Modern scientific techniques make it easier than ever to identify the source of an outbreak of food poisoning. The bad news is that the expense involved means these methods are unlikely to be used in anything other than the largest cases of food poisoning.

A GP treating a severe case of suspected food poisoning will often send stool samples for analysis. The laboratory will then isolate the bug responsible using a variety of techniques, including an analysis of the antigens (proteins) attached to the surface of the bacterium as well as DNA-typing. Further analysis will allow the microbiologist to identify the strain of bug involved so, for example, highly dangerous E. coli O157 can be separated from normal E. coli organisms.

A network of labs monitoring foodborne bugs shares much of this information so that patterns and possible sources of infection can be determined from an early stage.

Environmental health officers then examine the food history of the victim, and this might involve a visit to the food outlet where the person believes he got the food poison.

"These techniques allow us to get an accurate picture of the bug causing the disease and the types of food linked to the poisoning outbreak," says Wayne Anderson, chief food science expert with the Food Safety Authority of Ireland. "Now it's much harder for a business to say 'sorry, guv, it wasn't us'."

However, Anderson cautions, not all outbreaks are related to food. Many bugs, such as the winter vomiting virus, are transmitted person-to-person, though the blame wrongly falls on the dodgy sandwich or dirty salad.

In addition, 40 per cent of outbreaks that are alleged to have happened outside the home, for example in a restaurant, are actually traced back to the home. In all, only about 10 per cent of cases are reported.