Corps de l’article

Introduction

Over the past ten years in Canada, we have seen an increase in prosecutions against (mostly) men who fail to disclose their HIV-positive status to their sexual partners.[1] The Crown has relied on charges ranging in severity from nuisance[2] through to murder.[3] Most notably, in 1998, the Supreme Court of Canada held in R. v. Cuerrier that the failure to disclose one’s HIV-positive status, where this creates a significant risk of serious bodily harm to the complainant, constitutes fraud and thereby negates consent to sexual activity. When this is combined with the risk of transmitting HIV, the crimes of aggravated assault and aggravated sexual assault are made out.[4] The Court complicated this finding in R. v. Williams, where it held that, in order to establish aggravated assault or aggravated sexual assault, the Crown would have to prove that the complainant was HIV-negative at the time the accused failed to disclose his or her status. If this cannot be proven, then the proper verdict is attempted aggravated (sexual) assault.[5]

Many articles have been written on the pros and cons of the criminalization of nondisclosure,[6] and this case comment does not revisit that debate. Rather, it assumes that criminalization is here to stay, at least for the immediate future. The Supreme Court of Canada has considered this issue and upheld criminalization on three occasions without a single dissenting voice.[7] This comment focuses on how best to limit and define the contours of criminalization, looking specifically at the relevance of condom use and viral load.

R. v. Mabior[8] is but one of many cases dealing with serious criminal charges against an accused for failing to disclose his or her HIV status. While Mabior is only a trial decision, there are two difficult and important aspects of the decision that merit attention. First, Mabior is the only case in Canada to criminalize nondisclosure in the context of protected sex.[9] Second, Mabior is the only case to hold that, in the context of protected sex, a very low or undetectable viral load can sufficiently reduce the risk of serious bodily harm to preclude a finding of fraud negating consent. This comment will focus on these two issues and highlight the complexity of nondisclosure prosecutions. I argue that encouraging condom use is so important, and that the use of condoms reduces the risk of transmission so significantly, that the criminal law should distinguish between protected and unprotected sex in cases of nondisclosure. In contrast, however, I argue that viral load is not yet a manageable standard on which to base culpability.

I. The Facts in Mabior

The accused in Mabior was charged with ten counts of aggravated sexual assault and one count each of forcible confinement, invitation to sexual touching, and sexual interference. Only the aggravated sexual assault charges are the subject of this comment.[10] The evidence indicated that the accused sought out teenage runaways from vulnerable backgrounds, one as young as twelve years old at the time of the offence, by offering them drugs, alcohol and a place to stay. As of the date of trial, none of the complainants had tested positive for HIV.[11]

Mabior learned that he was HIV-positive in January of 2004. After receiving this information, he had extensive involvement with the public health system. Public health officials provided him with information regarding HIV and told him to practise safer sex and to disclose his HIV status to his sexual partners. Mabior was also warned of the potential criminal liability that could attach to nondisclosure. He received condoms from the health unit on a regular basis. However, he was diagnosed with gonorrhea and was a named as a contact person for chlamydia, facts which were used at trial to cast doubt on the assertion that he used condoms appropriately in sexual activity.[12]

Between 1 January 2004 and 31 March 2006, the accused had sexual intercourse with the nine complainants, sometimes with the use of a condom and sometimes without. In no case did he disclose his HIV status prior to sexual intercourse. He was, with one exception, compliant with his antiretroviral medication regime which brought his viral load down to an undetectable level from October 2004 to December 2005. However, neither public health officials nor his doctor ever suggested to Mabior that an undetectable viral load meant the virus was not transmissible.[13]

II. The Position of the Parties

The Crown took the position that nondisclosure should be criminalized regardless of whether the accused used a condom because of the “unreliability” of condoms. It argued that the fact that condoms had broken on three or four occasions with one complainant demonstrated that the accused was not using condoms properly. Further, the Crown pointed to the fact that there was no evidence that the virus could not be transmitted when a person’s viral load was undetectable. Even though that risk was very low, the potential consequences of transmission were so “lethal” that even a minimal risk was not one that a complainant should be expected to bear.[14]

The defence, in contrast, argued that protected sex should not be criminalized. Further, it was argued that it was highly probable that the accused could not have transmitted the virus when his viral count was low or undetectable. Defence counsel claimed that the accused knew that his viral load was low and that he knew there was little or no risk of transmission. The accused relied on the fact that none of the complainants had been infected to support his argument that there was a very high probability that he could not have transmitted HIV during the relevant time period.[15]

III. The Judgment

Justice McKelvey began her analysis by considering whether the Crown could establish that the accused had endangered the life of the complainant, a required element of the charge of aggravated assault. She found that this endangerment was established regardless of whether the accused used a condom. The finding of endangerment was based on the 20 per cent failure rate of condoms cited by the one expert witness at trial. In instances where the condom may have broken or fallen off, the trial judge found that this was the equivalent of no condom being used. The finding of endangerment was not negated during the period in which the accused had an undetectable viral load. The trial judge held that, although the risk was greatly reduced, the scientific evidence before her supported the position that HIV could still be transmitted when the accused’s viral load was undetectable, particularly if he or she had another sexually transmitted disease.[16]

In rejecting the relevance of viral load, Justice McKelvey stated that she was “prepared to follow the Supreme Court [in Cuerrier] in holding that the potentially lethal consequences of unprotected sexual contact leave room for no other conclusion than that endangerment of life has been substantiated.”[17] Justice McKelvey then considered whether Mabior’s conduct constituted fraud vitiating the consent of the complainants—an inquiry that might be seen as logically prior to the determination of endangerment. She referred to the elements of fraud from Cuerrier: deception and an attendant deprivation to the complainants. Justice McKelvey easily found the element of deception in the accused’s failure to disclose his HIV-positive status. The issue of deprivation focused on whether there was a significant risk of serious bodily harm to the complainants.

In the context of unprotected sex, this finding was straightforward and the trial judge found that nondisclosure constituted fraud negating consent to sexual activity. The analysis was more complex, however, in the context of protected sex. Given the expert evidence before her that condoms fail in 20 per cent of cases, the judge held that even in those circumstances where protection was used, a significant risk of serious bodily harm existed if medical evidence indicated that the accused was infectious.[18] However, where the accused’s viral load was very low or undetectable and the accused used a condom, these two factors together could reduce the risk of serious bodily harm such that there was no fraud vitiating consent:

With respect to the condom there can, of course, be failure, breakage or improper utilization. That being said, there was “a lower risk” when protection was utilized according to medical and scientific evidence. I am persuaded that the combination of an undetectable viral load and the use of a condom would serve to reduce the risk below what would be considered a significant risk of serious bodily harm.[19]

Thus, somewhat puzzlingly, the trial judge found that a low viral load in the context of protected sex could endanger the life of the complainant but that it could not create a significant risk of serious bodily harm for the purposes of fraud. Implicit in the judgment is the finding that the threshold for endangerment of life is lower than the significant risk of serious bodily harm required to negate fraud.

IV. Analysis

As mentioned above, there are two novel and important findings in Mabior that warrant discussion. First, the trial judge found that an accused who does not disclose his or her HIV-positive status to sexual partners can be convicted of aggravated sexual assault even when a condom is used. Second, she held that when an accused uses a condom and has an undetectable viral load, the risk of serious bodily harm is reduced to a point where it is no longer significant enough to constitute fraud. Both of these issues force one to confront the reasoning in Cuerrier. In broader policy terms, these findings raise difficult questions about the appropriate scope and purposes of criminal law and its relationship to our evolving scientific understanding of HIV/AIDS.

There is a range of approaches that could be taken to the criminalization of nondisclosure of HIV status. The courts could focus on the harm caused, imposing criminal liability only when the virus has actually been transmitted.[20] This approach is consistent with the criminal law’s focus on behaviour which causes harm to others. Alternatively, the courts could focus on the risk-taking behaviour of the accused and his or her moral blameworthiness, regardless of whether or not the virus was actually transmitted.[21] In Cuerrier, the Supreme Court of Canada took the latter approach and did not require transmission as an element of the offence of aggravated assault. In fact, neither of the complainants in Cuerrier tested HIV-positive by the time of trial. Rather, the concept of fraud was based on the degree of risk created by the accused’s conduct.[22]

The majority in Cuerrier was particularly concerned that fraud not be defined so broadly that any risk of harm (such as the emotional harm that may result from deceptive sexual practices), could negate consent to sexual activity and give rise to assault charges. They held that the deception must pose a significant risk of serious bodily harm in order to negate consent.[23] The Court conceptualized the duty to disclose in direct proportion to “the risks attendant upon the act of intercourse”: the greater the risk to the complainant, the more likely it is that the accused has a duty to disclose.[24] In light of the Court’s focus on risk, the critical question arising out of the facts of Mabior is how condoms and viral load tie into the analysis of significant risk.

A. The Use of Condoms

There are two components to the risk analysis in the determination of fraud. The first relates to the qualitative nature of the harm at issue—Cuerrier requires serious bodily harm. The second inquiry relates to the magnitude of the risk, that is, the likelihood that the harm will ensue—Cuerrier requires a significant risk. With respect to the first component, it is important to note that the risk encompassed by HIV transmission has changed since Cuerrier. The fact that HIV/AIDS is no longer necessarily fatal could change the risk calculation.[25] However, given the nature of HIV and its potential to cause AIDS, it is likely that the risk of transmitting HIV will be seen as a risk of serious bodily harm as long as no cure exists.

The second inquiry is more complex: what degree of risk of HIV transmission is significant enough to justify criminal responsibility? Arguably, because the potential harm is so serious, any risk of transmission should suffice. But this is not what the majority held in Cuerrier. The majority suggested, without explicitly deciding, that if an HIV-positive accused used a condom, the risk of harm might not be significant enough to warrant criminal liability and thus that there might not be a requirement to disclose one’s HIV-positive status:

To have intercourse with a person who is HIV-positive will always present risks. Absolutely safe sex may be impossible. Yet the careful use of condoms might be found to so reduce the risk of harm that it could no longer be considered significant so that there might not be either deprivation or a risk of deprivation.[26]

The concurring minority judgment of Justice McLachlin (as she then was) explicitly held that the use of a condom would negate fraud.

There are only a few lower court decisions that consider the issue of condom use, probably because charges are more likely to be laid in cases involving unprotected sex. In R. v. Edwards, the trial judge proceeded on the assumption that only unprotected sex could be subject to criminal liability.[27] Because he had a reasonable doubt as to whether the accused had used a condom, the trial judge did not find fraud negating consent. With respect to whether nondisclosure in the context of protected sex is criminal, the court in Edwards held that this issue should be left to Parliament.[28] Similarly, in R. v. Angnatuk-Mercier, the trial judge held that in order to convict the accused, “it must be established by the Crown beyond a reasonable doubt that unprotected sex with him took place.”[29]

Clearly, the trial judge in Mabior did not follow these cases. Her conclusion that nondisclosure negates consent in the context of protected sex is a first in Canadian criminal law. The trial judge relied on expert testimony that condoms have a failure rate of up to 20 per cent without really explaining what this figure means. Does it, for example, include breakage and improper use of condoms or does it refer to condoms not providing 100 per cent protection even when used correctly?

It is important to note that a 20 per cent failure rate of condoms does not equate with a 20 per cent chance of acquiring HIV from an act of sexual intercourse. In fact, the rate of transmission of HIV is lower than is generally assumed. Although different studies cite different numbers, it is estimated, for example, that in one act of unprotected vaginal intercourse where the male is infected with HIV, the risk that the female partner will acquire HIV may be as low as 1 in 1000. The risk in anal intercourse is approximately 1 in 50 for the receptive partner. When used correctly, condoms reduce the rate of transmission by up to 90 per cent, such that the risk to the receptive partner in anal intercourse drops to 1 in 500 instead of 1 in 50.[30] The Canadian AIDS Society considers unprotected anal and vaginal sexual intercourse to be “high risk behaviour[s]” in terms of the likelihood of the transmission of HIV, whereas it considers protected sex to be “low risk”.[31] The risk of transmission is further decreased where antiretroviral medications have succeeded in reducing the individual’s viral load to an undetectable level.[32] Thus, while a 20 per cent failure rate for condoms sounds high, the trial judge failed to examine what this figure means in terms of a risk of transmission of HIV: the key issue in determining whether fraud has been established. A 20 per cent failure rate of condoms does not quantify the risk of HIV transmission necessary to assess the significance of the risk.[33]

There are arguments for and against finding that the use of a condom negates any potential fraud. On the one hand, as discussed above, condoms do have a risk of failure. Condoms can fall off, break, or be used improperly. Proper use of condoms may also be more difficult where the parties are intoxicated, as they were in many of the sexual encounters at issue in Mabior. The question then becomes who should bear the risk of the condom not functioning properly, particularly if the virus is transmitted. Should the law require the infected partner to disclose his or her HIV-positive status even when using a condom so that the uninfected partner, the person at greatest risk, can decide whether he or she wants to bear the risk of the condom failing? Requiring disclosure even in instances of protected sex expands the scope of criminal liability for failure to disclose and prioritizes the autonomy of the complainant to choose what risks she or he is prepared to accept in the context of sexual activity.[34]

On the other hand, it is indisputable that, short of abstinence, the use of a condom is the best known way to prevent the transmission of HIV. Condom use is at the centre of public health efforts to stem the transmission of HIV.[35] Thus it is certainly arguable, as a matter of public policy, that the law should encourage individuals who may be HIV-positive to use condoms (and encourage others to insist on the use of condoms) with all of their sexual partners. Where possible, this argument goes, criminal law should be consistent with broad public health interests. Even the expert witness at trial disagreed with the trial judge’s criminalization of protected sex. The trial judge cited the expert evidence as follows:

There is no scientific justification to require HIV status disclosure if a condom is always used. There is a mutual responsibility for casual sex partners to be aware of the innate risks of non-monogamy and to ensure their own safety by adhering to consistent and correct condom use.[36]

Those who argue most forcefully against criminalization posit that it obscures the fact that both partners involved in a sexual encounter are responsible for taking precautions to prevent the transmission of disease.[37] Criminalization puts the burden entirely on the infected partner and masks the responsibility of uninfected partners to insist on condom use. In response to this argument, the Supreme Court of Canada explicitly held in Cuerrier that the infected partner has a much greater responsibility than the uninfected partner to prevent transmission.[38]

Studies suggest that up to 40 per cent of individuals who have tested HIV-positive do not disclose their status to any of their sexual partners.[39] One study found that 52 per cent of sexually active, HIV-positive men did not disclose their status to one or more sexual partners. Of these, a significant number engaged in unprotected sex, particularly with casual partners.[40] However, the incidence of nondisclosure must be seen in light of the fact that as many as two-thirds of all HIV transmissions occur before the infected individual knows that he or she is infected.[41] This fact has implications for the utility of criminal law in preventing transmission and also points to the importance of consistent condom use.

Some studies suggest that the use of condoms is decreasing as a result of the optimism over new treatment avenues for HIV/AIDS.[42] As indicated above, however, the use of condoms is crucially important if we are to curb the transmission of HIV. While everyone should disclose his or her HIV status to potential sexual partners, the criminal law should be reserved for the most egregious cases of nondisclosure in the context of unprotected sex. Those who use condoms are making efforts to avoid transmission of the virus. If the courts send the message that nondisclosure is criminal regardless of the precautions taken, this could decrease the incentive to use condoms. I recognize that this view interferes with the right of individuals to make their own choices about the degree of risk they are willing to accept in sexual activity. There may be a small number of potential complainants who acquire the virus and yet do not have access to the criminal justice system under this model because a condom was used. Having said this, however, criminal prosecution after the fact does little to undo any harm caused by the failure to disclose. The state’s most coercive power should be limited to the most blameworthy cases.[43]

Modifying sexual behaviour requires a multifaceted approach; criminalization is not, on its own, sufficient. A recent American study, for example, found that persons at high risk for HIV (both those who had been tested and those who had not) did not alter their sexual practices (i.e., disclosure and the use of condoms) based on whether they believed the law required disclosure or condom use, thus casting doubt on the broad deterrent value of criminalization.[44] Criminalizing nondisclosure regardless of protection will not protect society from persons such as Cuerrier[45] or Mabior, until after they have engaged in sexual intercourse and put potential complainants at risk. But the law serves a symbolic function as well as a deterrent function. Even where deterrence is difficult to prove, it is important that the law send messages that are consistent with well-accepted public health policy. The non-criminalization of protected sex, at the very least, does not discourage condom use.

B. Viral Load

With the development of new antiretroviral medications, it is now possible to reduce the presence of HIV in a person’s system to the point where it can no longer be detected by our current testing standards.[46] This does not mean that the virus is no longer present in the person’s immune system. Somewhat contradictory evidence was presented in Mabior as to the potential of infecting one’s partner when one’s viral load is undetectable. However, the preponderance of the evidence indicated that while an undetectable viral load decreases the likelihood of transmission significantly, it probably does not reduce it to zero, particularly if the accused has other sexually transmitted diseases.[47] Thus, this issue also raises questions about the threshold of risk. On the one hand, if one accepts that the use of a condom reduces risk such that criminal liability should not attach, why would the same rationale not apply to an undetectable viral load? Both factors significantly reduce the risk of HIV transmission. At what point does the risk of transmission become so low that it is no longer considered significant? The Supreme Court of Canada did not address this issue because, at the time of Cuerrier, we did not have the capacity to reduce a person’s viral load so significantly.[48]

The trial judge clearly acknowledged that there is a point at which the risk of HIV transmission is insufficient to meet the definition of fraud. She decided that this point is reached where the accused used a condom and he or she had an undetectable viral load at the time of the sexual contact. One can understand this reasoning; these two factors cumulatively produce a substantial reduction in the risk of transmission. However, the viral load issue is complicated given our current scientific understanding. We cannot yet establish the precise risk of transmission for someone with an undetectable viral load. One’s viral load paints a snapshot in time and does not rule out transmissibility altogether. How close in time to the act of intercourse would the finding of an undetectable load have to be to negate fraud? How often would one have to be tested in order to establish a pattern of undetectability? It is also important to note that we may develop new technologies to detect the virus at levels which are currently undetectable. Thus what is undetectable today may not be so in the future.[49]

If we accept the legal relevance of viral load, difficult questions arise regarding the burden of proof. Would an undetectable viral load constitute a defence such that the accused would have the burden of proving that his or her viral load was undetectable at the relevant time? The alternative would be to assign the Crown the burden of establishing beyond a reasonable doubt that the accused’s viral load was not undetectable at the time of the sexual activity in question. This burden of proof would be virtually impossible to meet, particularly where there were multiple acts of intercourse or where repeated test results were not available. Liability could then turn on whether and when the accused was tested and on the availability of those test results.[50] In addition, advances in scientific knowledge would likely impact several areas of HIV transmission prosecutions. For example, if we learn that the virus is not transmissible below a certain level, would proof of that viral level negate fraud and obviate the duty to disclose? If so, then surely the Crown would have to prove that the accused’s viral load was not at this level in order to obtain a conviction. In the only other case in which I was able to find a reference to viral load, the trial judge rejected its relevance, stating: “[I]t seems to be a fragile defence. All it reveals is the state of the blood tested on the day in question, not two weeks earlier, not two weeks later. … To rely on slips of paper from a lab seems fraught with hazard.”[51]

In addition to these practical obstacles to using viral load as a measure of risk in HIV nondisclosure cases, there are also significant policy concerns surrounding the use of viral load. Considering viral load in liability may open the door to HIV-positive individuals making their own risk-assessments about transmissibility and disclosure. For example, some literature suggests that men are less likely to disclose their HIV-positive status when they have a low or undetectable viral load.[52] The following quote reveals the dilemma:

Lowering viral load and keeping it low may reduce the likelihood that a seropositive person may infect a partner during sexual contact. However, as treatment options enable people with HIV infection to live longer and feel healthier, those people may become more sexually active. Those who believe that that low viral load renders them non-infectious may stop using condoms. ... Recent studies suggest that optimism about the new HIV therapies is associated with sexual risk-taking in MSM [men who have sex with men].[53]

It is understandable that the trial judge wanted to limit the potentially sweeping scope of criminal liability she extended to those who use condoms. However her means of achieving this end, by combining the use of a condom and low viral load, is problematic.

There is a compelling public policy rationale for encouraging condom usage. In contrast, there is less justification for including viral load in the risk calculation. Doing so could encourage people to make their own assessments about their infectivity before deciding on disclosure, thus potentially increasing the risk of transmission. Just as it is important to encourage the use of condoms as a public health measure, it is equally important not to send the message that people can assess their own viral loads and determine their own risk of transmission. When this concern is added to the problems created by imperfect scientific information, there is, at present, even less justification to employ viral load in the equation. It is possible that, as our understanding and ability to measure viral load develop, it will become a more manageable standard. But we have not yet reached that point.

In my view, Justice McKelvey should have held that condom use negates fraud. Such a finding probably would not have changed the outcome in Mabior given the dubious evidence of occasional and unreliable condom use. If the trial judge had limited criminal liability to unprotected sex, it would not have been necessary for her to consider the relevance of viral load because she only did so in the context of protected sex. The trial judge agreed that, in the absence of condom usage, a low viral load could not negate fraud.

Conclusion

It is clear that Mabior is a particularly troubling case and one that understandably evoked outrage in the trial judge. An HIV-positive man preying on vulnerable teenage girls and refusing to disclose his HIV-positive status makes a compelling case for prosecution. As the judge stated in convicting the accused:

The accused’s conduct was deplorable and despicable in all of the circumstances and must be condemned in the strongest possible terms. Those that are infected with HIV cannot inappropriately and indiscriminately engage in sexual relationships for their own pleasure without regard to the consequences to others.[54]

However, particularly egregious cases do not necessarily provide the best context in which to set broad criminal law policy. The outrageous nature of the facts in Mabior should not be invoked to extend the reach of criminal law to all potential accused who do not disclose their status, including individuals who genuinely attempt to protect their sexual partners by using a condom. Disclosure is extremely important and may be possible in the large majority of cases. Nevertheless, there may be reasons, such as intimate violence, that make disclosure extremely difficult or even dangerous for an individual. The social context of disclosing one’s HIV status should not be ignored.[55] Professor Ryan argues, for example, that where disclosure creates significant risks for the infected partner, the criminal law should view careful use of condoms as a substitute for disclosure.[56]

The vexing issues in this case demonstrate how unwieldy prosecutions for failure to disclose HIV can be. Tales of intoxication, condoms breaking and falling off, poor memory of whether a condom was used at all, and conflicting evidence about the transmissibility of the virus in someone who has an undetectable viral load all demonstrate the problems involved in such prosecutions.[57] The Manitoba Court of Appeal will soon have to decide whether viral load is relevant to liability and, if so, who bears the burden of proof of establishing the accused’s viral load at the relevant time. Similarly, the court will need to address whether the use of a condom sufficiently reduces the risk of bodily harm to preclude criminal liability. The answers to each of these issues will determine which cases can be prosecuted in the future. One hopes that the Manitoba Court of Appeal will take the opportunity to provide some certainty in this challenging area of criminal law.

Mabior demonstrates the difficulties in quantifying risk in the face of incomplete scientific knowledge about the factors that contribute to the risk caused by the accused’s conduct. There are no easy answers to the issues raised in Mabior. To argue against the prosecution of those who fail to disclose their status in the context of protected sex is not to suggest that nondisclosure is ever morally justified. But the criminal law is not coextensive with morality. When one balances, on the one hand, the limited utility of criminal prosecution in the context of protected sex with, on the other hand, the importance of promoting the use of condoms, the balance should come down in favour of the latter.