Hum. Reprod. Advance Access originally published online on June 13, 2006
Human Reproduction 2006 21(8):1956-1960; doi:10.1093/humrep/del082
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
OPINION |
Should ART be offered to HIV-serodiscordant and HIV-seroconcordant couples: an ethical discussion?
Women’s and Children’s Hospital—Research Institute, Adelaide, Australia
To whom correspondence should be addressed at: Women’s and Children’s Hospital–Research Institute, 72 King William Road, North Adelaide, Adelaide, Australia. E-mail: tamara.zutlevics{at}cywhs.sa.gov.au
 |
Abstract |
|---|
 Top Abstract IntroductionAcknowledgementsReferences |
|---|
Increasingly, fertility clinics are offering their services to human immunodeficiency virus (HIV)-serodiscordant couples where the woman is seropositive. In the case of HIV-seroconcordant couples, there remains a general reluctance to provide treatment. This attitude to seroconcordant couples is reminiscent of that once widely held towards serodiscordant couples when the risk of vertical transmission rates in pregnant women was greater than 1–2%. Due to recent advances in HIV clinical care and assisted reproduction technique (ART) procedures directed at reducing the risk of viral transmission during gamete transfer, where good healthcare is available, the current risk rate has fallen to 1–2%. This article deals with the ethical arguments of those who remain opposed to offering HIV-serodiscordant and HIV-seroconcordant couples access to ART. Until these arguments have been addressed, clinics providing ART to such couples cannot be assured that their practices are ethical.
Key words: assisted reproduction/ethics/HIV/seroconcordant/serodiscordant
|
Introduction |
|---|
|
TopAbstractIntroductionAcknowledgementsReferences |
|---|
In recent times, there has been a general shift in attitude regarding the ethics of providing human immunodeficiency virus (HIV)-serodiscordant couples with access to assisted reproduction technique (ART). Where once there was general opposition, now there is acceptance under some circumstances. This change in attitude has largely come about due to advances in HIV clinical care and ART procedures directed at reducing the risk of viral contamination during gamete transfer. Vertical transmission rates have fallen to 1–2% in pregnant women with RNA levels below 1000 copies/ml (Thornton et al., 2004
). Such improvements in HIV clinical care have led to affected women feeling increasingly able to pursue the plans they would have had, had they not contracted HIV. Consequently, whilst the number of women with HIV approaching clinics is in general below 1% of all clients, it is inevitable that the demand will increase, necessitating a carefully considered response by clinics.
The current risk level of vertical transmission, however, needs to be qualified and put into context. The 1–2% risk pertains to women who take antiretrovirals during pregnancy (Giles et al., 2004). Furthermore, whilst recent data indicate that a low viral load significantly reduces perinatal risk, ‘more than 90% of HIV infections in children are acquired from mother-to-child transmission, with most cases occurring during labour, delivery or via breast-feeding’ (Thornton et al., 2004).
In general, arguments for providing serodiscordant couples with access to fertility treatment are grounded in antidiscrimination and equity considerations. Fertility expert Mark Sauer is typical here arguing that
It is generally held that equal treatment should exist for all individuals under the law. ... Affected individuals are held to be without blame; therefore, society has an obligation to provide resources to help overcome disadvantaged situations (Sauer, 2003
It bears noting that support is most forthcoming when it is the male partner who is seropositive and sperm-washing techniques can be employed. A more cautious view is held when it is the woman who is seropositive. For HIV-seroconcordant couples, a general reluctance to provide access to fertility treatment remains. The European Society for Human Reproduction and Embryology (ESHRE) Taskforce on Ethics and Law recently concluded that at the present time ART should only be offered to serodiscordant couples ‘so that at least one parent is likely to be able to raise the child until adulthood’ (ESHRE Ethics and Law Task Force, 2004).
Yet even in the case of serodiscordant couples, there remain several forceful objections to the practice of providing access to ART. Until these arguments have been addressed, it cannot be assumed that providing either serodiscordant or seroconcordant couples with access to ART is ethical.
Clinical practice often outstrips ethical debate. For some, there is a tendency to view ethical debate as irrelevant to the real business of helping people and therefore redundant so long as the motives of those providing the service are beneficent. Such an attitude is of concern as it precludes resilient decision making, that is, decision making which is transparent, open to public scrutiny and founded on rational argument. Whilst the concern of this article is with the ethics of providing HIV-affected couples with access to ART, it is worth noting that some countries have instituted legislation such that those clinics refusing to provide services to such couples would be contravening the law. A good example is the 1990 Americans with Disability Act, which was seen to extend to persons who were HIV positive in a landmark case in the Supreme Court, Bragdon versus Abbott 1998 (Annas, 1998). Consequently, clinics formulating policies regarding their service provision to HIV-affected couples will, in addition to giving consideration to ethical concerns, need to be mindful of any legislative requirements.
In what follows, I hope to show that the arguments against providing either serodiscordant or seroconcordant couples with access to ART are mistaken and that we can therefore be confident that those clinics currently providing services to HIV-infected couples are acting ethically.
Arguments against providing ART to HIV-serodiscordant and HIV-seroconcordant couples
1. Savulescu’s best interests of the child argument
The following argument put forward by bioethicist Julian Savulescu is a restricted one, applying only to some ART procedures in the case of seropositive women. In fact, Savulescu argues for the moral necessity of providing serodiscordant couples, where the woman is HIV positive, with access to ART of a particular kind. If right, Savulescu’s argument casts doubt on the current practices of many fertility clinics, and would impose significant restrictions on clinics offering assistance to seropositive women. However, I will argue that Savulescu’s argument is flawed.
Savulescu begins by considering the case in which an HIV-serodiscordant couple uses ART; the woman is HIV positive and the man HIV negative. The couple produce a healthy embryo which is transferred to the woman. During the course of the pregnancy, the baby, which Savulescu calls ‘Gill’, becomes HIV positive. Savulescu then asks us to compare the situation, which produced Gill, with the following:
. . . the woman provides an egg that is then washed. Assume for simplicity that there is no chance of HIV being present in a single washed egg. Her partner’s sperm is inserted. The embryo is then transferred to another HIV negative female surrogate. The baby is born without HIV. Call her Gillian (Savulescu, 2003
Savulescu asks us to consider whether the first child Gill has been harmed by the actions of her parents and doctors. According to Savulescu the answer is yes.
Instead of the fertilized egg being transferred to her genetic mother’s womb, it could have been transferred to a surrogate who would have no chance of transferring HIV after embryo transfer. This would have eliminated or at least reduced the risk of Gill contracting HIV. Gill could have been Gillian. She could have been born without HIV (Savulescu, 2003
According to Savulescu, this argument implies that in cases where a couple is HIV serodiscordant and the woman is HIV positive, there exists a ‘moral obligation’ to use an HIV-negative surrogate to ‘maximally reduce the risk of any embryo of theirs going on to contract HIV’ (Savulescu, 2003). It is worth noting that, if it were valid, Savulescu’s argument would apply more broadly to all couples, seroconcordant or serodiscordant, where the woman is HIV seropositive, and indeed to couples where the best interests of the fetus would be better met by a surrogate. There are many medical conditions affecting pregnant women such that the health of their fetus would be better safeguarded by a surrogate. Insisting on surrogacy arrangements for these conditions would simply be impractical. Such pragmatic issues aside, there are several problems with Savulescu’s argument.
Firstly, the principal assumption underlying Savulescu’s argument is incorrect. To see this, we need to briefly consider Savulescu’s other case in which it is the man who is HIV positive. In this case, the man’s sperm is treated to reduce the risk of transmitting HIV, but it turns out without success. It is then inserted into a single oocyte, which is then transplanted into the woman. The resulting child, Hal, is born HIV positive. Again Savulescu asks whether Hal has been harmed by the actions of his parents and doctors. This time Savulescu’s answer is different.
They could have used sperm from an HIV-negative donor or avoided producing a child at all. Would either of these actions have been better for Hal? Of course not. In the first case, a different child (in 50% of cases it would be female) would have existed. In the second case, no child would have existed (Savulescu, 2003).
Either way, ‘Hal’ could not be said to benefit. At first glance, this line of reasoning does not seem to apply to Gill because Gill and Gillian are represented as being essentially the same person due to their genetic identity—something which Savulescu needs in order for his argument to go through. Therefore, by using a surrogate, Gill still exists. In contrast, by using different, or no, sperm, Hal cannot exist. We are entering some fairly complex philosophical terrain about the metaphysics of personal identity, but very simply, contrary to Savulescu, we cannot be so certain that Gill and Gillian are the same person, which is the assumption underlying Savulescu’s argument.
It may be the case that gestational environment is sufficiently important such that different environments produce different persons even though there is genetic identity, which would mean that Gill and Gillian are in fact different persons. We know that gestational environment has an impact on fetal outcome. For example, there is a correlation between maternal weight and birth weight. In the case of oocyte donation, birth weight follows the recipient, not the biological mother. Moreover, the case of identical twins shows that one can have genetic identity without personal identity; identical twins are clearly different persons. Hence, genetic identity, whilst crucial is not sufficient for personal identity. Certainly, biologists hold the view that phenotype is a product of genotype plus environment. What such considerations suggest is that the role played by gestational environment in personal identity is far from clear, in which case Savulescu’s argument that ‘Gill’ ought to be gestated by a surrogate may well be mistaken on metaphysical grounds alone. But there are other problems with Savulescu’s argument beyond these metaphysical concerns, which turn on the surrogacy relationship itself.
Psychologists Ciccarelli and Beckman (2005) recently reviewed the literature on studies examining the psychological aspects of surrogacy. What they found was a lack of empirical studies on the psychological and social aspects of contractual parenthood. Those that had been carried out were of limited sample size and generally lacking in controls. This makes it difficult to assess Savulescu’s surrogacy argument in terms of social impact.
Ciccarelli and Beckman, however, suggest that the empirical data, such as there are, do not tend to support the view that surrogacy arrangements are psychologically damaging for all involved (Ciccarelli and Beckman, 2005). Even so, they are careful to note that ‘navigating this rocky terrain in which few known ground rules exist is not easy’ (Ciccarelli and Beckman, 2005). Surrogacy arrangements can detrimentally impact upon the emotional well-being of the surrogate mother. In addition, Ciccarelli and Beckman note that occasionally women have come to regret their decisions to become a surrogate and that this dissatisfaction may increase over time as contact is reduced with the contracting couple (Ciccarelli and Beckman, 2005).
It therefore seems that we simply do not have enough empirical data to straightforwardly embrace Savulescu’s surrogacy suggestion; this is in addition to the other problems discussed above.
2. Seriousness of disease argument
Bioethicist Jeffrey Spike has long been opposed to HIV-positive couples being given access to fertility services. When the risk of seroconversion of the fetus was 25%, Spike and co-author Greenlaw argued that HIV-positive women should not have access to fertility services (Spike and Greenlaw, 1994). Spike continued to hold this position when the risk of transmission fell to 8% (Spike, 2003). With the risk falling to 1%, Spike began to shift his position. However, he remains opposed to the practice, putting the risk at 0.1% before he would completely shift his position. A contributing factor in his reconsideration of his initial position is the ‘risk of an uninfected woman seroconverting from attempting to get pregnant without IVF (which I could easily imagine happening more than once per 1000 couples)’ (Spike, 2003).
Spike’s concern is not unreasonable. In a recent study, Thornton et al. found that HIV-serodiscordant couples were prepared to engage in unsafe sex in order to conceive despite transmission risks (Thornton et al., 2004, 63). Another study cited by Shivas revealed that in a programme at Columbia University, 5% of patients interviewed predicted that ‘they would abandon condoms and attempt pregnancy with their partner if safer methods were unavailable’ (Shivas, 2003). Interestingly, in a study looking at actual behaviour (not predicted), Klein and colleagues reported that 80% of HIV-affected couples, who they surveyed, had engaged in unprotected intercourse to achieve a previous pregnancy (Klein et al., 2003).
Of course, it does not necessarily follow from individuals’ willingness to engage in reckless behaviour that we have an obligation to assist them in achieving their ends more safely. As Laura Purdy notes, ‘individuals should not be free to reproduce mindless of the consequences’ (Shivas, 2003). However, it is arguable that we have a moral obligation to meet the needs of people who cannot meet them themselves.
It is difficult to categorize reproductive ‘rights’. Nonetheless, whilst they are probably not in the same category as basic or fundamental needs such as adequate nutrition, shelter and basic healthcare, they are certainly not in the category of luxury items. Consequently, so long as the ethical checks and balances are in place, I believe we are obligated to help the ‘reproductively needy’. Clearly, HIV-affected couples fall into this category. Without reproductive assistance, such couples face serious risks to their health and the health of their fetus.
Despite being swayed by the risk-taking behaviour of HIV-infected couples denied access to ART, Spike is obviously uncomfortable with his shift in position recommending a policy of ‘testing both partners of a couple requesting IVF and excluding HIV-discordant couples where the woman is HIV positive but allowing it when the man is HIV positive’ (Spike 2003, 62).
Spike’s firm stance stems from his belief that until recently, the seriousness of HIV infection was on a par with diseases such as Cri-du-chat and Tay Sachs (Spike, 2003). Yet even before the 1–2% risk rate, HIV infection was regarded by some as a chronic illness rather than a deadly disease (Gallant, 2000). This statement is, however, in need of qualification. Highly active antiretroviral therapy (HAART), which is responsible for increased longevity, has now been used for around 7 years with good clinical outcomes. Nonetheless, longer-term data is unavailable and many of the studies pertain to adult rather than paediatric populations. Consequently, whilst we can say that the clinical outlook is promising for up to 7 years after commencing HAART, beyond this we need to be more cautious in our predictions.
However, even in the absence of extended longitudinal studies, equating HIV infection with Tay Sachs and Cri-du-chat ignores salient differences. The effects of Cri-du-Chat are extremely variable and despite significant health problems in some cases and behavioural difficulties, children with this condition are often characterized as ‘loving and sociable . . . with a great sense of humour who occupy a valued position within their families and communities’ (Cri-du-Chat Support Group of Australia, 2004) Therefore, likening HIV infection to Cri-du-Chat does not necessarily make a compelling case against conceiving a child who has a 1–2% risk of being born HIV positive. In contrast, Tay Sachs affords the child no real chance of a good life on even quite subjective conceptions of what a good life would entail. This is not the case for a child born HIV seropositive who is capable of enjoying a great variety of quality lives. The clinical care of a child born to a woman who is HIV seropositive is reasonably straightforward, involving regular blood tests until the child is 12 months old if all virological tests have been negative (Giles et al., 2004, 202). For children who continue to test positive after 18 months, and who show waning immunity, oral antiretrovirals are given. Depending on the particular HAART combination, the drug regime can be very strict requiring absolute diligence and complete compliance. Nonetheless, even on this worse case scenario, it is possible to live a life of quality for many years in a way that is precluded by Tay Sachs. Consequently, Spike’s seriousness of disease argument fails to make a case against providing HIV-affected couples with access to ART.
3. Use of others as a means argument
Using others solely as a means to one’s own ends is widely regarded as morally dubious. According to Spike, enabling HIV-affected couples to have access to ART allows them to use others as a means to their own ends. Spike puts the point in the form of a question: ‘is it fair to purposely create one baby whose life will be radically shortened and filled with suffering in order to create 99 healthy babies?’ (Spike, 2003). He likens this situation to the often-discussed example in moral philosophy of a train careering out of control towards a track where 10 people are strapped. An onlooker has the option of doing nothing so that the train kills the 10 people or throwing a switch so that the train changes course to another track where only one person is strapped. Many philosophers of a consequentialist ilk (those who believe morality lies in maximizing good outcomes) argue that the onlooker should throw the switch. Spike is obviously uncomfortable with this line of moral reasoning. He argues that offering HIV-affected couples access to fertility treatment can similarly be seen as a question of ‘whether to sacrifice one baby in order to save 99’ (Spike, 2003). In other words, if we offer ART to 100 women with HIV infection, we can expect one child to be ‘sacrificed’ to HIV. Consequently, fertility clinics should think very carefully before providing seropositive women with access to ART.
The cases are, however, disanalogous. In the track-switching example, the onlooker’s decision to throw the switch has a direct causal impact on the two groups of people, namely the death of one person and the rescue of 10. This is not the case for a decision to offer a particular HIV-serodiscordant couple access to fertility treatment. The decision has no causal impact on other fetuses. There are no such fetuses; it is a statistical correlation. Hence, the concern that one’s actions risk using another as a means does not apply in the case of HIV-infected couples accessing ART.
Another problem with Spike’s reasoning is that it may well extend to many choices pertaining to reproduction. For any one couple, there will probably be a similar risk of producing a child with some inherited genetic condition commensurable in some way to the clinical reality of HIV infection given currently available treatment. It is important to note that we do not need this risk to be associated with a single condition. It can be a one in 100 risk of producing a child with any condition similar in clinical consequences to HIV infection; where ‘C’ represents an adverse clinical condition, Pr (C1 or C2 or . . . Cn) = 1/100. In fact, one child in 40 is born with a significant persisting birth defect and about half of these children will have an intellectual disability. Clearly, the view that no one should attempt conception because they have a one in 100 chance of producing a child with a condition similar to HIV infection in terms of quality of life is untenable. Of course, it is debatable whether the majority of these conditions are in fact commensurable with HIV infection. Even so, within the context of current medical treatment, it is plausible to suggest that a significant number of congenital anomalies will pose a similar burden on quality of life as being born HIV positive. Consequently, identifying HIV infection as a special reason to deny couples access to ART is inequitable. Of course, it could be argued that any couple at risk of conceiving a child with a quality of life commensurable to that of an HIV-positive child should not be provided with access to ART. But this would constitute a radical departure from current views surrounding reproductive choices and would be rejected by many as highly unethical.
Finally, in general we allow parents to make their own decisions regarding the children they will have and the risks they will take in order to have those children. As Jan Hodgson and Merle Spriggs note
In genetic counselling facilitation of autonomous decision making is seen as a primary aim and respect for autonomy is used to justify a non-directive counselling approach whereby clients are free to make their own choices after being given all necessary information (Hodgson and Spriggs, 2005
This is obviously not sufficient to justify providing HIV-positive couples with access to ART, but it does carry some weight when considered in the light of other arguments discussed above.
4. Reproductive alternatives argument
Several writers have argued against providing HIV-infected couples with access to ART given that there are alternatives available for creating or adding to a family. Spike is typical here arguing that HIV-infected couples have other choices (Spike, 2003, 62). He argues that
While it won’t satisfy all couples, we as a society need more people willing to consider adoption. There are already many babies and children alive who are desperate for security and comfort and love. I would like to see some couples, whose lives have been dealt the unfair blow of finding one partner HIV positive, help fight this tragedy of unwanted babies and children (Spike, 2003
The implausibility of this argument lies in the fact that it threatens to extend more generally to those couples without HIV infection who require access to fertility treatment, and indeed to all couples wanting to have children. There are many institutions where children, and older people, reside without adequate care, and what they do is a terrible tragedy that particularly wealthy societies have an obligation to amend. Nonetheless, to lay the responsibility at the feet of HIV-infected couples in need of fertility treatment is simply inequitable. Furthermore, it bears noting that it is highly unlikely that an HIV-affected couple would be favourably considered by an adoption agency.
5. Orphan argument
An argument which has been directed against serodiscordant couples, but which is seen to particularly press in the case of seroconcordant couples, is the orphan argument. The worry is that children born to parents who are both HIV positive face an unacceptably high risk of being orphaned and that this in turn places their psychological and physical well-being at risk. As Spriggs and Charles (2003) note, some people believe that ‘when the medical and the social realities are considered a child born to someone with HIV is less likely to have a chance of a good life’.
It is generally the case that being orphaned is not a positive experience for children. Furthermore, losing both parents is likely to be more damaging to the child than if they were to lose only one parent. Nonetheless, as rightly argued by Savulescu, the risk of being orphaned, or indeed, actually being orphaned, is unlikely to be so detrimental to the child such that it would make non-existence preferable (Savulescu, 2003). Finally, it needs to be remembered that there have been periods in history where having either one or both parents into adulthood has been a minority occurrence. The idea that during such times couples should have refrained from procreation is clearly untenable.
In summary, there is an overwhelming desire amongst people to have children. And so long as the risks are low and managed appropriately, fertility clinics should not only continue to provide serodiscordant couples with access to ART but also expand their practices to include HIV-seroconcordant couples.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionAcknowledgementsReferences |
|---|
I thank the following people for helpful comments and discussion: Sandy Cleghorn, Deborah Coulter, Caroline Guerin, Eric Haan, Paul Henning, Neil Hotham, Jon Jureidini, Robert Norman, Ian Ravenscroft, Jeffery Robinson, Gareth Thomas, John Turnidge, the editors of this journal and two anonymous reviewers.
|
References |
|---|
|
TopAbstractIntroductionAcknowledgementsReferences |
|---|
Annas G. (1998) Protecting patients from discrimination—the American with Disabilities Act and HIV infection. N Engl J Med 339:1255–1259.
Ciccarelli J and Beckman L. (2005) Navigating rough waters: an overview of psychological aspects of surrogacy. J Soc Issues 61:21–43.
Cri-du-Chat Support Group of Australia. (2004) URL http://www.criduchat.asn.au./criduchat/what%20is%cdc.htm.
ESHRE Ethics and Law Task Force. (2004) Taskforce 8: ethics of medically assisted fertility treatment for HIV positive men and women. Hum Reprod 19:2454–2456.
Gallant J. (2000) Strategies for long-term success in the treatment of HIV infection. JAMA 283:1329–1334.
Giles M, Mijch A, Garland M, Grover S, Hellard S. (2004) HIV and pregnancy in Australia. Aust N Z J Obstet Gynaecol 44:197–204.[CrossRef][Web of Science][Medline]
Hodgson J and Spriggs M. (2005) A practical account of autonomy: why genetic counselling is especially well suited to the facilitation of informed autonomous decision making. J Genet Couns 14:289–97.[CrossRef][Medline]
Klein J, Pena JE, Thorton MH, Sauer MV. (2003) Understanding the motivations, concerns, and desires of human immunodeficiency virus 1-serodiscordant couples wishing to have children through assisted reproduction. Obstet Gynecol 101:987–994.[CrossRef][Web of Science][Medline]
Sauer M. (2003) Providing fertility care to those with HIV: time to re-examine healthcare policy. Am J Bioeth 3:33–40.[Web of Science][Medline]
Savulescu J. (2003) Assisted reproduction for HIV serodiscordant couples: the ethical issues in perspective. Am J Bioeth 3:53–57.[Web of Science][Medline]
Shivas T. (2003) HIV-1, reproduction, and justice. Am J Bioeth 3:63–64.[Web of Science][Medline]
Spike J. (2003) HIV-discordant couples and IVF: what is the question? Am J Bioeth 3:60–62.[Web of Science][Medline]
Spike J and Greenlaw J. (1994) Case study: ethics consultation. J Law Med Ethics 22:347–350.[Web of Science][Medline]
Spriggs M and Charles T. (2003) Should HIV discordant couples have access to assisted reproductive technologies? J Med Ethics 29:325–329.
Thornton A, Romanelli F, Collins J. (2004) Reproduction decision making for couples affected by HIV: a review of the literature. Top HIV Med 12:61–67.[Medline]
Submitted on September 13, 2005; resubmitted on December 12, 2005; accepted on January 17, 2006.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Ziebe, P. Devroey, and on behalf of the State of the ART 2007 Workshop Gr Assisted reproductive technologies are an integrated part of national strategies addressing demographic and reproductive challenges Hum. Reprod. Update, November 1, 2008; 14(6): 583 - 592. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Barreiro, J. A. Castilla, P. Labarga, and V. Soriano Is natural conception a valid option for HIV-serodiscordant couples? Hum. Reprod., September 1, 2007; 22(9): 2353 - 2358. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||