Improving the Quality of 'Personalized Medicine' Research and Practice

Through an Ethical Lens

George P Browman; Jochen Vollmann; Alice Virani; Jan Schildmann


Personalized Medicine. 2014;11(4):413-423. 

In This Article

Ethics, Quality & Technological Advances: Privacy, Data Sharing & Informed Consent

Genomic (and other 'omics – proteomics, metabolomics, etc.) analytical technologies are the laboratory enablers of 'PM' as we are defining it. But, as the technologies advance so too do other consequences that need to be considered when viewing PM from a systems' perspective.

Here, we address two particular issues related to technological advances and their ethical implications requiring further exploration for ensuring the quality of PM from a systems perspective.

The first of these, namely information-communications technologies, relates primarily to ethical concerns regarding privacy and data protection. The second related consideration concerns the dichotomy between individual versus societal rights in access to information.

Ethical Implications of Advanced Information/Communications Technologies & Data Sharing

Prompted by the Human Genome Project, several 'policy' recommendations have been and continue to be issued and updated by various research and professional organizations calling for the broad sharing of 'anonymized' individual genomic information embedded within large-scale databases for the benefit of society as a whole, and to advance scientific progress.[45–49] As early examples, the so-called 'Bermuda Rules' call for "open access by scientists to shared human genome sequence information in the public domain";[47] and, the Fort Lauderdale Rules observe that the "scientific community is best served by making results available immediately for free and unrestricted use".[50] Readers should note the use of the term 'rules' rather than 'guidance'. The Human Genome Organization (HUGO) states that "genomic databases should be viewed as a global public good";[46] and the European Society for Human Genetics declared that "it is an ethical imperative to promote access and exchange of [genomic] information".[45] The OECD goes further in its declaration that "operators of biobanks and data should strive to make data and materials widely available to researchers to advance knowledge".[49]

While early forms of the above recommendations acknowledged the social implications of these policies, as well as the role of informed consent and the need for new ethical frameworks,[51] these social and ethical considerations have received much more attention in later versions of these policies.

Against this background, in 2012 and 2013 Schadt et al.[52] and Gymrek et al. from the laboratory of Yaniv Erlich[53] demonstrated the unforeseen feasibility of confirming an individual study participant's identity from presumed anonymous public genetic data by cross-referencing to other publicly available databases. This possibility would never have been identified in consent documents up to that time.

As the field has evolved technologically and computationally, so too have the recommendations governing genomic data sharing (for example see[54]). Recently, the US NIH released a draft policy for sharing genomic data[55] asking for public input. Of interest, much of the input is dominated by research organizations concerned with data access issues for their members. Furthermore, such policies are often being confined to apply narrowly only to researchers who receive funding from the organization issuing the policy. If PM is to be pursued as an international endeavor, with broad information sharing as part of a systems approach, then harmonization of such policies across research organizations will be necessary. In addition, from an ethical perspective such organizations should ensure that all stakeholders' views are invited to ensure that domination from special interests is avoided.

Based on the language used in early versions of these policy statements, it would be reasonable to interpret these rules/recommendations for sharing genomic data as intended primarily to advance science for the benefit of society as a whole. But the appearance given is that societal benefits were placed ahead of the need to protect an individual's rights and freedoms. Yet, it is not uncommon for societal and individual rights to clash in many other areas of society (e.g., recent revelations of the scope of electronic surveillance by national intelligence agencies of ordinary citizens and national public officials in contrast to public government statements), with highly publicized cases within the field of genomics.[56,57]

Research ethics boards are charged with the responsibility of protecting and respecting the autonomy and well being of 'individual' research participants, balancing this conflict between what is best for society as a whole against the well-being of the individual. But, what realistic choices do we give individuals in the face of what are seen as unanimous and consistent recommendations from 'august' sources for information sharing that clearly has the potential to risk privacy breaches? Furthermore, to what extent will we limit current and future patients or research participants access to PM interventions if they choose not to agree to have such information released?

Under the research ethics principle of 'proportionate review'[58] – that is, the extent of review should be proportionate to the risks involved – the calls for wide sharing of supposedly anonymous personal/genomic data were understandable. But, now that we know the risks we may need to re-examine these policies for wide data sharing as PM evolves to involve widely linked data sources, including the incorporation of administrative data as part of a networked approach at a systems level. One possibility involves restructuring the traditional informed consent process, and instead adopting a dynamic consent model for genomic research whereby participants have the opportunity to choose how their data is used and the option to change those preferences, including preferred methods of re-contact, over time.[59,60] Or, as some suggest, we may need to take a more public health ethics approach to research and emphasize reciprocity and meeting public needs by asking those donating specimens to sign a Creative Commons universal waiver,[61] under which participants knowingly accept potential privacy breaches in the name of scientific research. Such a waiver, it is argued could allow for "unprecedented levels of commercial and academic creativity and collaboration".[62] Implementing regulations that prohibit queries that may lead to identification of a specific individual, limit access to public data, and strengthen IT-based security may also decrease the risk of personal privacy breaches.

Research protocols and consent documents already contain statements regarding mandatory (as opposed to optional) compliance with requests for releasing such information as a condition of access to experimental therapies, often where the participant has few other alternatives. These policy statements, by influential and respected organizations, also risk conferring upon researchers, especially those with a strong belief in the promise of PM, a sense of entitlement to coerce (albeit unwittingly) anxious and vulnerable research participants, and mislead patients into agreeing to such demands for sharing their personal information. In addition, there is ambiguity as to what extent modern consent documents need to improve clarity with respect to how such information can be used and the potential for misuse. This is of particular interest given reports of limitations in what research participants actually correctly recall about the information they were provided in informed consent documents regarding biobank participation.[63,64] At a minimum, organizations making recommendations about wide public sharing of potentially vulnerable personal information for the benefit of society ought to include public/patient representation in the process of creating and issuing such recommendations.

PM, Conflicting Ethical Principles & Societal vs Individual Rights

The four bioethical principles of Beauchamp and Childress (autonomy, beneficence, nonmaleficence and justice)[42] have recently been the subject of intense debate within PM circles with the publication of recommendations from the American College of Medical Genetics and Genomics (ACMG) for disclosure of incidental findings emanating from the clinical application of whole-genome sequence analysis,[65] although the same issues are relevant to the research setting, and just as controversial.[66]

The debate is fundamentally about the priorities we should attach to these principles. In general, ethicists have favored personal autonomy as a foundation of good ethical practice, leaving it up to clients to determine for themselves best courses of action. The autonomy principle implies that clients have a right 'not to know' about information that pertains to them. Research ethics boards hope to respect the primacy of this principle of autonomy through lengthy (and often too lengthy) consent forms.

In the case of PM the consent process may be particularly difficult or lengthy. Truly informed consent on the complex issues involved in genomics, such as disclosure of incidental findings, may be unrealistic in the short time frames usually offered to research participants, or to those approached for contributing tissue and/or data to biobanks. For example, multiple public deliberations on genomic biobanking have demonstrated that while the public can become truly informed as to the complexities of ethical, research, technical and other considerations relevant to biobank initiatives, this requires multiple days of information dissemination provided in a variety of formats.[67,68] In addition, genetic counselors routinely provide multiple sessions to truly inform and consent patients about whether genetic testing for hereditary cancer or neurological disease may be appropriate for them. Given this, it may be unrealistic to assume that patients faced with certain kinds of decisions related to PM will be provided enough time and information to fully appreciate the potential benefits and risks of genomic testing. Such risks include potential loss of privacy, genetic discrimination, ambiguous incidental findings, anxiety or familial discord.[69–72]

The integrity of the consent process is an important expression of respect for the autonomy of the research subject. The ACMG recommendations, however, assign priority to beneficence (acting in what is assumed to be in the person's best interests) ahead of autonomy, recommending that clients should not even be offered the choice of whether they want to know about an incidental finding that could be of clinical significance to them or to a family member. Arguments can be made on both sides of this debate, with critics of the ACMG recommendations upholding the more traditional view[73] that the recommendations do not respect an individual's autonomy. Of concern with respect to the development of the ACMG recommendations is that it does not appear that members of the public, research participants or nonexpert practitioners (such as general practitioners) who are key stakeholders were included in the process – an important ethical oversight. In the research domain various researchers of whom we are aware are modifying these recommendations to allow for changes of mind over time as personal circumstances for research participants change. Other remedies, such as stratified consent processes and reframing the language and concept of incidental findings, are also possible.[66]

The question then arises as to whether PM as a new paradigm is constructively raising such ethical conundrums that call for re-evaluation of fundamental ethical principles and how they should be applied. On the one hand, there is the view that genetic information is somehow fundamentally different from other kinds of health information ('genetic exceptionalism')[74] and ought to prompt such re-evaluation as expressed tangibly through special US legislation, such as the Genetic Information Nondiscrimination Act (GINA) 2008, that specifically addresses privacy and information protection standards on genetic information;[75,76] or, whether the problem of incidental findings is not new having been part of ongoing concerns, for instance in relation to HIV status and imaging research and practice.[74,77] On the other side, it is argued that there are few persuasive reasons to consider this issue as exceptional to genetics as a clinical/research field. There have been compelling arguments from policy and genetic counseling sources not to consider these issues as exceptional to genomics.[74,78,79] However, the increasing volume of genetic and other biological information, new informational and computational technologies and the global scale of research and economic competition put a new dimension to PM that raises exceptional challenges. This is true with regards to the setup of appropriate regulations that at present are confined to a national level and too weak to be able to govern the multinational developments in the context of PM. Furthermore, the enormous increase of information also poses exceptional challenges to researchers and clinicians to handle the information in a way that benefits the patient or research subject. Finally, as described above, ethical issues in PM may also be forcing a reconsideration of the structure and nature of the informed consent process. This reconsideration has important messages and implications for other areas of medicine, not just PM. For example, a re-examination of the consent process in PM is constructively raising issues and their solutions that ought to be applied more routinely in other practice and research areas.

The purpose of this section was to highlight areas of ongoing concern within the ethical domain of PM in order to raise awareness of impacts that ethical considerations can have on the quality of PM research and practice from a systems' point of view. That is not to say that these ethical issues necessarily have a universally acceptable remedy, but that failure to seriously address these issues in planning for PM transformations can undermine progress in other areas of PM in the context of a systems approach that depends on the confidence of those who interact with it.

From an ethical perspective the patient or research participant should be the dominant stakeholder in such considerations. Further deliberation of how these various issues are revealed to participants, through either informed consent processes, or other mechanisms of transparency (another ethical principle), including invitations to participate in these debates, should be part of ongoing planning discussions about how to incorporate ethical considerations into processes for maintaining and improving the quality of PM research and care.

Through the examples provided, we are prompting those concerned with the quality of PM research and care and who are involved in planning and promoting PM initiatives to consider the ethical implications in all areas of healthcare.