Ethical considerations for biobanking: should individual research results be shared with relatives?

The issue of returning individual research results (IRRs) is important and contentious. Indeed, Science magazine calls this, “…the most pressing issue in genetics today” [1]. At stake are several significant concerns, including public trust in, and support for, biomedical research, the wise use of limited research resources and the ethical obligations of the research community.

These concerns become even more complex and challenging with genetic biobank research. In this context, researchers must decide not only whether and how to share IRRs with donors, but also their relatives. After all, many genetic conditions have medical or personal significance for family members, and many ethical arguments for returning IRRs to donors might seem to extend to relatives as well [2].

This editorial briefly defends two important conclusions about these issues:

• ▪ The case for sharing IRRs with relatives is no stronger, and almost certainly weaker, than for donors;

• ▪ The case for returning IRRs to donors is not particularly convincing.

Taken together, these two claims support a strong presumption against sharing individual research findings with relatives.

Contrasting donors & relatives

The rationale for the first conclusion is straightforward. Outside of perhaps the general public itself, donors are the primary subjects to whom biobanks owe ethical consideration. Donors, not relatives, have explicitly (or, by not opting-out, implicitly) consented to their genetic materials being used by others. Donors’ interests are directly and powerfully affected by what happens in research. Their biological samples and health data reveal considerably more information about them than anyone else with whom they share partial genetic overlaps. Donors stand most fully and extensively to benefit from and/or be harmed by research. Hence, any obligations that biobanks have to donors’ relatives are weaker and narrower than to donors themselves.

The case against returning IRRs to donors

While many accept the first conclusion, far fewer accept the second. Indeed, if anything, the trend has been explicitly to reject it. Increasingly, bioethical opinion and international policies are “converging towards an ethical duty to return individual genetic research results to subjects, provided there is proof of validity, significance and benefit” [3]. Citing a wide variety of ethical principles, this emerging consensus holds that there are at least some, if not many, situations in which it would be unethical not to disclose certain research findings.

This claim has some intuitive appeal. Careful scrutiny, however, also reveals many weaknesses. Consider first, a ‘negative’ duty to return IRRs – that is, a duty to refrain from worsening the situation of others. Researchers have duties, for example, not to harm others and not to interfere with their ability to live according to their own values and preferences. However, such duties are fundamentally inapplicable in this context. While it may not make people better off, a policy of not returning IRRs can never, itself, make donors worse off. A donor’s prior health status, genetic profile and knowledge are left entirely unchanged – and, hence, neither improved nor worsened – by such a policy.

Consider then, a second kind of moral obligation – a ‘positive’ duty – to return IRRs. Such duties require actively contributing to certain outcomes for others, such as by assisting or empowering them in some way. Since this is precisely what is intended when sharing results, any duty to return IRRs will be positive in nature. However, serious weaknesses arise even here, regardless of whether the positive duties are general obligations that apply to all moral agents or special obligations that apply only to some moral agents with unique positions/roles.

The most familiar example of a general positive duty is beneficence, which requires acting for ‘the benefit of others’ [4]. While accounts differ over the specific demands of beneficence, much is left to the discretion of each individual moral agent. Deciding whether any specific beneficent action/policy is obligatory, then, requires weighing its likely benefits against other important interests and against the likely benefits of other specific actions/policies that could be chosen instead.

Qualified as such, a general positive duty of beneficence would hardly suggest, let alone require, that biobanks return IRRs. For one thing, a policy of returning IRRs may not produce overall net benefit. In many circumstances, returning IRRs could harm people by creating fear, anxiety, confusion and other aversive physical and emotional states. It could also threaten autonomy when fear, anxiety, confusion or faulty/incomplete information influences people’s subsequent decision-making. Additionally, returning IRRs might violate beneficence-based obligations to future patients by delaying the discovery and translation of generalizable knowledge. Also, even if returning IRRs actually yielded some net benefit for donors, this would not definitively establish a beneficence-based obligation to do so. To return results in an ethically responsible way would require a significant investment of monetary and human resources to create and maintain the necessary institutional infrastructure. Biobanks almost certainly could do more good for people (and save themselves a lot of hassle) by simply donating this money to a reputable charity instead.

Thus, if biobanks have an ethical duty to return IRRs, it is a special positive duty that arises from their unique position/role. Two frequently cited examples – the duty to warn/rescue and the duty of reciprocity – are worth considering briefly.

The duty to warn/rescue is a special obligation to prevent bad outcomes when uniquely positioned to do so, and when warning/rescuing requires only minimal sacrifice. It might be argued that biobanks meet the first criterion, given their unique knowledge of certain risks in people’s genetic profile. And if relevant IRRs are already known and available (i.e., no extra ‘hunting’ is required), then it might appear that the only sacrifices incurred are the costs of contacting donors.

Unfortunately, it is almost never a trivial task for biobanks to find relevant information, link it with a donor, find that donor and explain results in a meaningful and beneficial way. Each of these steps could be burdensome, if even possible at all. Indeed, imposing an obligation to return IRRs responsibly could threaten the very integrity and utility of biobanks. For one thing, it arguably transforms researchers into “fiduciaries of research participants” who “have a legal obligation to act in [their] best interests” [5]. This would fundamentally revise our conception of research and make the research community vulnerable to malpractice litigation and its multidimensional costs. Furthermore, to return IRRs responsibly, researchers must obtain valid consent from donors prior to collecting any samples and perhaps on an ongoing basis thereafter. This would fundamentally change the front end of biobank practice, perhaps drastically limiting its scale and utility. Also, even if it did not wholly transform the nature and practice of research as we know it, returning IRRs responsibly would certainly divert limited research resources from what is of greatest collective utility – swiftly deriving generalizable knowledge that can ‘warn/rescue’ even more people in the future (and in the more appropriate contexts of screening and clinical care). Since none of these sacrifices are minimal, any duty to warn/rescue will not require returning IRRs.

The second example, the duty of reciprocity, is a special obligation to return benefits to those from whom one has benefitted. Since biobanks benefit from having access to people’s genetic material and associated health-related information, perhaps they are obligated to benefit donors in return. And what better way to reciprocate than to share IRRs?

Despite its surface appeal, this proposal does not withstand scrutiny. First, even recognizing reciprocity as a genuine value, it is not clear that it is owed first or primarily to individual donors as such. At least with publicly financed research, reciprocity may be first and foremost owed to the general public. The best way to reciprocate, then, is by producing generalizable knowledge of collective benefit, rather than by using valuable limited resources to benefit only a relative few.

Second, even if reciprocity is owed to individual donors, returning IRRs is not a very compelling means of fulfilling this duty. As already noted, returning IRRs can sometimes ‘return’ harms instead of benefits. To the extent that this is likely, biobanks are well advised to consider safer and more sure-footed methods of benefiting donors. Moreover, the values of justice and equal treatment, if not reciprocity itself, require that any benefits provided to donors be given to all [6]. The selective return of IRRs flouts this requirement by only reciprocating benefits to some donors (e.g., those with analytically valid and clinically actionable findings). Just and equal reciprocation, then, would seem to require the universal return of IRRs. However, this would consume disproportionate quantities of limited research resources, while significantly increasing the risks of harming donors and sacrificing the interests of future patients and the general public. Thus, a reciprocity-based special duty to return IRRs (whether selectively or universally) appears fundamentally flawed.

The (amplified) case against returning IRRs to relatives

Each of the arguments for returning IRRs to donors face significant problems. If the first conclusion is true, then they are even less likely to support returning IRRs to relatives. For not only do these arguments face all the problems raised by returning IRRs to donors, they also encounter additional concerns in each case. Having not interacted at all with relatives, biobanks cannot possibly make them worse off by not contacting them. Indeed, if anything, they are far more likely to make relatives worse off by contacting them. Moreover, since relatives have only a partially overlapping genetic profile with donors, biobanks have less potential for benefitting them also. Any special duty to warn/rescue is clearly weaker given the added difficulty in tracking relatives down and obtaining prior informed consent. Also, any reciprocity owed to donors clearly does not apply in the case of relatives. Relatives have not themselves provided anything to benefit research and, hence, biobanks owe them nothing more than any other nondonor. Finally, even if these other considerations were not decisive, the threats to donor privacy and confidentiality posed by sharing their genetic information with others, even relatives, likely outweigh the limited benefits that might thereby be achievable. Ultimately, then, the arguments for returning IRRs to donors face crippling challenges that are only amplified when applied to relatives.

Conclusion

In summary, there is a strong presumptive case against returning IRRs to anyone involved in biobank research, but especially to relatives. Such a practice is neither obligatory nor advisable.

This can be a troubling conclusion in certain situations. One commonly discussed example is where an incidental discovery is clearly important to an individual’s health [7,8]. Surely then, the argument goes, there is an ethical obligation to disclose the results. Even in this situation, however, the practice of returning IRRs comes with such high costs – both moral and economic – that we should resist the temptation to (try to) benefit a few people now at the expense of many more people now and in the future.

Instead, we should support a clear division of labor between the domains of research, screening and clinical care. The over-riding mission of biobank research is to identify correlations between genotype and phenotypic expression of disease, not to benefit individual donors by sharing IRRs. Biobanks are not well suited to fulfill this latter task. To demand that they do would inevitably compromise their core mission. Their limited resources should be entirely devoted to the swift discovery and translation of generalizable knowledge that can then be incorporated into standard screening and clinical care. It is in these contexts, where well-established diagnostic and prescriptive procedures are devoted to the good of both individual people and larger populations, that sharing individual genetic information is most likely to realize its beneficent potential.