House of Commons - Agriculture

Select Committee on Agriculture Fourth Report

II. THE FOOD SAFETY PROBLEM

Microbiological hazards

25. Three bacterial micro-organisms, each with numerous sub-types, have managed to insinuate themselves tenaciously into the complex and lengthy UK food chain and are responsible for the majority of food poisoning cases: Salmonella (especially S.enteritidis and S.typhimurium); Campylobacter (especially C.jejuni); and verocytotoxin-producing Escherichia coli (VTEC) (especially E.coli O157:H7). Campylobacter comes ahead of Salmonella as the most common identified cause of human food poisoning[44]; E.coli O157, though much less frequent than its two rivals, more than compensates by its greater virulence, which leads to a higher proportion of deaths and long-term illnesses, especially haemolytic uraemic syndrome (HUS). The PHLS said that "patients with O157 VTEC infection are 20 times as likely to be admitted to hospital than those with other IID, the case fatality rate for O157 VTEC infection is 37 times that for other IIDs"[45].

26. A number of other pathogenic bacteria lurk in the background of the epidemiological picture. Human listeriosis, caused by the bacterium Listeria monocytogenes, associated mainly with soft cheeses and pâtés, and especially dangerous for the foetuses of pregnant women, peaked in the late 1980s, with 278 cases in 1988. Following withdrawal of contaminated imported pâté and the issuing of health warnings to vulnerable groups by the Chief Medical Officer the number of cases of listeriosis has fallen back to relatively low levels[46]. Foodborne botulism, caused by Clostridium botulinum, is thankfully rare in the UK: the last outbreak took place in 1989, with one death amongst 27 cases[47]. There are increasing reports, though still at a relatively low level, of viral foodborne infections. Indeed, small round structured viruses (SRSVs), mainly associated with shellfish, were the most common cause of food poisoning outbreaks in 1996, although the total number of cases of SRSV infectious intestinal diseases (IIDs) remains small[48]. Protozoan parasites such as Cryptosporidium parvum have recently been associated with outbreaks of foodborne illness[49].

27. In respect of Salmonella, particularly, but also, increasingly, Campylobacter, and E.coli O157, knowledge is being built up about the various characteristics of the bacteria which could be of assistance to control and eradication strategies throughout the food chain. Expert committees, notably the Advisory Committee on the Microbiological Safety of Food (ACMSF) and the Pennington Group, have produced a corpus of informed advice and recommendations for the food industry and the Government. The legislative foundations for effective control of food safety exist in the Food Safety Act 1990 and in the horizontal and vertical European Union food hygiene Directives[50]. And some inroads are being made: incidents of Salmonella in broiler breeder flocks and layer breeder flocks have declined since the early 1990s, as has the prevalence of Salmonella in raw retail chicken[51]. Dr Bernard Rowe, Director of the PHLS's Laboratory for Enteric Pathogens, argued that levels of food poisoning due to Campylobacter and Salmonella had remained reasonably steady between 1994 and 1996, with much of the rise in 1997 attributable to the warm weather in that year[52]. Marks and Spencer plc told us that, working together with their poultry supply chain, the level of Salmonella in their chickens had been reduced from over 40 per cent in 1995 to less than 10 per cent[53], and Safeway said that the average incidence of Salmonella in their poultry was 13 per cent[54].

28. Despite these encouraging developments, there is a long way to go before the British poultry supply chain can emulate its Norwegian and Swedish counterparts in virtually eliminating Salmonella. As Professor James pointed out, Sweden took "about 20 or 30 years" to cope with the Salmonella problem[55]. In respect of Campylobacter and E.coli O157, where the state of scientific knowledge is much more hazy, progress may well be more difficult, in the short term at least. Nevertheless, UK food safety policy must aim, as far as is reasonably achievable, at the elimination of these pathogenic bacteria from the food supply chain.

29. To a large extent, the observance of hygienic practices throughout the food chain provides common defences against all species of zoonotic organisms. However, measures taken to control one particular species may be unnecessary (or even useless) for another one. On our visit to a broiler producer near Newark we were told that success in reducing Salmonella levels had not been carried over to reduction in Campylobacter infection. For this reason, it is essential to understand as much as possible about each of the major pathogenic bacteria, to enable effective control measures to be devised and implemented. We summarize below the differences between the three main foodborne pathogens, as they are currently understood, indicating their implications for food safety policy.

Reservoirs of infection

30. In terms of foodstuffs, Salmonella is principally associated with poultry and eggs: S. enteritidis PT4, in particular, is invasive in chickens, and eggs and egg dishes and poultry are at the top of the list of suspected food vehicles in S.enteritidis PT4 outbreaks[56]. Reported incidents of S.typhimurium in cattle, sheep and pigs rose between 1990 and 1995, but have fallen substantially in the last 2 years[57]. One source of infection for Salmonella is animal feedingstuffs: the overall incidence of contamination is under 5 per cent, and, according to the Government, S. enteritidis and S. typhimurium, the two types most implicated in human foodborne illness, are "rarely isolated from animal feedingstuffs"[58]. Salmonella can also be transmitted vertically from parent bird to offspring. Campylobacter is also particularly associated with poultry: approximately 75 per cent of raw chicken carcases in England and Wales are Campylobacter-positive (compared to 30 per cent contaminated with Salmonellas)[59]. Unlike Salmonella, Campylobacter is not vertically transmitted in flocks; nor can it survive well on dry surfaces or in feedingstuffs[60]. It is, however, ubiquitous in the environment, and the Government claimed that "control of infection in extensively produced livestock is currently not a possibility"[61]. Once established in a poultry flock, Campylobacter spreads extremely rapidly[62]. E.coli O157 has been shown to infect a wide range of farm livestock species, including cattle, sheep, goats, horses and pigs. The methods of infection are not well understood, and nor is the correlation between the presence of the bacteria in the animal's gut and the shedding of it in faeces. The Government said that, with the current state of knowledge, there was "little scope for controlling the spread of infection or of eradicating infection"[63].

Disease in animals and humans

31. Although some Salmonella and Campylobacter serotypes cause illness in farm livestock, the majority do not, and animals and birds carrying the bacteria remain healthy[64]. E.coli O157 does not cause illness in livestock, and Professor Pennington told us that, in the past at least, this had caused problems in obtaining funding for research into the micro-organism, with neither agriculture nor health officials seeing it as their responsibility[65]. In humans, the dose of organisms able to cause disease ranges from as low as 40 in the case of E.coli O157, to 500 in the case of Campylobacter and between 100,000 and 10 million for Salmonella[66]. In humans, most types of bacterium colonise the gastro-intestinal tract to cause, either by the production of toxins or by direct infection, the classic food poisoning symptoms of diarrhoea, vomiting, abdominal pain and fever[67].

Effects of temperature, humidity, moisture

32. Bacteria have different levels of tolerance to environmental factors such as temperature, acidity and the degree of moisture present, with consequences for the effectiveness of measures to control them. Moreover PHLS work has shown that bacteria can modify themselves to become more virulent and tolerant of certain environmental factors, including those commonly encountered in food production processes. For example, the majority of wild type isolates of S. enteritidis PT4 and S. typhimurium DT104 show, according to the PHLS, "enhanced heat-, acid- and peroxide-tolerance, survive well in aerosols and on surfaces, and are more virulent in mice and more invasive in the reproductive tissues of laying hens"[68]. S.enteritidis PT4 is able to modify itself to protect against the slight acidity of raw egg mayonnaise (and hence against the acidity in the human stomach)[69]. There is some evidence that bacteria can adapt to withstand the effects of low and high temperatures, with implications for the effectiveness of domestic and chill-chain refrigeration and cooking practices for the control and elimination of bacteria[70].

Scientific knowledge

33. Scientific knowledge about Salmonella, its behavioural characteristics and its modes of transmission, is quite extensive, based upon the well-established typing (or "fingerprinting") methodology of the Laboratory of Enteric Pathogens[71]. Although some progress has been made with typing E.coli O157 strains[72], progress with Campylobacter typing has been slower, partly because, with such a high proportion of Campylobacter food poisoning cases being sporadic, it has been far harder than in the case of Salmonella to do the epidemiological work to link strains found in humans with those found in animals[73]. Professor Pennington concluded from this that "Salmonella will be a much easier task to sort out than Campylobacter and E.coli... With... Campylobacter and E.coli, we are in a much more difficult position in terms of reducing the food poisoning figures.... because we really still do not know enough about those bugs to have good focused control measures in there"[74].

34. The distinctive problems posed by Salmonella, Campylobacter and E.coli O157 lead us to the conclusion that the extension of the HACCP principle throughout the food chain (see paragraphs 56-59), welcome and valuable though it is, needs to be set in the context of specific strategies for controlling, reducing and eradicating particular micro-organisms from food. Such strategies will need to take into account levels of risk at all stages of the food-chain, adopting an holistic, panoramic view of the issues involved, the types of produce and the processing and handling methods used from plough to plate. We would propose that the Government, through the Food Standards Agency when it is established, should effectively apply the HACCP principle to the food-chain in its entirety, identifying the most appropriate points of intervention and control in the context of the methods of production and the characteristics of particular dangerous pathogens. The resources which the Agency allocates in this area should be distributed on the basis of that analysis. One example of the type of approach we have in mind would be the argument put forward by the PHLS that, in poultry production, intervention is more effective at the farm level than during slaughter and processing, whereas in red meat production the slaughterhouse, rather than the farm, should be the principal point of control for foodborne pathogens[75].

35. Virtually all the elements which need to be brought together into effective anti-microbial strategies have been enunciated by the Advisory Committee on the Microbiological Safety of Food in a succession of reports, and we would not wish to elevate our judgement above theirs. Here we simply highlight what seem to us to be the most urgent issues and objectives in relation to Salmonella, Campylobacter and E.coli O157. All these strategies need to be underpinned by focused research programmes.

Salmonella

(i) poultrymeat production: the aim should be eradication of Salmonella in broiler breeder flocks in the short term, combined with continuing pressure to reduce Salmonella contamination of raw chickens to below the ACMSF's target figure of 10 per cent and, eventually, to eliminate it altogether;

(ii) feedingstuffs: the PHLS doubted whether it would ever be possible to eliminate Salmonella entirely from feedingstuffs, given the possibility of contamination of farms[76], but Professor Georgala, Chairman of the ACMSF, cited the success of Sweden in support of the argument that "it should be made possible"[77] to make feed free of Salmonella. The Government has proposed the establishment of an Advisory Committee on Animal Feedingstuffs to support the work of the Food Standards Agency. We warmly welcome this decision, which should provide the necessary impetus to achieve wholly Salmonella-free feed in the future;

(iii) eggs: progress in reducing Salmonella contamination in eggs has been disappointing. About 1 in 600 eggs on retail sale were shown to be contaminated, either inside the egg or on the shell, in a 1995-96 survey, the same level as in 1991[78]. Mr Richard Carden of MAFF described this lack of progress as "rather curious", and drew the conclusion that closer attention needed to be paid to eggs than to chickens"[79]. The ACMSF has itself expressed concern and established an ad hoc Working Group to examine the implications of the survey's findings[80].

Campylobacter

(iv) research and typing: much greater information is required, through research, surveillance and serotyping, on the characteristics of Campylobacter: this was the main thrust of the ACMSF's 1993 Interim Report on Campylobacter[81]. Since that time the LEP has established a comprehensive Campylobacter reference service at Colindale for two NHS regions in England[82], with an emergency service on demand for the rest of England and Wales[83]. This is still not a fully national service, and Professor Georgala expressed concern about this, although he conceded that typing of Campylobacter was a very difficult matter[84]. He also expressed concern about the possible separation between veterinary and public health work on Campylobacter[85]:this is particularly worrying as the most urgent answer which needs to be found concerns the relationship between serotypes found in animals and those found in humans[86]. The swift establishment of a fully national Campylobacter typing service is a pre-requisite for the formulation of a coherent strategy to minimize the prevalence of the organism in the food chain.

E.coli O157

(v) research: as with Campylobacter, E.coli O157 (together with other verocytotoxin-producing E.coli) is a priority area for research to improve understanding of its epidemiology in farm livestock and its transmissibility and effects on humans. We support the call of the ACMSF's ad hoc Working Group set up to consider the implications of the Pennington Group report that research should be carried out throughout the UK, and not just in Scotland, into the carriage and prevalence of E.coli O157 in cattle[87]. This work needs to inform policy on hygienic practices in slaughterhouses, specifically in reduction of faecal contamination of carcases. Work also needs to be done to increase understanding of transmission of E.coli to humans via a wide range of foodstuffs[88].

(vi) cross-contamination: cross-contamination between raw and cooked meat products has been identified as a significant cause of foodborne E.coli O157 infection. For this reason the Pennington Group recommended the compulsory separation of raw and cooked meats in butchers and other premises under a licensing system[89]. Professor Pennington told us that he was broadly content with the way in which these and other recommendations were being taken forward[90]. We concur with the conclusions of the ACMSF's ad hoc Working Group that "the implementation of HACCP should be accelerated in high risk premises" and that an Industry Guide to Good Hygiene Practice should be developed for butchers and producers and retailers of cooked meats and cooked meat products[91].

(vii) the rise in the numbers of foodborne E.coli infections in humans urgently needs to be halted and reversed.

In its response to this Report, the Government should address the problem of formulating strategies against specific bacteriological hazards, as well as the detailed elements of those strategies which we have proposed.

36. The development of effective strategies against pathogenic micro-organisms in the food supply will naturally be far more detailed than the list of central concerns which we have given. As well as being informed by focused research programmes, these strategies will need to include comprehensive and well-structured surveillance schemes to record and monitor the microbiological status of foods on retail sale. Some informative data has been compiled over recent years by the PHLS, some under the EC Co-ordinated Food Control Programme and some in association with the Local Authorities Co-ordinating Body for Food and Trading Standards (LACOTS)[92]. Apart from repeated large-scale surveys of chickens and eggs, however, national food surveillance appears to be conducted in a relatively unscientific manner, often in the form of one-off exercises in response to particular concerns[93]. The PHLS also informed us that there were technical problems involved in the isolation of micro-organisms from foodstuffs, with the practical effect that "some previous surveys may have underestimated the prevalence of contaminated products"[94]. The establishment of a rigorous, scientific and statistically reliable national food surveillance system will be an essential component of an effective food safety policy in the future. Ideally these surveys should discriminate between foods produced according to different methods (one example amongst many would be battery and free-range eggs [95]) to provide information which can be used to better effect in influencing policy further down the food chain: they should also discriminate between UK-produced and imported foodstuffs. The results of such surveys must be published regularly.

Emergent threats

37. In their formulation of food safety policy the Government and the Food Standards Agency will need to retain the flexibility to respond to emergent microbiological hazards whether in the form of modified strains of existing bacteria, new micro-organisms, or changing associations of bacteria with particular foodstuffs. We have already referred (see paragraph 32) to the capacity of micro-organisms to adapt and evolve in response to environmental conditions. As Professor Humphrey of the PHLS put it: "They do evolve remarkably quickly. They have a lifecycle of 20 minutes and any change in any practice will ultimately result in a change in the bacteria associated with that practice"[96]. Control methods must remain robust enough to resist challenge from the most virulent and tolerant types of bacteria of all species. No one can predict when or if new micro-organisms, or new virulent strains of known bacteria, will emerge. A major food poisoning incident, resulting from such an emergent threat, in the early years of the Agency's existence could hardly be blamed on the Agency.

Antibiotic resistance

38. One example of bacteriological modification which has caused considerable concern in recent years is the development of resistance to antibiotics widely used in modern farming. As well as for therapeutic purposes, antimicrobials are used for prophylaxis, as feed additives, and as growth promoters[97]. Resistance in micro-organisms to antibiotics may, when humans are infected, render analogous medicinal antibiotics ineffective in combatting illness. Ninety-six per cent of Salmonella typhimurium DT104 isolated from humans are now resistant to more than one antibiotic, with most strains being resistant to at least 5[98]. The PHLS, citing the December 1996 outbreak involving a ciprofloxacin-resistant strain of S. typhimurium DT104, predicted that this could be the first of many such outbreaks given the widespread use of the related fluoroquinolone antibiotic eurofloxacin in turkey flocks[99].

39. Many organisations have conducted research on the question of antibiotic-resistant micro-organisms. The Soil Association submitted evidence to us, drawn from a report to be published this spring on antibiotic use in agriculture, calling for an end to the use of antibiotics as growth promoters[100]. The NFU is also carrying out a major study, and Mr Ian Gardiner, their Director of Policy, claimed that "initial soundings seem to indicate that the dangers from using antibiotics specifically in the animal population may be rather less than some people fear"[101]. Lax use of antibiotics in human medicine was alleged by Professor Pennington to have contributed to the growth in resistance amongst micro-organisms, although he also identified agricultural practices as problematic, and saw the whole issue as one of "very high priority"[102].

40. The Advisory Committee on the Microbiological Safety of Food has established a Working Group on Microbial Antibiotic Resistance which expects to complete its report in the first half of this year[103]. We consider the evidence of transfer of antibiotic-resistant micro-organisms from animals to humans through food to be approaching conclusiveness, and with the consequences of this potentially so serious, we favour a ban on the use of antibiotics in farming as growth promoters, and tighter restrictions on their use for subtherapeutic or prophylactic purposes. Every effort should be made to develop vaccines as alternatives to antibiotics for therapeutic purposes.

Viruses

41. Knowledge of the risks posed by Campylobacter and E.coli O157 may be patchy, but it is encyclopaedic in comparison with understanding of foodborne viral infections, another subject being considered by an ACMSF Working Group, expected to report shortly[104]. Professor Pennington said that "The viruses are a bit of a black box at the moment. We know they are out there and we know that they cause problems, we know that they are under-diagnosed and we know that our control methods are primitive in the extreme"[105]. Small round structured viruses are the most common cause of outbreaks of foodborne illness, though the numbers affected in each outbreak are low, as is the total number of cases caused by such viruses[106]. The PHLS explained that advances in molecular techniques would allow higher levels of detection of viruses in specimens, and that methods to detect and isolate viruses in shellfish required further development[107].

Contamination of fruit and vegetables

42. For meat, and for animal products such as eggs, the last line of defence against bacterial pathogens is thorough cooking. Increasing reports of illness caused by microbiological contamination of fruit and vegetables, many from abroad, are therefore alarming because these foods are often eaten raw. One particularly serious outbreak in Japan in 1996, which affected over 9,000 people and caused 11 deaths, was associated with E.coli O157 in radish sprouts. Subsequent research in Japan has shown that E.coli O157 bacteria can survive inside the tissue of radish plants grown from contaminated seeds. In such circumstances, it would be impossible to eliminate infection by washing[108]. Most experts link cases of contaminated fruit and vegetables with polluted irrigation water (abroad) and with the disposal of animal waste and sewage sludge on agricultural land, with any contamination of soil being subsequently transferred to crops. The Food and Drink Federation and the Chilled Food Association both drew attention to the threat to food safety from waterborne protozoa such as Cryptosporidium, especially in relation to fruit and vegetables eaten raw[109]; provisional figures for 1997 show an increase of 18 per cent over the previous year in the number of gastrointestinal infections in England and Wales attributable to Cryptosporidium[110]. In the UK, E.coli and Salmonella in vegetables have caused some food poisoning incidents, though to nothing like the extent of meat and animal products. The ACMSF has established an ad hoc Group to examine the issues surrounding the agricultural disposal of sewage sludge, and Professor Georgala spoke of "evidence from different countries... of a growing capacity of fruit and vegetables to act as a vehicle for contamination"[111]. The Environment Sub-Committee of the Environment, Transport and Regional Affairs Committee has recently inquired into the subject of disposal of sewage sludge on agricultural land, in the context of the potential increase in such disposal resulting from the EU's Urban Waste Water Treatment Directive requirement that all disposal at sea should end by the end of this year, and registered a number of concerns[112]. Safeguards exist under the Sludge (Use in Agriculture) Regulations and their associated Code of Practice. We see no proof yet of problems in respect of food safety, but the practices of disposal of sewage sludge and other organic waste on to agricultural land must be kept under close and continuous review in the light of the ACMSF's eventual findings.