustrate the possibility of the pres- ence of microbiological hazards in unpasteurized milk, as well as the need to develop effective interventions to control pathogens on the farm. Nine outbreaks (including some of those associated with raw milk product consumption) were associated with postpasteurization contamination of processed products. Post- pasteurization contamination usually results from lapses in cleaning and sanitizing procedures or from human food handling or processing errors that compromise product safety. The outbreak in 1988 brings into question the adequacy of current
242 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD TABLE 7.5 Outbreaks of Foodborne Disease Associated with Dairy Products Year Product and/or Source Organism 1985 Mexican-style soft cheese, illegally imported, raw milk suspected 1985 Mexican-style white cheese, environment and equipment grossly contaminated, even after clean-up; raw-milk delivery allegedly exceeded pasteurization capacity 1985 Pasteurized 2% milk; postpasteurization contamination; pipe cross-connection appears to have allowed raw milk to commingle with pasteurized Brucella Listeria ~ Salmonell 1988 Pasteurized milk; spores of Bacillus survived pasteurization and grew during B. cereus subsequent storage at refrigeration temperatures 1989 Mozzarella manufactured at a single plant, or cross-contaminated by a batch from S. Javiana that plant; low-level contamination of nationally distributed food product caused S. Oranier geographically dispersed foodborne outbreak that was difficult to detect 1992 Imported Irish soft unpasteurized cows' milk cheese (import into UK temporarily S. Dublin stopped, resumed after manufacturer decided to pasteurize milk used in production of cheese for export) 1994 Chocolate milk, leaking equipment, L. monocytogenes in plant environment, poor L. monoc sanitation, postpasteurization contamination, insufficient cooling 1994 1994 2000 Unpasteurized soft cheese cross-contaminated by chicken carcass (chickens dressed by cheese makers) Ice cream, contaminated through transport of pasteurized ice cream premix in tanker trailers that had previously carried nonpasteurized liquid eggs containing S. Enteritidis 1996 Formula dried milk for infants, international outbreak 1997 Raw milk, contaminated by cows at dairy of origin 1997 Mexican-style soft cheese made with raw milk 1998 Fresh cheese curds, unpasteurized, mislabeled as pasteurized 2000 Bottled pasteurized milk, possibly postpasteurization contamination from pigs via rinsing with untreated well water Fluid milk products; milk products formulated with skim milk powder bearing staphylococcal enterotoxin A 2000 2001 2002 S. Berta S. Enteriti Morbier cheese, one batch from a single processing plant incriminated (unpasteurized) Mexican-style soft cheese Raw milk obtained through cow-lease program, strategy used to circumvent legislation that prohibits sale of unpasteurized milk in this state 2002 Visit to dairy farm with E. coli-infected cows and calves 2003 Farmstead Gouda cheese; source under investigation S. Anatun E. cold O] S. Typhin E. cold O] Yersinia e Staphyloc enteroto' producec S. Typhin L. monoc' Campylob E. cold O E. cold O
CONTROLS FOR HAZARDS IN DAIRY PRODUCTS 'roducts 243 Organism Number of Cases Location Reference Brucella melitensis 9 TX Altekruse et al., 1998 nated, Listeria monocytogenes 145 CA Boor, 1997 n capacity Cation Salmonella Typhimurium 16,000 culture IL Ryan et al., 1987 confirmed; 168,791 to 197,581 cases estimated ring B. cereus 280 The Netherlands Van Netten et al., 1990 tch from S. Javiana, 164 WI, MN, MI, NY Hedberg et al., 1992 ct caused S. Oranienberg t rporarily S. Dublin 42 UK (South-east Maguire et al., 1992 production England) nt, poor L. monocytogenes 45 IL Dalton et al., 1997 ns S. Berta 82 Ontario Ellis et al., 1998 ix in S. Enteritidis 224,000 MN Hennessy et al., 1996 containing (estimate) S. Anatum 19 France, UK Threlfall et al., 1998 E. cold 0157:H7 6 OR Keene et al., 1997 S. Typhimurium DT104 54 WA Villar et al., 1999 E. cold 0157:H7 55 WI Durch et al., 2000 pigs Yersinia enterocolitica 10 VT, NH Ackers et al., 2000 aring Staphylococcal enterotoxin A, produced by S. aureus S. Typhimurium 14,700 Japan 113 Asao et al., 2002 De Valk et al., 2000 L. monocytogenes 3 NC Boggs et al., 2001 nt Campylobacter jejuni 5 WI CDC, 2002 E. cold 0157:H7 51 PA Crump et al., 2002 E. cold 0157:H7 11 Alberta CFIA, 2003
244 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD pasteurization practices for destruction of spore-forming organisms (e.g., Bacillus cereus) that can reproduce in fluid milk, particularly in those products that may be stored at refrigeration temperatures for extended times. In general, the presence of spore-forming organisms in raw milk that might not be destroyed by pasteuriza- tion has not been considered a significant public health risk (CFSAN, 2002~. Additional research, targeted at exploring the survival and outgrowth of spore- forming pathogens in conventionally pasteurized milk that is at refrigeration temperatures for more than 7 days, may be warranted. Finally, one large-scale outbreak in 2000 resulted from poor manufacturing practices in combination with reprocessing of past-code-date fluid milk products. Out-of-date cartons of fluid milk that had been delivered to a powdered milk processing plant were reportedly opened by hand and poured into vats that were not properly refrigerated. A subsequent power outage prevented the milk from being pasteurized for many hours. As a consequence, the milk was held for an extended period at tempera- tures permissive for bacterial growth. S. aureus was probably introduced into the milk during handling. This organism is predicted to have multiplied to levels necessary for enterotoxin production (> 100,000/mL) in the warm milk. After the electricity was restored, the milk was pasteurized, but conventional pasteuriza- tion conditions do not inactivate staphylococcal enterotoxin A. The milk was then dried into powdered milk ingredients. The resulting powdered milk ingredi- ents were used to formulate fluid milk products, which also were pasteurized. The presence and persistence of the enterotoxin from the powdered milk ingredi- ents in the pasteurized fluid milk products illustrates the toxin' s ability to with- stand conventional heat processing treatments and highlights the importance of preventing bacterial contamination of, and maintaining temperature control over, perishable food products. The committee concludes that the reduction in foodborne illnesses associ- ated with milk consumption in the United States is primarily a consequence of the near universal implementation of milk pasteurization for commercial fluid milk products, and also reflects the implementation of sanitation programs in process- ing plants that are designed to protect pasteurized milk from recontamination with pathogenic microbes. The committee further recognizes that despite the clear link that has been established between raw milk consumption and foodborne illnesses, some consumers continue to drink raw milk. The committee recom- mends that state and local authorities ban the sale of unpasteurized milk because of its inherent risks. Because most unpasteurized milk is sold or consumed at the farm, targeted educational programs that illustrate the hazards of raw milk con- sumption are warranted. FDA and state authorities should consider requiring clear and concise labeling to identify cheeses manufactured from unpasteurized milk to assist members of the public in making informed choices regarding food purchase and consumption.
CONTROLS FOR HAZARDS IN DAIRY PRODUCTS ECONOMIC AND ADMINISTRATIVE FEASIBILITY OF MILK PASTEURIZATION 245 Virtually all fluid milk processors that ship milk products via interstate commerce have invested in equipment for pasteurizing their product. Manufac- turers of cheese from milk that has not been pasteurized must hold the product for a minimum of 60 days at a temperature not less than 35°F. The expense of holding this inventory for the required time is part of the input cost of cheese manufactunng. As a measure for controlling numbers of bacterial pathogens, pasteurization of fluid milk and other products and holding times for cheeses are economically feasible and commonly applied. REFERENCES Ackers ML, Schoenfeld S. Markman J. Smith MG, Nicholson MA, DeWitt W. Cameron DN, Griffin PM, Slutsker L. 2000. An outbreak of Yersinia enterocolitica 0:8 infections associated with pasteurized milk. J. Infect Dis 181: 1834-1837. Altekruse SF, Timbo BE, Mowbray JC, Bean NH, Potter ME. 1998. Cheese-associated outbreaks of human illness in the United States, 1973 to 1992: Sanitary manufacturing practices protect consumers. J Food Prot 61: 1405-1407. AMS (Agricultural Marketing Service). 2002. Standardization Program Objectives. Online. U.S. Department of Agriculture (USDA). Available at http://www.ams.usda.gov/dairy/stand.htm. Accessed December 31, 2002. ANZFA (Australia/New Zealand Food Authority). 2000. Food Standards Code. Standard H1Milk and Liquid Milk Products. Online. Available at http://anzfa.gov.au/FoodStandardCode/code/ parth/Hl.htm. Accessed October 9, 2000. Asao T. Kumeda Y. Kawai T. Shibata T. Oda H. Haruki K, Nakazawa H. Kozaki S. 2002. An extensive outbreak of staphylococcal food poisoning due to low-fat milk in Japan: Estimation of enterotoxin A in the incriminated milk and powdered skim milk. Epidemiol Infect 130:33-40. Asher DM, Gibbs CJ Jr, Gajdusek DC. 1986. Slow viral infections: Safe handling of the agents of subacute spongiform encephalopathies. In: Miller BM, ed. Laboratory Safety: Principles and Practices. Washington, DC: American Society of Microbiology. Pp. 59-71. Boggs ID, Whitwam RE, Hale LM, Briscoe RP, Kahn SE, MacCormack IN, Maillard J-M, Grayson SC, Sigmon KS, Rardon JW, Saah JR. 2001. Outbreak of listeriosis associated with homemade Mexican-style cheese North Carolina, October 2000-January 2001. Morb Mortal Wkly Rep 50:560-562. Boor KJ. 1997. Pathogenic microorganisms of concern to the dairy industry. Dairy Food Environ Sanit 17:714-717. CDC. 2002. Outbreak of Campylobacter jejuni infections associated with drinking unpasteurized milk procured through a cow-leasing program Wisconsin, 2001. Morb Mortal Wkly Rep 51 :548-549. CFIA (Canadian Food Inspection Agency). 2003. Health Hazard Alert. Eyot Creek Farm Natural Farmstead Gouda Cheese May Contain Dangerous Bacteria. Online. Available at http:// www.inspection.gc.ca/english/corpaffr/recarapp/2003/20030128ee.shtml. Accessed March 12, 2003. CFIS (Canadian Food Inspection System). 1997. National Dairy Regulation and Code, Production and Processing Regulation October 1, 1997. Part 2. Processing. Online. Available at http:// www.cfis.agr.ca/english/regcode/prod_2ehtm#PartII. Accessed March 12, 2003.
246 SCIENTIFIC CRITERIA TO ENSURE SAFE FOOD CFSAN (Center for Food Safety and Applied Nutrition). 1998. Food Compliance Program. Domestic and Imported Cheese and Cheese Products. Online. Food and Drug Administration (FDA). Available at http://www.cfsan.fda.gov/~comm/cpO3037.html. Accessed September 30, 2002. CFSAN. 2000. Food Compliance Program. Milk Safety Program. Online. FDA. Available at http:// www.cfsan.fda.gov/~comm/cpl8003.html. Accessed September 30, 2002. CFSAN. 2002. Grade "A" Pasteurized Milk Ordinance. 2001 Revision. Online. FDA. Available at http://www.cfsan.fda.gov/~ear/pmoOltoc.html. Accessed September 30, 2002. CFSAN. 2003. Hazard Analysis and Critical Control Point. Online. FDA. Available at http:// www.cfsan.fda.gov/~lrd/haccp.html. Accessed March 17, 2003. Crump JA, Sulka AC, Langer AJ, Schaben C, Crielly AS, Gage R. Baysinger M, Moll M, Withers G. Toney DM, Hunter SB, Hoekstra RM, Wong SK, Griffin PM, van Gilder TJ. 2002. An out- break of Escherichia cold 0157:H7 infections among visitors to a dairy farm. N Engl J Med 347:555-560. Dairy Division. 2002. General Specifications for Dairy Plants Approved for USDA Inspection and Grading Service. Effective August 28, 2002. Online. AMS, FDA. Available at http:// www.ams.usda.gov/dairy/genpecs.pdf. Accessed December 31, 2002. Dalton CB, Austin CC, Sobel J. Hayes PS, Bibb WF, Graves LM, Swaminathan B. Proctor ME, Griffin PM. 1997. An outbreak of gastroenteritis and fever due to Listeria monocytogenes in milk. N Engl J Med 336:100-105. De Valk H. Delarocque-Astagneau E, Colomb G. Ple S. Godard E, Vaillant V, Haeghbaert S. Bouvet PH, Grimont F. Grimont P. Desenclos JC. 2000. A community-wide outbreak of Salmonella enterica serotype Typhimurium infection associated with eating a raw milk soft cheese in France. Epidemiol Infect 124:1-7. Douglas SA, Gray MJ, Crandall AD, Boor KJ. 2000. Characterization of chocolate milk spoilage patterns. J Food Prot 63:516-521. Durch J. Ringhand T. Manner K, Barnett M, Proctor M, Ahrabi-Fard S. Davis J. Boxrud D. 2000. Outbreak of Escherichia cold 0157:H7 infection associated with eating fresh cheese curds- Wisconsin, June 1998. Morb Mortal Wkly Rep 49:911-913. EEC (European Economic Community). 1994. Council Directive 92/46/EEC of June 16, 1992. Laying Down the Health Rules for the Production and Placing on the Market of Raw Milk, Heat- Treated Milk and Milk Based Products. Online. Available at http://europa.eu.int/comm/food/fs/ sfp/mr/mrO3_en.pdf. Accessed March 12, 2003. Ellis A, Preston M, Borczyk A, Miller B. Stone P. Hatton B. Chagla A, Hockin J. 1998. A commu- nity outbreak of Salmonella berta associated with a soft cheese product. Epidemiol Infect 120:29-35. Enright JB, Sadler WW, Thomas RC. 1957. Thermal Inactivation of Coxiella burnetii and its Rela- tion to Pasteurization of Milk. Public Health Monograph No. 47. Washington, DC: U.S. Govern- ment Printing Office. FDA. 2002. FDA Warns About Possible Enterobacter sakazakii Infections in Hospitalized Newborns Fed Powdered Infant Formulas. Online. Available at http://www.fda.gov/bbs/topics/ ANSWERS/2002/ANS01 146.html. Accessed September 30, 2002. Gilman HL, Dahlberg AC, Marquardt JC. 1946. The occurrence and survival of Brucella abortus in Cheddar and Limburger cheese. J Dairy Sci 29:71-85. Hammer BW. 1948. Dairy Bacteriology, 3rd ed. New York: John Wiley and Sons. Pp. 275-280. Hammer P. Lembke F. Suhren G. Heeschen W. 1995. Characterization of a heat resistant mesophilic Bacillus species affecting the quality of UHT milk. In: Heat Treatments and Alternative Methods. Proceedings of the IDF Symposium. Brussels: International Dairy Federation. Pp. 9-16. Hedberg CW, Korlath JA, D'Aoust JY, White KE, Schell WL, Miller MR, Cameron DN, MacDonald KL, Osterholm MT. 1992. A multistate outbreak of Salmonella Javiana and Salmonella Oranienburg infections due to consumption of contaminated cheese. JAMA 268:3203-3207.
CONTROLS FOR HAZARDS IN DAIRY PRODUCTS 247 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. The Investigation Team. N Engl J Med 334:1281-1286. Hillerton JE. 1997. Milk as a risk factor for BSE. In: Proceedings of the 36th Annual Meeting of the National Mastitis Council. Madison, WI: National Mastitis Council. Pp. 33-41. Johnson EA, Nelson JH, Johnson M. 1990. Microbiological safety of cheese made from heat-treated milk, Part II. Microbiology. J Food Prot 53:519-540. Keene WE, Hedberg K, Herriott DE, Hancock DD, McKay RW, Barrett TJ, Fleming DW. 1997. A prolonged outbreak of Escherichia cold 0157:H7 infections caused by commercially distrib- uted raw milk. J Infect Dis 176:815-818. Klijn N. Herrwegh AAPM, DeJong P. 2001. Heat inactivation data for Mycobacterium avium subsp. paratuberculosis: Implications for interpretation. JAppl Microbiol 91:697-704. Maguire H. Cowden J. Jacob M, Rowe B. Roberts D, Bruce J. Mitchell E. 1992. An outbreak of Salmonella Dublin infection in England and Wales associated with a soft unpasteurized cows' milk cheese. Epidemiol Infect 109:389-396. Mechor GD. 1997. Milk as a risk factor for Crohn's disease. In: Proceedings of the 36th Annual Meeting of the National Mastitis Council. Madison, WI: National Mastitis Council. Pp. 50-55. Olsen SJ, MacKinon LC, Goulding JS, Bean NH, Slutsker L. 2000. Surveillance for foodborne disease outbreaks United States, 1993-1997. Morb Mortal Wkly Rep 49:1-51. PHS (Public Health Service). 1940. Ordinance and Code. Public Health Bulletin #220, 1939 ed. Washington, DC: PHS. Reitsma CJ, Henning DR. 1996. Survival of enterohemorrhagic Escherichia cold 0157:H7 during the manufacture and curing of Cheddar cheese. J Food Prot 59:460-464. Ryan CA, Nickels MK, Hargrett-Bean NT, Potter ME, Endo T. Mayer L, Langkop CW, Gibson C, McDonald RC, Kenney RT, Puhr ND, McDonnell PJ, Martin RJ, Cohen ML, Blake PA. 1987. Massive outbreak of antimicrobial-resistant salmonellosis traced to pasteurized milk. JAMA 258:3269-3274. Ryser ET, Marth EH. 1987. Behavior of Listeria monocytogenes during the manufacture and ripen- ing of Cheddar cheese. J Food Prot 50:7-13. Threlfall EJ, Ward LR, Hampton MD, Ridley AM, Rowe B. Roberts D, Gilbert RJ, van Sommeren P. Wall PG, Grimon P. 1998. Molecular fingerprinting defines a strain of Salmonella enterica serotype Anatum responsible for an international outbreak associated with formula-dried milk. Epidemiol Infect 121:289-293. Van Netten P. van de Moosdijk A, van Hoensel P. Mossel DAA, Perales I. 1990. Psychrotrophic strains of Bacillus cereus producing enterotoxin. JAppl Bacteriol 69:73-79. Villar RG, Macek MD, Simons S. Hayes PS, Goldoft MJ, Lewis JH, Rowan LL, Hursh D, Patnode M, Mead PS. 1999. Investigation of multidrug-resistant Salmonella serotype Typhimurium DT104 infections linked to raw-milk cheese in Washington State. JAMA 281:1811-1816.
