April 2002, pages 112-115 Do you understand evidence? David Lyons looks at the reasons why prosecutions can fail due to evidential problems and offers some advice for those involved in food safety investigations and prosecutions

The Food Standards Agency in its Microbiological Foodborne Disease Strategy (July 2001) estimated that 88 per cent of general outbreaks of foodborne disease were associated with restaurants, hotels, pubs or bars, and caterers. However, quite often when prosecutions follow, they are for matters other than the sale of the suspect food, suggesting that, for a variety of reasons, prosecutions for the sale of food are not being attempted. It is possible that this is because proving that implicated food failed to comply with food safety requirements (Food Safety Act 1990, S8) is not as easy as proving a contravention of the Food Premises (General Food Hygiene) Regulations 1995.

It is also arguable that the public expects that where a business sells food that subsequently causes foodborne disease, it is prosecuted for the sale of the food, not for the condition of the walls. That said, gathering and presenting evidence to secure a conviction under Section 8 of the Food Safety Act 1990 is difficult, and it is a job not made any easier by the rules and laws relating to criminal evidence.

Some appreciation of these rules and laws, and consideration of them during outbreak investigations, can significantly improve the chance of a prosecution concluding successfully. The enforcement policy of many authorities includes a reference to the Guide for Crown Prosecutors. This makes it a pre-requisite for potential prosecutions to pass an evidential test before being proceeded with. Obviously, a case containing defective evidence should not pass this hurdle. Ultimately, the chances of a favourable outcome for the prosecuting authority are enhanced when the case comes before a court. If a case does not succeed, then it will fail on its merits (or lack of them), rather than due to a technical defect in the evidence.

Even investigations which, on the face of it, yield up fairly compelling evidence can unravel in all kinds of unforeseen ways because of evidential problems. For example, in the case of R -v- Kashioulis, [2001] All ER 139, a prosecution following over 150 linked cases of Salmonella enteriditis pt4 concluded with "not guilty" verdicts on all but one of the counts on the indictment. This included acquittal on five counts alleging breaches of the Food Safety Act 1990, Section 8(1). The single guilty verdict related to a failure to be registered under the Meat Products (Hygiene) Regulations 1994, resulting in a 12-month conditional discharge which was subsequently quashed on appeal - perhaps not the most favourable conclusion to a two-day Crown Court hearing. While a number of factors combined to defeat the Section 8(1) charges, in the final analysis the unreliability of the microbiological evidence appeared to have proved fatal.¹ Before considering the Kashioulis case in more depth, it is worth considering what exactly "evidence" is.

It is said that courts can only deal in probabilities, not certainties² and therefore, at its simplest, evidence is "any material which tends to persuade the court of the truth or probability of some fact asserted before it"3 or "something which may satisfy an inquirer of a fact's existence".⁴

Evidence is assessed in terms of three distinct but inter-related factors - relevance, admissibility and weight. Generally, evidence which is relevant to an issue is admissible and anything that is not sufficiently relevant is inadmissible. The vast body of the law relating to evidence covers the exception to this general rule.

Evidence will be relevant (according to Lord Simon in DPP -v- Kilbourne⁵), if it logically probative or disprobative of some matter which requires proof. This means that "relevant evidence" is that which proves a matter, or makes proof of it more or less probable. An example of this may be the use of a till receipt to prove that food has been sold. Thus, for an item to be admitted into evidence, it must be relevant - and an item shown to be irrelevant will be excluded.Consequently, an effective way to exclude evidence is to mount an argument that shows it to be irrelevant.

There is no test that can be applied to determine if an item is relevant or not, only logic, common sense and general experience can be applied. Admission of evidence depends to a large degree on "generalisations". Accordingly, evidence founded on questionable, incongruous or illogical generalisations will be deemed irrelevant. In preparing evidence, it is useful therefore to examine the generalisations on which it is based and assess how satisfactory they are and how readily they might be accepted by a "lay person" without further explanation.

The party putting the evidence forward bears the burden of showing that it is relevant. Expert evidence may be required to persuade a court to accept the more debatable or exotic generalisations. In the Kashioulis case, the generalisation that eating food contaminated with salmonella caused salmonella food poisoning hit a stumbling block when the question of infective dose was raised.

Although irrelevant evidence will always be excluded, it is not axiomatic that all relevant evidence will be admitted. Evidence may be inadmissible because it falls foul of some rule of law or evidence, for example the Police and Criminal Evidence Act 1984 or the rule against hearsay. Again the law is not clear-cut on the issue. Whether and when a particular item is in or out, will depend on the probative job it has to do. To discover the probative role for an item requires an examination of what the party tendering the evidence is actually seeking to prove and the relevance of the item to that goal.

Once the relevance of the item to the objective has been determined, an assessment of it can then be made in the light of the law and the rules of evidence to decide on whether it is admissible or not. Weight, although connected to relevance, is still distinctly separate from it. The generalisations used to prove an item relevant also operate to give it weight. Like relevance, the more questionable the generalisation used to support the item the less likely it is to be accepted as true and therefore the less weight will be attached to it.

Evidence can also be excluded on a weight argument; the line of reasoning being that an item's weight does not give it sufficient probative value when set against the prejudicial effect of it. In this sense, weight and relevance arguments are broadly similar.

As anyone who has observed criminal proceedings will be aware, evidence also comes in various forms to prove the facts in issue which are those facts that are material to the offence in question. In Kashioulis, a significant fact in issue was whether the food contaminated with salmonella was actually contaminated at the time of sale.

First there is direct evidence, or evidence of a fact in issue given by a witness who came to it by their own senses. It may also be evidence from a document or item of real evidence produced before the court. Circumstantial evidence are facts that are not in issue as evidence of facts that are in issue. For example, the presence of salmonella in food sampled after it has been sold as evidence that the food in question was in fact contaminated at the time of sale.

Circumstantial evidence can sometimes be thought of as a cable. Properly strung together, the individual strands can combine to support a case without the loss of any one being terminal. However, the further away circumstantial evidence is from the ultimate fact, the weaker it is and the more vulnerable it is to being used to prove a contrary conclusion. Circumstantial evidence is sometimes confused with collateral fact. Collateral facts are those that relate to some supplementary matter which may influence the probability of a fact in issue being accepted or not, for example, evidence relating to a witness' bias. Real evidence is evidence that can be presented for examination by the court, for example a piece of equipment, or sample of food. Documentary evidence relates not only to the contents of documents but also to other recordings, eg video tapes and computer printouts (if it can be shown the computer was working properly at the time). If an original document is produced, this is considered to be primary or original evidence. But if its contents are brought before a court in some other way, eg by someone testifying as to its contents, then this is secondary evidence.

The testimony of a witness may therefore introduce several types of evidence for the court's consideration. Testimony itself is statements made by witnesses, generally on oath, in court. They may be oral statements or more rarely written ones (affidavits).

An investigation into an infectious disease or food poisoning outbreak generates evidence in many, if not all of the aforementioned categories and deciding what to use can be problematic. The rules of evidence do not require the greatest quantity of evidence to be produced; in fact they require almost the opposite. The old common law rule of "best evidence" required a prosecuting authority to lay before the court the best evidence that could be presented. So, producing an item for examination was preferable to a photograph or a verbal description. In the past, evidence that was shown not to be best evidence was excluded. Although the rule has fallen into abeyance now, it is still a useful principle to bear in mind when gathering evidence in contemplation of a prosecution. Today, it is unlikely that an item would be excluded simply for falling foul of the best evidence rule. However, if it is not best evidence it opens a crack in the prosecution allowing a damaging, if not fatal argument to be made to exclude an item on the basis of relevance or weight. Depending on what facts a case is predicated, this may ultimately lead to it, too, being irreparably damaged. Best evidence was also one of the acknowledged justifications for courts refusing to accept "hearsay" evidence. A failure to grasp and understand the rules and issues surrounding hearsay evidence can strangle the clearest cut cases at birth. The definitive statement on hearsay was made by Lord Havers in R -v- Sharp⁶, who stated that "an assertion, other than one made by a person while giving oral evidence in the proceedings, is inadmissible of any fact asserted". Undoubtedly, the problem for many people is actually recognising hearsay when it is encountered. A useful formula for its more successful identification is, first, to look at any communication made outside the court, whether it is verbal or written communication, or communication by conduct or gesture. Second, consider the use it is being put to. If it is being used to prove the truth of what was said or written, then it is hearsay. If the purpose is to prove that a statement was made, then it is not hearsay.⁷

For example, an officer taking a food history from someone who has suffered food poisoning and then reading it out in court, is not evidence that the person consumed the food in question. It is, however, evidence that they had a conversation with the officer. For the court to accept the person consumed the food, the person would at least have to provide a signed written statement to that effect and potentially give evidence under oath of the facts asserted. As an area of law, hearsay is as complex as it is confusing and therefore only a basic appreciation can be offered here. There are, though, related issues also worth bearing mind when gathering and presenting evidence in food poisoning investigations. The rule against narrative (or rule against self-corroboration) provides that evidence cannot be given of the fact that a witness has made previous statements that are consistent with their testimony under oath. To do so might lend their evidence artificial or inflated weight.

It is often argued that it is the complexity of the exceptions to the rule on hearsay that give its reputation. Some exceptions to the hearsay rule that may be relevant to a food poisoning investigation include the fact that recording media, eg video tapes, that directly record an incident are not caught by it. Nor are documents produced by automatic recording machines, for example temperature logging equipment. That said, where there has been some human input in the system, for example the inputting of information into a computer, the output will be inadmissible unless properly proved.⁸

Therefore, the taking of food histories for compiling into a statistical analysis using a computer program will require by necessity that the information is gathered in the form of a witness statement. This may mean, in extreme circumstances, having those witnesses available to be examined and cross-examined in court. This represents a significant burden in the course of an investigation, but an essential one if a successful prosecution is to follow. In certain circumstances set out in the Criminal Justice act 1988, written hearsay may be admissible if for a variety of reasons a witness cannot attend court. This is a different issue to documentary evidence. In the course of an outbreak investigation, through the exercise of powers conferred under the Food Safety Act 1990, officers may accumulate a substantial body of documentary evidence from suspect premises and their suppliers. For the contents of a document to be admissible as evidence, then either the original document must be produced or a copy authenticated to the courts satisfaction must be produced.⁸ Proper authentication will usually mean the person who had control of the document and who made the copy giving the evidence of that fact, and the fact that it is a true copy of the original. Therefore, where an officer takes into their possession a document, it is probably prudent not to pass it on to others for copying.

In certain specified circumstances, oral testimony may be admitted as evidence of the contents of a document. This is another exception to the hearsay rule, and includes situations where the original has been lost or destroyed, or where there is some other impossibility or inconvenience, for example a document outside the jurisdiction.

Documents and reports etc, referred to by court acknowledged experts are not covered by the provisions relating to documentary evidence. Strictly speaking, the use of such material would constitute a breach of the rule against hearsay, but a strict interpretation would constrain experts from giving their opinion. Expert witnesses are generally skilled in some particular discipline and they give evidence on matters that are outside the court's experience. It is for the court in each particular case to decide whom they will recognise as an expert. Recognised experts are permitted to express an opinion on the facts of a particular case provided those facts themselves have been proved by admissible evidence.

In prosecutions under the Food Safety Act 1990, it is not unusual for officers involved in food law enforcement to be recognised as experts by the court. Certainly in Kashioulis, which serves as an interesting case study, this was the case. As mentioned earlier, it involved 150 linked cases of salmonella enteritidis pt4. All the cases were linked to the consumption of chopped liver. A sample taken from the delicatessen where the offending food was produced yielded a positive result for the same species of salmonella. With the outbreak several days old, a further sample of the chopped liver was obtained from the freezer of a family who had been ill and it too proved to be positive. Further positive samples of tuna salad, and an egg and onion sandwich filling (also produced by the delicatessen) were obtained from the fridge of another victim.

The conclusion looked clear - people had become ill by consuming food contaminated with salmonella produced by the defendant. The evidence being the 150 positive cases, the positive sample from the delicatessen, the positive sample from the freezer and the positive sample from the fridge. However, this was not as strong as it appeared at first glance. Proper food histories were not taken; indeed it appears that the idea of chopped liver being the source was already firmly rooted and therefore only confirmation as to whether this product was consumed or not was sought.

Given that salmonella enteritidis pt4 is the most common species of salmonella in England, and is associated with many other foods, the failure to take food histories and exclude other foods can be seen as depriving this evidence of substantial weight. Further, the presence of salmonella in the suspect food can only amount to circumstantial evidence, and expert evidence would probably be needed to convince a court of the link between its presence in the food and the illnesses caused.

Undoubtedly a key plank of the prosecution was the presence of the salmonella species in the food samples. However, this evidence was seriously undermined when the defence was able to demonstrate that the samples had not been taken in accordance with Approved Code of Practice No.7 ("sampling for analysis or examination").

The failure was not trivial, as the defence was able to successfully contest that the samples had been taken aseptically. This was done by having the officer demonstrate their sampling technique in the witness box. It did not help, either, that officers in cross-examination were, apparently, unable to adequately explain the difference between sterilisation and disinfection. Matters were further hampered by the fact that the positive sample from the delicatessen was the only one from 11 food samples and 23 environmental swabs. The information on the negative samples, which the defence would have access to by way of the provisions of the Criminal Procedures and Investigations Act 1996, would serve to dilute the effect and weight of the positive sample. All this provided ample opportunities for the defence experts to discredit the results and call in to question the expertise of the investigating officers. Thus, immediately calling into doubt the validity of the sample results and introducing an alternative and plausible explanation for the presence of the salmonella.

Furthermore, in respect of the samples obtained from the fridge and freezer, the prosecution again seems to have assumed that there was salmonella present before they were sampled, and therefore, the salmonella was present at the time of sale. Caution would seem to indicate that expert testimony should have been introduced to offer an opinion on this point and to discount other potential explanations, such as the food being cross-contaminated by something else in the fridge, or by someone else who had access to the fridge. Ultimately, these deficiencies opened the door for the defence experts to invite in reasonable doubt and thus bring the matter to a conclusion that favoured the defendant.

In the final analysis, the experts for the defence prevailed, but they certainly had their way eased by the weaknesses in the prosecution's evidence. That said, hindsight is a wonderful thing and while we can be critical of the manner in which the prosecution was prepared, we should be more concerned about learning from the experience to prevent the same mistakes from recurring in future.

David Lyons,
Group Manager (Public Safety),
Middlesbrough Council

References

1. Dr Slim Dinsdale, "Microbiological results from foods" New Law Journal, 26 May 2000.
2. Allen C, Practical guide to evidence, 1999.
3. Murphy P, Murphy on evidence (5th edition) 1995.
4. Tapper C, Cross and tapper on evidence (9th edition) 1999.
5. [1973] AC 729.
6. (1988) 86 Cr App R 274.
7. Subramaniam -v- Public Prosecutor [1956] 1 WLR 965.
8. R -v- Coventry JJ ex parte Bullard (1992) 95 Cr App R 175.
9. Criminal Justice Act 1988 - Section 27.