Ten years after the human genome project, the potential for personalized medicine lies not in a single genome but in many1. Before physicians can discern features of disease that are specific to individuals, researchers need to catalogue the enormous range of genomic and phenotypic variation in human populations. Collections of biospecimens that are accompanied by data on medical history, behaviour and health outcomes are crucial to this task1,2,3. If built well, these biobanks will facilitate research on health and disease for decades to come, but they require significant investment up front. This includes getting large numbers of people to submit biospecimens and personal health information.

Building such participation has not been easy. The US National Cancer Institute's (NCI's) Office of Biorepositories and Biospecimen Research found in 2009 that 80% of researchers have difficulty acquiring quality biospecimens, and that this compromises research findings4. It is understandable that people hesitate to participate, as controversies have eroded public trust in scientists who collect specimens for research. A 2009 lawsuit against the Texas Department of State Health Services alleged that researchers had violated the rights of newborns by not seeking parental consent to use their bloodspots in research, and led to the destruction of 5 million samples (see Nature 475, 156–158; 2011). The 2010 book The Immortal Life of Henrietta Lacks by Rebecca Skloot raised questions about researchers' obligations towards Henrietta Lacks, who became an unwitting research subject when her cancer cells were used to create an important research tool: the HeLa cell line. In 2010, Arizona State University in Tempe settled a lawsuit stemming from the allegedly unauthorized use of blood samples from the local Havasupai tribe (see http://go.nature.com/kcuacr for more).

A participant's involvement in research doesn't have to end the day a sample is collected. Credit: NATIONAL PSORIASIS FOUNDATION

Partly in response to these controversies, and for the first time in 20 years, the United States has proposed substantial changes to 'the Common Rule', the regulations that govern research involving human subjects5. The 90-day public comment period closes on 26 October. This is an opportune moment to reconsider the role and rights of participants in biomedical research.

Current practices in managing biobanks tend to see the public as little more than a resource for mining data and materials, and as a potential source of resistance. Participants provide information or tissues with little or no knowledge of the researchers' priorities, goals or expected outcomes. Barriers are erected. Materials and information are 'de-identified' to protect people's identities. Participants neither see how their donations are used, nor what the research produces.

Because success will depend on earning the public's trust, scientists must also trust the public.

As they stand, the proposed changes to the Common Rule5 risk further widening the divide between researchers and donors.

The changes encourage blanket consent — asking donors to authorize virtually everything, with opt-out checkboxes for predefined categories of research that might pose “unique concerns”, such as creating a cell line5. Scientists could use the samples for additional projects without seeking re-consent. The changes are intended to ensure that the scope of authorized research is clear to all parties, thereby circumventing the ambiguities that were at the heart of the Havasupai case. But this is achieved by telling donors next to nothing about how and for what purpose their donation will be used. Although this may reduce administrative complexity, we believe that it will decrease, not increase, public involvement in biobanks and fail to deliver on its ethical aim of better protecting participants' rights.

Most people prefer to have a say in how their donations are used. In a study in which donors were asked whether they would make their de-identified data available for future research, 90% said it was 'important' that they were asked for permission — yet the vast majority (86%) gave their consent6.

We need an alternative approach, in which donors are made partners by staying connected to research. Partnership is a win–win approach: it will build trust, make research better and faster, and generate large, diverse cohorts with longitudinal data1,2,3. By staying connected with participants (for example through the Internet), scientists can seek additional information and assemble cohorts for specific studies without the administrative hassles of the current system.

Pathways to partnership

Instead of letting failures and controversies drive reform, we can learn from projects that have developed innovative ways of connecting participants with research. Disease advocacy organizations have constructed biobanks by mobilizing disease communities to make rare biospecimens available to researchers. They have also demanded a more active role in shaping biomedical research priorities, offering banked materials to encourage the research they care about. The advocacy organizations reject the assumption that the asymmetry in scientific expertise between researchers and participants justifies asymmetry in research governance. Working with scientists, these organizations have enhanced research: they have uncovered genes responsible for Huntington's disease and multiple myeloma; discovered biomarkers for spinal muscular atrophy; and coordinated clinical trials for children with ataxia telangiectasia. One study7 found that 91% of these organizations assist in participant recruitment, 75% collect data for research, 45% support a registry or biobank and 56% have assisted with study design. Because the rules that govern most mainstream biobanks preclude these activities, many advocacy organizations have created their own biobanks — such as the Genetic Alliance Biobank, which researchers can only access by applying to an advocacy organization.

But disease advocacy organizations are primarily motivated to find cures for specific diseases, not to bolster general, exploratory research. Therefore, resources such as the Genetic Alliance Biobank, although valuable, do not achieve an important goal of biobanking — to build diverse collections in order to reveal factors such as genes and biomarkers that cut across disease categories or express variably in the same individual as disease develops2,3. However, companies such as PatientsLikeMe in Cambridge, Massachusetts, and 23andMe in Mountain View, California, demonstrate that individuals who lack the strong motivation and resources of disease advocates will nevertheless participate in exploratory biobanking, if the terms are right.

Credit: SOURCES: PATIENTSLIKEME; 23ANDME

Both companies offer benefits to individuals — personalized genomic information in the case of 23andMe and insights into disease conditions with PatientsLikeMe — while building collective resources for research. Both biobanks have developed rapidly and with remarkably low overheads by using the web (see 'People power'). PatientsLikeMe has built an extensive repository of detailed phenotypic and health data, crowd-sourced from a community of more than 115,000 participants. 23andMe now holds biosamples and genomic data from in excess of 120,000 individuals, more than 87% of whom have opted to participate in research. These biobanks are powerful tools. Researchers can assemble cohorts in hours rather than months, and rapidly acquire follow-on data from participants with minimal administrative hassle. When a large multi-year study funded by the US National Institutes of Health revealed a gene related to Parkinson's disease, 23andMe nimbly replicated the finding within an hour, using genomic data it already had in its then much smaller 30,000-person database, which included more than 3,000 people with Parkinson's8.

Through a web portal, participants control the information they provide, see how it is put to use and are regularly updated on what new findings mean for them. These companies have achieved trust, while still protecting privacy. The result is that participants have chosen to give more than the minimum. They are rewarded by witnessing scientific progress in process — including what new knowledge means for them as individuals.

These companies, like advocacy organizations, operate on the fringes of biomedicine and are regarded with suspicion by some, who doubt the validity of self-reported data collected online, and fear the risks of direct-to-consumer genetic testing. But the rapid growth of the companies' biobanks suggests that they have found a promising way to achieve trust while advancing science.

A better infrastructure

Rather than encourage a one-size-fits-all consent process5, the Common Rule should be revised to give participants the choice to remain involved with research. For instance, donors could track the studies to which they are contributing; make choices about the level of privacy protection for specific studies; receive medically significant incidental findings (possibly via their physicians); and have the option to contribute additional information for research on diseases they care about.

More choice and greater accountability can be easily and efficiently achieved using the Internet. With a secure web portal, participants can log in to receive personalized information and track research using their samples. Under current management practices, researchers who need more information from participants, such as whether they experienced specific side effects from a certain drug, must navigate a complicated administrative process to recontact people. If a secure web portal existed, however, participants could be efficiently contacted at virtually no cost to seek additional data, reconsent or even to be notified of clinical trial opportunities — all without compromising privacy. And blanket consent with no future contact would still remain an option, but as one choice among many. Along with revising rules, the US government should take the lead in building a web-based infrastructure.

Making biobanking more effective is not simply a problem of administration or building systems. We must also rethink the ethical foundations that underlie research. The bioethical principle of autonomy as 'respect for persons' should not be narrowly construed in the context of biobanking to mean only freedom from coercion — in effect, the right to sign a content form. It should also entail a respect for the ability, willingness and right of participants to share in imagining the futures to which research aspires. If human subjects are asked to give material from their bodies for research, they should also be treated as competent to govern the material's future uses9.

Because success will depend on earning the public's trust, scientists must also trust the public — to understand what matters for science, and to participate seriously in furthering the acquisition of knowledge. Such an approach has proved highly productive at the margins. It should be brought into the mainstream.