---

Invasive listeriosis, toxoplasmosis, hepatitis A and haemolytic uraemic syndrome are much less common than gastroenteritis, but each can lead to clinically serious illness (although there is a range from mild to serious for most of them). Information on the clinical characteristics of each of the illnesses and details of estimates are given in Appendix A. A summary of estimates of incidence, health service use and time off from activities and work is shown in Table 3.1.

Basic incidence data for invasive listeriosis, hepatitis A and haemolytic uraemic syndrome were obtained from notifiable disease data. We assumed that the notified numbers of invasive listeriosis were half those that actually occur, and adjusted under-reporting by a factor of 2.⁶ For listeriosis there is a significant danger to the unborn foetus if the mother is infected, even if the mother is not very unwell. Each materno-foetal pair is counted as one illness.

Incidence data for toxoplasmosis was not available for Australia and was based on data from the United States of America (Mead et al. 1999).

For hepatitis A, under-reporting was assumed to be a factor, as in previous studies overseas (Mead et al. 1999). The National Notifiable Diseases Surveillance System collects data on both laboratory-confirmed cases and GP notifications for ‘probable cases’ of hepatitis A, but it is likely that a significant number of cases are not reported. People with mild illness are unlikely to present to a doctor. In addition, some people may not have a definitive diagnosis because hepatitis A antibodies were not detected. The number of notifications was therefore doubled6 before adjusting for a fraction ‘due to foodborne transmission’.

Haemolytic uraemic syndrome estimates were derived from notified data and the Australian Paediatric Surveillance Unit (APSU).

Proportion of illnesses that are foodborne

There are few data on the proportion of these illnesses due to foodborne transmission. Nine foodborne disease experts were asked for their opinion (May 2004) and, for haemolytic uraemic syndrome, supplementary data were obtained from the APSU (2004) and the Victorian surveillance system (J Gregory, OzFoodNet, DHS, 2004, pers. comm.). The results are summarised in Table 3.2. There was considerable variation in opinion for toxoplasmosis, hepatitis A and especially haemolytic uraemic syndrome. Data available for haemolytic uraemic syndrome from the APSU and outbreaks in Victoria lend support to the estimate of around 50%.

Table 3.2 Proportion foodborne: Delphi results

Disease	Mean	Standard deviation	95% CrI
Listeriosis, materno-foetal	98	3	92–100
Listeriosis, others	98	3	92–100
Toxoplasmosis	35	18	0–71
Hepatitis A	10	7	0–24
Haemolytic uraemic syndrome, children 0–15 years	50	26	0–100
Haemolytic uraemic syndrome, adults 15 years or over	56	29	0–100

Footnotes
6. Not all infections causing hospitalisations are correctly diagnosed or reported. To account for this under-reporting we doubled the number of cases, as in Mead and others 1999.

prev page| TOC |next page