Most food poisoning is never diagnosed, so someone has to go and ask

FoodNet determines the incidence of laboratory-confirmed infections for
the bacteria commonly transmitted through food. That count is reliable and it
is not the number of people who got ill.

Most diarrhoeal illnesses are undiagnosed and therefore never reported.
Someone who is ill for two days and does not see a doctor, or sees one and is
not tested, never enters the surveillance system at all.

What the survey is for

The FoodNet Population Survey is a periodic survey of **randomly selected
residents** of the FoodNet surveillance area, and it answers three questions
the laboratory data cannot:

1. How often people get acute diarrhoeal illness
2. How often they seek care for it
3. How often they are exposed to things linked to diarrhoeal illness

The second question is the bridge between the two datasets. Knowing what
fraction of ill people seek care — and what fraction of those get tested — is
what lets a confirmed case count be scaled into an estimate of the real
burden.

What it asks

The most recent survey ran in 2018–2019, using modernised and expanded
methods: multiple modes of administration and multiple sampling frames.

Participants were asked about basic demographics, **episodes of diarrhoea or
vomiting**, and exposures across these categories:

Asking about exposures in the general population — not just among people who
fell ill — is what makes it possible to say whether an exposure is actually
associated with illness, rather than simply common.

Source: Centers for Disease Control and Prevention.