Battle of the bugs
For 12 years, I have been making Lanark Blue cheese from raw ewes' milk, provided by a flock of 400 dairy ewes kept on our farm 1,000ft above sea level in the upper Clyde Valley. Our staff of six have always understood that hygiene is a priority - ours was the first cheese farm to receive accreditation with the Food from Britain quality scheme for hand-made cheese, and we follow a systematic testing programme for both milk and cheese.
Yet in December 1994, Clydesdale District Council environmental health officers (EHOs) told us that the Lanark Blue cheese they had sampled showed the presence of the bacterium Listeria monocytogenes (Lm) - at levels higher than 1,000,000 colony forming units (cfu) per gram in some cases.
The EHOs said they feared a nationwide outbreak of human listeriosis - carrying a 30-40% mortality risk - from consumption of Lanark Blue. Because of this, they insisted on a product recall, and sought our agreement to shut production down until we could satisfy them as to a total absence of Lm. A public health warning for Lanark Blue cheese was issued throughout the UK.
Over the course of the subsequent year, one court hearing after another ensued as we desperately fought to prevent the EHOs condemning the entire stock of cheese. Finally in December 1995, the sheriff at Lanark issued his judgment: the cheese was fit for consumption after all.
While the legal battle was going on, it received keen press interest, and the public support was amazing. We received 1,000 letters, and donations to our fighting fund of more than £30,000. Some of this support stemmed from a perception of a little guy being bullied by mindless bureaucrats, but mainly the interest arose from the issues at stake.
Proper risk assessment
The 1990 Food Act allows a magistrate - or, in Scotland, a sheriff - to condemn a food if the balance of evidence suggests this food is unfit for consumption. The compilation of the evidence upon which this decision is based is known as risk assessment. It includes looking at epidemiological factors: signs of human illness. In our case, there were no such signs. Furthermore, the sheriff found that the EHOs' tests were unreliable, but that the tests carried out for us - showing an absence or low levels of Lm - were reliable.
Another important microbiological factor was the pathogenicity of the specific strain of bacteria involved. The sheriff found that no evidence of pathogenicity was offered. Also, he accepted that our standards of production were good.
Putting their case, the EHOs accepted the lack of evidence against the cheese, but argued that it might cause harm. This boils down to trying to persuade the sheriff to discard the balance of evidence principle enshrined in the Food Act, and adopt, in its place, what might be termed the "precautionary" principle.
This is the same argument involved, I believe, in the debate over BSE - in which, of course, the precautionary principle won the day.
If such a "better safe than sorry" approach had been accepted by the sheriff, no food containing bacteria would now be safe from the risk of EHO attempts to condemn it. All bacteria are "potentially" harmful; it is impossible to prove a negative in microbiology - that is, that an organism will never harm anyone.
Is raw milk cheese safe to eat?
Food lovers agree the best cheese is made from raw milk because it is rich in the flavour-producing microflora that pasteurisation kills.
However, many public health officials believe raw milk cheese should be made illegal because it is intrinsically dangerous. This belief prompted the sheriff to describe the head of the Listeria Reference Unit of the Public Health Laboratory Service (PHLS), who gave evidence as an expert witness against Lanark Blue, as "lacking in objectivity, and finding it necessary to support a view at all cost rather than approaching matters in a measured and balanced way as a scientist."
Having rejected the PHLS prejudice against raw milk cheese, the sheriff accepted the evidence of our witness, Jean-Jacques Devoyod, a French scientific expert on cheese. He argued that raw cheese is intrinsically safer than pasteurised cheese. This is partly because fresh raw milk contains substances which prevent pathogens from multiplying. In addition, the raw milk cheese-making process involves the production of acetic acid (as well as lactic acid), an inhibitor of pathogens. Raw milk cheese also contains the bacteria enterococci, which are hostile to Lm.
Nonetheless, raw milk cheese may sometimes be capable of causing illness, with three situations giving rise to this possibility.
One is process contamination, where dirt is located somewhere in the production line. However, any responsible cheese-maker has systems in place to control this hazard. Devoyod demonstrated that the risk of a raw milk cheese being affected is significantly lower than that with a pasteurised one, because the pasteurised milk provides the ideal environment for pathogens to invade and multiply.
Another risk is starter activity failing. If this occurs, it will show up clearly in the records that all cheesemakers keep - our records showed starter activity had been consistently good.
Lastly, there is the risk of illness among the animals that produce the milk for a raw milk cheese. In our case, the sheriff accepted our veterinary evidence that our sheep had not suffered any listeric illness. We did not deny that Lm had sometimes been present in the milk and therefore - at low levels - in the cheese. However, this does not mean this cheese is harmful, any more than Lm in human milk - not an uncommon occurrence - will threaten babies' health.
Testing, testing
The Lanark Blue case raises the issue of the standard of the tests that EHOs use for food samples, and the calibre of the advice they receive on interpreting these tests.
Microbiology is an inexact science, particularly when it deals with the complex microflora present in a blue cheese made from raw milk. In the medium used for detecting Lm, a variety of other bacteria will appear, and mistakes are possible unless laboratory staff have sufficient experience in testing this type of cheese.
In our case, the sheriff found that the EHO tests used on our Lanark Blue were biologically implausible. In addition, he accepted that failures of laboratory protocol would explain why the tests had gone wrong.
For information relevant to our samples, the EHO expert advisers derived their knowledge from the Listeria Reference Unit of the PHLS, which says all strains of Lm must be assumed to be harmful and that, whenever Lm occurs above a minimal level in a food, condemnation of the food is appropriate.
This view is based on data from the PHLS practice of serotyping bacterial strains. Serotyping is an expensive and skill-intensive technique, restricted in the UK to the PHLS. Strains of virtually all sero-groups have been associated in the past with human illness and therefore, the argument runs, all strains are potential pathogens.
However, serotyping as a method was developed in the 19th century, and microbiology has moved on since then. Molecular biology now shows that serotyping actually tells us nothing about the pathogenic potential of, for example, an Lm strain. On the other hand, we do now know that many - perhaps most - Lm strains are harmless.
Furthermore, new methods of strain identification will identify possible strain pathogenicity. The PHLS does not have the facilities to apply these new methods.
As a result of the Lanark Blue case, and the exposure of the serotyping technique's inadequacy, questions must be asked about the quality of PHLS advice. I believe that the PHLS's ignorance of raw milk cheese microbiology is profound.
Where now?
From a personal standpoint, it is obviously a huge relief to be vindicated by the sheriff's judgment of December 1995. In February this year, the sheriff granted our applications for expenses, on the grounds of a common-law entitlement to expenses in a situation where Clydesdale District Council had been out to win at all costs and whatever the evidence. So eventually - we have not been paid a penny yet - we will be out of pocket by about £20,000, rather than the hundreds of thousands of pounds which the dispute cost the parties involved.
From a wider perspective, though, the situation for many specialist food producers remains worrying, not least because of the decision to abandon the principle of looking at the evidence, in favour of applying the "precautionary" principle - as in the BSE problem.
Microbial standards
There are directions from Brussels designed to impose the concept of "microbial standards" on the food laws of member states. These standards are already incorporated into the UK 1990 Food Act, with a temporary exemption granted to traditional cheeses.
However, if - as seems likely - this exemption is withdrawn in the near future, a Lanark Blue-type of dispute between a cheese-maker and an EHO would not even reach a court of law. Condemnation of the cheese could be automatic upon the detection of Lm at any level in an EHO test. The need for the EHO to offer evidence indicating an unacceptable health risk would not arise - pleas of cheese-makers on this point have been ignored by ministers.
A further cause for anxiety is that there is no appeal out of a food condemnation order, or any indication that the Government will consider changing the law to allow for an appeals procedure.
In fact, the defence of Lanark Blue was only possible - and ultimately successful - because at an early court hearing (February 1995), our magistrate refused to allow us to cross-examine the evidence of the PHLS. This refusal allowed us to seek judicial review of the decision to condemn the cheese. Had the magistrate not made a mess of things - in judicial terms - at that early hearing, we would have been beaten.
