Incidence of foodborne illness

Foodborne pathogens, chemicals and physical contaminants

While there are some foodborne pathogens, such asSalmonella, which have a high profile in the community, there is a large range of other agents which cause food poisoning. The majority of these are listed in Figure 2.

Figure 2 Foodborne pathogens and chemicals of public health importance ⁹

Pathogenic bacteria

Bacillus cereus
Brucella spp
Campylobacter jejuni
Clostridium botulinum
Clostridium perfringens
Escherichia coli spp
E. coli  enterotoxigenic (ETEC)
E. coli  enteropathogenic (EPEC)
E. coli 0157:H7 (EHEC)
E. coli  enteroinvasive (EIEC)
Listeria monocytogenes
Mycobacterium bovis
Salmonella typhi and paratyphi
Salmonella (non-typhi) spp
Shigella spp
Staphylococcus aureus
Vibrio cholerae 01
Vibrio parahaemolyticus
Vibrio vulnificus
Yersinia enterocolitica

Viruses

Hepatitis A virus
Hepatitis E virus
Norwalk virus group
Poliovirus
Rotavirus

Trematodes

Clonorchis sinensis
Fasciola hepatica
Fasciolopsis buski
Opisthorchis felineus
Opisthorchis viverrini
Paragonimus westermani

Protozoa

Cryptosporidium spp
Entamoeba histolytica
Giardia lamblia
Toxoplasma gondii

Cestodes

Diphyllobothrium spp
Echinococcus spp
Taenia solium
Taenia saginatum

Nematodes

Anisakis spp
Ascaris lumbricoides
Trichinella spiralis
Richuris trichiura

Natural toxins

Marine biotoxins
Ciguatera poisoning
Shellfish toxins
Scombroid poisoning/histamine
Tetrodotoxin (pufferfish)
Mushroom toxins
Mycotoxins (eg aflatoxins)
Plant toxicants
Pyrrolizidine alkaloids
Phytohaemagglutinin (red kidney beans)
Grayanotoxin (honey intoxication)

Chemicals

Pesticides
Toxic metals (cadmium, copper, lead, mercury, tin, zinc)
Polychlorinated biphenyls
Radionuclides

Chemical contaminants

As listed in Figure 2, chemical contaminants in food include natural toxicants, pesticide residues and environmental contaminants such as mercury, cadmium, radionuclides and polychlorinated biphenyls. However, there is also a range of nutrients and food additives (such as Vitamin A, zinc, sodium nitrite) which are potentially toxic. Although safe levels of such chemicals are set out in theFood Standards Code, failure to observe the stated levels can result in food poisoning.

Environmental contaminants cannot be regarded in the same way as food additives since they are not intentionally added to food. Their levels cannot necessarily be adjusted to provide the same margin of safety which that can be achieved with a food additive. Thus, assessments are conducted on the basis of the lowest achievable level in commodities and foodstuffs. Maximum permitted concentrations (MPCs) are generally established on the basis of the lowest achievable levels of contamination consistent with protection of public health and safety.

Some naturally-occurring chemicals in plants or marine organisms can have acute health consequences as well as potential long-term effects. Assessing the acute effects of many naturally-occurring chemicals has been through a trial-and-error approach throughout human history. The most serious chemical contaminants are well documented and, in most cases, are not used as food. There are, however, many which are consumed at a low level throughout life and a better understanding of the risk associated with their consumption is needed. These include both chemicals occurring naturally in food, eg, glycoalkaloids in potatoes, and those formed during cooking or food processing, eg, heterocyclic amines.

Physical contaminants

Physical contaminants are generally considered to be extraneous matter or foreign objects and include physical matter not normally found in food which may cause illness (including psychological trauma) or injury to an individual (Corlett 1991). As such, fish bones can be considered a physical hazard when they are present in a boneless fillet, but not when they are in a whole fish or a cutlet.

Foreign matter which most commonly occurs in food includes:

Material	Likely sources
Glass	light fixtures, bottles, jars, gauge covers
Wood	pallets, boxes, buildings
Metal	machinery, wire
Stones	dried fruits, vegetables, grains
Rubber	seals, packaging
Plastic	packaging
Bone	vegetables, nuts, grains,
Insects and rodents	all foods
Faecal pellets	all foods
Personal effects	staff.

Any foreign material can be a safety hazard if it can choke the consumer. This is particularly important for foods which may be consumed by children. Physical hazards can also cause lacerations, damage teeth and puncture the gastro-intestinal tract. There is a nil tolerance of most physical hazards in food, either on the basis of safety (such as glass fragments and wire) or suitability (insect parts). In Australia, the dried fruit industry has a nil tolerance for contamination with  spiky weed seeds because of the potential damage if ingested.

For a number of primary products, some physical contamination may be tolerated provided that further processing, through sieving or the use of flotation tanks for example, will remove offending material.

Physical contamination of food was the reason for over 40 per cent of food recalls in Australia during 1997 and there can be an extreme financial liability to the offending company.

In 1988, the New South Wales Supreme Court awarded almost $500,000 in damages to a consumer who had eaten a toothpick in a can of beans and as a consequence suffered a perforated bowel and peritonitis. ¹⁰

Physical contamination can introduce both chemical and microbiological hazards into the product. This can pose a significant risk when the contamination occurs after all processing steps which would normally control these hazards.

Summary

Chemical, physical and microbiological hazards can affect the safety of food. These hazards need to be countered by the food industry to ensure that only safe food is sold. Preventative measures such as food safety programs based on Hazard Analysis and Critical Control Points (HACCP) can help fo od businesse s m inimise t he potentia l for contamination.

Worldwide increases in foodborne illness

Surveillance and monitoring by a number of countries indicates that foodborne illness is increasing around the world (WHO 1997; United States General Accounting Office 1996; Henson 1996; Sockett 1993). For example, in the United Kingdom formally reported cases of food poisoning increased from 12,763 in 1980 to 72,790 in 1993 (Henson 1996). In Australia, notification rates for both campylobacteriosis and salmonellosis (common foodborne illnesses) have continued to increase annually (Crerar et al 1996).

Salmonellainfections have been steadily increasing in Australia since 1992 as described in Figure 3.

Figure 3 Second quarter case rates per 100,000 ofSalmonellainfection (National Enteric Pathogens Surveillance Scheme 1998)

Period	ACT	NSW	Vic	Qld	SA	WA	Tas	NT	Total
2nd quarter 1998	5.7	6.7	5.0	25.5	13.5	10.8	4.0	68.9	1,959
2nd quarter 1997	4.7	6.0	7.3	13.7	7.2	11.3	4.5	44.6	1,527
2nd quarter 1996	4.0	3.6	4.4	18.4	5.7	9.3	6.8	54.7	1,356
2nd quarter 1995	5.4	4.5	4.7	13.3	8.2	9.5	6.1	52.3	1,307
2nd quarter 1994	5.1	4.7	4.7	14.2	6.0	10.4	5.5	52.1	1,305
2nd quarter 1993	4.0	4.7	2.6	13.3	4.6	16.1	5.5	56.8	1,126
2nd quarter 1992	2.8	3.9	4.9	15.7	6.8	11.4	8.0	63.9	1,208

Although there is an increasing incidence of foodborne illness, only a small proportion of food poisoning cases are formally reported (Hennessey et al 1996; Crerar et al 1996; Kaferstein 1997; United States General Accounting Office 1996).

A number of factors are responsible for the increasing risk of foodborne illness both in Australia and internationally. These include:

" changes to human demographics;

" changes to the food supply including new animal husbandry practices;

" new and emerging pathogens;

" improved methods of detection; and

" consumer preferences.

Changes to human demographics

In developed nations, the proportion of the population with a heightened susceptibility to severe foodborne infections is increasing (Altekruse and Swerdlow 1996; Desmarchelier 1996; Gerba et al 1996, United States General Accounting Office 1996). Those at-risk include the elderly, the very young, pregnant women and immuno-suppressed people including cancer and transplant patients.

In the 20th century, the median age of populations in developed countries has increased, and continues to do so. In addition, advances in medical technology have extended the life expectancy of people with chronic illnesses. There is an increasing proportion of immuno-suppressed people in the community as a result of the use of immuno-suppressive drugs and diseases such as AIDS.

In the United States, 20 per cent of the population is considered to be in this at-risk group (Gerba et al 1996). This is expected to increase significantly early in the next century.

These people are at increased risk because a much smaller infective dose of foodborne microorganisms can cause illness. The severity of an illness can often be much greater than for those people not in the at-risk group (Smith 1998). For example, the case fatality rate from outbreaks of foodborne illness is 10 times higher in nursing homes than in the general population (Gerba et al 1996). Of the 496 people poisoned by eating food contaminated withEscherichia coli0157:H7 in Scotland in 1996, the 18 who died were aged at least 68 years (Pennington Group 1997). Similarly, pregnant women suffer from a case mortality ratio from hepatitis infections 10 times higher than the general population during waterborne disease outbreaks (Gerba et al 1996).

Changes in food production

In developed countries such as Australia, there have been a number of structural changes in the food supply system. These include intense farming practices, more extensive food distribution systems and the increasing move towards minimally processed food. If potential hazards associated with these practices are not recognised and controlled, they can contribute to an increased risk of foodborne illness.

Animal husbandry

Modern intensive animal husbandry techniques, introduced to maximise production, have led to the emergence of zoonoticSalmonellaserovars andCampylobacterspp as the most often reported causes of foodborne illness in Australia (Crerar et al 1996). The main reservoirs of these bacteria are the gastro-intestinal tracts of animals and birds both domestic and wild (Baird-Parker 1990; Mulder 1996).

Transmission of pathogenic bacteria, introduced onto farms through contaminated feed, new stock and wildlife, is increased by crowding animals together (Murray et al 1995). Non-typhoidalSalmonellaserovars are now endemic in meat and poultry animals (Desmarchelier 1996; Troutt and Osburn 1997). In addition, the conditions and stress associated with transporting animals to slaughter and dietary changes prior to slaughter can increase carriage rates and shedding ofSalmonella(Baird-Parker 1990).

Changes to animal husbandry through feeding sheep offal to cows led to the outbreak of bovine spongiform encephalopathy (BSE) in Britain and Europe. Ingestion of beef from animals infected with BSE is believed to cause Creutzfeldt-Jakob disease in humans.

The incorrect therapeutic use of anti-microbial agents in human and animal populations creates conditions that favour survival of resistant bacteria. Resistant strains ofSalmonella typhimurium(such as DT104) have become the predominantSalmonellaserovar in the United Kingdom (Threlfall 1998).

Contamination of products

Safe disposal of manure from large-scale animal and bird production facilities is a growing problem. Without proper disposal, manure may serve as a reservoir for pathogens. These may find their way into the food chain when animal and bird production facilities are located close to horticultural operations or inappropriately treated manure is used to fertilise crops. Contaminated crops including tomatoes, melons, alfalfa sprouts, raspberries and juices have been the source of recent outbreaks of foodborne illness in the United States (United States General Accounting Office 1996). In Australia, faecal contamination of oranges in a packing house was the source of aSalmonellaoutbreak from orange juice which affected over 480 people.

Many hazards present in the food chain are introduced at the animal production stage and are spread during slaughter and processing (Ahl and Buntain 1997). Faecal contamination of meat during slaughtering remains an ongoing problem because many farm animals carry pathogens such asSalmonellaand enterohaemorrhagicEscherichia coli(EHEC) in their gut (Vanderlinde et al 1998).

Extensive food distribution systems

There is a trend, both in Australia and internationally towards wider geographic distribution of food from large centralised processors. This increases the potential for rapid, widespread distribution of contaminated product.

In 1985, 185,000 people in Chicago contracted salmonellosis from contaminated pasteurised milk distributed by one dairy plant. ¹¹

In one outbreak in 1994, over 200,000 people across the United States contracted salmonellosis from contaminated icecream made in Minnesota and distributed nationally. ¹²

Industry consolidation and mass distribution of foods increase the risk that breakdowns in food safety will cause further widespread outbreaks of foodborne disease. Outbreaks can be difficult to detect as the increase in cases may not be apparent against the background level of illness (Tauxe 1997). Often it is a fortuitous concentration of cases in one location or the identification of an unusual strain that leads to the identification of a widespread outbreak.

Changes in food technology

With changes in food technology have come a range of new food products such as fresh cut salads, sous-vide, and other 'ready to eat' and pre-prepared foods. Foods such as sous-vide (precooked dishes kept under vacuum) can harbour less common pathogens such asClostridium botulinum.Products that have no bacterial 'kill step' before eating can cause foodborne illness if subject to temperature abuse.

The increased use of refrigeration to prolong the shelf life of food has contributed to the emergence ofListeria monocytogenesas a food hazard (Smith and Fratamico 1995).L. monocytogenesis a hazard for pregnant women and other previously discussed 'at-risk' populations. Unlike most foodborne pathogens, it can grow at refrigeration temperatures and become a health risk.

A multi-state outbreak of listeriosis from contaminated hot dogs and deli meats in the United States, resulted in six deaths, two miscarriages and more than 50 cases of illness between August 1998 and January 1999. ¹³

Horticultural products as vectors for pathogens

Despite apple juice being implicated in an outbreak of typhoid fever as early as 1922 (Paquet 1923 cited by Parish 1998), the risk of foodborne illness from horticultural crops has generally been considered to be low. However this risk perception has changed more recently with outbreaks of enterohaemorrhagicEscherichia coliassociated with lettuce (Wang et al 1996), cantaloupe and watermelon (del Rosario and Beuchat 1995), apple juice (Centers for Disease Control and Prevention 1997) and bean sprouts. There have also been major outbreaks as a result of contamination of raspberries with the parasiteCyclospora cayetanensis(Herwaldt et al 1997) and with Hepatitis A (Reid and Robinson 1987).

Contaminated orange juice caused an outbreak of salmonellosis in South Australia in March 1999 which affected over 480 people. A number of other outbreaks associated with orange juice are cited by Parish (1997) and concern in the United States led the United States Food and Drug Administration (FDA) to require fruit juice processors to implement HACCP-based food safety programs to minimise the potential for contamination. In addition, if the juice is not pasteurised the FDA requires a warning label on the product (Lewis 1998).

In the United States, unpasteurised fruit juice must be labelled: 'WARNING: This product has not been pasteurized and, therefore, may contain harmful bacteria that can cause serious illness in children, the elderly and persons with weakened immune systems.'

The potential for further outbreaks of foodborne illness from horticultural crops is great. Beuchat (1995) lists over 30 horticultural crops from which pathogens includingSalmonella, E. coli0157, Shigella, Yersinia enterocolitica, Listeria monocytogenesandCampylobacter jejunihave been isolated.

Changes in eating patterns

Previously unrecognised microbial foodborne hazards have emerged as a result of changes in food consumption (Collins 1997). Fresh fruit and vegetable consumption has increased in the United States by 50 per cent from 1974 to 1994. Such produce is susceptible to faecal contamination and, when consumed without a bacterial 'kill step' such as cooking, can cause foodborne illness. There have been a number of significant outbreaks of foodborne illness associated with fresh fruit and vegetables (United States General Accounting Office 1996).

The time between processing and consumption of food is increasing. This increases the opportunity for contamination and time/temperature abuse of the product and hence the risk of foodborne illness.

The percentage of spending on food eaten away from home has increased in recent decades. Fast food restaurants and salad bars, rare 50 years ago, are today a primary source of food consumption for many Australians. In Australia, it is estimated that the number of food service outlets has grown 57 per cent with Australians spending 30 per cent of their food budget on take away food and dining out (BIS Shrapnel 1997). In Australia, New Zealand, the United States and the United Kingdom, around 60 per cent to 80 per cent of foodborne illness arises from the food service sector (Crerar et al 1996; Auckland Healthcare 1998; Bryan 1980; Cowden at al 1995; Todd 1985).

Around sixty to eighty per cent of foodborne illness arises from the food service industry.

Studies in Wales have shown that a history of eating out in the week before the onset of gastroenteritis was associated with a significantly increased yield ofSalmonellaandCampylobacter(Palmer et al 1996). Consequently, with the increasing trend towards take away food and dining out, there is the likelihood of increasing numbers of people contracting foodborne illness.

'Health consciousness' by consumers has resulted in preferences for fresh minimally processed foods with lowered salt and fat contents. However, processes such as salting, acidifying and canning were specifically designed to ensure safety and to prolong the product shelf life (Desmarchelier 1996).

International travel

International travel has increased dramatically during the 20th century. Five million international tourist arrivals were reported worldwide in 1950 and this number is expected to increase to 937 million by 2010. Travellers may become infected with foodborne pathogens uncommon in their home country thus complicating diagnosis and treatment should symptoms present after their return. There are significant economic and trade implications in the movement of foodborne illnesses through international travel (Cartwright and Chahed 1997).

New and emerging pathogens

Three of the four most significant foodborne pathogens in the United States,Campylobacter,Listeriaand enterohaemorrhagicE. coli(EHEC), were unrecognised as causes of foodborne illness 20 years ago (United States General Accounting Office 1996).

Strains of EHEC have caused serious outbreaks of foodborne illness throughout the world, including Australia. A virulent strain,E. coli0157:H7 was first identified as a cause of human illness in 1982 following two outbreaks of bloody diarrhoea in the United States. An outbreak ofE. coli0157:H7 from contaminated meat in Scotland in 1996, affected 496 people and killed 18 (Pennington Group 1997).

In 1995, a different strain of EHEC,E. coli0111, was responsible for a major outbreak of foodborne disease in South Australia and the death of one child. The 1995 South Australian Coroner s Report determined that approximately 150 people contracted illness from contaminated mettwurst. Twenty-three children were diagnosed as suffering from haemolytic uraemic syndrome (HUS), a condition which, if left untreated, will result in a large percentage of patients dying from kidney failure (Chivell 1995).

Food linked to cases of EHEC poisoning include mettwurst, radish sprouts, milk, cheese, apple cider, orange juice, lettuce, hamburger patties and raw meat. Some EHEC strains can tolerate low pH environments and are better able to survive the acid in the stomach. This means an infectious dose can be as low as a single bacterium (Chivell 1995). In addition, they can survive in high acid foods, like fresh apple juice, yoghurt and fermented meats (Doyle 1991), which were not previously considered as high risk sources of such infection.

Campylobacter jejuniwas first identified as a human pathogen in 1973 and is now the most commonly reported bacterial cause of foodborne illness in the United States (Altekruse et al 1999). Unlike other foodborne pathogens,C. jejuniis often associated with sporadic infections and can lead to long-term effects including reactive arthritis and Guillain-Barré syndrome. Handling raw poultry and eating undercooked poultry have been identified as major risk factors for contracting campylobacteriosis.

Several serovars ofSalmonellahave emerged as serious foodborne pathogens in the last few years. In the 1980s,Salmonella enteritidisbecame a major foodborne pathogen in the northern hemisphere (Smith and Fratamico 1995). The ability of this serovar to infect the ovaries and oviduct of hens permits its transmission into developing eggs. In Canada during 1988 92Salmonella enteritidisaccounted for the largest number of foodborne outbreaks, cases and deaths (Bean et al 1996).

Salmonella typhimuriumDT104 which is multi-drug resistant has become the dominantSalmonellaserovar in the United Kingdom during the last few years (Threlfall 1998). As well as being highly virulent, it too can survive low pH and be infectious in very low numbers.

Other pathogens which have more recently been recognised as significant causes of human illness areListeriamonocytogenes, Arcobacter butzleri, Helicobacter pylori, Cryptosporidium parvumandCyclospora(Meng and Doyle 1997).

Recognition of factors affecting the virulence and survival of pathogens

As well as bacterial strains developing acid tolerance as discussed above, it is also recognised that bacterial survival can be enhanced by the nature of the food. Foods high in fat such as icecream, chocolate and peanut butter can protect bacterial cells from stomach acid. This dramatically reduces the number of bacteria required to form an infectious dose. Studies in the United States following the 1994 outbreak ofSalmonella enteritidisin contaminated icecream have shown that ingesting just 28 of the bacteria was sufficient to cause food poisoning (Vought and Tatini 1998). In an Australian outbreak in 1996, a concentration of 20 cells ofSalmonella mbandakain 100 grams of peanut butter was sufficient to cause food poisoning.

Stomach acidity is an important barrier in preventing the survival and growth of many bacterial pathogens. The increasing use of medications such as antacids that reduce the acidity of the stomach increases the likelihood that pathogens will survive and cause food poisoning.

Improved methods of detection

Scientific advances have vastly improved the ability of laboratories to detect and isolate pathogens. The advent of molecular biology techniques has made it possible to characterise EHEC strains such asE. coli0157:H7 andE. coli0111. Electron microscopy is important for identifying foodborne viral pathogens.

New techniques to concentrate very low numbers of pathogens to levels where diagnostic techniques can be applied has been essential in isolating a number of waterborne pathogens such asCryptosporidiumandGiardia.

Coordination of national surveillance systems and use of pulsed-field electronic surveillance networks has also led to faster identification of pathogens and clusters that would otherwise have gone unreported.

Under-reporting of foodborne illness

Background

Reliable data are not available on the actual incidence of foodborne illness in Australia. In fact, it is estimated that less than 1 per cent of cases is captured in existing notification schemes in Australia and overseas (Hennessey at al 1996; Crerar et al 1996; Kaferstein 1997; United States General Accounting Office 1996; Todd 1989).

There are many reasons for this including:

" the short duration of many foodborne illnesses which means that those affected are unlikely to visit a doctor or report the incident to their local health authority. A study in Wales indicated that only one in 26 people who suffered from an acute episode of gastroenteritis consulted a general practitioner (Palmer et al 1996);

" outbreaks are far more likely to be reported and investigated than sporadic cases even though there is evidence that sporadic cases cause far more illness than do recognised outbreaks (Veitch and Hogg 1997);

" the difficulty in determining the cause of a reported foodborne illness due to the absence of either stool or food samples;

" when tested the person affected may no longer be excreting the causative pathogen even though they still exhibit the symptoms of foodborne illness;

" the very low infective dose (less than one organism per gram) associated with a number of pathogens makes it difficult to identify the causative organism;

" difficulty in detecting viruses. While studies have shown viruses are responsible for around 30 per cent of cases of foodborne illness (Crerar et al 1996; Ryan 1997; Djuretic et al 1996), their detection is very difficult. This is because specimens have to be collected soon after disease onset and specialised diagnostic techniques are required (Cliver 1997; Crerar et al 1996). Such techniques are not available in many States (Crerar et al 1996).

" not all foodborne illnesses are notifiable to the National Notifiable Diseases Surveillance System (NNDSS) which collects data from all States and Territories. Non-notifiable diseases include those caused byBacillus cereus,Clostridium perfringens,E. coliandStaphylococcusspp; and

" many local and State and Territory health authorities are not collating data on the number of actual or suspected foodborne illnesses reported and investigated.

FoodNet Active surveillance in the United States

The Foodborne Diseases Active Surveillance Network (FoodNet) was established in the United States in 1995 to provide better data on the incidence of foodborne illness (United States Department of Health and Human Services 1998). FoodNet undertakes active surveillance for foodborne diseases and related epidemiological studies and currently covers 6 per cent of the United States population.

The FoodNet Report for 1997 indicated:

" 360 million cases of diarrhoeal illness occur per year and there were 28 million medical consultations;

"Campylobacterwas the most frequent cause of illness;

"E. coli0157:H7 infections were more common in northern states and that undercooked minced beef was the principal source of infection; and

" hospitalisation associated with foodborne illness was an important public health burden withListeriainfection resulting in the highest rate of hospitalisation and causing nearly half of the reported deaths.

An indication of the depth of under-reporting which occurs in the absence of a robust monitoring and surveillance strategy, is provided in Figure 4. This figure, prepared by FoodNet shows the burden of foodborne disease, usingSalmonellaas an example, and typifies the under-reporting by conventional reporting methods.

Figure 4 Under-reporting of foodborne salmonellosis in the United States

2,090 cases reported	Active surveillance
2,090 cases tested	Laboratory survey
10,460 cases sought care	Physician and population survey
70,150 cases in FoodNet	Population survey
1,400,000 cases in US

Until Australia implements active surveillance schemes such as FoodNet, the real number of cases of foodborne illness will be difficult to estimate.

Foodborne illness in Australia

Australia is by no means exempt from foodborne illness and its associated costs. Some examples include:

" the 1991 Norwalk virus outbreak from contaminated orange juice that affected over 4,000 people across Australia including many air travellers;

" the 1995 outbreak in South Australia from mettwurst contaminated withE. coli0111 that caused serious illness in 23 children and the death of one;

" the 1996Salmonella mbandakaoutbreak due to contaminated peanut butter manufactured in Victoria which affected over 60 people including many children;

" the 1996 QueenslandSalmonella heidelbergoutbreak due to a contaminated sauce on a Qantas flight which affected over 500 people;

" the 1997 outbreak of Hepatitis A in New South Wales associated with oysters which affected over 700 people; and

" the 1997 outbreaks of salmonellosis which affected over 800 people in Victoria and caused two deaths; and

" the 1999 outbreak of salmonellosis from faecally contaminated orange juice which affected over 480 people in South Australia.

While these are headline-grabbing incidents magnified through media attention, the real problem is far less obvious. Every day across Australia, an average of 11,500 people succumb to a variety of foodborne illnesses which are not mentioned in the media nor otherwise reported.

Estimated number of cases of foodborne illness in Australia

Reliable data are not available on the actual incidence of foodborne illness in Australia. In fact, it is estimated that less than one per cent of cases is captured in existing notification schemes.

ANZFA has used Australian survey data to estimate the incidence of foodborne illness. Although the data are imperfect, the use of a number of different surveys conducted at different times and places, provides greater confidence in the validity of the data as they all lead to similar estimates. Subsequent comparison of the Australian estimates has confirmed that the figures are not dissimilar to overseas data.

It is estimated that there are over 4 million cases of foodborne illness every year in Australia.

The 1996 Australian Bureau of Statistics survey

A 1996 s tud y into the inciden ce of foodborne i llness in Tasman ia reported that 23,100 people contracted food related illness in a three month period (Australian Bureau of Statistics 1996). Assuming there is no seasonality in this data, there woul d be 92,400 cases in T asmani a over a 12-month period. In 1996 Tasmania had a population of 475,000 and the total Australian population was 18,312,000 (Australian Bureau of Statistics 1998). Extrapolating the Tasmanian annual rate of illness Australia-wide on a population basis indicates a national incidence of 3,562,166 (say 3.5 million) cases of foodborne illness per annum.

However these figures would underestimate the annual national incidence since the study was undertaken in winter when the incidence of foodborne illness is lower . In a ddition, cool climate states such as Tasmania generally have a lower annual incidence of foodborne illnesses than do warmer States and Territories in the north of Australia as described in Figure 3.

If the reduced incidence in Tasmania is conservatively assumed to be 75 per cent of the national rate, then the national incidence of foodborne illness would be 4.2 million cases per year.

The 1997 98 Water Quality Study

In 1997 98, the Department of Epidemiology and Preventative Medicine, Monash University and the Melbourne Water Corporation under the auspices of the Cooperative Research Centre for Water Quality and Treatment conducted a randomised double blind trial which assessed the level of gastrointestinal illness in selected Melbourne households.

The results from the pilot study indicate that the rate of gastroenteritis was 1.04 cases per perso n per year ( Fairley et al 1 999 ). It is k now n that the level of pe rso n-to-p erso n t ransmission of pathogens such asSalmonellais around 10 per cent (Stöhr and Hoppe 1995) which means the remaining 90 per cent of cases result from contaminated food, water and animal-to-human contact.

Data from the United States indicate that 30 to 35 per cent of intestinal infectious diseases result from contaminated food (Archer and Kvenberg 1985). Discussions with Australian researchers indicate a range of 30 per cent to 60 percent (Veitch 1997 pers comm). A conservativ e estimate of 30 per cent of gastroenteritis re sultin g from co ntaminate d food would i ndicate that an annual incidence of foodborne illness is around 5.4 million cases per year.

The 1989 and 1995 National Health Surveys

The 1989 and 1995 National Health Surveys estimated that 174,900 and 165,100 cases of  diarrhoea enteritis (respectively), occurred in Australia in the two weeks preceding the survey.

Extrapolated to a year, the 1989 data indicate approximately 4.6 million cases or 0.27 episodes per person per year. The 1995 data equates to 4.3 million cases per annum with an average of the two studies of 4.45 million cases per year. As discussed above, it is conservatively estimated that 30 per cent of diarrhoeal illnesses are foodborne giving a total of 1.34 million cases of foodborne diarrhoeal enteritis a year.

However, it should be recognised that a number of foodborne illnesses do not involve diarrhoea but instead have symptoms such as nausea, vomiting and lethargy. These foodborne illnesses are not included in these estimates and are in addition to cases of diarrhoea enteritis. Data from a United States study indicate that around 50 per cent of foodborne illnesses result in non-diarrhoeal symptoms such as vomiting and fever (Monto and Koopman 1980). As such, of the 1.34 million cases of diarrhoeal enteritis, there may be an additional 1.34 million cases of non-diarrhoeal illness: a total of 2.7 million cases a year.

Data from the National Notifiable Diseases Surveillance System

The only national data collected in Australia on foodborne illnesses are for those diseases required by law to be notified to the NNDSS. Notifiable diseases that may be foodborne include campylobacteriosis, salmonellosis, shigellosis, hepatitis A, listeriosis, typhoid fever and yersiniosis. There are many more organisms that cause foodborne illness but are not notifiable. Viruses for example are not notifiable but are estimated to cause between 30 per cent and 40 per cent of foodborne illnesses (Crerar et al 1996; Ryan 1997; Djuretic et al 1996). Figure 5 shows the total number of notifiable diseases that may be foodborne that were notified to the in 1997.

Health professionals acknowledge that the number of notified cases of foodborne disease significantly underestimates the actual number of cases. Before active surveillance was adopted, surveillance systems used in the United States were similar to those used in Australia. Such systems were shown to detect fewer than 1 per cent of cases of salmonellosis during an outbreak (Hennessy et al 1996). Underestimation of the number of cases of foodborne disease is likely to be of this magnitude in Australia (Crerar et al 1996). If the number of cases reported in 1997 was 23,102 (Figure 5), then the total number of cases of notifiable disease would be at least 2.3 million. There are other common non-notifiable diseases including viruses which are estimated to cause 30 per cent to 40 per cent of cases of foodborne illness (Stolle and Sperner 1997). When non-notifiable bacteria such asStaphylococcusspp,Clostridium perfringensandCampylobacterin New South Wales are taken into account, the real incidence is likely to be above four million cases per year.

Figure 5 Cases of notified diseases that may be foodborne(National Centre for Disease Control 1999)

Disease	Cases
Campylobacteriosis	11,848
Salmonellosis	7,004
Shigellosis	799
Hepatitis A	3,076
Listeriosis	53
Typhoid fever	77
Yersiniosis	245
Total	23,102

Summary

The incidence of foodborne disease in Australia was estimated using data from four sources which are summarised in Figure 6.

Figure 6  Estimates of the incidence of foodborne illness in Australia

Survey	Estimated annual incidence of foodborne illness
Australian Bureau of Statistics survey (1996)	4.7 million
CRC for Water Quality and Treatment (1997 98)	5.4 million
National Health Surveys (1989 and 1995)	2.7 million
Extrapolations from the number of cases notified to the NNDSS	4.0 million

For the purposes of this report an average figure of 4.2 million cases per annum will be used. This equates to a person contracting foodborne illness on average once every four years in Australia. Expressed in another way, the rate of foodborne illness in Australia is 220 cases per 1,000 population. In comparison, estimated rates for New Zealand are up to 190 cases per 1,000, the United Kingdom 190 cases per 1,000 and the United States 175 cases per 1,000.

 

Estimated number of cases of foodborne illness in Australia every year	4.2 million
Estimated number of new cases of foodborne illness in Australia every day	11,500
Number of meals eaten in Australia every year	20.8 billion
Percentage of meals in Australia which cause foodborne illness	0.02%

Cost of foodborne illness

Breakdown of costs

There are no Australian data which accurately quantify the total cost of foodborne illness to the Australian community. Overseas data show the average cost per case is between A$1,000 and A$2,000 in these countries. The cost estimates take into account direct socio-economic costs including losses incurred by society, either as economic costs through production losses resulting from sickness-related absenteeism or costs to the affected individual and his/her family from illness-related expense. However the cost estimates of between A$1,000 and A$2,000 in these countries are not readily transferable to Australia.

The factors which contribute to this cost are summarised in .

Figure 7 Economic factors associated with a foodborne illness(Todd 1989) continued

	Action involving cost or loss	Type of costs	Groups affected by cost or loss
Medical care	1. Physicians visit	Direct	Ill persons, insurance companies or government insurance plans
	2. Physicians at Hospitals	Direct	As above
	3. Surgery	Direct	As above
	4. Intensive care	Direct	As above
	5. Hospitalisation board and bed	Direct	As above
	6. Inoculation with antisera or antitoxin	Direct	As above
	7. Other medication	Direct	As above
	8. Convalescence (nursing costs, convalescent homes)	Direct	As above
Travel	1. Travel of family and friends to visit hospitalised persons	Direct	As above
Investigation of illness	1. Epidemiological team	Direct	Local, state/provincial and federal governments
	2. Laboratory analysis of food, environmental and clinical samples	Direct	As above
	3. Warning of public	Direct	As above
	4. Administrative costs	Direct	As above
Loss of productivity	1. Illness of salaried employer	Direct	Employer
	2. Illness of wage earners or self employed persons	Direct	Employee or employer
	3. Infected persons prevented from working	Direct	Employee or employer
	4. Patient staying at home to look after Ill or Infected children	Direct	Employee or employer
	5. Claims by Ill or Infected employees	Direct	Insurance companies or workers' compensation
	6. Illness of students, home makers	Indirect	Ill persons and family
	7. Rehabilitation with occupational retraining	Indirect	Ill persons, employer and government
	8. Long-term effects	Indirect	Ill persons
Emotional loss	1. Pain, grief and suffering	Indirect	Ill persons, family and friends
	2. Inability to work at previous occupation	Indirect	Ill persons and family
	3. Leisure time loss	Indirect	Ill persons and those looking after them
	4. Death	Indirect	Family and friends, employer and government (loss of tax revenues)
Loss of business	1. Loss of sales	Direct	Food company, eg restaurant, processor
	2. Recall and destruction of food	Direct	Food company
Legal action	1. Successful law suits	Direct	Food company, eg restaurant, processor
	2. Legal costs	Direct	Food company or Ill person

Direct costs: those that are measurable and apply directly to the problem

Indirect costs: those that are difficult to measure or are not directly associated with the problem

Estimation of the cost of foodborne illness in Australia

Foodborne illness costs the Australian community over $2.6 billion every year.

Cost to consumers

For consumers, a bout of foodborne illness means enduring unpleasant symptoms as well as costs associated with medical expenses and time off work both for the sufferers and carers. There can also be long-term health consequences from foodborne illness (Bunning et al 1997, Lindsay 1997, McDowell and McElvaine 1997, Council for Agricultural Science and Technology 1994). Archer and Kvenberg (1985) estimate that chronic illness are likely to occur in two to three per cent of cases of foodborne infection.

Salmonellosis can lead to the condition of reactive arthritis and campylobacteriosis to the severe illness of Guillain-Barré Syndrome (GBS). GBS causes increasing paralysis over one to four weeks and recovery is gradual taking one to two years. Some 20 per cent of sufferers are significantly disabled and around 2 per cent die (Buzby and Roberts 1997). In the United States, conservative estimates of the cost to the community (including production losses) of foodborne GBS is US$142 million to US$1 billion (Buzby and Roberts 1997).

Foodborne illness also costs consumers indirectly. Recent food poisoning outbreaks have received considerable media attention which reduces consumer confidence in the safety of food and can also a ffect eating habits. For example, the 1997 hepatitis A outbrea k a ssociated with oysters severely affected not only oyster sales across Australia but seafood sales in general.

Cost to government

There are significant costs to all levels of government from foodborne illness in Australia. Each case of foodborne illness which involves a medical consultation, costs the government through the Medicare rebate. Hospitalisation and pathology tests add to this burden. Many of the children affected by the HUS outbreak in 1995 will require long-term medical treatment including kidney transplants and dialysis. When antibiotics or other pharmaceuticals are prescribed, there are additional costs to government through the Pharmaceutical Benefits Scheme.

As well as treatment costs, government also faces the expense of investigating sporadic cases and outbreaks of foodborne illness as well as any follow-up action required. Major outbreaks such as the 1997 hepatitis A and salmonellosis outbreaks put substantial demands on health authorities during their investigations, including pathology, communication, recall coordination and follow-up action. There are also costs associated with lost productivity and its impact on revenue.

Cost to industry

Food poisoning in Australia is an increasing burden on the food industry. In addition to the immediate loss of sales by the food business identified as causing the illness, customer loyalty can be affected in the longer term. This can be exacerbated by word-of-mouth communication by afflicted consumers.

Salmonellacontamination of pork rolls in Victoria in 1997 resulted in closure of the business,cost the bakery owners $16,000 in finesand led to an insurance settlement of $750,000.

Food recalls and withdrawals are also expensive undertakings for businesses with loss of stock, loss of consumer confidence and potentially bad publicity. Contamination of minced beef with enterohaemorrhagicEscherichia coliin the United States in 1997 resulted in the recall of 12 million kilograms of product equivalent to 100 million hamburgers.

Litigation, particularly in the form of class actions, is becoming a feature of the more publicised outbreaks of foodborne disease in Australia and can result in substantial financial outlay. Closure of food businesses, bankruptcy and prosecutions for manslaughter have all been associated with recent outbreaks.

Internationally, filing of a lawsuit is becoming an immediate response to outbreaks of foodborne illness with negotiations leading to multi-million dollar settlements (Morton 1998). In New York State, there are around 20,000 legal suits per annum, some 40 per cent successful (Todd 1987).

There are also costs associated with re-establishing goodwill and market share. This is true not only for the company involved but also in the flow-on effect to whole industry sectors and to Australia s international reputation as a supplier of safe food.

An outbreak and subsequent recall of product associated with a New Zealand supermarket delicatessen counter cost the outlet around $1.5 million in recall costs, consultants fees, reparation, product loss and lost custom. Over the next six months the supermarket chain as a whole lost $3 million in reduced sales.

Even if a business is not responsible for a food poisoning incident, it can expect a downturn in sales as consumer confidence in a particular sector is affected. This can result in significant loss of sales in both domestic and international markets. Following the Garibaldi mettwurst incident, mettwurst sales throughout Australia fell by 40 per cent. There have been reports of 400 to 500 smallgoods producers going out of business as a result.

It is estimated that the downturn in trade following this incident and theSalmonellaoutbreaks in Victoria in 1997 cost the Australian smallgoods industry over $400 million.

The 1996 Salmonella contamination of peanut butter cost the company involved over $55 million. 14 The effect flowed on to another company which lost $100,000 in sales of its peanut butter muesli bars. 15

Perception by overseas markets that Australian products are safe is essential to maintaining Australia s food exports. This can be damaged very easily even when Australia is not at fault. Exports of Australian beef to Japan were affected by the international reaction to BSE (mad cow disease) in the United Kingdom, irrespective of the fact that the Australian herd was free of the disease. Detection of pathogens like enterohaemorrhagicEscherichia coliin Australian exports or linking Australian products to an outbreak of foodborne illness, would be expected to have a marked effect on Australia s export earnings.

Quantifying the cost of foodborne illness in Australia

Access Economics undertook an analysis based on the cost of 1,000 people falling ill each day. Given that 46 per cent of the population is employed, this would mean that 460 people of those affected would be absent from work. Based on the latest average weekly earnings, this would represent a cost of $121 per day or for the 460 people afflicted: $55,660 per day. If one-half of the people not employed but sick result in someone not working (ie caring for the sick), the proportion would rise to 73 per cent and the daily cost to $88,330.

In terms of other costs, if there are 1,000 people affected on any one day, around 25 per cent would go to a medical practitioner (Veitch and Hogg 1997): the cost of the visit and medication would be about $55, giving a daily cost of $13,750. If around 10 per cent of the people affected needed to be hospitalised (keeping in mind that the very young, the elderly and other vulnerable groups are particularly susceptible to complications from food poisoning) this would generate additional costs for the community of around $55,000 per day.

Bringing these three direct costs of foodborne illness together gives an average cost of about $157 per person affected per day. The debilitating symptoms of foodborne illness can last anything from a day to several weeks and even many months or years for term health consequences as discussed previously. On the basis that each illness lasts two days, thedirectcost of each illness would be around $315. The sensitivity of this estimate to changing assumptions can be illustrated. For example, if the same number of people fall ill each day with the same proportion hospitalised but the stay in hospital is four rather than two days (and others are still affected for two days), the average cost of each illness would rise to around $430.

As discussed, there are other costs associated with foodborne illness such as product recalls costs, litigation, loss of consumer confidence in a product, costs borne by the public sector and so on. It is much more difficult to estimate these but international studies suggest they might total as much again as the directly measurable costs. Taking the $315 as a base it would suggest that the cost of each illness is around $630. The most recent Canadian data indicate the average cost per case of foodborne illness in Canada is around C$1,000 (Mayers and Couture 1999).

As previously established there are 4.2 million cases of foodborne illness every year in Australia. With an average cost of $630 per case, this equates to a total cost of foodborne illness in Australia of over $2.6 billion every year.

However, there are potential additional costs which are not included in this figure: for example, the potential impact on our food export markets. This was clearly demonstrated by the BSE (mad cow disease) incident in the United Kingdom and its impact on exports of British meat, estimated to cost more than $10 billion.

As discussed, perception by overseas markets that Australian products are safe is essential to maintaining Australia s food exports. In 1995 96, export of high risk category commodities totalled close to $7 billion, compared to those in the medium and low risk