Patients care not only about the potential consequences of being identified based on genetic information, but also about feeling respected in their role as a research participant. [© drizzd - Fotolia.com]
Three months after examining issues associated with researchers using data from whole-genome sequencing, the Presidential Commission for the Study of Bioethical Issues turned its attention to the other side of the balance it is striving to strike, by discussing issues associated with protecting the privacy of the individuals whose genomes are under study.
At the dawn of genomics more than a decade ago, during the human genome study, the field saw very open and broad data sharing. Eventually, individuals’ genomic information was put out into databases available on the internet to anybody who could access them.
Against that backdrop, Russ B. Altman, M.D., Ph.D., of Stanford University and co-authors called for “new policies or creative technical approaches” in a 2004 Science paper, after noting that surprisingly small amounts of individual genomic sequence data such as SNPs are identifiable, though efforts to understand disease susceptibility or therapeutic opportunity require access to large genomic datasets.
In a Nature Reviews Genetics paper published last year, David W. Craig, Ph.D., of The Translational Genomic Research Institute (TGen) and co-authors found that meta-analysis across multiple genome-wide association studies (GWASs) with combined cohort sizes of tens of thousands of individuals often uncovered many more genome-wide associated loci than the original individual studies. “Even sharing summary-level data such as allele frequencies inherently carries some degree of privacy risk to study participants,” the study cautioned.
“These studies have really led to kind of a shift in policy away from sort of sole emphasis on making genetic data publicly available to more restricted data-sharing policies and the establishment of what we call controlled access to scientific databases,” Amy L. McGuire, Ph.D., associate director of research for the Center for Medical Ethics and Health Policy at Baylor College of Medicine (BCM), told the commission.
The result has been a 180-degree swing from total open data to a near exclusive focus on risk-benefit calculation and on protecting human subjects and protecting patients from potential harm. Dr. McGuire presented a more balanced perspective before the commission, calling for identifying the benefits of making that data available to research participants and patients, so they can make better-informed decisions about how they want their data shared. Dr. McGuire was one of four speakers who discussed data privacy and its potential applications in genomics during the Presidential Commission’s ninth meeting.
Ways to Protect Privacy
Dr. McGuire cited the importance of perceived risks and benefits and the resulting unease with which many patients view data sharing. While that reflects the perception that harm can come from being identified on the basis of genomic information, she said, addressing that fear alone won’t dispel the perception among patients: “They also care, maybe more, about feeling respected in the role they’re in, feeling respected as a research participant.”
Dr. McGuire led a research team that published results in Genetics in Medicine last year from their study of consent given to researchers by 323 adults recruited into one of six genome studies at BCM. The adults were randomly assigned into one of three groups: a traditional consent level allowing public data release, a “binary” level with restricted release, and a no-release level.
Before being debriefed by researchers, 83.9% of patients favored public release, 6.6% the restricted release, and 9.6% no release. Following debriefing, however, the majority allowing public release slid to a slim 53.1%, while restricted release leaped to 33.1%, and no-release rose to 13.7%.
Among issues the commission will consider is how broadly or narrowly should researchers secure consent from their patients? Can both be used depending on the circumstance? Dr. McGuire called for more transparency by researchers in explaining to patients how data will be used but added that transparency alone was not sufficient for securing their consent.
The balancing act will also have to account for growing scrutiny by patients of how data collected from them is being used by researchers. Madison Powers, Ph.D., senior research scholar, Kennedy Institute of Ethics and a philosophy professor at Georgetown University, noted that some broadly restrictive privacy rights in genetics and medicine are needed for:
- preserving unforced deliberation and decision about life matters “of profound and pervasive effect”
- avoiding social stigma
- controlling personal information about health prospects
- restricting information about intimate conduct
- preventing legal discrimination
Agreeing with those values significantly was Dr. Gutmann, who is also president of the University of Pensylvania. She added a sixth value: stopping misuse of genetic information by people with conflicting interests.
Because potential future misusers may be researchers, Dr. Powers added, “I’m all for science to be able to store and retrieve and utilize in the future but with a very cautious attitude toward re-consenting.” He added that broadly restrictive privacy rules are especially valuable in situations where one’s livelihood including access to healthcare could be at grave risk based on genetic data disclosure.
Giving Patients Control
“There is a growing population of people who want something more granular than that,” countered Greg Biggers, a council member of the Genetic Alliance. “Do you think we can get more granular than that?” Dr. McGuire asked.
Biggers, who is also CEO of Genomera, replied, “absolutely.” Genomera has developed a personal health collaboration platform that facilitates sharing by individuals of genomic and phenotypic information. “People store their data in this system in the cloud. But each individual controls where their data can flow.”
While Biggers had no moral objection to researchers obtaining consent through blanket authorizations, he expressed how social media is changing the expectations of patients. “As people are finally beginning to become comfortable with granular settings on Facebook, so that not every photo is shared with everyone, similar things are beginning to happen in health,” Biggers said. “There is a new tide coming into the harbor of health research, a tide of empowerment for the people that doctors call patients.”
Social media and smartphones, he noted, have sped up the sharing of data and securing of patient consent that once required creating entire organizations in the 1990s and developing customized technology tools in the 2000s. “Where we’re headed is this concept of a network of networks that choose to cooperate, that federate together to get research done. And it’s beginning to happen.”
That desire to federate satisfies not only the agendas of patients but especially those of researchers eager to aggregate larger volumes of data. The commission’s challenge will be balancing the interests of both groups by ensuring any patient-protective recommendations on data sharing that it may make don’t add undue burdens in time or logistics that discourage new research.
Also addressing the commission was Leonard D’Avolio, Ph.D., associate center director for biomedical informatics with the U.S. Department of Veterans Affairs’ Massachusetts Veterans Epidemiology Research and Information Center (MAVERIC) in Boston. The 130-person multidisciplinary R&D center is one of three VA sites involved with the agency’s recently launched Point of Care Research Program, which aims to develop a better alternative to either costly, time-consuming randomized clinical trials or bias-prone observational studies.
Dr. D’Avolio said the VA has opted not to share data with the public, “because our responsibility to the veteran is to maintain their privacy first and foremost”—but added that researchers with “valid” scientific questions can access the data after undergoing VA’s consent process.
With research interest in veterans high as so many return from Iraq and Afghanistan, however, Washington will need to develop policies that protect veterans’ identities while promoting data sharing by investigators and patient groups with a stake in future treatments. Developing recommendations that enable that and other research is one way the commission can prove valuable.
Another important issue is setting parameters for patient consent. Dr. McGuire’s study offers three variations that are worth study by the commission. So too is the discussion by the speakers on authorizing consent based on circumstance, though just what those circumstances might be will be up to the commission to determine.