In a recent blog post, Maximilian Hardy of 9 Bedford Row warned that reckless transmission of disease is a crime, and those diagnosed with coronavirus should be wary of the legal consequences that follow. This is sensible advice, but may not go far enough. This blog post explores when a person who has symptoms of coronavirus but has not been diagnosed may find themselves subject to criminal liability. It does not attempt a comprehensive analysis of the possible offences, but addresses three key issues.

First, can a person commit s. 20 GBH by recklessly transmitting a disease they do not know they have? Second, what modes of disease transmission will amount to a battery, and so ground a conviction of assault occasioning actual bodily harm? Third, can a person face criminal liability even without transmitting the virus, or perhaps having it at all, under the offence of criminal public nuisance?

I. Recklessness for s. 20 GBH

A person who intentionally or recklessly causes another to suffer grievous bodily harm commits an offence under the Offences Against the Person Act 1861, s. 20. Where D infects V with coronavirus, and V suffers severe symptoms, D causes V to suffer grievous bodily harm. This section asks whether D does so recklessly where D fails to self-isolate after developing symptoms of coronavirus.

D is reckless if they foresee harm, yet unjustifiably go on to take the risk of it (Cunningham [1957] 2 QB 396R v G [2003] 3 WLR 1060). For GBH, D only needs to foresee some harm, not the specific harm which occurs (Mowatt [1968] 1 QB 421). So long as D realises that their symptoms mean they might have coronavirus, and knows that their conduct poses a risk of infecting another, D foresees harm. The questions are (1) whether the fact that D only suspects, but does not know, that they are infected precludes recklessness, and (2) whether D’s conduct is unjustifiable in light of the risk.

The CPS states, in the context of recklessly transmitted STIs, that the prosecutor must show that the defendant knew, or would have known if they did not turn a blind eye, that they were infected. This view is shared by academic literature[1], including literature cited with approval in the Court of Appeal.[2]There have been successful convictions for reckless transmission where the defendant knew they were infected (e.g. Konzani [2005] EWCA Crim 706) and where the defendant turned a blind eye (Adaye, reported inThe Times, January 10, 2004[3]), but none where the defendant only suspected they might be infected. It might therefore be presumed that D must either know they are infected, or turn a blind eye. If so, many cases of coronavirus transmission by individuals who have symptoms but have not been diagnosed will be excluded. Only a person whose symptoms leave no room for any diagnosis but coronavirus can be characterised as turning a blind eye. Anything else and D only suspects they have, or might have, coronavirus which is not enough.

However, there is reason to doubt that a conviction for recklessly transmitting disease under s. 20 requires knowledge or turning a blind eye. First, this requirement is not stated in any case. The strongest authority is the reference to John Spencer’s article, which takes this view, in Dica [2004] QB 1257.[2] But this is not part of the ratio in Dica, so is not binding precedent. Second, the definition of recklessness leaves little space for a rigid ‘knowledge’ requirement. If knowledge were a requirement, it would have to fall under either the ‘foresight’ or ‘unjustifiable risk-taking’ elements of recklessness. It cannot fall under ‘foresight’ as it is possible to foresee a risk of transmitting disease though you only suspect you are infected. It must therefore fall under the ‘unjustifiable risk-taking’ element, and amount to an assertion that it will never be unjustifiable to take a risk of transmitting a disease unless you know, or would know if you were not turning a blind eye, that you are infected with that disease. Perhaps ordinarily it will never be unjustifiable for a person to risk disease transmission when they only suspect they are infected, because of the ‘novel and significant positive obligations’ which would otherwise be imposed.[4] But in extraordinary circumstances, what is ordinarily justifiable may become unjustifiable. The issue of knowledge should therefore be assessed as a factor feeding into whether D’s conduct is justifiable, rather than dealt with by a general rule which is both unsupported by precedent and analytically unconvincing.

No case has set out in full the factors to be taken into account in assessing justifiability. Professor John Spencer QC has suggested a number of factors, including how serious the harm would be if it happened, how likely the harm is to occur, and what precautions would be necessary to avoid the risk.[5] Whether D is reckless would therefore turn on the facts of the particular case, and require evidence of the risk of transmission created by D’s precise conduct, coupled with evidence of the likelihood of severe harm resulting from that transmission. However, three general points can be made. First, it is not enough to make D reckless that PHE guidance says D ought to have self-isolated. Something may be unjustifiable from a broad public health perspective, while justifiable from the perspective of the likelihood of individual harm resulting. Second, that PHE guidance is nonetheless part of a changed social context, which must feed into the assessment. When society has decided that significant restrictions on individual liberty are a price worth paying to stem the spread of disease, the concern that considering D reckless in these circumstances severely impinges on D’s liberty loses weight. Third, it becomes easier to characterise D as reckless by more narrowly describing D’s conduct. If we say ‘D was reckless by failing to self-isolate’, personal liberty considerations become more weighty, as we imply that the alternative to D’s conduct was effective home-arrest. If we instead say ‘D was reckless by not coughing into their arm’, or ‘D was reckless by failing to wash their hands before leaving home’, those concerns are diminished. However, this could only be done if it could be shown that D was aware that that particular conduct posed an increased risk of transmitting the virus.

In my view, there will be cases where a person is reckless, and so commits s. 20 GBH, in transmitting coronavirus to another, who suffers severe symptoms, after failing to self-isolate. The significant change in context represented by the lock-down justifies departing from the normal view that actual knowledge of infection, or close to it, is necessary, so that where a person’s conduct is particularly egregious (i.e. not just going for a walk, but straying close to others and coughing, or coughing into the air instead of into their arm) they will be acting recklessly. But, as the above discussion makes clear, there is significant uncertainty about where the line would be drawn.

II. Assault occasioning actual bodily harm

Not every transmission of coronavirus will result in grievous bodily harm. Where the harm is lesser, but still serious, it will amount to actual bodily harm. But it is no crime simply to cause another to suffer actual bodily harm. To commit an offence, D must cause that harm by means of an assault or battery: s. 47 Offences Against the Person Act 1861. This section asks what modes of disease transmission will amount to a battery.

There is a battery if D unlawfully touches V. Direct physical contact, for example a handshake, is plainly touching. Though everyday touching is not a battery (Collins v Wilcock [1984] 3 All ER 374, per Goff LJ), in our socially-distanced lives casual touching is no longer a part of everyday life. And V’s consent to a handshake may be ineffective if they do not know D might be infected with coronavirus (by analogy with Dica [2004] QB 1257). If so, the touching would be unlawful and amount to a battery. Against this, D might argue that all, or most, of us are potentially infected, as there is normally a period of two to fourteen days between contracting the virus and developing symptoms. Hence V does consent to a risk of disease transmission by accepting a handshake, or other contact, from D. The cases do not address whether consent must be to the precise risk, or simply to the existence of any risk which is more than negligible. However, though the basic principle behind D’s argument seems sound, it should be rejected as inapplicable here. Consent to a risk of transmission from someone who shows no symptoms of coronavirus is substantially different from consent to a risk of transmission from someone who does show symptoms. Hence, unless D discloses their symptoms, direct physical contact between D and V would be a battery.

Likewise, there is a battery if D coughs directly onto V. It is less clear whether there is a battery where D coughs onto a surface which V later touches, or into air which V later breathes. A battery may be indirect, as in DPP v K [1990] 1 All ER 331, where the defendant poured acid into a hand-drier later used by the victim. There is an obvious analogy between leaving acid in a dispenser and leaving a virus on a surface or in the air. However, the courts should reject any attempt to characterise the latter as a battery. It strains language too far to characterise leaving a virus on a surface which V later touches as touching V. If this is D touching V, then surely any person who catches the virus from V is also touched by D, however indirectly. The word ‘touching’ is reduced to meaninglessness if stretched so far.

Hence, there is a battery where D transmits the virus to V through direct physical contact, or coughing onto V. But there is no battery, and hence no offence under s. 47, where D transmits the virus indirectly.

III. Criminal Public Nuisance

Some who have symptoms of Covid-19 are not infected with coronavirus. But we are advised to self-isolate even without certainty that we are infected, because the risk that we might be is judged too great. If D’s failure to self-isolate will only result in an offence where it transpires D is infected, and D transmits the virus to another while out, then there is an element of chance involved. From a preventative perspective, this might be considered undesirable. It is therefore worth asking whether any other event imposes liability in these circumstances. Criminal public nuisance may do so.

It is a criminal public nuisance to do an act not warranted by law which endangers the life, health, property or comfort of the public (Goldstein [2004] 2 All ER 589 at [3]). Exposing the public to the risk of disease may be a criminal public nuisance. For example, in Vantandillo (1815) 4 M & S 73, a woman committed a criminal public nuisance by carrying her smallpox-infected child through a public highway. However, there are two difficulties in applying the offence here.

First, there is little clarity about which diseases can ground a public nuisance. Smallpox has done so (Vantandillo), as has glanders (Henson (1852) Dears 24). The indictment in Henson described the disease as “contagious, infectious, and dangerous”. The indictment in Vantandillo used the same phrasing, while the judgment of Le Blanc J simply referred to a “contagious disorder”. Today, it is safe to expect that if a person were prosecuted under public nuisance for exposing the public to a risk of disease, the court would clarify that there must be a high threshold of infectiousness and dangerousness. To do otherwise would effectively place under house large numbers of people who fall ill. However, coronavirus surely satisfies even a high threshold. The wide-spread measures taken by countries around the world testify to the threat it poses. 

Second, it is not clear that a person, D, who fails to self-isolate when not certain they have coronavirus, necessarily endangers the health of the public. If D is infected then they endanger public health by going out, as they create a risk of disease transmission. If D is not infected then they cannot create any risk of transmission. But at the time, and perhaps afterwards depending on how precisely antibody tests can tell us when a person was infected, we cannot know whether D is infected. Hence, we know only that by going out they might create a risk of transmission.

In ordinary language, D endangers public health by going out when they might be infected. But if they are not infected, D could argue that though they riskedendangering public health, they did not actuallyendanger public health, as nobody could have been infected by their going out. Though the former is probably closer to the ordinary meaning of ‘endanger’, there are two reasons to prefer the latter approach. First, the only criminal law authority is in cases involving an actual risk of disease transmission (e.g. VantandilloHenson). Second, cases on tortious public nuisance establish that potential creation of a risk is not sufficient. In cases on smallpox-hospitals, the possibility that smallpox might transmit through the air could not found a tort (Attorney General v Corporation of Nottingham [1904] 1 Ch 673;Withington Local Board of Health v Corporation of Manchester [1893] 2 Ch 19). There was only a nuisance where evidence of actual public health impact showed that a risk was created (Hill v Metropolitan Asylum District (1879) 4 QBD 433). 

This is not conclusive. But a court should err towards caution in expanding the scope of criminal offences. As Lord Bingham has stated, in the context of public nuisance, “existing offences may not be extended to cover facts which did not previously constitute a criminal offence” (Goldstein at [35]). There is a substantial difference between creating a risk of infection and risking creating a risk of infection. Given this, and the tort law authority, it is unlikely that the offence of public nuisance would be extended to cover this situation.

This leaves minimal room for the offence of public nuisance to apply. First, it must be shown that D was infected with coronavirus at the relevant time. Second, public nuisance requires that the defendant knew or ought to have known that their conduct would cause a nuisance (Shorrock[1994] QB 279). Where D’s symptoms, for example, are nothing more than a persistent cough, it is impossible to say that they ought to have known they were infected and hence that they would cause a public nuisance by failing to self-isolate. D will only commit an offence where their symptoms are such that they ought to know they areinfected, not that they might be infected. This is a very high threshold to clear.


[1] S Ryan, ‘Reckless transmission of HIV: knowledge and culpability’ (2006) Criminal Law Review981 at 982-984.

[2] JR Spencer, ‘Liability for Reckless Infection: Part 2’ (2004) 154 New Law Journal 448 at 471, described as “illuminating” by Judge LJ in Dica [2004] QB 1257 at 1272.

[3] Discussed by S Ryan, n 1, at 984.

[4] M Weait, ‘Criminal Law and the Sexual Transmission of HIV: R v Dica(2005)’ Modern Law Review121 at 131, though note he makes this point to further his argument that even turning a blind eye should not be sufficient.

[5] JR Spencer QC, ‘Liability for reckless infection – part 1’ 154 New Law Journal (2004) 384.