Caballero B. (ed.) Encyclopaedia of Food Science, Food Technology and Nutrition. Ten-Volume Set

Подождите немного. Документ загружается.

be encountered in severity and duration. An add-

itional factor is whether there are certain populations

that are more susceptible to the development of a

particular chronic condition. These factors apply in

varying degrees to all of the sequelae listed in Table 3,

but many of the complications either occur very

rarely, or are associated with pathogens that are

themselves rare. Arthritis can be a relatively frequent

complication of foodborne illness and is associated

with some of the more common bacterial pathogens

in food. In addition, arthritic sequelae are one of the

more complex examples in terms of severity, dur-

ation, and susceptible populations. For these reasons,

arthritic sequelae will serve as a paradigm for exam-

ining various economic and clinical considerations

involved in developing cost estimates.

Arthritis Connection to Foodborne Illness

0024 Arthritic sequelae associated with foodborne illness

are organized into three entities: Reactive Arthritis

(ReA), Reiter’s syndrome (RS) and Ankylosing Spon-

dylitis (AS) (Table 4). The incidence of these three

conditions is increased in individuals that possess

the human Class I histocompatability antigen HLA-

B27. This association was one of the first documented

examples of genetic predisposition to disease. The

interest focused on this connection, the mechanisms

involved, and the diseases themselves have resulted in

numerous research and review articles. The genetic

connection and a commonality of symptoms (joint

inflammation) have led experts to group the three

conditions together as ‘HLA-B27-related spondylo-

arthropathies.’

0025 ReA is characterized by the inflammation of one

or more joints and generally occurs 1–3 weeks

following acute foodborne illness. Viable organisms

of the causative pathogen cannot be isolated from

the inflamed joints, making it a sterile arthritis, but

there are reports that bacterial antigens or DNA

may be present in inflamed joints.

0026 RS is a complicated form of ReA that includes joint

involvement as well as at least one extra-articular

feature such as inflammation of the urethra (pri-

marily in males) and/or uveitis (eye inflammation).

In some cases, no preceding infection can be identi-

fied. Those cases that have an identifiable preced-

ing infection are sometimes called Reactive RS.

Infections leading to RS can be derived from food

(e.g., Salmonella and Yersina) or from sexually

transmitted organisms such as Chlamydia spp. Be-

cause RS associated with foodborne illness is simi-

lar to ReA with respect to articular features and

may only have one extra-articular feature, the term

ReA will be used to capture both entities.

0027AS is the most chronic and debilitating of the spon-

dyloarthropathies, characterized by progressive

degeneration of the lumbar spine. AS can take

many years to develop, and in most cases, it is

difficult to make a specific association with a pre-

ceding foodborne infection. The relationship of AS

to foodborne illness derives from its strong link to

HLA-B27, its inclusion in the spondyloarthropa-

thies, and studies of foodborne disease outbreaks

in which patients have been followed for several

(many) years. These studies have provided evidence

that ReA or RS can progress to AS in some individ-

uals and support the theory that the sponsylo-

arthropathies are a spectrum of related diseases.

Clinically, this relationship has been useful in the

early diagnosis of AS, when doing a long-term

follow-up of foodborne outbreaks. Progression to

diagnosed AS is not the only potential chronic out-

come of ReA, which can sometimes lead to recur-

rent or chronic arthritis of joints other than the

spine. The frequency with which these various

scenarios occur has a great bearing on the potential

economic consequences of foodborne illness.

0028Before questions surrounding the development of

long-term arthritic sequelae can be addressed, the

frequency of primary joint inflammation (ReA) after

foodborne illness must be estimated. An assessment

of the circumstances surrounding the initial episode

of illness is also relevant to development of ReA and

long-term sequelae. Most of the available informa-

tion in this area comes from outbreaks of foodborne

illness. However, most outbreak investigations focus

on the acute, gastrointestinal phase of infection,

whereas follow-up surveys to detect arthritic and

other sequelae are rare. There is even less information

tbl0004 Table 4 HLA-B27-associated spondyloarthropathies

Condition Temporalrelationship toinfection HLA-B27 association Prognosis

Reactive arthritis 1–3 weeks after enteric illness 30–80% Usually self-limiting

Reiter’s syndrome Follows enteric or sexually transmitted

infection (1–3 weeks)

60–80% May be self-limiting non-articular

inflammation

Ankylosing spondylitis Difficult to link 90þ% Progressive

2678 FOOD POISONING/Economic Implications

on the occurrence of ReA following sporadic food-

borne illness. (Sporadic cases are foodborne illnesses

that have not been identified as part of an outbreak.)

Clinical information gathered from outbreaks has

suggested that several factors influence the incidence

and severity of ReA. These include the duration or

severity of diarrhea, which is influenced by the viru-

lence of the pathogen and the number of pathogens

ingested. The primary host factors appear to be the

presence of HLA-B27 or other antigenically related

antigens from the HLA-B cross-reactive group of

antigens. With respect to factors that influence pro-

gression from ReA to chronic joint problems, includ-

ing AS, severity of initial joint symptoms, and

presence of HLA-B27 are the most prominent. Recent

studies on AS have also suggested that other host

genetic components related to the production of in-

flammatory mediators such as interferon and tumor

necrosis factor play a role as well. With respect to the

frequency with which primary and chronic joint

problems occur, there is considerable variation in

published reports.

Disease-outcome Tree – Node 1

0029 Our goal was to develop a flow chart or disease-

outcome tree for arthritic sequelae that expressed

the probabilities of various short term and chronic

outcomes based on available clinical literature. This

tree was developed to distribute the cases into differ-

ent categories that are likely to have different average

costs. Several common variables from published

reports of both short- and long-term sequelae were

used to develop the tree. These included the duration

of patient follow-up, the size of the initial outbreak,

response rate to follow-up surveys, the pathogen in-

volved, and bias within the affected population. With

regard to potential bias in the populations under

study, the incidence of HLA-B27 varies among

national and ethnic groups and tends to be higher

among Northern European groups where many

outbreaks have been studied for arthritic sequelae.

Studies from these countries were included, because

to exclude them would have omitted several valuable

reports. When the data were compiled, there was no

obviously higher incidence of Salmonella-associated

ReA among the European versus North American

outbreaks (Table 5).

0030Numerous studies or case reports involving small

numbers of patients were not included, because these

studies provided no basis for determining an initial

incidence of ReA. Most of the information on the

initial incidence of arthritis following infection after

Yersinia enterocolitica, Shigella flexneri, and Cam-

pylobacter jejuni was derived from review articles.

In general, original reports involving the incidence

of ReA following outbreaks of salmonellosis with

various serovars were more frequent and were also

included. These data were incorporated into the first

node of the disease-outcome tree (Figure 2). Only the

four pathogens above were considered, because they

occur with sufficient frequency in the USA to be

captured by FoodNet active surveillance. In Europe,

tbl0005 Table 5 Initial incidence of reactive arthritis following foodborne illness

Source Type (cases) Location Incidence of ReA (%)

Keat (1983) Review Multiple outbreaks 2–3 (2.5)

Maki-Ikola and Granfors (1992) Review Multiple outbreaks 1.2–7.3 (3.5)

Yu and Thomson (1994) Review Outbreaks Multiple 6–10 (8.0)

Simonet (1999) Review Multiple outbreaks 0.5–15 (4.3)

Ike et al. (1986) Outbreak (565) Salmonella – milk US survey 30% response rate 2.3

Inman et al. (1988) Outbreak (260) Salmonella Canada survey 7.3

Thomson et al. (1992) Outbreak (83) Salmonella – luncheon Canada survey 92% response rate 8.2

Locht et al. (1993) Outbreak (126) Salmonella – dinner Sweden survey 90% response rate 15

Samuel et al. (1995) Outbreak (919) Salmonella – restaurant US survey 33% response rate 14

Thomson et al. (1995) Outbreak (423) Salmonella Canada survey 97% response rate 6.4

Mattila et al. (1998) Outbreak (210) Salmonella Finland survey 91% response rate 11.5

Full recovery

Reactive

arthritis

Progressive/

recurrent

arthritis

Full recovery

Acute FBI

92%

(85−97.7)

16%

(18−80)

(2.3−15)

40%

(20−82)

92%

3.2%

1.9%

1.4%

100%

Mild or intermittent

joint pain

40% (20−46)

Severe/chronic joint

pain

30% (5.9−48)

30% (14−32)

Sacroiliitis/Spondylitis

Disease Outcome Tree of Arthropathies*

*Following exposure to bacterial foodborne pathogens

fig0002Figure 2 Arthritis–foodborne illness disease outcome tree.

FOOD POISONING/Economic Implications 2679

Yersinia pseudotuberculosis has been associated with

outbreaks of foodborne illness featuring a high inci-

dence of ReA. This organism was not included, be-

cause cases of infection with this pathogen are rare in

the USA, with no confirmed cases from 1996–1999,

based on FoodNet surveillance. The clinical picture of

ReA with all of the four included pathogens appears

to be similar. In looking at a number of outbreaks, the

overall incidence of initial arthritis also appears to

be similar among the above four organisms, but the

incidence of arthritis may vary widely between differ-

ent outbreaks involving the same pathogenic species.

This is probably due to variation in pathogen viru-

lence, dose, host population, or even the effectiveness

of tracking arthritis. For this reason, in developing

the disease-outcome tree, each report or review was

treated as a distinct entity and given equal weight,

rather than pooling cases from different outbreaks.

The reports contributing to the first node of the tree

give a range of 2.3–15% for the initial incidence of

arthritis following foodborne illness. This results in

an average of approximately 8% of foodborne ill-

nesses that result in ReA.

Disease-outcome Tree – Node 2

0031 The second node of the disease-outcome tree distin-

guishes between ReA that resolves completely and

does not recur after the initial episode, and ReA in

which the joint inflammation shows any form of

progression or periodic recurrence. This node is

based on studies that report some progression, or

lack of resolution of joint disease, but do not measure

or define progression to a particular clinical state.

This includes studies with relatively short follow-up

times of less than 1 year. Compared with studies on

the initial incidence of ReA, studies involving

progression or recurrence were fewer and involved

outbreaks of salmonellosis more often than other

pathogens (Table 6). Therefore, most of the data

used for the second node of the disease-outcome tree

were also derived from studies involving Salmonella.

However, our purpose was to quantify, within reason-

able bounds, the possible sequelae of foodborne in-

fection from self-limiting to chronic joint disease. To

focus exclusively on Salmonella would have resulted

in omission of two important long-term follow-up

studies with Shigella and Yersinia enterocolitica.

0032It is conceivable that there is some difference in

the methodology and criteria for detection of ReA

between older studies and more current reports. For

example, a large outbreak of Shigella in 1948 in

Finland reported a very low 0.2% incidence of arth-

ritis, yet, upon follow-up of 100 of these cases 20

years later, 80% had shown some progression with

32% showing signs of AS. The figure of 80% pro-

gression is quite high compared with other studies

and may have resulted from initial recognition of

arthritis in only the most severe cases. These cases

would be the most likely to progress to chronic status.

Other review articles reported the proportion of cases

that progress to chronic status using a qualitative

approach, based on the clinical experience of the

authors. In considering the percentage of cases in

which initial ReA does not completely resolve, or

progresses, the second node of the tree gives a range

of 18–80%. This results in an average of 60% of cases

that do not evidence full recovery.

Disease-outcome Tree – Node 3

0033The third node of the tree measures the progression to

chronic joint disease, based on the data from relevant

studies, making a distinction between AS and other

forms of chronic joint inflammation. Fewer data

points were available for this node. In the various

studies, there was considerable variation in the

number of subjects, the duration of follow-up, and

the clinical criteria and terminology for progression

and for severity. Chronic symptoms were categorized

in these reports on the basis of which joints were

affected, radiological criteria, patient histories, or a

combination of these criteria. If progression was

severe enough to cause radiologically detectable

tbl0006 Table 6 Progression of reactive arthritis

Study Follow-up (cases)

Anyprogression (%) Severe/chronic (%)

Sairanen et al. (1969) 20 years (100) 80 32

Leirisalo and Suoranta (1988) 10 years (85) 51.8 5.9

Maki-Ikola and Granfors (1992) >1 year (164) 78 18.3

Thomson et al. (1992) 18 months (6) 33 na

Locht et al. (1993) 6 months (17) 47 na

Yu and Thomson (1994) na (review) 66 33

Thomson et al. (1995) 5 years (27) 66 29.6

Leirisalo et al. (1997) 5–22 years (50) 60 40

Mattila et al. (1998) 4 months (22) 18 na

Original number of ReA cases.

2680 FOOD POISONING/Economic Implications

spinal damage, these cases tended to be identified as

sacroiliitis or spondylitis (AS). These cases were dis-

tinguished in a separate category because AS is readily

diagnosed as a distinct entity in the HLA-B27-

associated spondyloathropathies. Two additional cat-

egories were used to capture other types of chronic or

recurrent symptoms. These included cases in which

symptoms manifested as periodic flare-ups, and cases

in which there was chronic joint pain not associated

with the back. Approximately 40% of cases in this

node involve minor or recurrent symptoms. Sixty per

cent of the cases are more severe or chronic joint

inflammation, divided approximately equally be-

tween those that show radiologic evidence of spine

and lower back inflammation and those having

inflammation of other joints.

Conclusions

0034 The disease-outcome tree is a useful tool for eco-

nomic evaluation that establishes estimates for the

likelihood of various arthritic sequelae following an

outbreak of foodborne bacterial illness. The path-

ways are based on reported patterns of clinical dis-

ease. The estimates are necessarily inexact, because of

limitations of the available data. The estimates are

based on incidence data that were all derived from

outbreaks of foodborne illness. It is possible that the

incidence of arthritis following sporadic cases might

be different than the incidence associated with out-

breaks. The use of outbreaks might be biased toward

more virulent pathogen strains. At present, there is no

information on the incidence of ReA following spor-

adic foodborne illness. An active surveillance system

for foodborne illnesses such as FoodNet may prove to

be an excellent mechanism from which to determine

the incidence of ReA in sporadic illness and to con-

duct long-term follow-ups for progressive or chronic

joint disease. As more definitive data become avail-

able, the disease-outcome tree can be updated and

modified.

003 5 The next issue is the economic valuation method

used. The Economic Research Service’s annual food-

borne illness cost estimates use a cost-of-illness ap-

proach for illnesses and a willingness-to-pay (WTP)

approach for premature deaths. CFSAN has routinely

used the value of QALYs lost in its regulatory impact

analyses of food safety regulations. To date, no stand-

ardized method has been chosen by the economic

profession. Ongoing research is developing WTP esti-

mates for morbidity outcomes. In comparison with

cost-of-illness approaches for chronic sequelae, such

as arthritis, both the WTP and QALY estimates can

substantially increase the economic valuation. The

reason is the more comprehensive nature of these

estimates that include an implicit valuation for pain

and suffering, lost leisure time, and other implicit

costs that are not included in the cost of illness esti-

mates based on medical costs, productivity losses,

and other out-of-pocket costs.

0036In the disease-outcome tree developed here, around

8% of the cases of salmonellosis can be expected to

develop reactive arthritis. Forty per cent of the react-

ive athritis cases will fully recover. The remaining

60% can be expected to have recurring episodes of

varying severity (40% mild/intermittent joint pain,

30% severe joint pain, 30% spondylitis). In summary,

the disease-outcome tree encompasses all the possible

lifetime consequences of those in the US population

who have been exposed to foodborne pathogens. The

tree predicts that on average, 92% of acute cases will

have a full recovery; 3.2% will have full recovery

with reactive arthritis symptoms for a few weeks;

1.9% will have lifelong mild, intermittent joint pain;

1.4% will have a lifelong progression to severe joint

pain; and 1.4% of cases will have a lifelong progres-

sion to spondylitis. Valuation of avoiding just the pain

and suffering associated with these arthritic outcomes

can be expected to be significant. An early study by

Thompson suggested that those afflicted would give

up 22% of their household income to be rid of

rheumatoid arthritis. The disease-outcome tree en-

compasses all the possible lifetime consequences of

those in the U.S. population who have been exposed

to foodborne pathogens. The life-long nature and

severity of many chronic sequelae, such as arthritis,

indicate that these will make substantial contribu-

tions to the estimates of the cost of foodborne illness.

See also: Campylobacter: Properties and Occurrence;

Escherichia coli: Food Poisoning; Food Poisoning by

Species other than Escherichia coli; Food Poisoning:

Classification; Tracing Origins and Testing; Statistics;

Listeria: Listeriosis; Salmonella: Properties and

Occurrence; Shigella; Yersinia enterocolitica:

Properties and Occurrence

Further Reading

Council for Agricultural Science and Technology (CAST)

(1994) Foodborne Pathogens: Risk and Consequences,

Task Force Report 122 (cochairs, Foegeding PM and

Roberts T). Ames, IA: CAST.

Economic Research Service (2001) United States Depart-

ment of Agriculture http://www.ers.usda.gov/briefing/

FoodborneDisease/.

Food and Drug Administration (2001) Hazard Analysis and

Critical Control Point (HAACP); Procedures for the

Safe and Sanitary Processing and Importing of Juice;

Final Rule, 66 Federal Register 6138–6202, esp. pp.

6179–6194 (January 19).

FOOD POISONING/Economic Implications 2681

Food and Drug Administration (1999) Preliminary Regula-

tory Impact Analysis and Initial Regulatory Flexibility

Analysis of the Proposed Rule to Require Refrigeration

of Shell Eggs at Retail and Safe Handling Labels, 64

Federal Register, pp. 36516–36539 (July 6).

Food and Drug Administration (1995) Procedures for the

Safe and Sanitary Processing and Importing of Fish and

Fishery Products; Final Rule, 60 Federal Register, pp.

65096–65202, esp. pp. 65179–65192 (December 18).

Locht H, Kihlstrom E and Lindstrom FD (1993) Reactive

arthritis after Salmonella among medical doctors: Study

of an outbreak. Journal of Rheumatology 20: 845–848.

Mattila L, Leirisalo-Repo M, Pelkonen P et al. (1998)

Reactive arthritis following an outbreak of Salmonella

bovismorbificans infection. Journal of Infection 36:

289–295.

Mead PS, Slutsker L, Dietz V et al. (1999) Food related

illness and death in the United States. Emerging Infec-

tious Diseases 5: 607–625.

Sairanen E, Paronen I and Mahonen H (1969) Reiter’s

syndrome: a follow-up study. Acta Medica Scandinavica

185: 57–63.

Samuel MP, Zwillich SH, Thomson TD et al. (1995) Fast

food arthritis – a clinico-pathologic study of post-

Salmonella reactive arthritis. The Journal of Rheumatol-

ogy 22: 1947–1952.

Simonet ML (1999) Enterobacteria in reactive arthritis:

Yersinia, Shigella, and Salmonella. Revue du Rheuma-

tisme (English edition) 66(1, Supplement): 14S–19S.

Thomson GTD, Chiu B, Derubeis D, Falk J and Inman RD

(1992) Immunoepidemiology of post-Salmonella react-

ive arthritis in a cohort of women. Clinical Immunology

and Immunopathology 64: 227–232.

Thomson GTD, DeRubeis DA, Hodge MA, Rajanaygam C

and Inman RD (1995) Post-Salmonella reactive arthritis:

late clinical sequelae in a point source cohort. American

Journal of Medicine 98: 13–21.

Viscusi WK (1993) The value of risks to life and health.

Journal of Economic Literature 31: 1912–1946.

Yu DTY and Thomson GTD (1994) Clinical, epidemi-

ological and pathogenic aspects of reactive arthritis.

Food Microbiology 11: 97–108.

FOOD SAFETY

R Rooney, World Health Organization, Geneva,

Switzerland

P G Wall, Food Safety Authority of Ireland, Dublin,

Ireland

Background

0001 Contaminated food is one of the most widespread

public health problems of the contemporary world

and causes considerable morbidity and mortality.

Globally, millions of people are affected by foodborne

disease. Food poisoning can be very serious in vulner-

able groups such as the elderly, infants, young chil-

dren, pregnant women, and immunocompromised

individuals. The proportion of vulnerable groups

such as the elderly and immunocompromised is

increasing in industrialized countries.

0002 A number of high-profile food-safety emergencies

and scares, in recent years, have shaken consumer

confidence in the production of food and have

focused attention on the way food is produced, pro-

cessed, and marketed. These incidents have included

an epidemic of bovine spongiform encephalitis (BSE)

in cattle, possible links between BSE and variant

Creutzfeldt–Jakob disease (CJD) in humans, and

the discovery of dioxin contaminated animal feed

entering the food chain. New foodborne pathogens

such as Escherichia coli O157:H7, Cyclospora

cayentanensis, and Listeria monocytogenes have

also emerged. Infection with Escherichia coli

O157:H7 can be fatal in young children, the elderly,

and the immunocompromised, and within the last 10

years, serious outbreaks have occurred in the USA,

Scotland, and Japan. Listeria monocytogenes can

result in meningoencephalitis and/or septicemia in

newborns and adults and abortion in pregnant

women. Parasites such as Cyclospora cayentanensis

and Cryptosporidium parvuum have emerged as a

serious health threat for immunocompromised

individuals, particularly AIDS patients.

0003Chemicals in food have also caused concern. Veter-

inary medicines, growth promoters, pesticides, and

other man-made chemicals are widely used in the

food industry. Improper use of these substances in

primary production can result in unsafe residues in

food. Chemicals are also added to foods during pro-

cessing in the form of food additives, flavorings, and

colors. Many consumers are worried about the long-

term impact of mixtures of chemicals and chronic as

well as acute effects on vulnerable groups.

0004Food safety is a sensitive political and economic

issue as well as an important public health issue. This

2682 FOOD SAFETY

paper aims to highlight the factors that have contrib-

uted to food-safety problems over recent decades such

as intensive farming, globalization of the food

industry, and changes in consumer demands. The

biological, chemical, and physical hazards in food

and possible control and prevention strategies for

the industry will be discussed.

Food Safety Hazards

0005 The term contaminant covers harmful substances or

microorganisms that are not intentionally added to

food. Contaminants may enter the food accidentally

during growth, cultivation, or preparation, accumu-

late in food during storage, form in the food through

the interaction of chemical components, or may be

concentrated from the natural components of the

food. There are three main types of hazard: bio-

logical, chemical, and physical.

Biological Hazards

0006 Biological agents of concern to public health include

pathogenic strains of bacteria, viruses, helminthes,

protozoa and algae, and certain toxic products they

may produce. Foodborne infections are caused when

microorganisms are ingested, and these can multiply

in the human body. Intoxications result when micro-

bial or naturally occurring toxins are consumed in

contaminated foods. Microorganisms or toxins may

be introduced directly from infected food animals or

from workers, other foods, or the environment during

the preparation or processing of food. Poisonous sub-

stances may also be produced by the growth of bac-

teria and molds in food. Table 1 lists pathogenic

organisms of public health importance, which may

be transmitted through contaminated food.

Chemical Hazards

0007 Foodborne toxicants can be categorized according to

their chemical nature. Some foodborne toxicants are

inorganic, e.g., lead and arsenic, whereas others are

organic, e.g., saxitoxin and polychlorinated biphe-

nols (PCBs). The main categories of man-made

chemicals that occur in foods are shown in Table 2.

Chemical hazards may result from pollution arising

from industrial and other human activities (e.g., lead,

mercury, cadmium, and PCBs), from agricultural prac-

tice (e.g., pesticides, fertilizers, and residues resulting

from improper use of veterinary drugs in food-produ-

cing animals) or from food processing and packaging

(e.g., nitrosamines and certain polycyclic aromatic

hydrocarbons). These contaminants may present a

potential hazard for human health if exposure

exceeds tolerable levels.

Physical Hazards

0008Physical hazards are foreign matter, including dirt

and glass, that accidentally get into food They can

be the result of environmental contamination during

production, processing, storage, packaging, and trans-

port, or from fraudulent practices. The potential for

tbl0001Table 1 Examples of main biological hazards in foods

Bacteria (spore-forming) Viruses

Clostridium botulinum Hepatitis A and E

Clostridium perfringens Norwalk virus group

Bacillus cereus Rotavirus

Bacteria (nonspore-forming) Protozoa and parasites

Brucella abortis Cryptosporidium parvum

Brucella suis Diphyllabothrium

Campylobacter spp. Entamoeba histolytica

Pathogenic Escherichia coli

(E. coli O157:H7, EHEC, EIEC,

ETEC, EPEC)

Giardia lamblia

Ascaris lumbricoides

Taenia solium

Listeria monocytogenes Taenia saginata

Salmonella spp. (S. typhimurium,

S. enteritidis)

Trichinella spiralis

Shigella (S. dysenteriae)

Staphylococcus aureus

Streptococcus pyogenes

Vibrio cholera

Vibrio parahaemolyticus

Vibrio vulnificus

Yersinia enterocolytica

Source: FAO (1998) Food Quality and Safety Systems. A Training Manual on

Food Hygiene and the Hazard Analysis and Critical Control Point (HACCP)

System. Rome: Food and Agricultural Organization of the United Nations,

with permission.

tbl0002Table 2 Examples of chemical hazard in foods

Industrial/environmental pollution Additives

Polychlorinated biphenyls (PCBs) Feed additives

Agricultural chemicals Veterinary drugs

Pesticides Food additives

Fertilizers Vitamins and minerals

Antibiotics Contaminants

Growth hormones Lubricants

Cleaners

Toxic elements and compounds Pest control chemicals

Coatings

Paints

Inorganic chemicals

Lead

Chemicals migrating

from packaging

Tin Plasticizers

Mercury Vinyl chloride

Zinc Printing/coding inks

Cadmium Adhesives

Arsenic Lead

Cyanide Tin

Source: FAO (1998). Food Quality and Safety Systems. A Training Manual on

Food Hygiene and the Hazard Analysis and Critical Control Point (HACCP)

System. Rome: Food and Agricultural Organization of the United Nations,

with permission.

FOOD SAFETY 2683

ionizing radiation to have long-term health effects,

not only on the people living nearby but also the

health of the ecosystem, is considerable. Following

the nuclear reactor failure at Chernoble (Ukraine) in

1986, food contaminated with radionucleides with

long half-lives, such as cesium 137, is the major

source of exposure for people living in this area.

Table 3 lists examples of physical hazards in food.

Food Safety Emergencies

0009 The ultimate objective of the food industry and gov-

ernment regulators is to insure that food reaching the

consumer is safe and wholesome. Feed manufactur-

ers, farmers, and food operators have the primary

responsibility for food safety. Government regulators

and consumer associations also have a role to play.

However, food-safety emergencies over the last two

decades have exposed weaknesses in the food indus-

try from farm to fork. They have shown that consider-

able progress is needed to come to grips with the

preventive aspects of foodborne disease.

0010 Many factors have contributed to recent food

emergencies. The food production chain is becoming

increasingly complex because of mass production.

Modern farming practices are intensive and can result

in microorganisms contaminating a large number of

crops or infecting a large number of animals. Healthy

animals may carry pathogens that cause disease in

humans. Examples include Salmonella spp., Listeria

spp., and some strains of Escherichia coli and Cam-

pylobacter. Animals may be infected from feed, from

other animals, or from the environment.

Food distribution systems have expanded dramat-

ically over recent decades. This has lengthened the

food chain adding opportunities for contamination

to occur. Ingredients may come from many different

countries and be combined in a single product. The

Food and Agricultural Organization (FAO) of the

United Nations estimates that over 500 million

tonnes of food move in international trade, valued

at around $400–500 billion, each year.

0012Trends show that consumers are relying increas-

ingly on convenience and ready-prepared foods. Such

foods have few barriers to microbial growth, such as

salt and preservatives, and simple mistakes could

easily result in an outbreak of foodborne disease.

World Trade

0013The Food and Agriculture Organization (FAO) and

the World Health Organization (WHO) have comple-

mentary food-safety mandates to protect the health of

consumers, to prevent the spread of disease and to

ensure that the procedures followed in food trade are

fair. The Codex Alimentarious Commission (CAC),

which manages the joint FAO/WHO food-standards

program, sets international standards for food. It was

established in 1961 and 1962 by the FAO and WHO

and aims to protect the health of consumers and

ensure fair practices in food trade.

0014The World Trade Organization (WTO) rules pro-

vide a framework for the application of food-safety

measures in international trade. The standards and

guidelines of the CAC are recognized by the WTO as

the international benchmark standards, guidelines,

and recommendations to be used as the basis of

harmonizing food safety measures affecting human

health and world food trade.

0015International standards influence trade as well as

food safety, and national governments are under

pressure to ensure that their regulatory programs are

comprehensive and conform to these global stand-

ards. Owing to the complexity of the food-

production chain, the traditional role of government

as the food-safety inspector has changed, with the

food industry taking more responsibility for the

implementation of quality-assurance programs

throughout food production, distribution, and retail.

However, a strong national food control agency is

required to bring about cooperation between govern-

ment departments and industry, and to set standards

and targets for health protection.

Risk Analysis

0016The concept of risk analysis is gaining acceptance

internationally as a very important component of

national food control programs, and its principles

are being integrated more fully into the work of

food safety regulators and policy-makers. This prin-

ciple consists of three components: risk assessment,

risk management, and risk communication.

tbl0003 Table 3 Examples of physical hazards

Material Sources

Glass Bottles, jars, light fixtures, utensils, gauge

covers, etc.

Wood Field sources, pallets, boxes, building

materials

Stones Fields, buildings

Metal Machinery, fields, wire

Insulation Building materials

Bone Improper processing

Plastic Packaging, pallets, equipment

Personal effects Employees

Source: FAO (1998) Food Quality and Safety Systems. A Training Manual on

Food Hygiene and the Hazard Analysis and Critical Control Point (HACCP)

System. Rome: Food and Agricultural Organization of the United Nations,

with permission.

2684 FOOD SAFETY

0017 Risk assessment is a science-based investigation

that can provide an estimate of the probability and

impact of adverse health effects attributable to poten-

tially contaminated foods. The aim is to identify, in

some quantitative or comparable way, the relation-

ship between hazards and actual exposure to harm.

The detail and complexity of a given risk assessment

will vary, depending on the availability of time, re-

sources, and scientific information. A full risk assess-

ment may not always be carried out, because data

may be incomplete, and resources may be inadequate.

However, decisions that are made by well-focused

assessments, using whatever data are available, will

usually be more reliable and transparent than subject-

ive judgement. Risk assessment links into the other

two components of risk analysis, the others being risk

management and risk communication. The stages of

the risk analysis process are closely linked and cannot

be separated easily. Consideration of the ways in

which a risk might be managed is necessarily part of

the assessment and communication.

0018Risk assessment consists of four steps: hazard iden-

tification, hazard characterization, exposure assess-

ment, and risk characterization (Figure 1). In hazard

identification, the association between disease and

the presence of a pathogen or contaminant in food

is documented. Hazard characterization involves

obtaining quantitative information about the hazard,

including, where possible, information on dose–

response relationships. Exposure assessment esti-

mates the intake or exposure to a chemical or

pathogen, in terms of its magnitude, duration, and

frequency, for the general population, for subgroups

or for individuals. Risk characterization is the inte-

gration of hazard identification, hazard characteriza-

tion, and human intake/exposure assessment. It

assesses the likelihood of a particular event. This is

the framework adopted by the CAC, the international

standard setting body for foods in international trade.

0019Risk assessment is already fairly well developed for

chemical hazards. The estimation of chemical health

risks often relies on data obtained in the customary

Description of

food safety

problem and

context

Hazard identification

What agents/chemicals are present in the

food and capable of causing adverse effects?

Exposure assessment

What is the likely

frequency and level of

consumption of the

hazard?

Hazard

characterization

What is the nature of

adverse effects?

Dose−response

assessment

Risk

characterization

Integration of exposure

assessment and hazard

characterization.

Risk

estimate

Probability and severity of

illness attributable to the

pathogen/chemical source:

e.g. no of illnesses per year.

Uncertainty: What important

data or knowledge is missing?

Variability: What variable

factors influence the

magnitude of the risk?

fig0001 Figure 1 Steps of food-safety risk assessment. Adapted from Lammerding AM and Fazil A (2000) Hazard identification and exposure

assessment for microbial food safety risk assessment. International Journal of Food Microbiology 58: 147–157, with permission.

FOOD SAFETY 2685

rodent feeding assays using standardized safety

margins. Such assessments call for an extrapolation

of response data derived from animal studies to

humans. The setting of advisory standards, such as

acceptable daily intakes (ADIs) for food additives,

pesticides, and veterinary residues in food, is a key

component in the overall risk assessment of many

chemicals subject to regulatory control. Quantitative

risk assessment is an emerging tool in the field of

microbial food safety. Estimation of microbiological

hazards presented by foods is more complex, because

of biological diversity and variability. However, such

estimations are usually more accurate than estimates

on adverse effects from chemicals in foods. This is

because incidents are frequent, in sharp contrast to

those resulting from exposure to chemicals. This

allows for a better risk assessment.

Risk Management and Risk Communication

0020 Food scares such as salmonellae in eggs, Escherichia

coli O157 in beef and BSE have all damaged consumer

confidence in regulatory agencies. Such problems

were compounded when government departments

had the dual role of promoting farming and the food

industry as well as protecting the consumer. Such roles

resulted in a conflict of interest. A regulatory agency

should ideally be independent of trade and economic

interests and have consumer interests as its primary

function. Although it is the primary responsibility of

the food industry to comply with the laws and regula-

tions, the regulatory agency should provide an inte-

grated inspection system from farm to fork. This may

best be achieved by establishing a national food stand-

ards agency that has responsibility for all sectors of the

food chain. Many countries have established such

agencies and coordinate the work of all government

departments involved in food safety. Such coordin-

ation overcomes the problems of duplication of

work and conflicts of interest.

0021 A foodborne disease surveillance program is also an

essential part of a food control program. Surveillance

is necessary to identify foodborne contaminants, their

sources, and their modes of transmission in order to

determine the best short- and long-term control meas-

ures. Interpretation of trends and investigation of out-

breaks can help identify the mechanisms by which

contamination and disease transmission occurred.

When an outbreak occurs or a chemical contamin-

ation incident is identified, it is vital to learn from

the investigation what went wrong in order to devise

strategies to prevent similar events in the future.

Quality Assurance

0022 Food quality and safety can only be insured through

the application of quality-control systems throughout

the entire food chain. They should be implemented at

farm level with the application of good agricultural

practices and good veterinary practices at production,

good manufacturing practice at processing, and good

hygiene practices at retail and catering levels. One of

the most effective ways for the food sector to protect

public health is to base their food management pro-

grams on Hazard Analysis Critical Control Point

(HACCP).

0023HACCP consists of seven principles that outline

how to establish, implement, and maintain a quality-

assurance plan for a food establishment. It is a system-

atic approach to the control of potential hazards in a

food operation and aims to identify problems before

they occur. Control is proactive, since remedial action

can be taken. The HACCP principles have inter-

national recognition, and the seven principles are

listed in Table 4.

0024HACCP plan adoption has greatly improved the

food industry’s ability to design programs to insure

the safety of food. It was introduced in the early

1970s and has become the premier system for

preventing foodborne hazards, particularly those of

microbiological origin. The basic process in creating

HACCP plans is that the main hazards associated

with a food product are determined, and key steps

in the production, processing, distribution,

marketing, and preparation of food are controlled

tbl0004Table 4 Seven principles of HACCP

Principle 1

Conduct a hazard analysis. Prepare a list of steps in the

process where significant hazards occur and describe the

preventive measures. Assess the likelihood of occurrence of

the hazard(s) and identify the measures for their control.

Principle 2

Identify the Critical Control Points (CCPs) in the process.

Determine the points, procedures or operational steps that can

be controlled to eliminate the hazard(s) or minimize its (their)

likelihood of occurrence.

Principle 3

Establish critical limits for preventative measures

associated with each identified CCP.

Principle 4

Establish CCP monitoring requirements. Establish

procedures from the results of monitoring to adjust the process

and maintain control.

Principle 5

Establish corrective actions to be taken when monitoring

indicates a deviation from an established critical limit.

Principle 6

Establish effective record-keeping procedures that

document the HACCP system.

Principle 7

Establish procedures for verification that the HACCP

system is working correctly.

Source: Mortimore S and Wallace C (1997) HACCP: A Practical Approach.

London: Chapman & Hall, with permission.

2686 FOOD SAFETY

within preestablished limits to insure the safety of

food. A program of monitoring, verification, and

record keeping is then implemented to insure that

the system functions correctly.

0025 The Codex General Principles of Food Hygiene rec-

ommend a HACCP-based approach to enhance food

safety. This management tool is internationally recog-

nized as essential to ensure food safety and is recom-

mended by governments, industry, and consumers.

Conclusion

0026 Food safety in the early twenty-first century is an

international challenge requiring close cooperation

between countries in agreeing standards and in setting

up transnational surveillance systems. The lessons of

the past two decades are plain to those engaged in the

food industry. No longer can farmers grow just what

they want or use technical aids to farming without

taking into account the effect on the quality of the

food produced. Food will always present some risk,

and it is the task of the food industry to keep the level

of risk to the minimum, which is practicable and

technologically feasible. It should be the role of offi-

cial bodies and the food industry to use risk analysis

to determine realistic and achievable risk levels for

foodborne hazards and to base food-safety practices

on the practical application of the results of these

analyses, thus continuously improving the safety of

food and thereby lowering the disease risk. By pro-

viding effective food-control, countries not only pro-

vide public health benefits but enable themselves to

participate in international food trade with greater

confidence.

0027 The lasting solution to the problems, which have

shaken the food industry in recent years, can only be

overcome if all those involved – farmers, food proces-

sors, wholesalers, transporters, retailers, caterers,

scientists, regulators, and government – work to-

gether towards common goals.

See also: Bovine Spongiform Encephalopathy (BSE);

Contamination of Food; Escherichia coli: Occurrence

and Epidemiology of Species other than Escherichia coli;

Food and Agriculture Organization of the United

Nations; Hazard Analysis Critical Control Point;

Legislation: Codex; Listeria: Listeriosis; Quality

Assurance and Quality Control; Risk Assessment;

World Health Organization; World Trade Organization