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

Подождите немного. Документ загружается.

Throughout the 1990s, the annual number of cases in

England and Wales fluctuated, but there is an overall

increasing trend. By 1997, the number of reported

cases was over 1000.

0029 Reports from Scotland are far higher than those

from England and Wales. In part, this may be due to

its higher profile there. For example, in 1989, E. coli

O157 was one of a list of 32 informally selected

organisms which were made ‘reportable infections’

by laboratories in Scotland, in addition to the statu-

torily notifiable diseases, and this coincided with a

doubling (at least) of the recorded incidence in a

12-month period. However, during 1992 and 1993,

and against the background of heightened awareness

and increased screening, the incidence dropped to

about half of the 1991 level (2.3 per 100 000 com-

pared to 4.0 in 1991). The subsequent marked in-

crease, in 1994, also cannot be attributed to changes

in reporting rates or screening, suggesting that

Scotland does have a higher incidence than the rest

of the UK and that there are real significant variations

in the incidence of VTEC from year to year.

0030Figure 5 shows the trends in SRSV infections. Viral

gastroenteritis is much more common than these data

suggest. Most cases are not investigated through la-

boratories. When they are, diagnosis requires electron

microscopy of feces taken within the first hours of the

onset in order to have a high chance of success. Viral

gastroenteritis can be foodborne from contamination

of any food by dirty hands, or airborne contamin-

ation following vomiting or from food contaminated

from sewage such as oysters.

0031Figure 6 shows the laboratory reports of Listeria

infection, which is usually foodborne in origin.

People affected are almost exclusively the elderly,

pregnant women, or immunosuppressed patients.

The peak in cases in 1987–89 coincided with an

outbreak from pa

produced in Belgium and led to

health advice for the public and considerable work by

the food industry to address the risks from cook-

chilled foods.

The Reporting Pyramid

0032In order to gain a better appreciation of the true

incidence of gastrointestinal infections, a national

study, the Infectious Intestinal Disease (IID) in Eng-

land study, was commissioned by the Department of

Health. This provides the best data to date of the true

incidence of enteric pathogens in the country. The

objectives of the study were: first to estimate the

number, and microbiological causes, of cases of intes-

tinal infectious disease people presenting to their GPs

and from whom stool specimens are routinely sent

for laboratory examination; second, to compare

these data with the data from the national reporting

surveillance system; and third, to estimate the preva-

lence of asymptomatic infection with these agents.

200

400

600

800

1000

1200

1982 1985 1988 1991 1994 1997 2000

fig0004 Figure 4 Reports of E. coli O157 in England and Wales.

500

1000

1500

2000

2500

1986 1988 1990 1992 1994 1996 20001998

fig0005 Figure 5 Reports of SRSV infections in England and Wales.

100

150

200

250

300

1983 1986 1989 1992 1995 1998

fig0006Figure 6 Laboratory reports of Listeria infection in England and

Wales.

2668 FOOD POISONING/Statistics

Seventy practices were selected to be representative of

the socioeconomic characteristics of the area and to

represent an urban and rural mix. Cases of infectious

intestinal disease were defined as persons with loose

stools or significant vomiting lasting less than 2 weeks

in the absence of a known non-infectious cause and

preceded by a symptom-free period of 3 weeks.

Vomiting was considered significant if it occurred

more than once in the 24-h period and if it incapaci-

tated the case or was accompanied by other symptoms

such as cramps or fever. One component of the study

was to draw at random a cohort of people registered

with the 70 GPs, who were followed up for a period of

6 months. These volunteers agreed to fill out diary

cards every week and return them to the GPs. These

cards stated whether the person had suffered gastro-

intestinal (GI) illness or not. If someone developed an

illness, a stool sample was submitted.

0033 Of particular interest in this study was the estimate

of the carriage of potential pathogens in otherwise

healthy people. Ten out of 2264 controls (0.4%) were

found to be excreting Salmonella, 16 out of 2264

(0.7%) were excreting Campylobacter, but none out

of 2264 were excreting E. coli 0157. These data

confirm that VTEC is a rare infection in this country

at the moment.

0034 The IID study was able to quantify the reporting

pyramid referred to above using the cohort compon-

ent of the study. For all IIDs, it was estimated that for

every case reported to the CDSC, six patients were

investigated by routine laboratory tests, 23 presented

with GI symptoms to their GP, and there were 136

actual cases in the community. For Salmonella cases,

however, the ratios were much smaller. For every case

reported to CDSC, it was estimated that there were

2.3 cases presenting to their GP and 3.2 actual cases

in the community (see Figures 7 and 8).

Outbreak Reports

0035In addition to the reporting of individual cases, na-

tional surveillance centers receive reports of food-

borne disease outbreaks. A general outbreak is

Reported to CDSC

2.3

(1.4−4.3)

3.2

(14−12.0)

1.2

(1.0−1.7)

3.6

(2.4−5.8)

7.6

(3.6−17.4)

1.5

(1.2−2.0)

Positive by routine

laboratory investigation

Presenting to the GP

Community case

(a) (b)

Reported to CDSC

Positive by routine

laboratory investigation

Presenting to the GP

Community case

fig0007 Figure 7 Reporting pyramids for (a) Salmonella and (b) Campylobacter.

Reported to CDSC

248

(30−2,026)

1,562

(140−17,424)

1.4

(1.3−1.6)

Positive by routine

laboratory investigation

Presenting to the GP

Community case

fig0008Figure 8 Reporting pyramid for SRSV.

FOOD POISONING/Statistics 2669

defined as ‘affecting members of more than one

private residence or residence of an institution.’ Gen-

eral outbreaks are distinguished from family out-

breaks affecting members of the same private

residence only. In 1992, an improved systematic sur-

veillance system for outbreaks was developed by the

PHLS in England and Wales in response to the Rich-

mond’s Committee recommendations. The CDSC is

made aware of outbreaks from laboratory reports,

consultants in communicable disease control, envir-

onmental health officers, and others. Outbreaks are

then followed up using a standard request form that

documents basic details. Participation in the surveil-

lance scheme is entirely voluntary. Figure 9 shows the

number of outbreaks of Salmonella enteritidis PT4 in

England and Wales for 1992–97.

Enter-net

0036 An international surveillance network (Enter-net) for

human gastrointestinal infections involving 15 coun-

tries of the European Union and the USA has been set

up (see Figure 10). The aim of the network is to

conduct international surveillance of human salmon-

ellosis and verocytotoxin producing E. coli 0157.

This project began first of all with Salmonella cases

in 1994 and was funded by the EU. The network

consisted of the microbiologists in charge of the na-

tional reference laboratories in participating coun-

tries and epidemiologists responsible for national

surveillance. The minimum dataset is collected for

each isolate and reported to a central unit.

0037A dynamic database has been created, which shows

that trends in salmonellosis declined throughout

Western Europe from 1996 (see Figure 11). As a result

of this network, international outbreaks were recog-

nized. Individual countries identifying apparently

isolated incidents were linked together to identify

wider outbreaks. One of these outbreaks involved

12 cases of E. coli 0157 in Denmark, England,

Wales, Finland, and Sweden in visitors to Fuerterven-

tura in the Canary Islands.

Other Sources of Statistics

Registrations of Deaths

0038Most countries have statutory deaths registrations

systems that seek to record the causes of death.

These are of limited use in foodborne diseases, since

University/College

0 20406080100

Shop

School

Residential Institution

Pub

Private house

Other

Mobile Retailer

Hotel/Guest House

Hospital

Holiday Camp

Hall

Farm

Community

Canteen

Armed Services Camp

Restaurant/Cafe

fig0009 Figure 9 Salmonella enteritidis PT4 outbreaks in England and Wales, by place of outbreak, from 1992 to 1997.

2670 FOOD POISONING/Statistics

such diseases rarely are primary causes of death, and

even when they are contributory causes, the registra-

tion data may not systematically record this.

Hospital Admissions Data

0039Many countries have statistics on the numbers of

patients admitted or discharged from hospital by dis-

ease. Once again, these data are of very limited use in

developed countries, since foodborne disease rarely

causes admission to hospital. Even when it does, the

microbiological causes may not be recorded.

Clinical Reporting Networks

0040When a new foodborne disease emerges, special

reporting systems may be set up, such as the UK

collaborative study of childhood hemolytic uremic

syndrome (HUS), which began at the end of February

1997 and is conducted by the British Pediatric Sur-

veillance Unit (BPJU), the PHLS, Scottish Centre for

Infection and Environmental Health (SCIEH) and the

Scottish 0157 Reference Laboratory. Pediatricians

report cases of clinically diagnosed HUS directly to

the PHLS, CDSC, or SCIEH by telephone at the time

of the diagnosis and also follow this up with a stand-

ard report card to the BPSU. Clinical and epidemi-

ological data are collected directly from pediatricians.

Routine feces and serum samples are sent for stand-

ard examination and isolation and subtyping of

VTEC. During the first two years of the study, data

January

1000

2000

3000

4000

5000

6000

7000

8000

9000

1995

1996

1997

1998

February March April May June July

Number of cases

August September October November December

fig0011 Figure 11 Salmonella enteritidis cases in Western Europe from 1995 to 1998.

fig0010 Figure 10 Enter-net participating countries.

FOOD POISONING/Statistics 2671

were collected from 193 clinically confirmed cases,

and five children were reported to have died. One

hundred and eighty-six cases presented with a diar-

rheal illness, and 156 of these suffered bloody diar-

rhea. Stool and/or serum specimens were obtained

from 185 cases, and of these, 165 were shown to

have suffered VTEC infection.

See also: Campylobacter: Properties and Occurrence;

Escherichia coli: Food Poisoning; Food Poisoning:

Tracing Origins and Testing; Listeria: Properties and

Occurrence; Salmonella: Properties and Occurrence

Further Reading

Adak GK, Long SM and O’Brien SJ (2002) Trends in indi-

genous foodborne disease and deaths, England and

Wales: 1992 to 2000. Gut 51(6): 832–841.

Advisory Committee on the Microbiological Safety of Food

(2001) Second Report on Salmonella in Eggs. London:

The Stationery Office.

Berkelman RL, Stroup DF and Buehler JW (2002) Public

Health Surveillance. In: Detels J, McEwen J, Beaglehole

R and Tanaka H (eds) Oxford Textbook of Public

Health. The Scope of Public Health, Vol. 1, 4th edn.,

pp. 759–781. Oxford University Press.

Cowden JM (2002) Foodborne infectious risks: do we need

a wide system of data collection and survey? The lessons

learned from the study of infectious intestinal disease in

England. Rev Epidemiol Sante Publique 50(1): 89–92.

Fisher IS and Gill ON (2001) International surveillance

networks and principles of collaboration. European

Surveillance 6(2): 17–21.

Food Standards Agency (2000) A Report of the Study of

Infectious Intestinal Disease in England. London: The

Stationery Office.

Frost JA, Gillespie IA and O’Brien SJ (2002) Public health

implications of campylobacter outbreaks in England and

Wales, 1995–1999: epidemiological and microbiological

investigations. Epidemiology and Infection 128(2):

111–118.

Herikstad H, Yang S, Van Gilder TJ et al. (2002) A popula-

tion based estimate of the burden of diarrhoeal illness in

the United States: FoodNet, 1996–97. Epidemiology and

Infection 129(1): 9–17.

Palmer SR, Parry S, Perry D et al. (2000) The role of

outbreaks in developing food safety policy: population

based surveillance of salmonella outbreaks in Wales

1986–1998. Epidemiology and Infection 125(3): 467–

472.

Parry SM and Palmer SR (2002) E. coli. Environmental

Health Issues of VTEC 0157. Clay’s Library of Health

and the Environment. Spon Press.

Economic Implications

R B Raybourne and K M Williams, Center for Food

Safety and Applied Nutrition, Laurel, MD, USA

T Roberts, Economic Research Service, Washington,

DC, USA

Arthritis Working Group

Background

0001Foodborne diseases are caused by ingesting bacteria,

fungi, parasites, viruses or their toxic metabolites

through contaminated food or water or through

person-to-person contact. Each year, microbial

pathogens cause an estimated 76 million cases of

foodborne illness, including 5200 deaths in the USA,

according to the study by Mead et al. at the Centers

for Disease Control and Prevention (CDC). The Eco-

nomic Research Service (ERS) in the United States

Department of Agriculture and the Center for Food

Safety and Applied Nutrition (CFSAN) in the Food

and Drug Administration have estimated the annual

human illness costs for a number of causes of food-

borne illness. The ERS estimates that the total cost of

five major bacterial pathogens is $6.9 billion annu-

ally. CFSAN estimates a cost of $27.9 billion for nine

major causes of foodborne illness. In their estimates

for human foodborne illness costs, both ERS and

CFSAN include medical costs, productivity losses

from missed work, and an estimate of the value of

premature deaths. CFSAN also includes an estimate

of the cost of pain and suffering due to illness.

0002The vast majority of foodborne illnesses are classi-

fied as ‘acute.’ These are usually self-limiting and

of short duration, although the cases can range

from mild to severe. Gastrointestinal problems and

vomiting are common acute symptoms of many food-

borne illnesses. Deaths from acute foodborne illness,

while rare, are more likely to occur in the very

young (including the fetus), the elderly, or patients

with compromised immune systems (such as those

suffering from AIDS and cancer).

The other participants in the Arthritis Working Group of the Risk

Assessment Consortium (RAC) are: ERS – P Frenzen; Centers for Disease

Control and Prevention – J Sobel and S O’ Connor; and Food and Drug

Administration – A Jessup, C McGlaughlin, C Nardinelli, and R Scharff. The

RAC’s members are United States government agencies with food safety

research and management responsibilities. The RAC enhances

communication and coordination among the member agencies and promotes

the conduct of scientific research that facilitates risk assessments. The RAC’s

web page is: http://www.foodriskclearing house.umd.edu/

Risk_Assessment_Consortium.htm.

2672 FOOD POISONING/Economic Implications

0003 Complications of foodborne illness, called chronic

sequelae, can occur in any part of the body including

the joints, nervous system, kidneys, or heart. These

chronic illnesses may afflict the patients for the

remainder of their lives or result in premature death.

For example, Campylobacter infections are estimated

to be responsible for 20–40% of Guillain–Barre

syn-

drome (GBS) cases (a major cause of paralysis unre-

lated to trauma) in the USA. About 1.5% of E. coli

O157:H7 disease patients develop hemolytic uremic

syndrome (HUS), which usually involves red blood

cell destruction, kidney failure, and neurological

complications, such as seizures and strokes.

0004 We will focus on a group of chronic sequelae of

foodborne infections composed of reactive arthritis,

Reiter’s syndrome, and ankylosing spondylitis, known

collectively as the spondyloathropathies. We will use

these chronic sequelae as an in depth example of how

cost estimates are developed. These arthritis-like con-

ditions occur in the aftermath of acute infection with

several genera of Gram-negative Enterobacteriaceae

including Salmonella spp., and all involve at a min-

imum some form of joint inflammation. The reported

incidence of postacute joint inflammation can vary

widely. The degree of severity and duration of arthritis

is also highly variable. However, in the most severe

or chronic cases, the consequences may be life-long,

and there is potential to significantly increase the

economic impact beyond the initial cost of acute

foodborne illness.

Exposure to Pathogens and Associated

Costs

0005 Actions by the food industry, consumers, and the

public health sector all influence how food is pro-

duced, marketed, prepared, and consumed. These

actions influence the probability that a food item

contains pathogens. People who consume contamin-

ated food have some probability of becoming ill.

Foodborne illness generates costs that are borne by

the food industry, households whose members become

ill, and/or the public health sector (Figure 1). A full

accounting of the costs of all sectors of the economy

would include estimates for all the costs listed in

Table 1. Depending on the purpose of the study,

different cost categories can be selected, though

medical costs are those most commonly estimated.

Cost Estimates for Foodborne Disease

0006 The Centers for Disease Control and Prevention esti-

mates that out of a total of 76 million cases of food-

borne illness each year in the USA, as many as 62

million cases are of unknown origin. Thus, in 82% of

cases of foodborne illness, the disease or agent has not

been identified, and costs have not been estimated.

ERS and CFSAN have estimated costs for the most

common identifiable causes of foodborne illness. As

the pioneer of cost estimates for US foodborne illness,

ERS has estimated costs for five common foodborne

pathogens, likely to be associated with meat and

poultry. As the agency with the responsibility for

regulating the safety of the most food groups, CFSAN

has estimated the cost of almost 20 causes of food-

borne illness. Table 2 illustrates the estimated costs of

the most common foodborne illnesses that have been

studied by the two agencies.

0007For each agent or disease that causes foodborne

illness, ERS and CFSAN estimate an annual cost of

illness based on the estimated number of cases, the

expected severity of illness, and whether we expect an

acute illness to result in chronic sequelae. Estimated

costs include medical costs, lost productivity costs,

other illness-specific costs (such as special education

and residential-care costs), and an estimate of the

value of premature deaths. CFSAN also measures

quality-adjusted life year (QALY) losses to value the

pain and suffering associated with foodborne illness.

Costs have not been estimated for the vast majority of

complications associated with foodborne illnesses.

The cost estimates do, however, cover the following

chronic complications in our cost estimates: paralysis

following Campylobacter spp. infections (called the

Guillian-Barre

Syndrome), kidney failure following

E. coli O157:H7 infections (HUS), and chronic dis-

ability or impairment following congenital and new-

born infections from Listeria monocytogenes.

CFSAN estimates also include costs for arthritic com-

plications that result from illnesses due to Salmonella

spp. and Shigella spp.

0008Medical costs ERS and CFSAN use similar methods

to estimate medical costs. Disease-outcome trees are

developed for each illness to put the medical data in

perspective. The total number of cases are divided

into different levels of severity of disease:

0009cases who recover without seeking medical care,

0010cases who visit a physician and recover,

0011cases who are hospitalized and recover,

0012cases who develop chronic complications (can

overlap with an acute survivor category), and

0013those who die prematurely because of their illness.

For the different severity groups, the percentage re-

covering fully, partially recovering, or dying prema-

turely is estimated and entered into the disease

outcome tree. (The disease outcome tree approach

will be illustrated later when discussing the medical

evidence connecting foodborne illness to arthritis.)

Medical costs are estimated for physician and

FOOD POISONING/Economic Implications 2673

hospital services, supplies, medications, and special

procedures unique to treating the particular food-

borne illnesses. Such costs reflect the average number

of days or treatments of a medical service, the average

cost per service or treatment in the USA, and the

number of patients receiving such service or treat-

ment.

0014 ERS estimates of nonmedical costs For the 2000

cost estimates, ERS revised the methodology for

valuing premature deaths due to foodborne illness.

ERS previously assigned all deaths the same value

regardless of age at time of death, based on infor-

mation about the wage premiums for high-risk occu-

pations (the ‘labor market approach’). Now, the

updated cost estimates use information about the

age distribution of deaths to adjust this value to

account for age at death.

0015In essence, the labor market approach values the

economic cost of premature deaths based on the risk

premium revealed by the higher wages paid for dan-

gerous jobs. Under this approach, the value of a stat-

istical life equals $6.5 million in August 2000 dollars

after updating Viscusi’s original 1990 estimate of

$5.0 million to account for inflation. The labor

market approach assumes that risk preferences

observed in job choices are indicative of risk prefer-

ences for food safety.

0016ERS modified the labor market approach by taking

the age distribution of deaths from each pathogen

Foodborne disease, exposure, and types of costs

Food production,

marketing and preparation

Food industry

production

Household

production

Public health

regulations

Presence of a

hazard in food

Incidence of damage

Exposure to a hazard

via consumption of food

Food industry

Product recall

Plant closings

and cleanup

Product liability cost

Reduced product

demand

Insurance

administration

Households

Estimates of the value

of premature death

Medical cost

Income or productivity

loss

Pain and suffering

Leisure time cost

Child care cost

Risk aversion cost

Travel cost

Self protection costs

Home modifications

Vocational and

physical rehabilitation

Public health sector

Disease surveillance

cost

Cost of investigating

outbreak

Cost of cleanup

Public education

Implementation of

HACCP systems

Cost of damage

Monetizable

cost

Nonmonetizable

cost

fig0001 Figure 1 Foodborne disease, exposure, and types of costs.

2674 FOOD POISONING/Economic Implications

into account, in effect treating the value of life as an

annuity paid over the average US life span at an

interest rate of 3.0%. Following age adjustment, the

assumed cost of each death ranges from $8.9 million

for individuals who died before their first birthday to

$1.7 million for individuals who died at age 85 or

older.

0017Since the five microbial pathogens have different

health outcomes for different age groups, adjusting

for the age of death raises the cost of some foodborne

illnesses and lowers the cost of others. For example,

the annual cost of foodborne illness caused by Sal-

monella spp. decreases to $2.2 billion, because over

two-thirds of the deaths from salmonellosis occur

among people over 65 years of age. Adjusting food-

borne illness costs for E. coli O157:H7 by age at time

of death increases the estimates by $200 million, be-

cause most deaths are children under the age of five.

tbl0001 Table 1 Social costs of foodborne illness

Costs to individuals/households

Human illness costs:

Medical costs:

Physician visits

Laboratory costs

Hospitalization or nursing home

Drugs and other medications

Ambulance or other travel costs

Income or productivity loss for:

Ill person or fatal case

Caregiver for ill person

Other illness costs:

Travel costs to visit ill person

Home modifications

Vocational/physical rehabilitation

Child care costs

Special educational programs

Institutional care

Lost leisure time

Psychological costs:

Pain and other psychological costs

Risk aversion

Averting behavior costs:

Extra cleaning/cooking time costs

Extra cost of refrigerator, freezer, etc.

Flavor changes from traditional recipes (especially

meat, milk, egg dishes)

Increased food cost when more expensive but safer

foods are purchased

Altruism (willingness to pay for others to avoid illness)

Industry costs

Impact of pathogens on animal production costs:

Morbidity and mortality of animals on farms

Reduced growth rate/feed efficiency and increased time to

market

Costs of disposal of contaminated animals on farm and at

slaughterhouse

Increased trimming or reworking at slaughterhouse and

processing plant

Lost productivity due to illness among workers caused by

handling contaminated animals or products

Increased meat product spoilage due to pathogen

contamination

Control costs for pathogens at all links in the food chain

New farm practices (age segregated housing, sterilized

feed, etc.)

Altered animal transport and marketing patterns (animal

identification systems, feeding/watering)

New slaughterhouse procedures (hide wash, knife

sterilization, carcass sterilizing)

New processing procedures (pathogen tests, contract

purchasing requirements)

Altered product transport (increased use of time/

temperature indicators)

New wholesale/retail practices (pathogen tests, employee

training, procedures)

Risk assessment modelling by industry for all links in the

food chain

Price incentives for pathogen-reduced product at each link in

the food chain

Outbreak costs:

Herd slaughter/product recall

Plant closings and cleanup

Regulatory fines

Product liability suits from consumers and other firms

Reduced product demand because of outbreak:

Generic animal product – all firms affected

Reduction for specific firm at wholesale or retail level

Increased advertising or consumer assurances following

outbreak

Impact of outbreaks on tourism industry

Regulatory and public health sector costs for foodborne pathogens

Disease surveillance costs to:

Monitor incidence/severity of human disease by foodborne

pathogens

Monitor pathogen incidence in the food chain

Develop integrated database from farm to table for

foodborne pathogens

Research to:

Identify new foodborne pathogens for acute and chronic

human illnesses

Establish high-risk products and production and

consumption practices

Identify which consumers are at high-risk for which

pathogens

Develop cheaper and faster pathogen tests

Risk assessment modelling for all links in the food chain

Outbreak costs:

Costs of investigating outbreak

Testing to contain an outbreak (for example, serum testing

and administration of IG in persons exposed to

Hepatitis A)

Costs of cleanup

Legal suits to enforce regulations that may have been

violated

Other considerations

Distributional effects in different regions, industries, etc.

Equity considerations, such as special concern for children

Some industry costs may fall with better pathogen control, such as

reduced product spoilage, possible increases in product shelf-life, and

extended shelf-life permitting shipment to more distant markets or

lowering shipment costs to nearby markets.

In adding up costs, care must be taken to assure that product liability

costs to firms are not already counted in the estimated pain and suffering

cost to individuals. However, the legal and court expenses incurred by all

parties are social costs.

FOOD POISONING/Economic Implications 2675

0018 One difficult issue is assigning a value to the

productivity losses associated with individuals who

become ill and are unable to return to work, or for

those illnesses that result in a lifetime of disability

(such as prenatal exposure to Listeria monocyto-

genes). ERS currently measures the productivity

losses due to nonfatal foodborne illness by the value

of forgone or lost wages, regardless of whether the

lost wages involved a few days missed from work or a

permanent disability that prevented an individual

from returning to work. Using the value of lost

wages for cases resulting in disability probably under-

states an individual’s willingness to pay to avoid dis-

ability because it does not account for the value

placed on avoiding pain and suffering. The willing-

ness to pay measure derived from labor market stud-

ies that ERS uses to value a premature death is not an

appropriate measure of willingness to pay to avoid

disability because it measures the higher wages paid

to workers to accept a higher risk of premature death,

not disability. Methods have been suggested to adjust

willingness to pay to reduce the risk of premature

death downward to estimate willingness to pay to

avoid disability, such as the approach based on meas-

uring QALYs. As yet, there is no consensus among

economists about how to use these methods to value

willingness to pay to avoid the disability, pain, and

suffering associated with foodborne illnesses. ERS’s

conservative estimates of the annual costs due to

foodborne illness (particularly the chronic conditions

associated with Campylobacter spp.) would be

substantially increased if willingness to pay to avoid

disability, pain, and suffering were also taken into

account.

0019CFSAN estimates of nonmedical costs Medical

costs are a small fraction of the total social cost of

foodborne illness. In addition to the cost of medical

care, economists are concerned with the loss of prod-

uctivity and the loss of well-being (through pain and

suffering) that accompany foodborne illness. CFSAN

routinely estimates these nonmedical costs of food-

borne illness.

0020CFSAN was one of the first government agencies to

use the labor market approach to estimating the eco-

nomic cost of premature death. Economists have long

favored this method, because it is based on observed

tradeoffs that individuals make between dollars and

risk. An unadjusted figure of $5.0 million per prema-

ture death is generally used in CFSAN analyses. As

noted above, this value does not take into account

economic changes that have occurred since 1990.

tbl0002 Table 2 Economic costs of foodborne illness in the USA: estimated annual costs due to selected US foodborne pathogens, 2000

Pathogen Estimatedannual foodborne illnesses

ERS costs

CFSANcosts

Cases Hospitalizations Deaths Billion 200 0 dollars

Bacterial infections

Campylobacter spp. 1 963 141 10 539 99 1.2 –

Salmonella 1 341 873 15 608 553 2.4 17.2

Clostridium perfringens 248 520 41 7 – 0.2

Shigella 89 648 1 246 14 – 0.3

E. coli O157:H7 62 458 1 843 52 0.7 2.2

E. coli, non-O157 STEC

31 229 921 26 0.3 –

Listeria monocytogenes 2 493 2 298 499 2.3 2.5

Parasitic infections

Giardia lamblia 200 000 500 1 – 0.2

Cryptosporidium parvum 30 000 199 7 – 0.1

Viral infections

Norwalk-like viruses 9 200 000 20 000 124 – 5.3

Hepatitis A 4 170 90 4 – 0.1

To t a l s

ERS total 3 401 194 31 209 1 229 6.9 –

CFSAN total 11 179 162 41 825 1 261 – 28.1

As these new estimates of foodborne illness costs are based on new data and improved methodologies for valuing these costs, the estimates presented

here are not directly comparable with earlier ERS and CFSAN estimates of the costs of foodborne disease.

Data from the Centers for Disease Control and Prevention published in Mead PS, Slutsker L, Dietz V et al. (1999) Food-related illness and death in the

United States. Emerging Infectious Diseases 5: 607–625.

The ERS total estimated costs include specific chronic complications in the case of Campylobacter (Guillain–Barre

syndrome), E. coli O157:H7 (hemolytic

uremic syndrome), and Listeria monocytogenes (congenital and newborn infections resulting in chronic disability or impairment). Estimated costs for

Listeria monocytogenes exclude less serious cases that do not require hospitalization.

The CFSAN total estimated costs include specific chronic complications in the case of Salmonella (arthritic conditions), Shigella (arthritic conditions and

hemolytic uremic syndrome), and E. coli O157:H7 (hemolytic uremic syndrome). CFSAN uses QALYs to estimate the value of pain and suffereing.

STEC stands for Shiga toxin-producing E. coli .

2676 FOOD POISONING/Economic Implications

Economists at CFSAN are currently working on an

adjustment factor based on both nominal and real

changes in society’s valuation of health. As a result,

the figures that are reported in Table 2 are likely to

be an underestimate of the true value of foodborne

illness.

0021 The value of productivity losses and pain and

suffering are measured by placing values on quality

adjusted life years (QALYs). This method uses a two-

step procedure for valuing health losses. In the first

step, the effect of a condition on health is estimated to

be between zero (well-being in the full health state)

and one (death). For example, a QALY loss of 0.14

for arthritis means that for every day of suffering with

arthritis, the affected individual has a level of well-

being 14% lower than they would have had in the

absence of arthritis. In the next step, the value of a

QALY is estimated and multiplied by expected QALY

losses to calculate the value of health losses. QALYs

are designed to measure the loss of well-being both

from symptoms and from activity limitation. Conse-

quently, both pain and suffering and productivity

values are captured by this measure.

0022 ERS and CFSAN use different methods of measur-

ing the nonmedical costs of illness. The net effect of

the differing approaches is that CFSAN estimates for

the value of foodborne illness include a value for pain

and suffering, whereas ERS estimates do not. The

inclusion of a value for pain and suffering is demon-

strated in Table 2. CFSAN estimates of the cost of

foodborne illness are uniformly larger than estimates

made by ERS. In the case of E. coli O157, CFSAN

based its medical cost estimates on estimates calcu-

lated by ERS. Nonetheless, the inclusion of QALY

losses boosts CFSAN estimates of economic loss due

to E. coli O157 to $2.2 billion, as opposed to $0.7

billion for ERS.

Chronic Sequelae of Foodborne Illness

0023A listing of the association of foodborne pathogens

with chronic sequelae (Table 3) illustrates that many

or most of these agents are associated with some type

of complication. Of particular interest is the fact that

several bacterial pathogens are associated with arth-

ritic sequelae. In addition to identifying the various

disease conditions, determining the economic impact

of chronic sequelae associated with foodborne illness

requires not only a systematic method of estimating

costs but also some understanding of the clinical

aspects of these entities. This is particularly

true with respect to the degree of variation that may

tbl0003 Table 3 Chronic complications associated with foodborne pathogens

Bacterialand parasiticinfection transmittedby foods Complications/sequelae

Bacterial infections

Aeromonas hydrophila enteritis Bronchopneumonia, cholecystitis

Brucellosis Aortitis, epididymo-orchitis, meningitis, pericarditis, spondylitis

Campylobacteriosis Arthritis, carditis, cholecystitis, colitis, endocarditis, erythema nodosum, Guillain–

Barre

syndrome, hemolytic uremic syndrome, meningitis, pancreatitis,

septicemia

Escherichia coli (EHEC-types) enteritis Erythema nodosum, hemolytic uremic syndrome, seronegative arthropathy,

thrombotic thrombocytopenic purpura

Q-fever Endocarditis, granulomatous hepatitis

Salmonellosis Aortitis, cholecystitis, colitis, endocarditis, epididymo-orchitis, meningitis,

myocarditis, osteomyelitis, pancreatitis, Reiter’s disease, rheumatoid

syndromes, septicemia, splenic abscesses, thyroiditis, septic arthritis

(sickle-cell anemic persons)

Shigellosis Erythema nodosum, hemolytic-uremic syndrome, peripheral neuropathy,

pneumonia, Reiter’s disease, septicemia, splenic abscesses, synovitis

Vibrio parahaemolyticus enteritis Septicemia

Yersiniosis Arthritis, cholangitis, erythema nodosum, liver and splenic abscesses,

lymphadenitis, pneumonia, pyomyositis, Reiter’s disease, septicemia,

spondylitis, Still’s disease

Parasitic infections

Cryptosporidiosis

Severe diarrhea, prolonged and sometimes fatal

Giardiasis

Cholangitis, dystrophy, joint symptoms, lymphoidal hyperplasia

Taeniasis Arthritis, cysticercosis (T. solium)

Toxoplasmosis Encephalitis and other central nervous system diseases, pancarditis, polymyositis

Trichinosis Cardiac dysfunction, neurologic sequelae

Suspected of being foodborne or waterborne.

Waterborne.

Source: Council for Agricultural Science and Technology (CAST) (1994) Foodborne Pathogens: Risks and Consequences. Ames, IA: CAST.

FOOD POISONING/Economic Implications 2677