Richard A Sprenger
managing director Highfield.co.uk Ltd

The aims of the Food Standards Agency
The Food Standards Agency have stated that they will reduce levels of food poisoning by 20% by the year 2006. If the actual levels of food poisoning in the community could be reduced by 20% in 5 years this may help promote public confidence, and indicate an improvement in food safety standards. However, the current "food poisoning" statistics have very little relevance to levels of food poisoning in the community. The statistics include
several organisms classed as "foodborne" which may be transmitted in many ways, other than in food, including person to person spread, contact with animals, airborne or from the environment. So even if the actual levels of, so called, food poisoning are reduced it may have nothing to do with improvements in food safety.

The unreliability of notifications of food poisoning
Because of the difficulties in obtaining reliable information on levels of food poisoning in the community most people tend to use the statistics generated by the notification of food poisoning to reflect increases or reductions in actual levels of food poisoning. Unfortunately these statistics are totally unreliable as an indication of food poisoning levels. They include a large number of suspect food poisoning cases, i.e. people suffering from diarrhoea and vomiting and also cases of viral gastroenteritis and campylobacter. The vast majority of cases of viral gastroenteritis are not foodborne and a large number of cases of campylobacter are also not foodborne. These statistics would be described more accurately as cases of gastroenteritis or cases of infectious intestinal disease.

It is extremely regrettable that we do not collect statistics of confirmed cases of food poisoning, as these would be of much more use.
In the absence of statistics relating to actual confirmed cases of food poisoning i.e. where food poisoning is suspected AND a food poisoning organism is identified in the faeces, vomit or blood of an ill person, then we need to consider other statistics to measure hygiene standards in commercial premises.

Most reliable current statistics
Probably the two most reliable sets of statistics currently available are the laboratory isolations of salmonella obtained from specimens of ill people and the number of outbreaks of food poisoning from commercial premises.

Outbreaks from commercial premises
Fortunately or unfortunately the number of confirmed outbreaks of food poisoning is relatively small, around 100 per year in England and Wales. As would be expected the majority of these outbreaks are from catering premises as it is notoriously difficult to detect outbreaks of food poisoning from manufacturing or retail outlets. This usually only occurs when an unusual strain of, for example, salmonella is involved such as Salmonella napoli, or a serious organism such as E. coli O157.

What is extremely interesting is that despite the tremendous increase in the number of people eating meals out and the improved methods of detection and investigation over the last 30 years the number of identified outbreaks has gradually reduced despite the increases in salmonella isolates and reported cases of gastroenteritis in the 1980's and 1990's.

Salmonella isolations
Although the number of salmonella isolations will include some isolates from carriers and non-foodborne cases, it is generally recognized that these are in a small minority and the vast majority of isolates are from actual cases of salmonella food poisoning.

Given that the number of salmonella isolates dropped by 25% in 1998 and a further 25% in 1999, due to the vaccination of chickens, the question must be asked "what is the significance to food safety of a drop in food poisoning of 20% by the year 2006?"

Changes in administration affecting trends in notifications
Anyone who has followed trends in food poisoning statistics knows that they tend to be cyclical and trends are affected by many things, which have nothing to do with improving or deteriorating standards of food safety. In fact the biggest increase I recall was in 1974 when the reported cases of food poisoning (diarrhoea and vomiting) doubled from 4000 to 8000. At that time this appeared to be ignored. There was no adverse publicity, no public outcry and no panic measures. The only event of significance at the time was the reorganistion of local government and I suspect this resulted in a more efficient reporting of suspect cases of food poisoning; it certainly had no obvious relationship with food safety standards. Administrative changes in the PHLS relating to family outbreaks resulted in similar distortions.

Meaningful statistics related to food safety
If the Food Safety Standards Agency is looking for ways of improving food safety they first need to establish meaningful statistics. As regards food poisoning this means collecting information on actual confirmed cases and recording each individual organism separately. This would ensure that if an increase was due to, for example, viral gastroenteritis, we know, in terms of food safety, this is unlikely to be as significant as increases in levels of salmonella.

Alternative indicators of food safety
As the number of confirmed food poisoning cases and food poisoning outbreaks is relatively small and/or unreliable, objective information is difficult to obtain and we need to look for alternative indicators of food safety standards.

For example, the Food Standards Agency could require each local authority annually to randomly select 5 high-risk food premises. These premises could be inspected to determine the number of "contributory factors present which most commonly resulted in food poisoning", for example:
High-risk food stored at ambient temperature for longer than 4 hours
Food left to cool for longer than 2 hours
Failure to cook or reheat food above 75°C
Food obtained from non-approved sources
Incidents of cross-contamination
Infected food handlers involved in high-risk food preparation
Management failures

The Food Standards Agency would be able to analyse the information and determine whether standards of food safety were improving or deteriorating. Furthermore, they would obtain information on the particular areas that enforcement officers should concentrate on in future inspections.

If anyone is interested in discussing this paper or has any other proposals for objectively measuring food safety standards I should be pleased to hear from them on 0845 2260350 or by e-mail: Richard@highfieldpublications.com

Back to papers <