Enhancing foodborne disease surveillance across Australia in 2001: the OzFoodNet Working Group

In 2000, the OzFoodNet network was established to enhance surveillance of foodborne diseases across Australia. OzFoodNet consists of 7 sites and covers 68 per cent of Australia's population. This report is the first Annual report of the OzFoodNet Working Group, and was published in Communicable Diseases Intelligence Vol 26 No 3, September 2002. This report can be viewed in 5 HTML documents and is also available in PDF format.

Page last updated: 03 October 2002

A print friendly PDF version is available from this Communicable Diseases Intelligence issue's table of contents.


Incidence of foodborne disease

National foodborne disease incidence

This section documents trends in the incidence of enteric diseases in OzFoodNet sites. OzFoodNet epidemiologists provide regular summaries of foodborne disease incidence from notifiable disease datasets. The OzFoodNet data on sporadic disease are a subset of the information reported to the National Notifiable Diseases Surveillance System (NNDSS), but are more detailed and allow interpretation at the state, territory or public health unit levels. NNDSS annual reports should be consulted for national notification rates of foodborne diseases. OzFoodNet provides a national picture by recording details of outbreaks and clusters occurring across jurisdictional boundaries. Improved communication and cross-jurisdictional investigations provide important information about the food handling practices that have led to food contamination and the causes of foodborne disease.

Interpreting the data

It is important to recognise that subtle differences between the three sources of data used in this report, OzFoodnet, NNDSS and the National Enteric Pathogens Surveillance System (NEPSS) can make interpretation difficult. Some of the inherent limitations of the data include:
  • Data in the surveillance systems may come from different information sources, e.g. the proportion of notifications received from medical practitioners varies from jurisdiction to jurisdiction.
  • Each surveillance system will have different delays in receipt and processing of reports, which can affect the total number reported in any time period.
  • Where the surveillance data are reported, the reporting date is often different, e.g. sometimes the 'date of onset of symptoms' is used, while at other times reports will relate to the 'date of specimen collection', or the 'date of receipt of notification'. In this report, the 'date of receipt of notification' is also used, except for historical comparisons where we use the 'date of onset'.
Managers of the various surveillance schemes may still be cleaning data at the time of reporting. This cleaning will involve checking for accuracy of information on the database, and removing duplicate entries.

The data reported usually reflect a complex mix of biases that are inherent in public health surveillance. One bias that particularly affects surveillance data is ascertainment bias, i.e. some groups of the population are more likely to be detected as cases by the surveillance system. In notifiable disease datasets it is common to have an over representation of younger children, people who are elderly or immunocompromised, and people who are severely affected by the illness. This is usually because these patients are more likely to seek medical attention, and doctors are more likely to conduct tests on these patients. The data are rarely representative of the true burden of infection in the community, or the gradient of symptoms associated with infection.

The states and territories have differing approaches to surveillance, which may be reflected in the data. This also impacts the way that different jurisdictions choose to report data, such as outbreaks or clusters.

Some diseases are not notifiable in certain jurisdictions, e.g. individual cases of Campylobacter infection are not notifiable in New South Wales.

OzFoodNet data

OzFoodNet reports surveillance data for several bacterial pathogens and summary information from outbreaks potentially related to food and water. In this report, data are reported by the date of receipt of notifications at the health agency, unless specified. Historical comparisons use date of onset of symptoms for comparative purposes. Summary data for OzFoodNet sites on notified cases are shown in Appendix 1.

The Hunter OzFoodNet site supplied data for all of New South Wales. These data were used where possible for reporting total figures. Data for the Northern Territory are not reported unless specified.
Rates were calculated using the Australian Bureau of Statistics estimated resident populations for 2001. Where appropriate, we directly standardised regional rates of disease within jurisdictions by age to estimated resident population for Australia, 2000.

Top of page

Salmonella

In 2001, OzFoodNet sites reported 6,607 cases of Salmonella infection, which represented an increase of 2.1 per cent over the mean of the previous 3 years.* The overall rate of Salmonella notification in OzFoodNet sites was 34.1 cases per 100,000 population, and ranged from 23.1 cases per 100,000 population in the Hunter region to 59.8 cases per 100,000 population in Queensland (Figure 3).

Figure 3. Crude notification rates of salmonellosis, 1998 to 2001, by site and year

Figure 3. Crude notification rates of salmonellosis, 1998 to 2001, by site and year

* In this report, historical comparisons use date of onset of patient's symptoms or nearest equivalent for analysis.


Overall, notification rates of salmonellosis for 2001 were increased in the states of Tasmania (26.2%), Western Australia (20.7%) and the Hunter Health Area, New South Wales (17.1%) when compared with the 3-year mean rates for 1998-2000. There were moderate declines in the number of notifications of Salmonella in Queensland (-12.4%), South Australia (-6.3%), the Australian Capital Territory (-5.5%), and in Victoria (-2.8%) from the 3-year mean values.

OzFoodNet sites reported that the ratio of males to females was approximately 1:1, and ranged from 1.2:1 in Victoria to 0.9:1 in Tasmania. The median age of cases ranged between 18:21 years at all OzFoodNet sites, except for Queensland where the median age was 9 years. There were no major changes in the median ages of salmonellosis cases from 2001 to 2000.

Rates of salmonellosis are highest in northern areas of Australia, with the highest rates in the Kimberley region.4 Western Australia reported that the Kimberly region had a rate of 559 cases per 100,000 population. OzFoodNet sites reported that notification rates increased from south to north along eastern Australia (Figure 4).

Figure 4. Standardise rates of Salmonella notifications in OzFoodNet regions in eastern Australia, 2001, by date of notification

Figure 4. Standardise rates of emSalmonella/em notifications in OzFoodNet regions in eastern Australia, 2001, by date of notification

† Notifications were analysed by date of receipt at the health department. Rates were directly standardised to the Australian Bureau of Statistics estimated resident population for Australia in 2000.


During 2001, there were 520 notifications of Salmonella Typhimurium phage type 135 to OzFoodNet sites (including New South Wales) making it the most common infection (Table   1). There were 330 notifications of Salmonella Typhimurium phage type 9, which has been a common phage type for many years. South Australia recorded the emergence of Salmonella Typhimurium phage type 126, which had previously been rare in this state. The incidence of this phage type also increased in other Australian jurisdictions during 2001, particularly New South Wales and Queensland.

Top of pageTable 1. Numbers, rates and proportions of top five Salmonella infections, 2000 to 2001, by site

OzFoodNet site
Top five Salmonella infections
Salmonella type
(serovar & phage type)
2001 Rate 2001* Proportion (%) 2000 Ratio
ACT Typhimurium 9
10
3.2
12.8
31
0.3
Stanley
5
1.6
6.4
1
5.0
Bovismorbificans 14
4
1.3
5.1
0
-
Paratyphi B bv Java Dundee
2
0.6
2.6
1
2.0
Enteritidis RDNC 11
2
0.6
2.6
0
-
Hunter Typhimurium 135
15
2.8
12.2
10
1.5
Typhimurium 126
9
1.7
7.3
3
3.0
Typhimurium 64
9
1.7
7.3
14
0.6
Birkenhead
5
0.9
4.1
9
0.6
Typhimurium U290
3
0.6
2.4
0
-
New South Wales Typhimurium 135
202
3.1
11.9
115
1.8
Typhimurium 9
133
2.0
7.8
138
1.0
Typhimurium 126
98
1.5
5.8
56
1.8
Birkenhead
87
1.3
5.1
73
1.2
Infantis
41
0.6
2.4
25
1.6
Queensland Virchow 8
177
4.9
8.2
189
0.9
Saintpaul
164
4.5
7.6
184
0.9
Typhimurium 135
137
3.8
6.3
118
1.2
Birkenhead
130
3.6
6.0
102
1.3
Aberdeen
81
2.2
3.7
52
1.6
South Australia Typhimurium 126
110
7.3
18.0
5
22.0
Typhimurium 9
49
3.3
8.0
26
1.9
Typhimurium 108
31
2.1
5.1
11
2.8
Typhimurium 64 var
21
1.4
3.4
0
-
Infantis
19
1.3
3.1
8
2.4
Tasmania Mississippi
98
20.8
59.0
73
1.3
Typhimurium 9
11
2.3
6.6
22
0.5
Typhimurium 135
5
1.0
3.0
5
1.0
Infantis
3
0.6
1.8
4
0.8
Saintpaul
2
0.4
1.2
2
1.0
Victoria Typhimurium 9
127
2.6
11.4
186
0.7
Typhimurium 135
96
2.0
8.6
70
1.4
Typhimurium 4
79
1.6
7.1
37
2.1
Typhimurium 170
73
1.5
6.5
36
2.0
Virchow 34
35
0.7
3.1
60
0.6
Western Australia Typhimurium 135
80
4.2
9.0
68
1.2
Saintpaul
45
2.4
5.1
42
1.1
Chester
31
1.6
3.5
12
2.6
Muenchen
23
1.2
2.6
29
0.8
Stanley
21
1.1
2.4
5
4.2

* Rate per 100,000 population.
† Proportion of total Salmonella notified for this jurisdiction.
‡ Ratio of the number of reported cases in 2001 compared with 2000.


Certain Salmonella serovars were localised to specific geographical areas in Australia. During 2001, Salmonella Birkenhead was the fourth most common serovar for both New South Wales and Queensland. This elevated notification rate relates to an endemic focus of Salmonella Birkenhead in northern New South Wales and south-eastern Queensland. In Tasmania, 59 per cent (98/166) of Salmonella reports were the Mississippi serovar, which is rarely reported anywhere else in Australia. The notification rate for Salmonella Mississippi in Tasmania was 20.8 cases per 100,000 population, which was the highest specific rate of any serovar in OzFoodNet sites.

During 2001, NEPSS recorded 6,912 cases of Salmonella and documented specific epidemiological changes. The most notable of these changes was the emergence of Salmonella Typhimurium 126 across Australia. NEPSS also detected increases in Salmonella Stanley, Salmonella Typhimurium 170, Salmonella Typhimurium DT 104, and other serovars. NEPSS collaborated with state and territory health departments and OzFoodNet on a regular basis and participated in several joint investigations and routine teleconferences. NEPSS is a valuable Australian resource due to the data they collect on human and non-human sources of Salmonella and other enteric pathogens.

The rates of salmonellosis in Australia are higher than in the United States of America, but lower than in New Zealand (personal communication, Michael Baker, ESR, New Zealand, 1 August 2002).3 In 2001, the FoodNet active surveillance system in the United States of America reported an incidence of 15.1 cases per 100,000 population compared to 64.7 cases per 100,000 population in New Zealand. It is difficult to compare the true incidence between countries due to the different healthcare systems and cultural settings.

Campylobacter

In 2001, OzFoodNet sites reported 15,815 cases of Campylobacter infection, which equated to a rate of 125 cases per 100,000 population however, data was not available for New South Wales, and the Hunter Health Area, New South Wales. This notification rate represented a 20.6 per cent increase over the mean for the previous 3 years. The increase was consistently observed in each quarter of 2001, with the highest rates in spring.

Rates of campylobacteriosis increased in all sites, ranging from 3.4 per cent in Queensland to 63.2 per cent in Western Australia (Figure 5). The increased rate in Western Australia is partly attributable to the introduction of voluntary laboratory notifications in 2000 for the first time. Geographically, the rates of Campylobacter infection were higher in southern parts of Australia in contrast to the rates observed for Salmonella infections (Figure 6). The north south trend was less marked for Campylobacter infections, but this phenomenon has been observed in other countries.5 The highest rate of Campylobacter infection was 174 notifications per 100,000 population in South Australia.

Top of pageFigure 5. Crude notification rates of Campylobacter infection, 1998 to 2001, by site and year

Figure 5. Crude notification rates of Campylobacter infection, 1998 to 2001, by site and year

Figure 6. Standardised rates of Campylobacter notifications in OzFoodNet regions in eastern Australia, 2001, by date of notification*

Figure 6. Standardised rates of emCampylobacter/em notifications in OzFoodNet regions in eastern Australia, 2001, by date of notification

* Notifications were analysed by date of receipt at the health department. Rates were directly standardised to the Australian Bureau of Statistics estimated resident population for Australia in 2000.


Sites reported a slight predominance of males (range of male to female ratio: 1:2-1.3:1) amongst notified cases. The median age of cases ranged from 26 to 32 years. Six outbreaks were reported due to Campylobacter in 2001. Apart from an overall increase in rates, OzFoodNet sites did not record any significant changes in the epidemiology of Campylobacter infections from 2001 to 2000.

There are substantial differences in rates of campylobacteriosis between countries. The rate in USA FoodNet sites is 13.8 cases per 100,000 population and 271.5 cases per 100,000 population in New Zealand (personal communication, Michael Baker, ESR, New Zealand, 1 August 2002).3 It is difficult to determine whether these represent true differences in community incidence of the disease, or relate more to healthcare access, laboratory testing procedures and surveillance modalities.

Listeria

OzFoodNet sites reported 61 cases of listeriosis in 2001, which represents a notification rate of 0.3 cases per 100,000 population. This was an increase of 4 per cent compared to the mean of the previous 3 years. Western Australia had the highest notification rate amongst OzFoodNet sites and the incidence increased in Queensland over the last 3 years (Figure   7).

Figure 7. Crude notification rates of Listeria infections in OzFoodNet sites, 1998 to 2001, by site and year

Figure 7. Crude notification rates of Listeria infections in OzFoodNet sites, 1998 to 2001, by site and year

The majority of cases during 2001 were reported in elderly people who were immunocompromised. OzFoodNet sites reported that the median age of non-pregnancy associated cases ranged from 60 to 86 years. Thirteen per cent (7/54) of non-pregnancy associated cases died. In Queensland, the outcome of 47 per cent (8/17) cases was unknown. Sites reported six maternal foetal infections during 2001, which equated to a rate of 2.3 cases per 100,000 births. (Births data from AIHW National Perinatal Statistics Unit for 1999 and includes live births and foetal deaths.6). The foetus or neonate died in three of these cases, giving a neonatal mortality rate of 50 per cent.

Top of page

Yersinia

The CDNA agreed to stop reporting notifications of Yersinia infections to the NNDSS as of January 2001. The main reasons for this were the apparent decline in incidence and the absence of identified outbreaks. In May 2001, the Victorian Government revised regulations governing reporting of infectious diseases, at which time they removed yersiniosis from the list of reportable conditions. Currently, no other Australian jurisdiction has amended legislation to remove yersiniosis from lists of reportable conditions.

In 2001, OzFoodNet sites reported 71 cases of yersiniosis, which equated to a rate of 0.6 cases per 100,000 population. The overall rate was 50 per cent of the mean of the previous 3 years. The reasons for this decline in yersiniosis are unclear, but follow similar trends in other countries. Queensland reported 75 per cent (53/71) of all cases and had the highest rate of 1.5 cases per 100,000 population (Figure 8). The rates of yersiniosis were similar in all 3 Queensland Health zones, and ranged from 1.1 cases per 100,000 population in the Central zone to 1.9 cases per 100,000 population in the Northern zone.

Figure 8. Crude notification rates of yersiniosis, 1998 to 2001, by site and year

Figure 8. Crude notification rates of yersiniosis, 1998 to 2001, by site and year

The median age of cases of yersiniosis ranged from 6 to 26 years in different sites. In the two jurisdictions with the majority of cases, South Australia and Queensland, males were more common than females with a male to female ratio of 1.8:1 and 1.5:1 respectively.

The decrease in Yersinia notifications may be due to changes in laboratory testing practices rather than a true decline in incidence. Despite the declining rates of this disease, it is important for health agencies to continue surveillance for yersiniosis. The rates of yersiniosis in neighbouring New Zealand are 11.5 per 100,000 population, which is significantly higher than Australia (personal communication, Michael Baker, ESR, New Zealand, 1 August 2002).

Shigella

OzFoodNet sites reported 326 cases of shigellosis during 2001, which equated to a notification rate of 2.6 cases per 100,000 population. The rate of notification decreased by 23 per cent from the mean of the previous 3 years and only Tasmania observed an increase in the 3-year period (Figure 9). The median ages ranged from 20-43 years. Males were more commonly reported from all sites, except for Tasmania and Western Australia. There were no reported outbreaks of shigellosis or confirmed links with food. In Australia, the majority of shigellosis infections are thought to be due to person-to-person transmission, or are acquired overseas.

Figure 9. Crude notification rates of shigellosis, 1998 to 2001, by site and year

Figure 9. Crude notification rates of shigellosis, 1998 to 2001, by site and year

Top of page

Typhoid

OzFoodNet sites reported 67 cases of typhoid infection during 2001. This represents an overall notification rate of 0.3 cases per 100,000 population, which was similar to previous years. The highest rates were reported in Western Australia (Figure 10). Where travel status was known, sites reported that 92.5 per cent (37/40) of cases of typhoid had recently travelled overseas. Fifty-five per cent (22/40) of these cases had recently returned from Indonesia.

Figure 10. Crude notification rates of typhoid, 1998 to 2001, by site and year

Figure 10. Crude notification rates of typhoid, 1998 to 2001, by site and year

Shiga-toxin producing E. coli

OzFoodNet sites reported 47 cases of shiga-toxin producing E. coli (STEC) infection during 2001. The notification rate of 0.2 cases per 100,000 population was a 15 per cent increase over the mean rate for the previous 3 years (Figure 11). South Australia (26 cases) and Queensland (13 cases) reported the majority of cases. The median age of cases ranged from 10-28 years and females were more commonly infected than males in Queensland, South Australia and Victoria. All of the cases appeared to be sporadic.

Figure 11. Crude notification rates of shiga-toxin E. coli, 1998 to 2001, by site and year

Figure 11. Crude notification rates of shiga-toxin E. coli, 1998 to 2001, by site and year

Haemolytic uraemic syndrome

The highest rate of STEC infections was in South Australia, due to the specific testing of bloody stool (both microscopic and macroscopic) for the presence of shiga-toxin or the gene coding for production of the toxin. The majority of reports in South Australia were detected by polymerase chain reaction. Only 12 per cent (3/26) of cases in South Australia were reported to be due to E. coli O157. Victoria reported that two out of 4 cases were due to E. coli O157, while Queensland reported that four out of 10 cases were due to this serovar.

There were 5 cases of haemolytic uraemic syndrome (HUS) reported during 2001, corresponding to an overall rate of 0.02 cases per 100,000 population. There were 2 cases reported in New South Wales, and one case reported in each of Victoria, South Australia and Queensland (Figure 12). The median age of cases was 16 years (range 2-53 years) and the male to female ratio was 1:1. One case, an 83-year-old male, died giving a case fatality rate of 20 per cent.

It is likely that there is substantial under-reporting of this serious disease. The Queensland site reported that there were 21 patients recorded in hospitalisation statistics for the financial year 2000/01, compared to only 3 cases for the same time period on the notification dataset. There is very little known about the notification fraction for diseases potentially due to food, and this is an area of future work for OzFoodNet.

Figure 12. Numbers of notifications of haemolytic uraemic syndrome, 1998 to 2001, by month of onset and site

Figure 12. Numbers of notifications of haemolytic uraemic syndrome, 1998 to 2001, by month of onset and site


This article was published in Communicable Diseases Intelligence Volume 26, No 3, September 2002

Communicable Diseases Intelligence subscriptions

Sign-up to email updates: Subscribe Now