October 20 2011
Ethics and de-identified genomic plus phenotypic (health) data
Image source: Amazon
During the annual Cerner Health Conference last week, several clients came up to discuss my previous post and associated essay on de-identified health data. In my essay, I basically agree with Marc Lappé and others who have asserted that medical ethics as it exists and is practiced today is different from—and historically antagonistic toward—public health ethics (‘communitarian ethics'). This point of view is covered at some length in a paper by Eric Meslin and Ibrahim Garba in the September issue of Human Genetics.
The clients who visited with me substantially agreed with the idea that there is presently an imbalance between individualistic ethics and communitarian ethics. They asked what resources might be available to correct this imbalance, particularly with respect to genomics-based research and predictive/preventive tests or algorithms based on these that might bear upon not only the person who was tested but also on their relatives and on members of their community or groups of people (ones currently alive, plus those who may be born in the future) who have similar genetic and other features.
One of the clients noted that Kathy Hudson's paper in the New England Journal of Medicine last month repeats the misconception that true de-identification of biologic specimens or genomic data is not always possible, because a small number of genetic markers can uniquely identify a participant. “Even aggregate data (e.g., pooled data from several hundred subjects) may not be safe,” Hudson says.
I advised them that that statement is only true if someone wishing to identify a participant (a) independently possesses some identified tissue from the individual that can be or has been sequenced or tested for markers that are in the other pooled data, or (b) independently has access to some identified, different data from the individual where those markers or sequences exist, such that the pooled de-identified data can be matched with certainty against the independent identified data. Hudson's conjectured inferring someone's identity with certainty from observable phenotypic characteristics (Shurin 2008) is therefore extremely remote.
“Study participants are asked to consent to future studies with unspecified aims, broad data sharing policies, and ongoing uncertainties regarding confidentiality protections and the potential benefit of incidental genomic research findings. Because more research is conducted under these new conditions, the very nature of the researcher-subject relationship is shifting and will require new governance mechanisms to promote the original goals of informed consent.” Gail Henderson, 2011
Consenting for secondary-use was another topic that clients visiting with me were keen to learn more about. In that regard, you may like to look at the paper by de Vries and colleagues. They point out differences in genomics research that affect the consenting process. Genomics and other biomarker studies are “hypothesis-generating,” rather than “hypothesis-testing” like most conventional studies. The scope and endpoints are unknown or only partially known at the beginning, so to do "narrow" consent either forecloses upon pursuit of hypotheses that are generated or forces re-consenting.
Genomics studies are undertaken in a climate of collaboration, with multiple secondary uses of the same information and biobanked specimens over an indefinite period of time. In the course of multiple secondary uses, sharing may extend to additional collaborators (both academic ones and commercial ones) who were not party to the original process (and were therefore not mentioned in the initial informed consent materials under which consent was obtained), whereas the investigational center(s) and sponsors in conventional studies are known to each study subject at the time of consenting. Conventional studies—with some exceptions, such as the Framingham Heart Study, generally have a crisply defined beginning and end.
“One ‘size' [with regard to genomics study ethics and secondary uses of data and specimens] does not fit all, and the development of appropriate responses to these ethical issues will need to be conducted on a case by case basis. What will link international genomics research projects, though, are the more general principles of justice, ownership and the fair distribution of resources.” Jantina de Vries, 2011
I will now summarize some of the recommendations I made to the clients who met with me at CHC last week. Links are appended at the bottom so you and they can access these resources easily and pursue your review and policy-making efforts on your own.
First of all, I strongly recommend conducting multiple pre-study workshops among the participating investigational centers that will be involved in any genomics-based study, to flush out and address differences in IRBs' interpretations of ethics requirements that the genomics study presents. This would be similar to what was famously done in the multinational MalariaGEN project. Each time a new secondary-use study that would re-use the materials collected in the original study is proposed, conduct a new round of workshops among the newly expanded set of centers to achieve transparency and consensus among the IRBs.
Secondly, the 2010 paper by Malin and coworkers at Vanderbilt is a wonderful survey of the issues and contains valuable suggestions. You will want to take a look at that paper. Other ideas, such as Contreras's notion of “latency” or several months or years “embargoing” of information sharing may also be applicable where the privacy or sensitivity of certain information has a time-scope that is shorter than the person's lifetime.
Narrow consent documents describe the benefits and risks of a specific study, whereas a broad consent may ask participants to agree to any number of future studies that will use their samples or data. In the case of NIH's dbGaP, because no patient identifiers are to be submitted (although they will be retained at the institutions where primary data collection occurred), the HHS Office of Human Research Protections (OHRP) opined that investigators who access dbGaP data must be secondary users who are not conducting ‘human subjects research' as defined by federal regulations. OHRP does not require that secondary users obtain IRB consent in order to access the data, nor do secondary users have to re-seek informed consent from the original subjects, who are de-identified in dbGaP.
However, because some data sets may contain information that is conjecturally (re-) identifiable (and would be re-identifiable if the secondary users were to obtain tissue or identified genomic information in some other way, from a set of people that happened to include subjects from the primary study), the primary study investigators may require that secondary users obtain approval from their own IRBs. That's what the Framingham SHARe did, for example. SHARe includes data from the Framingham Heart Study, which was based in a medium-sized community in Massachusetts. Each NIH Institute with data in dbGaP is creating a Data Access Committee to review applications and a Data Use Certification that secondary investigators must sign to gain access to a dataset. In fact, the NIH has established ongoing policies and procedures to oversee GWAS policy implementation across the agency and to monitor whole genome association data-use practices. Cerner's own data-use policies and procedures emulate those promulgated by NIH.
Finally, I suggested to the clients that the ethics of uncertainty has in the past been one-sidedly addressed (e.g., Loewy 1987) from the vantage point of the paternalistic clinician/expert—the possessor of facts or prognostic information—with regard to the provider's/researcher's duties to inform her passive, child-like patient/recipient and how to convey to that recipient the imprecision with which the diagnosis or prognosis is known or, alternatively, with regard to complying with the patient's request not to be informed.
Instead, in the future the ethics of uncertainty should, I believe, explicitly address things from the vantage points of other stakeholders, including the person (the one whose [genomic] information has been measured) and the person's family and community. In other words, these other stakeholders are bearers of rights and duties as well, and, so long as they are members of a family and community, they do not have the right to treat the group with contempt and unilaterally breach their duty of care.
In other words, the act that opter-outers “commit” by their opting-out is not a crime but a tort—those who capriciously withhold their consent for secondary uses of their de-identified information for public health and family and other valuable purposes. Heinz Kohut refers to such acts (or non-acts, or failures to act) as ‘breaches in empathy'. By far, the best account to-date of these ideas is in Peter Gerhart's book Tort Law and Social Morality. Future policy-making regarding human research subjects protections and genomics-based study ethics would do well to consider the ideas set forth by Gerhart, as a means of reconciling individualistic ethics and ‘other-regarding' communitarian/public health ethics.
Please feel free to add comments or contact me if you have criticisms, suggestions or would like to discuss further. Thank you!
Douglas McNair, MD, PhD, is president of Cerner Math, Inc., and one of three Cerner Engineering Fellows and is responsible for innovations in decision support and very-large-scale datamining. McNair joined Cerner in 1986, first as VP of Cerner's Knowledge Systems engineering department; then as VP of Regulatory Affairs; then as General Manager for Cerner's Detroit and Kansas City branches. Subsequently, he was Chief Research Officer, responsible for Cerner's clinical research operations. In 1987, McNair was co-inventor and co-developer of Discern Expert®, a decision-support engine that today is used in more than 2,000 health care facilities around the world. Between 1977 and 1986, McNair was a faculty member of Baylor College of Medicine in the Departments of Medicine and Pathology. He is a diplomate of the American Board of Pathology and the American Board of Internal Medicine.
Anker M. The Ethics of Uncertainty: Aporetic Openings. Atropos, 2009.
ASCO. Policy statement update: Genetic testing for cancer susceptibility. J Clin Oncol. 2003;21:2397-406.
Avineri S, de-Shalit A, eds. Communitarianism and Individualism. Oxford Univ, 1992.
Badiou A, Zizek S. Philosophy in the Present. Polity, 2009.
Bell D. Communitarianism and Its Critics. Oxford Univ, 1993.
Clayton E, Ross L. Implications of disclosing individual results of clinical research. JAMA 2006;295:37-8.
Contreras J. Information access: Prepublication data release, latency, and genome commons. Science. 2010;329:393-4.
de Vries J, et al. Ethical issues in human genomics research in developing countries. BMC Med Ethics. 2011;12:5-14. (discusses MalariaGEN example)
Etzioni A, ed. New Communitarian Thinking: Persons, Virtues, Institutions, and Communities. Univ Virginia, 1995.
Gerhart P. Tort Law and Social Morality. Cambridge Univ, 2010.
Goldstein D. Growth of genome screening needs debate. Nature. 2011;476(7358):27-8.
Greely H. The uneasy ethical and legal underpinnings of large-scale genomic biobanks. Ann Rev Genomics Hum Genet 2007;8:343-64.
Held V. The Ethics of Care: Personal, Political, Global. Oxford, 2007.
Henderson G. Is informed consent broken? Am J Med Sci. 2011;342:267-72.
Hudson K. Genomics, health care, and society. NEJM 2011;365:1033-41.
Kohut H. The Chicago Institute Lectures. Routledge, 1996.
Krawczak M, et al. Is the NIH policy for sharing GWAS data running the risk of being counterproductive? Investig Genet. 2010;1:3-6.
Langanke M, et al. Comparing different scientific approaches to personalized medicine: Research ethics and privacy protection. Per Med. 2011;8:437-44.
Lappé M. 'Ethics and public health,' in Maxcy-Rosenau Public Health and Preventive Medicine. 12 e. Last J, ed. Appleton-Century-Crofts, 1986, pp. 1867–77.
Lemke A, et al; GRRIP Consortium. Broad data sharing in genetic research: Views of institutional review board professionals. IRB. 2011;33:1-5.
Loewy E. The uncertainty of certainty in clinical ethics. J Med Humanit Bioeth. 1987;8:26-33.
Lunshof J, et al. From genetic privacy to open consent. Nat Rev Genet 2008;9:406-11.
Lin Z, et al. Genomic research and human subject privacy. Science 2004;305:183.
Malin B, et al. Technical and policy approaches to balancing patient privacy and data sharing in clinical and translational research. J Investig Med. 2010;58:11-8.
McGuire A, et al. To share or not to share: A randomized trial of consent for data sharing in genome research. Genet Med. 2011 Jul 22.
Meslin E, Garba I. Biobanking and public health: Is a human rights approach the tie that binds? Hum Genet. 2011;130:451-63.
NIH. Policy for Sharing of Data Obtained in NIH-Supported or Conducted Genome-Wide Association Studies (GWAS), 2007. (http://grants.nih.gov/grants/guide/notice-files/NOT-OD-07-088.html )
NIH. Notice on Development of Data Sharing Policy for Sequence and Related Genomic Data, 2009. (http://grants.nih.gov/grants/guide/notice-files/NOT-HG-10-006.html )
NIH. GWAS: Genome Wide Association Studies. (http://grants.nih.gov/grants/gwas)
NIH. Implementation Guidance and Instructions for Applicants: Policy for Sharing of Data Obtained in NIH-Supported or NIH-Conducted GWAS, 2008. (http://grants.nih.gov/grants/guide/notice-files/NOT-OD-08-013.html)
NIH. dbGaP. (http://www.ncbi.nlm.nih.gov/sites/entrez?db=gap)
Noddings N. Caring: A Feminine Approach to Ethics and Moral Education. 2e. Univ California, 2003.
Pettersen T. Comprehending Care: Problems and Possibilities in The Ethics of Care. Lexington, 2008.
Putnam R. Bowling Alone: The Collapse and Revival of American Community. Touchstone, 2001.
Shurin S. The genome-wide association studies data sharing policy. Clin Transl Sci. 2008;1:91.
Slote M. The Ethics of Care and Empathy. Routledge, 2007.
Speicher M, et al. Effect of genome-wide association studies, direct-to-consumer genetic testing, and high-speed sequencing technologies on predictive genetic counselling for cancer risk. Lancet Oncol. 2010;11:890-8.
Tronto J. Moral Boundaries: A Political Argument for an Ethic of Care. Routledge, 1993.
Wilcken B. Ethical issues in genetics. J Paediatr Child Health. 2011;47:668-71.
Zawati M, et al. Closure of population biobanks and direct-to-consumer genetic testing companies. Hum Genet. 2011;130:425-32.