Food poisoning cases are relatively common; medical professionals who make mistakes are at risk of adverse headlines; drug dealers cause moral outrage when their customers die, or they fail to seek medical attention for friends who have overdosed. Increasingly, the legal options include criminal prosecution.

This article looks beyond general negligence principles and considers how criminal liability has taken shape in recent cases. An analysis of some case examples shows a distinct lack of legal principle and wide factual tests for juries, which leave businesses, healthcare professionals and the nefarious open to criminal prosecution for manslaughter with little certainty as to outcome. Lord Atkin in Andrews v DPP [1937] AC 576, said:

‘of all crimes manslaughter appears to afford most difficulties of definition, for it concerns homicide in so many and so varying conditions’ .

We conclude that the drive for factual tests has not made the task of determining liability any simpler in medical, food and drug related deaths.

Medical cases

The House of Lords decision in R v Adomako [1995] 1 AC 171 was the leading case on manslaughter by gross negligence. In R v Rose [2017] EWCA 1168, where an optometrist failed to conduct a proper eye exam on a young boy who died four months later of acute hydrocephalus secondary to undiagnosed papilloedema, the four stage Adomako test was expanded into five stages:

(a) duty of care;

(b) negligent breach;

(c) reasonably foreseeable that the breach gave rise to a serious and obvious risk of death;

(d) breach caused the death;

(e) circumstances of the breach were truly exceptionally bad and so reprehensible as to justify the conclusion that it amounted to GN and required criminal sanction.

The point in paragraph (c) on the risk of death is not new, but the emphasis on a serious and obvious risk of death being reasonably foreseeable, based on what the defendant knew at that time, is significant. Controversially, the fact that Rose failed to comply with her statutory responsibility to conduct a proper eye exam, and thus saw no evidence of papilloedema, was of critical relevance on this revised formulation. This test (reasonable foreseeability of a serious and obvious risk to the reasonable person in the defendant’s position) appears intended to mitigate the potential breadth (and harshness) of gross negligence manslaughter. It has also been separated from the gross negligence element (para e) when the Adomako test previously approached both as part of same question.

Has R v Rose changed or just re-formatted the common law? This remains a moot point, but it appears to narrow the offence and arguably makes successful prosecution of medics a tougher task. What is meant by serious and obvious? And why are both needed? Jurors might (not unreasonably) think that obvious risks of death are serious and that serious risks should be obvious. There has been little clarification, beyond noting that ‘serious’ qualifies the nature of the risk of death as something much more than minimal or remote. An obvious risk is one that is present, clear, and unambiguous. It is immediately apparent, striking and glaring rather than something that might become apparent on further investigation. In every case, the factual matrix will be critical and there remains no ‘bright line’ test of liability.

Food cases

The complexities in food cases were recently highlighted in R v Rebelo (No1) [2019] EWCA Crim 633, followed by R v Rebelo (No2) [2021] EWCA Crim 306. Those appeals from the trial and retrial related to the supply (by post) of diet pills in capsule form made from an industrial chemical. The indictment alleged a food regulatory offence and both unlawful and dangerous act manslaughter (UDAM) and gross negligence manslaughter (GNM). The evidence was that the appellant supplied the capsules by post following an online order from the deceased. The chemical was not combustible so was not dangerous per se, but experts gave evidence that it was unpredictable and thus dangerous, if ingested. Experts also gave evidence that the deceased was vulnerable as someone suffering from an eating disorder from her medical notes, and that she lacked capacity to make sensible decisions about her eating. Her computer records showed that she had done considerable research, knew exactly what she was taking and that she was highly intelligent so appeared to understand the risks. Text messages tragically recorded her dying communications suggesting she had taken an erroneous dose.

Put shortly, the Court of Appeal in Rebelo (No 1) accepted that posting such an item (unlike throwing concrete from a bridge – see DPP v Newbury [1977] AC 500) is not UDAM that caused her death. The failure to comply with a food regulation was not sufficient. The Court of Appeal, led by Leveson P, rejected the prosecution’s suggestion that causation was proved if the deceased lacked capacity and accepted defence submissions to apply the classic definition of autonomy – whether she was ‘free, informed and deliberate’ such that the ‘chain of causation’ from the supplier was broken. This was explained and reaffirmed in Rebelo (No2) with Dame Victoria Sharp P now in the Chair, following Sir Brian Leveson’s retirement. Mr Rebelo was convicted at retrial of GNM and his subsequent appeal was dismissed.

This does not end the complexities in food safety cases: R v Kuddus [2019] EWCA Crim 837 involved a restaurant owner and chef who prepared a meal containing peanut proteins for a girl who identified this allergy on the takeaway order. In both Rose and Kuddus, the Court of Appeal concluded that neither defendant knew about circumstances which made the risk of death serious and obvious – even though they should have known. This poses a key question – is gross negligence an objective or subjective form of fault? In theory it is supposed to be objective – after all, that is the whole point of negligence liability. In reality, of course, the common law has created something of a hybrid model with recklessness in play in terms of assessing gross negligence. Rose and Kuddus are clear examples where absence of recklessness remains critical. Ultimately, the reasonably foreseeable serious and obvious risk of death (to someone in the defendant’s position) really does dilute the objective nature of gross negligence.

Drug cases

In R v Broughton [2020] EWCA Crim 1093 a 24-year-old woman died at a music festival having taken a ‘bumped up’ dose of 2C-P, a Class A drug, supplied by her boyfriend, the appellant. It was alleged that he was grossly negligent in failing to obtain the on-site medical assistance and that this failure caused her death. His conviction was quashed on the basis that the case should have been withdrawn from jury as the evidence could not establish causation to the criminal standard. There was a single expert on causation who estimated the chances of survival with medical care at 90% but that survival could not be guaranteed. The court’s approach is set out in paras 23 and 100:

‘The prosecution must prove to the criminal standard that the gross negligence was at least a substantial contributory cause of death. That means that the prosecution must prove that the deceased would have lived in the sense that life would have been significantly prolonged.’

‘In the context of causation in this very sad case the task of the jury was to ask whether the evidence established to the criminal standard that, with medical intervention as soon as possible after Louella’s condition presented a serious and obvious risk of death, she would have lived. ’

Requiring the Crown to prove that the deceased ‘would have lived’ appears, on the face of it, to place the bar impossibly high. Perhaps, the unstated view in Broughton was that the dominant cause of death was ingesting the drugs, and not the failure to obtain assistance. Further, that the drugs supplied by the appellant were less likely to cause death than heroin, as in the case of R v Evans [2009] EWCA Crim 650, an otherwise similar scenario where the conviction was upheld. It also reflects the continuing unease with manslaughter by omitting to seek assistance on the basis that omissions do not start causal processes but rather allow other causal processes to unfold. However, in Broughton, the appellant supplied the drugs, so this point falls away.

In Rebelo (No1) the Court of Appeal sought to assist the lower courts with a potential formulation (see para 76). In Rebelo (No2) the court held that this was not intended to be prescriptive. The principle is that the breach of duty must be a substantial and operative cause of death and the breach of duty would not be a cause of death if the deceased had or might have made a fully free, voluntary, and informed decision. The expression ‘fully free, voluntary and informed’ meant her ‘ability to assess the risk and understand the consequences relating to the toxicity of the substance and her appreciation of the risk to her health’. This is described as a balancing exercise in order to decide whether the prosecution had established that a defendant’s breach of duty was a substantial and operative cause of death, even if it was not the sole such cause; that, only if the deceased’s decision to take the capsules was a fully free, voluntary and informed decision, or might have been, would her death, as a matter of law, have been caused by her free choice because, in those circumstances, the defendant had only set the scene for her to make that decision, but he would not have caused her death.

Finally, a note on alcohol consumption. The Court of Appeal held in R v Field [2021] EWCA Crim 380 that it was open to a jury to convict of murder, notwithstanding that the deceased had voluntarily consumed the alcohol and/or medication that caused his death. Kennedy (No.2) [2007] UKHL 38 was distinguished because the deceased did not have an informed appreciation of the ‘truly perilous nature of what was occurring’ (per Fulford LJ at [60]) and the deceased was induced into a dangerous situation in the company of an individual with murderous intent. The distinction in some cases will be hard to draw and it remains to be seen if there will be a temptation to charge murder with GNM alternatives which will further complicate this area of law.

Issues of liability will continue to test us

A range of authorities were discussed in each of the appeals we have looked at above, and it seems that the law is intended to be settled. However, leaving such complex issues as questions of fact, in the huge range of circumstances in which it might be said that a person has failed to use reasonable care and skill, puts numerous professionals at risk of prosecution when dealing with vulnerable people, even ones they have never met. At the same time, the bracket for seemingly callous members of the public, certainly in drug and alcohol related cases, seems narrower. These are issues of liability which will continue to test trial advocates, judges, and juries for some significant time to come and, in the absence of clarity of law, charging criminal offences seems more likely when a death occurs. Lord Atkin’s words remain true – outcomes are still notoriously difficult to predict.