House of Commons - Agriculture

Select Committee on Agriculture Fourth Report

II. THE FOOD SAFETY PROBLEM

14. In this section of our Report we examine the underlying issues relating to microbiological food safety which need to be addressed by the Government, both before and after the establishment of a Food Standards Agency. The proposals for an FSA itself, particularly in terms of its structure, accountability, guiding principles, functions, powers and resources, as set out in the Government's White Paper, are the subject of Section III of this Report. Overall judgements about the desirability of an Agency, and the form it should take, clearly hinge on the extent to which it will be able to achieve improvements in the safety and quality of the nation's food supply. Many of the recommendations contained in this section of our Report, therefore, point to particular policies or tasks which, while they need to be set in train now by the Government, will be of a high priority for the Agency when it is established. Likewise, many of the recommendations which we make in Section III of this Report, about the Agency's structure and functions, have to be seen in the context of the underlying food problems which the Agency must successfully resolve.

The incidence of food poisoning

15. The available statistical information about the incidence of food poisoning appears to be fairly comprehensive, but is widely acknowledged to be deficient in a number of respects, most notably in substantially understating the actual level of food poisoning in the UK. Because reported cases, which appear in the official statistics, are the tip of an iceberg of unknown size[11], it is only possible to guess at the true incidence of food poisoning. Dr Eileen Rubery, Head of the Department of Health's Health Aspects of Environment Division, said that "under-ascertainment" was "somewhere between ten and one hundred fold"[12]. One consequence is that it is never possible to be sure that trends in reported illness mirror trends in actual illness, or whether they simply reflect such factors as changes in propensity to notify cases and investigate outbreaks of illness. While some would argue that increases in reported cases are largely a result of greater public awareness of food poisoning[13], in the absence of clear evidence it is probably most prudent to assume that the trend in actual food poisoning cases is rising roughly in line with reported cases.

16. A substantial amount of statistical information on food poisoning cases and levels of microbiological contamination of food is contained in the written evidence which we received from the Government and from the PHLS[14]. Much information is also contained in the October 1997 report by the Parliamentary Office of Science and Technology (POST) on Safer Eating: Microbiological Food Poisoning and its Prevention. We recommend these sources for Members and others wishing to inform themselves in detail of the underlying trends in food poisoning in the UK. For the purposes of this Report, we draw attention to those issues which appear to us to be of most significance and concern.

17. The most recent figures for food poisoning notifications, for 1997, show a continuation of the overall increase which has been observed since the 1980s. In England and Wales provisional figures for 1997 show a rate of 179.6 per 100,000 population, up from 160 in 1996 and from 58.3 ten years previously, in 1987[15]. There has been a slight, though probably not statistically significant, decrease in notifications in Scotland, from 199.6 per 100,000 population in 1996 to 198.6 in 1997, and a small rise in Northern Ireland, from 87.5 in 1996 to 91.8 in 1997[16]. For the first time, in 1997, there were more than 100,000 food poisoning notifications in the UK. The PHLS informed us that each year hundreds of people in the UK are hospitalized with food poisoning and "between 100 and 200 may die"[17].

18. There has been some speculation about the reasons for the apparent divergences in rates of food poisoning between different parts of the UK, and in particular the relatively high rate in Scotland (most marked in terms of E.coli O157 food poisoning) and the relatively low rate in Northern Ireland. Some witnesses offered hunches. Mr Richard Carden, Head of MAFF's Food Safety and Environment Directorate, said that "it may have something to do with lifestyle and the amount of eating out and the things that people eat"[18]. Professor Hugh Pennington, on the other hand, felt that, in relation to Scottish levels of E.coli O157 it was worth looking at the hypothesis that there might be higher levels of carriage of the pathogen in cattle in Scotland[19]. He also pointed out that there were unexplained regional variations in E.coli O157 incidence within Scotland itself[20]. The Government's existing research programme into pathogenic micro-organisms does not appear capable of providing much illumination on this subject[21]. We recognize that there may be a mixture of factors at play: agricultural and food industry practices, social and cultural norms about diet and food preparation, and, possibly, differential reporting and investigation rates for foodborne illnesses. Nevertheless, the trends are well-established and pronounced, and it cannot be the case that Northern Ireland is simply more fortunate, and Scotland less fortunate, than England and Wales. We recommend that the Government explore means of devising research projects to provide a fuller understanding of the different rates of food poisoning in the constituent parts of the UK. The absence of clear conclusions on these discrepancies is inexcusable, and must be remedied if the Agency's work is to be underpinned by a real understanding of the incidence and causes of food poisoning in the UK.

19. Before food poisoning cases can enter the national notification statistics, they must surmount a number of hurdles, and at each hurdle an unknown number of cases falls by the wayside. First, many people, probably the large majority, do not visit their doctors, usually because their symptoms are mild or pass quickly. Secondly, though doctors throughout the UK are statutorily required to notify cases of food poisoning (in England and Wales, to the local authority; in Scotland to the area Health Board; and in Northern Ireland, to the local Health and Social Services Board[22]), it is thought that there is a substantial degree of under-reporting by doctors[23]. The best estimate of the extent of under-reporting is provided by data from the GP Sentinel Practice Scheme, which records the number of GP consultations for infectious intestinal disease (IID). In 1997 there was a mean annual incidence for such consultations of 2,420 per 100,000 population, twenty-four times the figure for formal food poisoning notification by doctors and thirteen times the total number of notifications (which include cases ascertained by other means than through GP notification)[24]. Though the ratio of consultations to notifications has decreased substantially over the last ten years, indicating that GPs are probably more likely now to notify cases, a wide margin of under-notification is still apparent. This hampers efforts to understand the extent of the problem of foodborne illness in society. Dr Rubery told us that the Government had "actually taken quite a number of steps like writing to GPs or encouraging GPs to notify more frequently"[25]. Under-notification by GPs is only one part of the wider jigsaw of under-ascertainment of infectious intestinal disease in the UK, but it is an important part, and the Government must both ensure that GPs are advised of developments in foodborne illnesses and their symptoms, and continue to press GPs to meet their responsibilities in respect of notification. This task will be a crucial one if the Agency's work is to be informed by accurate statistical information and not just by hunches and guesswork.

20. Complementary to the GP notification system is the voluntary laboratory reporting system, under which laboratories throughout the UK report faecal isolates of pathogenic organisms, following analysis of specimens referred by doctors, to their respective surveillance centres (in England and Wales, to the PHLS's Communicable Disease Surveillance Centre (CDSC); in Scotland, to the Scottish Centre for Infection and Environmental Health (SCIEH); in Northern Ireland, to the Department of Health and Social Services)[26]. Drawing attention to the variability in reporting levels, the Government also pointed out that laboratory-confirmed infections gave a "significant underestimate of the true incidence of foodborne infections since not all cases seek medical attention and only a proportion of these have a specimen submitted for analysis"[27]. The PHLS was critical of the laboratory reporting system's lack of coherence and consistency, claiming that "laboratories in different regions vary greatly in their reporting efficiency with a concentration of poorly reporting laboratories in certain regions"[28]. Professor Brian Duerden, Deputy Director of the PHLS, argued that "laboratory notifications should become a statutory responsibility"[29]. We agree. Such measures are essential in improving public health and food safety policy.

21. In addition to formal notifications and laboratory confirmations, local authorities ascertain cases by other means, from outbreaks in the community, for example, or as a result of direct contacts from members of the public or owners of food businesses. For general outbreaks, defined as "an outbreak affecting members of more than one private residence or residents of an institution"[30], the CDSC has developed a surveillance scheme in England and Wales which involves the dispatch of a questionnaire to the local authority for completion and return on the conclusion of the investigation of the outbreak[31]. Implication of foodstuffs in such outbreaks may be based upon microbiological analysis, statistical probability or more circumstantial evidence. The PHLS pointed out that because of the voluntary nature of the surveillance scheme, there was significant variability between local authorities both in the rigour with which outbreaks were investigated and in reporting levels[32]. In concert with the SCIEH, the CDSC is undertaking a Department of Health-funded project to seek to improve consistency and comparability of approach between local authorities in investigating outbreaks[33].

22. The Government informed us that much valuable information about the level and causes of food poisoning in England would result from its study of infectious intestinal disease, which was set up in 1993[34]. Giving evidence to us on 11 November, Dr Rubery expressed the hope that the IID study would shortly be available[35] and that it would provide "really useful definitive data"[36]. Towards the end of our inquiry we asked for an update of progress in completing this study, and the Government said that "detailed analysis and careful interpretation" was still required before publication, although it was confident that it would be able to publish a report on the findings "later this year"[37]. The Government confirmed the point, set out in its original memorandum, that the findings of the study supported the view that the current reporting system "significantly underestimates the true magnitude of infectious intestinal disease" in England[38]. The publication of this report should be extremely important in understanding the true prevalence of food poisoning in the UK.

23. It is relatively easy to diagnose weaknesses in the UK's notification and reporting systems in relation to foodborne illness, but important to recognize that it will never be possible to establish a fully comprehensive and accurate picture, given the nature of the problem. Even for the minority of cases which present to a GP, it is rarely possible to link illness to a particular foodstuff. This is especially so as the great majority of food poisoning cases are sporadic, rather than part of a general outbreak. Moreover, an unknown proportion of infectious intestinal disease is not associated with food. We also have no reason to dispute the Government's contention that "by international standards the UK has one of the better systems" for reporting food poisoning cases[39].

24. With such imprecise data, it is difficult to arrive at an assessment of the overall costs to society of such foodborne illness. The IID study referred to above (see paragraph 21) is intended to provide updated information on this subject[40]. Until then, the best available study, based upon data collected in 1988-89 and expressed in terms of 1992 prices, estimates that costs to the public sector were approximately £70 million a year. This figure, which Dr Rubery advised us had to be "treated with considerable caution", omits estimates of costs to industry and the family. The Chartered Institute of Environmental Health cited a number of rather higher figures, up to between £0.5 billion and £1 billion a year at 1988 prices[41]. Purely for indicative purposes, we sought from the Government details of the cost of smoking-related illnesses and road accidents: the former are estimated to cost the NHS in England between £1,400 million and £1,700 million a year, and the latter £490 million a year in casualty-related medical and ambulance costs. It is important to have some idea of the economic costs to society of food-related illness, both in absolute terms and in relation to other public health risks, to inform the allocation of resources to deal with the problem. Major food safety issues such as the BSE epidemic may also impose huge costs on the economy in an attempt to prevent widespread illness and deaths. At the same time, the consumption of food is a biological necessity, and not directly comparable with smoking or driving. As Dr Rubery said, public concern and public perception, as well as the effectiveness of action which can be taken, are also very important factors in determining policy[42]. We also take to heart the comments of Professor Philip James that "there is ten times more ill health... which comes from the inappropriate nutritional quality of the diet compared even with infection"[43].