Results, cont

Gastrointestinal and foodborne disease outbreaks

During 2004, OzFoodNet sites reported 1,085 outbreaks of gastrointestinal illness affecting 28,461 persons (Figure 13). One hundred and eighteen of the outbreaks were due to consumption of contaminated food or water giving an overall rate of 5.9 foodborne outbreaks per million population. During 2004, there was one outbreak of zoonotic origin that occurred throughout the year. This outbreak of antibiotic resistant Salmonella Paratyphi biovar Java was related to contact with tropical fish.

The mode of transmission of the remaining outbreaks was either unknown or were most likely due to person-to-person transmission. Sites conducted 92 investigations into outbreaks or clusters where the mode of transmission was not determined, or a foodborne source was not identified. Person-to-person transmission was suspected as the cause of 874 outbreaks affecting 25,363 persons. The rates of non-foodborne outbreaks were reasonably consistent across different jurisdictions and ranged between 20.6 per million population in Queensland to 74.7 per million population in Tasmania (Figure 14). The majority of person-to-person outbreaks occurred in aged care facilities (593 outbreaks; 19,295 people affected) and hospitals (140 outbreaks; 3,423 affected). Norovirus was confirmed as the aetiological agent for 398 outbreaks spread from person-to-person that affected 13,842 people.

Figure 13. Foodborne and gastroenteritis outbreaks reported by OzFoodNet sites, Australia, 2004, by suspected mode of transmission ( n =1,085 outbreaks)

Figure 14. Number and rates of non-foodborne gastroenteritis outbreaks,* Australia, 2004, by OzFoodNet site

* Includes outbreaks spread from person-to-person or of unknown mode of transmission and investigations of clusters of infections other than those caused by Salmonella, ( n =924 outbreaks).

Foodborne disease outbreaks

In 2004, 118 foodborne disease outbreaks affected 2,076 persons, resulting in 116 hospitalisations and two associated deaths (Table 7). A summary description of each outbreak is shown in Appendix 2.

New South Wales reported the largest number of outbreaks, which represented 36 per cent (43/118) of all outbreaks reported (Table 7). The reporting rates of foodborne outbreaks for different OzFoodNet sites ranged from 1.0 outbreaks per million population in Western Australia to 15.5 outbreaks per million population in the Australian Capital Territory. The majority of outbreaks occurred in summer and autumn (Figure 15).

Figure 15. Outbreaks of foodborne disease, Australia, 2001 to 2004, by selected aetiological agents

Table 7. Outbreaks of foodborne disease in Australia, 2004, by OzFoodNet site

State	Number of outbreaks	Outbreaks per million population	Mean number of cases per outbreak	Number affected	Hospitalised	Deaths
ACT	5	15.5	58.8	294	2	0
NSW	43	6.4	14.8	635	45	0
NT	2	10.1	7.0	14	2	0
Qld	27	7.1	9.4	254	20	0
SA	17	11.1	9.0	153	10	2
Tas	1	2.1	57.0	57	0	0
Vic	21	4.3	26.2	550	37	0
WA	2	1.0	59.5	119	0	0
Total	118	5.9	17.6	2,076	116	2

Aetiological agents

The most common agent responsible for foodborne disease outbreaks was Salmonella, which was responsible for 31 per cent (36/118) of outbreaks (Table 8). These outbreaks affected a total of 679 persons with a hospitalisation rate of 12 per cent (79/679). S. Typhimurium was responsible for 81 per cent (29/36) of Salmonella outbreaks. Norovirus caused 14 outbreaks of foodborne illness, with a low hospitalisation rate of only 0.4 per cent (2/500). There were eight outbreaks of 'suspected toxin' poisoning, which included seven outbreaks suspected to be due to Clostridium pefringens and one outbreak of suspected histamine poisoning due to 'butterfish' consumption.

In 2004, there were seven small outbreaks of ciguatera, all of which occurred in Queensland. There were five outbreaks of campylobacteriosis, which was more than in previous years. There were single outbreaks each of Bacillus cereus, rotavirus, mixed toxins from B. cereus and Staphylococcus aureus and Listeria infection. In an outbreak of listeriosis two cases died, although it is unclear whether Listeria infection was the major contributing factor to the deaths. Thirty-five per cent (41/118) of outbreaks were of unknown aetiology.

Table 8. Aetiological agents responsible for foodborne disease outbreaks showing number of outbreaks and numbers of persons affected, Australia, 2004

Agent category	Number of outbreaks	Number of people affected	Number of people hospitalised	Mean size of outbreak
Bacillus cereus	1	6	0	6.0
Clostridium perfringens	3	128	1	42.7
Campylobacter	5	58	4	11.6
Ciguatoxin	7	24	3	3.4
Listeriosis	1	2	2	2.0
Norovirus	14	500	2	35.7
Rotavirus	1	14	0	14.0
Salmonella other	7	80	5	11.4
Salmonella Typhimurium	29	599	74	20.7
Suspected toxin	8	209	2	26.1
Mixed toxins	1	16	0	16.0
Unknown	41	440	23	10.7
Total	118	2,076	116	17.6

Top of page

Food vehicles

There was a wide variety of foods implicated in outbreaks of foodborne disease during 2004 (Table 9), although investigators could not identify a vehicle for 47 per cent (55/118) of outbreaks. Contaminated fish was the most common food vehicle, seven of which caused ciguatera poisoning. Fish was responsible for nine per cent (10/118) of outbreaks, followed by seafood and mixed meat dishes each responsible for six outbreaks. Poultry, cakes, pizza, oysters and egg dishes were also common causes of outbreaks.

Table 9. Categories of food vehicles implicated in foodborne disease outbreaks, Australia, 2004

Food category	Number of outbreaks	Per cent	Number affected	Number hospitalised
Cakes	4	3.4	82	10
Custard	1	0.9	43	17
Dessert	1	0.9	4	0
Dips	1	0.9	14	0
Eggs	1	0.9	4	0
Fish	10	8.6	52	8
Mixed dish	5	4.3	63	1
Mixed meat dish	6	5.2	191	2
Oysters	4	3.4	35	1
Pizza	4	3.4	108	8
Pork	1	0.9	27	1
Poultry	6	5.1	188	3
Salad	1	0.9	28	3
Sandwiches	3	2.6	270	0
Seafood	6	5.2	45	10
Suspected eggs	2	1.7	19	6
Suspected poultry	2	1.7	24	2
Suspected red meat	1	0.9	5	5
Suspected water	1	0.9	7	0
Vegetable dish	1	0.9	6	0
Unknown	57	49.1	861	39
Total	118	100.0	2,076	116

Outbreak settings

The most common settings where food was prepared was at restaurants and caf é s (36%), followed by commercial caterers (14%), takeaway venues including nationally franchised fast food chains (13%), and private residences (11%) (Table 10). Contaminated primary produce was responsible for 7 (6%) outbreaks. Five outbreaks occurred in association with foods prepared in aged care facilities. Four outbreaks each were due to foods prepared in bakeries and in hospital settings. The settings where foods were consumed were similar to where it was prepared. Restaurants and cafés (36%) were the most common venues, followed by private residences (20%), catered functions (9%) and community settings (7%).

Table 10. Categories of settings where food was prepared in association with foodborne disease outbreaks, Australia, 2004

Setting prepared	Number of outbreaks	Number affected	Number hospitalised
Aged care	5	75	4
Bakery	4	82	10
Café	2	17	3
Camp	1	5	0
Commercial caterer	16	683	15
Contaminated primary produce	7	58	9
Grocery store/delicatessen	2	30	0
Hospital	4	42	7
Institution	2	52	17
National franchised fast food	7	83	11
Private residence*	14	157	6
Restaurant	40	558	27
Takeaway	8	30	1
Other	1	27	1
Unknown	5	177	5
Total	118	2,076	116

* Includes one outbreak where food prepared included food prepared by takeaway stores.

Top of page

Investigative methods and levels of evidence

States and territories investigated 41 outbreaks using retrospective cohort studies and nine outbreaks using case control studies. Forty-four per cent (18/41) of outbreak investigations using cohort studies were of unknown aetiology. Twenty-four per cent (10/41) of investigations using cohort studies were Salmonella outbreaks. Fifty per cent of investigations of toxin and suspected toxin outbreaks used cohort studies. Fifty-six per cent (5/9) of outbreak investigations using case control studies were due to Salmonella. Sixty outbreaks relied on descriptive information to attribute a foodborne cause or identify a food vehicle, while investigators did not collect individual patient data for eight outbreaks.

To attribute the cause of the outbreak to a specific food vehicle, investigators obtained analytical evidence from epidemiological studies for 15 outbreaks. Sixty-six per cent (27/41) of cohort and 50 per cent (4/8) of case control studies did not identify a significant association between illness and a specific food vehicle. Microbiological evidence of contaminated food was found in 10 outbreaks, with a further four outbreak investigations obtaining both microbiological and analytical evidence. Investigators obtained analytical and/or microbiological evidence for 33 per cent (12/36) of Salmonella outbreaks (Appendix 2).

Large outbreaks (>50 persons affected)

Six outbreaks affected 50 persons or more in 2004. Two were due to norovirus, two due to Salmonella, one due to C. perfringens, and one due to suspected C. perfringens intoxication. The food for two of these outbreaks was prepared at commercial caterers, with a third using a mixture of food prepared at homes and takeaway food. Outbreaks also resulted from the food prepared at a restaurant and a bakery. The sixth outbreak was a large community-wide outbreak of Salmonella Typhimurium 12 in New South Wales that was associated with chicken prepared in a variety of settings. A variety of foods were implicated in these large outbreaks, including: bakery products, chicken, sandwiches, and dishes containing chicken including pizza.

The two large outbreaks of norovirus were related to preparation of foods that required considerable handling. The food vehicles implicated in the two outbreaks were sandwiches containing salmon and egg fillings in one large outbreak involving a commercial caterer, and contaminated bakery products in the second outbreak. In both outbreaks, the investigation identified food handlers who had worked while ill with gastroenteritis.

One of the outbreaks of Salmonella was due to serotype Typhimurium phage type 9 at a pizza restaurant in Melbourne. Cases continued to occur after an initial cleaning of the facility. Several foods were positive for S. Typhimurium 9, along with swabs of food preparation areas. Cases occurring early in the outbreak were associated with pizza, whereas those occurring after the initial cleaning were associated with dishes containing chicken, including pasta, risotto and pizza. The cause of the outbreak was suspected to be due to cross contamination in the kitchen due to poor hygienic practices, including cooking chicken on trays in a pizza conveyor belt. After a second clean up of the restaurant there were no new cases or positive food samples.¹⁶

New South Wales conducted a major investigation into a state-wide increase of Salmonella Typhimurium 12 during February. A case series investigation of 40 cases interviewed with hypothesis generating questionnaires identified that cases reported a high consumption rate of fruit and vegetables when compared to previous studies. To examine this hypothesis, a case control study used community-based controls recruited randomly using the electronic white pages, and cases with other Salmonella Typhimurium phage types as controls. Forty-one cases—48 S. Typhimurium controls and 203 community controls—were recruited. Consuming chicken breast prepared in the home was identified as a risk factor for S. Typhimurium 12 infection (Odds Ratio 4.6, p<0.1). New South Wales reported 141 cases as part of this outbreak, making it the largest outbreak of salmonellosis during 2004.

Top of page

Medium outbreaks (20–49 persons affected)

There were 22 outbreaks affecting between 20 and 49 persons. Seven of these outbreaks were due to Salmonella, including an outbreak of Salmonella Stanley in a school in Victoria. Food vehicles were only identified for three of these Salmonella outbreaks, which were custard, salad rolls containing red onion, and roast pork.

There were four outbreaks of norovirus, three of which occurred at restaurants and a fourth that occurred at a mass catered event. One norovirus outbreak in New South Wales affected 24 people following consumption of locally-grown oysters. No further illness was identified and norovirus was not detected in samples of oyster meat. There were two outbreaks of campylobacteriosis, one of which was suspected to be related to a barbecue meal at an aged care facility in Victoria, while the other was suspected to be due to a meal of chicken in New South Wales .

Victoria reported four outbreaks of suspected C. perfringens affecting between 20–49 people, only one of which could be confirmed by the presence of heavy growth of C. perfringens and the presence of toxin in stool samples. One of these outbreaks was in an aged care facility, while another was in a hospital. The other two suspected toxin related illnesses were associated with restaurants.

There were three outbreaks associated with imported foods with potential international implications. These outbreaks were all due to contaminated Individually Quick Frozen oysters. All outbreaks were small affecting a total of 11 people. Two occurred in Queensland and one in the Northern Territory. Investigations confirmed that these oysters from Japan were the same as those implicated in December 2003 in outbreaks in the Northern Territory and Western Australia. Oysters from all outbreaks were tested but norovirus was not detected in any of these three batches. Investigation of the outbreak that occurred in Queensland in October 2004 implicated oyster meat from the same batch as the outbreak in the Northern Territory that occurred a year earlier, although it was a smaller size oyster.¹⁷ This particular batch had reportedly been withdrawn from sale earlier in 2004, and had been shown by laboratory tests to be contaminated with norovirus.

Cluster investigations

A cluster is defined as an increase in infections that are epidemiologically related in time, place or person where investigators are unable to implicate a vehicle or determine a mode of transmission for the increase. An example is a temporal or geographic increase in the number of cases of a certain type of Salmonella serovar or phage type. Another example is a community-wide increase of cryptosporidiosis that extends over some weeks or months. In this report, there were a small number of outbreaks of different pathogens where the mode of transmission was unknown, that have been classified as a cluster.

During 2004, states and territories conducted 54 cluster investigations. These clusters affected 622 persons with 51 cases hospitalised. Seventy-eight per cent (42/54) of these investigations related to clusters of Salmonella. Salmonella clusters affected 473 persons with 46 cases hospitalised. S. Typhimurium was responsible for 55 per cent (23/42) of cluster investigations, with phage types 135/a (6 investigations) and 170/108 (5) being the most common. Of the remaining 19 investigations, there were 18 other different Salmonella serovars involved. There were 11 clusters due to pathogens other than Salmonella, with Campylobacter, Cryptosporidium, Shiga toxin-producing E. coli and Shigella causing two each and one investigation into a cluster of hepatitis A. No aetiology was identified for three cluster investigations.

OzFoodNet sites compared investigations into concurrent increases in several Salmonella serotypes that occurred across multiple jurisdictions. These included Salmonella serotypes Singapore, Typhimurium 12a, Typhimurium 170/108, and Paratyphi B biovar Java.

The cluster investigation into cases of Salmonella Paratyphi B biovar Java was part of a national case series to investigate the association with tropical fish aquariums. Eighteen cases infected with this multi-drug resistant serotype were investigated. In the month prior to illness, 85 per cent (11/13) of cases with aquarium/tanks had contact with sick or dead fish.

The true number of clusters investigated is difficult to determine, as the figures do not include all cluster investigations conducted in Public Health Units or local government areas. Jurisdictions have different definitions of 'cluster' and triggers for investigating clusters to fit with staff resources and local priorities.

Top of page

Risk factors for infection

During 2004, OzFoodNet identified several important risk factors for foodborne illness as a result of outbreak investigations and from preliminary results of case control studies. These included risks due to the following foods and settings for foodborne disease.

Eggs

Sites continue to report outbreaks associated with the consumption of egg-based products, such as eggs, salad dressings, cakes and desserts. There were three outbreaks suspected to be caused by eggs, and a further three associated with desserts and cakes where eggs were suspected as the source of Salmonella. In one outbreak of Salmonella Typhimurium 126 infections in Victoria, illness was associated with consumption of one brand of organic eggs. In several of these outbreaks, investigators were unable to trace implicated eggs back to a single farm. There is a need to identify potential interventions, and a review of quality assurance in the industry may be appropriate. The restaurant and catering industries need to be made aware of the potential risks of using raw unpasteurised eggs in sauces, dressings and desserts.

Chicken and poultry

During 2004, outbreaks of poultry-associated salmonellosis continued to occur, including a major outbreak of Salmonella Typhimurium 12 in New South Wales. Poultry is consumed by approximately 80 per cent of people each week. To make our food supply safer, it is important to consider ways to reduce the burden of infections in the community due to the consumption of poultry.

Oysters and fish

The four outbreaks associated with oysters during 2004 showed their potential to cause outbreaks of human illness. Three of these outbreaks were due to contaminated imported oyster meat from a single estuary system in Japan. In 2004, the Australian Quarantine and Inspection Service restricted the importation of these products from this growing area. Importation from certain growing areas in Korea were also restricted, as oysters from this area had caused norovirus outbreaks in New Zealand.

There were 10 outbreaks due to fish during 2004, making it the most common food vehicle. The majority of these were small outbreaks of ciguatera poisoning in Queensland. Many outbreaks of ciguatera relate to fish caught by amateur fishermen, but one of these outbreaks was associated with coral trout eaten at a restaurant. Ciguatera can be a severe illness and there is a continuing need to educate amateur fishermen about ciguatera including the risks associated with fishing in known ciguatera areas and consuming large ocean water fish.

Settings

There were several settings where food was prepared or consumed that were identified as high risk for foodborne disease, which included:

Bakeries

The four outbreaks occurring in bakeries in 2004 revealed the need for assessment of food safety issues in these premises. Three of the outbreaks were associated with cakes, some of which were filled with cream or custard. Two of the outbreaks were caused by Salmonella Typhimurium, while one was unknown and another was due to norovirus. Epidemiological investigation of these outbreaks often does not uncover the real source of contamination, as there is a time lag between food consumption and the recognition of the outbreak. Food safety agencies may need to consider the development of hazard reduction plans for these facilities to prevent further outbreaks.

Restaurants and catered events

Outbreaks in this sector constituted 49 per cent (58/118) of outbreaks. A variety of pathogens caused these outbreaks, including Salmonella, C. perfringens, norovirus and ciguatera. Outbreaks involving restaurants and commercial caterers are more readily recognised, as the meals are often served to large numbers of persons. A wide range of food vehicles were responsible for outbreaks in this sector. Clearly there is a need to continue to monitor the causes of outbreaks in restaurant and catering settings to identify potential gaps in food safety practices.

Hospitals and aged care

People resident in aged care facilities and patients in hospital are at particular risk for foodborne disease, which is shown by the nine outbreaks that occurred during 2004. Four of the outbreaks were suspected or confirmed Clostridium perfringens outbreaks, while two were due to Campylobacter infection, one due to Salmonella Typhimurium 126 var, one due to Listeria monocytogenes O1 and one unknown. The majority of these outbreaks indicate problems with preparation and handling of foods for residents. The outcomes for patients in these settings are often more adverse, as these sub-populations are more susceptible to serious foodborne disease. The food supplied to hospital patients and persons in institutions should be comprehensively monitored. In addition, there is a need to ensure that patients at risk for infection should not be fed high-risk foods in hospitals.

Top of page

Surveillance evaluation and enhancement

Continuous improvement of surveillance is important to ensure that foodborne illness is investigated rapidly and effectively. To improve surveillance it is necessary to evaluate and compare practices conducted at different sites.

National information sharing

In 2004, all jurisdictions contributed to a fortnightly national cluster report to identify foodborne illness occurring across state and territory boundaries. The cluster report supplemented information sharing on a closed list server, teleconferences and at quarterly face-to-face meetings.

Outbreak reporting and investigation

During 2004, the Australian Capital Territory site reported the highest rate of outbreaks of foodborne disease (15.5 outbreaks per 100,000 population). The rates of reporting foodborne Salmonella outbreaks ranged between 1.6–5.0 outbreaks per 100,000 population. New South Wales investigated the largest number of foodborne disease outbreaks (43 outbreaks; 6.4 per 100,000 population). Victoria and Queensland investigated 13 Salmonella clusters each, giving rates of 3.4 and 2.6 per million population respectively.

States and territories conducted 50 analytical studies (cohort or case control studies) to investigate foodborne disease outbreaks or clusters of suspected foodborne illness. Investigators used analytical studies for 42 per cent (50/118) of foodborne disease outbreaks, which was similar to previous years. Queensland conducted four case control studies to investigate outbreaks of foodborne infections during 2004, which was the most for any jurisdiction. Every jurisdiction reported conducting at least one cohort study. New South Wales conducted 40 per cent of all cohort studies.

Completeness of Salmonella serotype and phage type reports

There was considerable improvement in the completeness of Salmonella available on state and territory surveillance databases between the years 2000 to 2004 (Figure 16). Overall 98.4 per cent (7,671/7,798) of Salmonella notifications on databases contained either serotype or phage type, which was an increase of 7.3 per cent from 2000 and 1.5 per cent from 2003.

Only 76.9 per cent (39/48) of phage type information was reported for S. Hadar, which was a decline from the previous year (Figure 17). Phage typing information was available for 91.9 per cent (37/40) of S.  Heidelberg and 92.6 per cent (418/449) of reports for S. Enteritidis in 2004. The largest increase in completeness between 2000 and 2004 was reported for S. Heidelberg (23.3%) and S. Bovismorbificans (17%).

South Australia had the highest proportion of complete Salmonella notification (100%), while four sites reported 98 per cent or higher. Western Australia reported the lowest rate of completeness with 91.9 per cent. New South Wales reported the largest improvement with 19.1 per cent improvement, when compared to 2000 figures.

Figure 16. Proportion of Salmonella infections notified to State and Territory health departments with serotype and phage type information available, Australia, 2000 to 2004

Figure 17. Proportion of Salmonella infections for six serotypes notified to State and Territory health departments with phage type information available, Australia, 2000 to 2004

 

This article was published in Communicable Diseases Intelligence Vol 29 No 2, June 2005.