Debating the goals of HIV testing

In the early years of the bioethical debate over AIDS-related matters, occurring primarily in the United States and Western Europe in the mid 1980s (e.g. Hastings Center Report, 1985), the issue of voluntary versus compulsory HIV testing was predominant. As the point of testing lies in the subsequent use of knowledge of a given test's outcome, at the time many people questioned the value of widespread screening when no effective life-sustaining treatments yet existed. As AIDS was, and still is, a disease that predominantly affects socially marginalised or economically disadvantaged groups in developed countries like the United States (CDC, 1994b; CDC, 2010), many at-risk people also rightly questioned the value of knowing that they were infected with a stigmatising and (then) uniformly fatal illness.

In hindsight, such concerns were not unfounded. During the early years of the HIV epidemic in the United States, discrimination against people living with HIV/AIDS was common, not only in employment and housing but even within the health-care profession. In a 1991 survey of over two thousand health-care providers, for example, nearly half of respondents indicated that they would not treat people with HIV infection if given a choice (Gerbert, Maguire, Bleeker, Coates and McPhee, Reference Gerbert, Maguire, Bleeker, Coates and Mcphee1991). The stigma associated with HIV/AIDS also carried over to uninfected members of at-risk communities. In the same 1991 survey, 35 per cent of respondents would ‘feel nervous among a group of homosexuals’ and 55 per cent would feel uneasy having intravenous drug users as patients (Gerbert et al., Reference Gerbert, Maguire, Bleeker, Coates and Mcphee1991). Even today, in high-prevalence developing countries, similar concerns have been raised about the stigmatisation of people known to be or thought to be HIV infected (Schüklenk and Kleinschmidt, Reference Schüklenk and Kleinsmidt2007). Particularly in sub-Saharan Africa, where the burden of HIV/AIDS falls disproportionately on women, disclosure of positive test results can result in job and housing discrimination or worse, as the widely reported murder of Gugu Dlamini tragically demonstrated (UNAIDS, 2000; Cameron, Reference Cameron2005).

During these early debates, however, the counter-argument was made that only people who know about their HIV status are able to make informed choices about preventing the transmission of the virus to others. Some of the strongest proponents of this argument came, surprisingly enough, from the HIV/AIDS community. At the first national meeting of people living with HIV/AIDS in the USA, a document called the ‘Denver Principles’ was adopted which stated that ‘people with AIDS have an ethical responsibility to inform their potential sexual partners of their health status’ (National Association of People with AIDS, 2011). However, this counter-argument begged several pertinent questions, not least of which was why, in an era where at least people belonging to groups with a higher-than-average risk of acquiring HIV were strongly advised to practise safe sex, knowledge of one's infection status should matter to begin with.

Eventually, concerns about HIV-related discrimination and stigma won out, and public health officials began to treat HIV very differently from other sexually transmitted illnesses or other public health threats, a situation that came to be known as ‘HIV exceptionalism’ (Bayer, Reference Bayer1999). For example, mandatory testing proposals were largely abandoned in favour of voluntary approaches. Moreover, in many countries HIV was not subject to traditional public health practices like contract tracing or partner notification. Support for this approach to HIV has diminished recently, but many of the ‘exceptionalist’ policies and procedures that guide public health practice in the United States and Europe still remain in place. This is despite the fact that many of the socio-economic, political and biomedical circumstances that originally favoured an exceptional approach to HIV no longer exist.

Consider, for example, the question of mandatory testing for HIV. The general consensus among bioethicists is that, all other things being equal, mandatory testing for any disease is only acceptable when successful treatment is available. Even that can reasonably be questioned, however, given that adults in most countries have the legal right to decline medical care, even life-sustaining treatment. The basic ethical principle known as respect for individual autonomy gives a competent patient the right to refuse treatment even if that refusal will lead to their death, a position that has been repeatedly affirmed in court.Footnote ¹ Acting contrary to the wishes of competent patients can easily lead to charges of assault against the attending health-care professionals. There are, however, limits to individual autonomy, particularly when the health and wellbeing of others is also put at risk (Gostin, Reference Gostin2000; Wynia, Reference Wynia2007). In many public health campaigns, such as mandatory school-age vaccinations or quarantine and/or forced treatment of individuals during outbreaks of infectious disease, mandatory treatments are both ethically justifiable and legally enforceable.

As far as testing is concerned, we can legitimately argue about the level of required efficacy of a given treatment, but it is uncontroversial that people need to benefit significantly from the intervention in question to justify pressuring or even forcing people to undergo testing. With respect to voluntary versus compulsory HIV testing, the earliest debates occurred in the mid-1980s when no effective AIDS treatments existed. Representative of the mood at that time is a paper in which Ronald Bayer and colleagues argued ‘there is no demonstrable public health benefit that justifies universal mandatory screening, given the invasion of privacy involved’ (Bayer, Levine and Wolf, Reference Bayer, Levine and Wolf1986, p. 1770). This line of reasoning, widely accepted by health officials responsible for established HIV/AIDS testing policies in most Western countries, accepts the argument that there is no valid paternalistic reason to implement mandatory testing unless there is a clear benefit to infected individuals of knowing their infection status. Furthermore, if there is no benefit for society at large in knowing that certain individuals are infected, there is also no good public health justification for mandatory testing.

The rejection of mandatory HIV testing by Bayer and his colleagues thus was, and still is, conditional upon not having efficient drugs that: (1) can prolong the AIDS-free period; (2) can improve the quality of life of a person at risk or living with AIDS; and (3) can prevent the transmission of HIV to others. None of this holds true any longer. As described above, current HIV treatment regimes are capable of preserving and substantially improving the quality of life of people with AIDS. Equally as important, these treatment regimes can also prevent the transmission of HIV to others, as illustrated by studies in Taiwan and Uganda (Fang et al., Reference Fang, Hsu, Twu, Chen, Chang and Hwang2004; Donnell et al., Reference Donnell, Baeten, Klarie, Thomas, Stevens and Cohen2010). Perhaps then, it is time to reopen the debate over voluntary versus mandatory screening for HIV in the United States and Western Europe, as was done in limited fashion for pregnant women.

Mandatory HIV testing of pregnant women

As alluded to previously, not only does antiretroviral treatment reduce overall AIDS-related morbidity and mortality among HIV-infected women (e.g. Gange et al., Reference Gange, Barron, Greenblatt, Anastos, Minkoff and Young2002), but these drugs are also highly effective at preventing mother-to-child transmission of the virus (CDC, 1994a; Sperling et al., Reference Sperling, Shapiro, Coombs, Todd, Herman and McSherry1996; Wade et al., Reference Wade, Birkhead, Warren, Charbonneau, French, Wang, Baum, Tesoriero and Savicki1998). However, because of the physical side-effects associated with these drugs, it is preferable to determine the HIV status of pregnant women or screen newborn children for HIV exposure rather than treat uninfected women and unexposed children unnecessarily. In the United States, many states have opt-out testing of women, in which all pregnant women are routinely tested for HIV unless they refuse. The adoption of the opt-out approach led to significant increases in the number of pregnant women getting tested for HIV (CDC, 2003). In accordance with the United States Centers for Disease Control and Prevention recommendations that HIV testing be part of the routine panel for prenatal screening tests for all pregnant women, consent for HIV testing is presumed in twenty-two states unless women expressly decline to be tested. Ten states also have programmes that screen newborns for exposure to HIV, with two states (New York and Connecticut) mandating screening of all newborns (Henry J. Kaiser Foundation, 2009). Such programmes have proven remarkably effective in reducing rates of mother-to-child transmission of HIV. In New York, for example, rates of perinatal HIV infection have declined 70 per cent now that mandatory screening of all newborns is required (Birkhead et al., Reference Birkhead, Pulver, Warren, Klein, Parker, Caggana and Smith2010).

Screening newborns for exposure to HIV rather than mandating testing of pregnant women would seem to achieve the public health goal of reducing mother-to-child transmission of HIV while still respecting the autonomy of individual women. However, it is impossible to test the HIV status of the newborn without also uncovering the mother's status. Mandatory screening of newborns is thus a way of conducting mandatory HIV testing of women by proxy, which raises a number of ethical concerns. For example, does the goal of achieving the best possible treatments for newborns justify what is essentially mandatory HIV testing for pregnant women? Does the mother's interest in privacy outweigh the newborn's interest in optimal treatment?

This line of reasoning has met with opposition from some feminist thinkers and activists in developed countries, who criticised the allegedly ‘maternal ideology [according to which] good mothers engage in acts of self-sacrifice and self-abnegation, always putting the interests of their children before their own’ (Zivi, Reference Zivi2005, p. 349). Yet despite this opposition from feminist thinkers, the testing and treatment of pregnant women and newborns, even when mandatory, should not be seen as self-sacrificing. Such screening and treatment also preserves the mothers' lives. Furthermore, this particular feminist argument views this issue as yet another instance of the traditional conflict between foetal and maternal interests. However, this interpretation is misguided, as it is directed at the interests of the future child rather than the foetus. The voluntary and informed decision to bring a child into being carries with it a moral obligation towards reasonably improving the newborn's chances of living a life worth living. One of us has argued that if abortion is reasonably available to such women, and they voluntarily choose to carry the foetus to term, then a moral case can be made that the decision to decline abortion while also refusing HIV testing and treatment, if clinically indicated, counts as harm to others (Schüklenk and Kleinsmidt, Reference Schüklenk and Kleinsmidt2007). We do support, certainly for low HIV prevalence countries, the implementation of opt-out approaches to testing.

The changing face of AIDS – incentivising HIV testing and treatment

In this section we describe the ethical and policy considerations that led historically to the current status quo in terms of the criminalisation of HIV transmission. We show that the empirical basis these sorts of arguments were based on has changed sufficiently to support a call for radical policy and regulatory changes.

On the background of what ethical analysis did today's policy response evolve? In the mid-1980s, libertarians such as Richard Mohr and liberals like Patricia Illingworth vigorously argued against what is today's legal status quo. They believed that using the criminal law in the context of policing HIV transmission was mistaken. Mohr rightly noted that AIDS' ‘mode of contagion assures that those at risk are those whose actions contribute to their risk of infection, chiefly through intimate sexual contact and shared hypodermic needles’ (Mohr, Reference Mohr1987, p. 38). His argument for non-intervention in cases of unsafe sexual encounters among autonomously acting consenting adults was this: ‘If independence – the ability to guide one's life by one's own lights to an extent compatible with the like ability on the part of others – is, as it is, a major value, one cannot respect that value while preventing people from putting themselves at risk through voluntary association’ (p. 39). Obviously, even if one agrees with his interpretation, the option of justifying interventions by health authorities through paternalistic strategies remains, and does not look prima facie unreasonable. After all, society often interferes to prevent harm to self – legislation to enforce the use of seat belts in cars is a good example of this. However, Illingworth maintained that at least within a liberal framework ‘it would be wrong to adopt policies which would interfere with individual liberty’ (Illingworth, Reference Illingworth1990a, p. 11). The volenti non fit iniuria principle supports this point of view.

As is often the case in pluralistic societies like the United States and Western Europe, there is no general consensus that HIV infection primarily constitutes harm to self, not even in cases where the individuals are perfectly aware of the risks they take and have voluntarily agreed to engage in unsafe sexual encounters with individuals of unknown HIV infection status. Ronald Bayer conceded that HIV infections in voluntary sexual encounters were a matter of ‘private choices, choices that will be made in the most intimate of settings beyond the observation of even the most thoroughgoing surveillance’ (Bayer, Reference Bayer1989, p. 11). He maintained, however, that any interpretation which would assume that the transmission of AIDS ‘between consenting adults [belongs] to the private realm alone and therefore [is] beyond the legitimate concern of the state would make a serious mistake’ (p. 12). Bayer continues this argument by stating that HIV infection must always be considered as harm to others and should be interpreted as a ‘public health’ matter. The reasoning behind this interpretation was that the chances of keeping the disease under control were good when only a few responsibly acting individuals are infected and that the odds turn against societal interests when the level of infection reaches certain saturation levels. It thus would be problematic to ignore the collective harmful impact of individual acts which lead to HIV infections.

Society typically interferes with individual choices by means of criminal law only when there is a need to prevent and/or punish harm inflicted on others. Far more controversial is the use of criminal law to prevent competent adults from harming themselves. In the early years of the AIDS pandemic, legislators in many countries had to address the question of whether HIV infection that is acquired during unsafe sexual encounters between consenting adults was a case of harm to self, or whether it should be seen as a case of harm to others that should be punished through the criminal law. At issue was the fact that most infections that are the result of sexual encounters occur as a consequence of voluntary acts among consenting adults. In many jurisdictions today, criminal law is employed as a means of reducing HIV transmission by deterring infected people from engaging in unsafe consensual sex with uninfected partners. In the United States, thirty-four states have laws that explicitly criminalise exposing an uninfected individual to HIV through sex, shared needles or other routes. Similar laws exist in other Western countries, including Austria, Canada, Denmark, Finland, France, Germany, the Netherlands, Russia, Sweden, Switzerland and the United Kingdom (GNP+ and the Terrence Higgins Trust, 2005). Several high-profile cases or convictions under these laws have been reported in the media recently, including the 2010 conviction of German pop star Nadja Benaissa (BBC, 2010). What are prosecuted today under existing HIV transmission statutes are usually either cases where someone is charged with transmitting HIV, or cases where an infected person is charged with engaging in unsafe sex with willing partners who were unaware of his or her HIV status (Schüklenk, Reference Schüklenk2008). In the latter case what is usually punished is subjecting someone to the risk of serious bodily harm or something similar.

However, given the changing nature of the AIDS pandemic, and the rapid improvements in drugs that can be used to treat and even prevent the spread of HIV, a re-analysis of public health policies and laws on HIV transmission, testing and treatment is called for. Early HIV detection and treatment is now demonstrably critical to long-term health and survival (Sax and Baden, Reference Sax and Baden2009). The fact that HIV-infected people who are on antiretroviral therapy are half as likely to die as those who remain untreated (HIV-CAUSAL Collaboration, 2010; Malta et al., Reference Malta, Bastos, Da silva, Pereira, Lucena, Fonseca and Strathdee2009) means that testing is no longer simply a sub-optimal means of curtailing risk behaviour. Rather, it is also a means of providing critical, time-sensitive, life-preserving information. Furthermore, antiretroviral therapy renders many HIV carriers non-contagious (Fang et al., Reference Fang, Hsu, Twu, Chen, Chang and Hwang2004; Vernazza et al., Reference Vernazza, Hirschel, Bernasconi and Flepp2008; Donnell et al., Reference Donnell, Baeten, Klarie, Thomas, Stevens and Cohen2010). Recent mathematical modelling suggests that the widespread use of testing and immediate treatment could shift the AIDS epidemic towards elimination within a decade, and reduce the overall prevalence of HIV to below 1 per cent even in the world's worst-affected regions before 2050 (Granich, Gilks, Dye, De Cock and Williams, Reference Granich, Gilks, Dye, De cock and Williams2009). The proven effectiveness of the treatment-and-prevention paradigm thus has significant policy implications for current public health debates about HIV testing and treatment practices, as well as ongoing legal debates about HIV criminalisation laws.

Legal commentators have argued and many courts have held that HIV-positive people have a duty to disclose their serostatus to their prospective sexual partners, regardless of the circumstances. The rationale is that willingness to engage in unsafe sex, in the absence of such materially relevant information, does not constitute valid consent. Thus, the unsafe sex act is transformed into a case of sexual assault. In R v. Cuerrier,Footnote ² for example, the Supreme Court of Canada found that:

However, empirical research has shown that ‘the majority of communication in sex is non-verbal’ (Davies, Hickson, Weatherburn and Hunt Reference Davies, Hockson, Weatherburn and HUNT1993, p. 49). Hence, a duty to disclose could not reasonably involve a duty on the side of the HIV-positive person to say explicitly that he is infected because this would violate the basic ‘rule of conduct’ in many situations that lead to sexual encounters. Illingworth pointed out that a policy of disclosure may also have the unintended consequence of uninfected people taking fewer precautions to avoid infections. A fictitious example of hers is quite compelling: ‘If Sam believes that it is Bob's moral responsibility to inform him of his antibody status, he may construe Bob's silence as an invitation not to practice safe sex’ (Illingworth, Reference Illingworth1990a, p. 42). A duty of this sort clearly requires everyone to be aware of his or her antibody status as well as to comply with the rules of this alleged moral duty to disclose. Otherwise the public health objective cannot be achieved. But, as we mentioned earlier, 80 per cent of HIV infected people worldwide are unaware of their serostatus because of a lack of access to screening facilities or because of policies and laws that deincentivise testing and treatment.

Of the millions of HIV infections worldwide, only a very few cases of HIV transmission have been prosecuted. Although those that have been prosecuted usually include cases of egregious, uncontroversially unethical and harmful wrongdoing by those convicted (Schüklenk, Reference Schüklenk2008), current criminalisation laws generally do not require proof of intent to transmit HIV for prosecution under these statutes (Center for HIV Law & Policy, 2010). Furthermore, the only successful defence to prosecution under these statutes is either a claim of disclosure (given the background of one-on-one sexual encounters this is often hard to prove) or ignorance of one's HIV-positive status (Center for HIV Law & Policy, 2010). Thus, current criminalisation laws may actually exacerbate the very problem they were intended to prevent. In addition to the social stigma associated with a positive result, HIV testing can also make a person criminally liable for normally legal conduct: consensual sexual activity.

Although mandatory HIV testing is pretty much non-existent in liberal democracies, and voluntary testing practices lack sufficient incentives to encourage most individuals to seek testing, those who know themselves to be HIV-infected are legally obliged to disclose their HIV status to long-time, casual and even anonymous sexual partners. By contrast, a member of a high-risk group who participates in large numbers of high-risk unsafe sexual encounters but who does not know that he is infected (even if he has strong reason to assume so, based on behaviour) is under no obligation to disclose such pertinent information. Someone else who displays identical risk behaviour but decides to get tested, and knows himself to be infected must disclose his HIV status to his sex partners, or face the threat of criminal prosecution. There can be no reasonable doubt that this policy state of affairs is a deterrent to HIV testing.

Quite conceivably at-risk individuals may be discouraged from seeking HIV testing because of fears of legal prosecution, despite the fact that testing and treatment could not only preserve the lives of those infected but could also reduce the likelihood that they will transmit the virus to others. It would be valuable to investigate to what extent such policies prevent at-risk individuals from getting tested. Current clinical evidence shows that it is of paramount importance to detect HIV infection and to initiate treatment early in order to improve infected people's health outcomes. This course of action would have two desirable effects. First, infected patients would be able to live their lives longer and with a much higher quality of life than if they delay or forgo HIV testing and care altogether. Second, effective treatment would soon render these patients non-infectious to all intents and purposes. In the absence of a working preventive HIV vaccine, this would take society closer to a treatment regime that happens to be vaccine equivalent. For the first time in the history of AIDS, individual and societal interests seem to align with regard to HIV testing policy matters. Unfortunately, however, policies have not kept up with the changing clinical landscape. Drastic changes are called for.

Policy propositions worthy of serious consideration

One such policy proposal is an alteration of existing HIV transmission statutes to prevent the prosecution of those who not only seek testing but also effective antiretroviral treatment. After all, such individuals are considerably less infectious than individuals who remain untreated, either by choice or by ignorance of their serostatus. In fact, according to many studies, highly active antiretroviral treatment is as effective in terms of reducing the risk of HIV transmission as using condoms. Effective condom use has in the past been considered a reasonable defence by some courts. Given that effective treatment can essentially render a person non-infectious, it seems unreasonable to continue to require HIV-infected people to disclose their serostatus to sexual partners if they have undertaken the necessary precautions to make HIV transmissions exceedingly unlikely events. A few jurisdictions have recognised this point, and no longer require HIV-positive individuals to disclose their serostatus if they use condoms and/or are on medication. The Manitoba Court of Appeal, for instance, recently acquitted an HIV-positive man who used condoms and was on antiretroviral medication, concluding that he posed no serious risk to the women he had intercourse with despite the fact that he did not disclose his HIV status.Footnote ⁴

The Manitoba Court's ruling is a positive step, but the problem at this point in time is that such decisions are too often left to the vagaries of poorly informed juries or individual judges. Codifying the argument that treated individuals, for the most part, do not present an HIV risk to their partners through a change in law (or in charging guidelines) would provide an additional incentive for those at risk of HIV infection to get tested and treated. Patients with undetectable HIV viral loads should be under no obligation to disclose their HIV status, and policy changes such as the following should be implemented as a matter of urgency.

In particular, prosecutorial charging guidelines could be changed to stipulate that people who know that they are HIV infected but who are able to demonstrate that they had undetectable viral loads resulting from their adherence to antiretroviral treatment regimes will not be prosecuted. They should be under no obligation to disclose their HIV status, because they are essentially unable to pass the virus on to sexual partners. Changing charging guidelines would offer the evident benefit of not having to change existing legislation aimed at deterring infected people who could pass the virus on to sexual partners from doing so. Public health authorities as well as community-based organisations serving high-risk communities should aggressively publish this information to relevant constituencies in order to impact positively on HIV testing and treatment uptake.

A second policy proposal worthy of serious consideration is the elimination of many of the current policies that restrict access to HIV testing or that require written consent and extensive pre- and post-test counselling as part of the screening process. In many developing countries the lack of available counselling staff prevents people at risk from getting tested in a timely fashion (e.g. Bwambale, Ssali, Byaruhanga, Kalyango and Karamagi, Reference Bwambale, Ssali, Byaruhanga, Kalyango and Karamagi2008). Similar barriers to HIV testing also exist in the developed world. In one study of primary care patients at medical centres run by the United States Veterans Administration, a majority of survey respondents found that pre-test counselling was an unnecessary and time-consuming process that did little but perpetuate the idea of HIV as ‘exceptional’ (Bokhour, Solomon, Knapp, Asch and Gifford, Reference Bokhour, Solomon, Knapp, Asch and Gifford2009). Another study conducted in the United States, where different states have different counselling and written consent requirements, found that people were less likely to be tested for HIV if they lived in states that required written informed consent prior to routine HIV testing (Ehrenkranz, Pagán, Begier, Linas, Madison and Armstrong, Reference Ehrenkranz, Pagán, Begier, Linas, Madison and Armstrong2009). Some policy-makers have resisted loosening the consent and counselling requirements for routine HIV testing, largely because of concerns that it would compromise patient decision-making and autonomy, but these fears have proven to be largely unfounded (Cohan, Gomez, Greenberg, Washington and Charlebois, Reference Cohan, Gomez, Greenberg, Washington and Charlebois2009).

Although these policies were sensible when AIDS was a terminal illness without known treatment, and may still be sensible in countries where HIV stigmatisation remains a serious problem, in most North American and Western European countries HIV/AIDS is now a somewhat destigmatised chronic illness that can be controlled by means of existing medication. Of paramount importance, both for public health reasons as well as for health objectives to do with maximising individual patients' quality and length of life, is to implement policies that ensure easy access to HIV testing. It is unclear to us what rationale there could be, given the dramatically changed clinical reality, for putting unnecessary hurdles in the way of people's easy access to HIV testing. It is unclear what the public health rationale for current policies in countries like Canada, Germany and the United Kingdom is, where health authorities maintain strict control of access to HIV testing. Reliable rapid HIV testing can be undertaken by anybody anonymously, for instance in the safety of their own homes, or in a sex-on-venue establishment prior to sexual intercourse or in any number of other environments. In the absence of the inevitable premature death that a positive test result signalled in the mid-1980s, there is no good reason that we can think of for why access to testing should remain controlled by health-care professionals. Quite to the contrary, sensible public health policies would demand that everything feasible is done to give people the easiest access to life-preserving health information. Undergoing compulsory counselling, and knowing that one's sexual partners will be informed, undermines those objectives.

Problems

We believe that the case we have made for fairly radical policy changes is sound. Possible counter-arguments could include that we do not know to what extent current policies deter people at risk of HIV infection from getting tested in a timely fashion, or from getting tested at all. Conceivably the public health objectives we outlined might not be achievable even if our ideas were implemented. Two responses to this concern: (1) We agree that research investigating the possible impact of such policy changes is desirable. (2) However, even if there were only a negligible impact on the public health front, it would still hold true that the current unfair legal treatment of HIV infected people with an undetectable viral load ought to change.

Another possible counter-argument could go along the lines that there might be infected people who indeed have been on antiretroviral therapy, and who indeed have had an undetectable viral load but who have not been treatment compliant, resulting in periods of time when their viral load was sufficiently high as to render them infectious again. We think that this can be dealt with efficiently by way of ensuring that viral load determinations occur in intervals sufficiently short to catch such exceptional cases. Equally, it is known today that the risk of viral failure declines with the duration of antiretroviral therapy, irrespective of adherence levels of individual patients (Lima et al., Reference Lima, Bangsberg, Harrigan, Deeks, Yip, Hogg and Montaner2010).