For many large-scale research networks, sharing data is at the heart of their value proposition. Pooling data to generate large data sets has the potential to increase scientists’ confidence in statistical findings, decrease duplication of efforts, attract talented researchers to their fields and drive scientific discovery—in short, sharing data can make for better science.
The benefits of sharing data are now well-accepted and have spurred the formation of many data-sharing initiatives. However, at the time that the Malaria Genomic Epidemiology Network (MalariaGEN) was first founded, there was no map or formula for how to build a living, breathing data-sharing community.
“The starting point for MalariaGEN was a research consortium to investigate human immune responses to malaria. We knew that building a core resource of clinical, epidemiological and genomic data was essential to achieving our scientific goals. For a variety of very good reasons, this wasn’t going to happen spontaneously,” recalls Dr. Dominic Kwiatkowski, CGGH Director and founder of MalariaGEN.
The network needed to deeply engage with the process of developing of an ethical framework and, eventually, to collectively agree on policies for data ownership and data sharing, both internally and externally.
Establishing internal policies that facilitate data-sharing
To help establish this dialogue and to support the ethical components of MalariaGEN’s research, Kwiatkowski reached out to Professor Michael Parker, Director of the Ethox Centre at the University of Oxford. This marked the beginning of a longstanding partnership, as Parker became a founding member of both MalariaGEN and the CGGH, overseeing the ethics programme for both.
One important aspect of this work was determining an approach to data ownership and sharing that was suited to the unique nature of MalariaGEN which is large, diffuse (partners work in more than 20 countries), and characterised by varying capacity amongst collaborators, particularly those working in resource-limited environments.
At the time, this was somewhat uncharted waters. Existing policies and frameworks had largely been developed for smaller, bi-lateral collaborations being undertaken in high-income settings.
“We were operating in an area beyond the guidelines and regulations. We had to take a framework and literature about ethical issues that was developed in one setting and sensitively determine what were the ethically salient features of the other setting to figure out how we identify and respond to those issues, and what we think is the appropriate way forward,” recalls Dr Susan Bull, who served as the CGGH Senior Ethics Coordinator (2008-2012) and continues to work with the Ethox Centre.
Bull was appointed to support MalariaGEN alongside Dr Jantina de Vries who led the work to establish MalariaGEN’s informed consent guidelines and worked with MalariaGEN partners to obtain the required ethics approvals. Building this dedicated team, under Parker’s leadership, was a noteworthy investment of resources and signalled a deep commitment to tackling the ethical issues in a meaningful way.
The ethics team, working with MalariaGEN’s many collaborators, dug in. “It’s about generating a particular kind of discussion. Just as scientists might argue about the merits of particular strategies in relation to genomic analysis, for example, so ethical problems get talked about in a similar kind of way,” explains Parker, who recalls the process as hard but rewarding work.
“It requires communication, a lot of collaborative relationship building, and getting people to talk to each other and work things through. The way we did it with the data sharing policy, for example, was we developed a number of early drafts that were circulated, we had meetings, got people together, and worked on subsequent versions until we reached a point where consensus had been achieved," he adds.
Out of this dialogue, MalariaGEN generated a deep understanding of the challenges facing researchers operating in low- and middle-income settings and the ethical issues that arise from the unique nature of genomic studies.
This informed MalariaGEN’s policies for data ownership and sharing. A key collective decision was that partners would retain ownership of their samples and the resulting data. Specific projects were defined up front with clear guidance about how the samples would be used. Partners then chose to contribute to one or more of these discreet projects.
Every partner knew up front what data would be generated on their samples, how it would be pooled with other data contributions, and what joint analysis would be done. Any new uses, for example those made possible by advancing technologies, would need to be agreed with the relevant contributors.
Debating key ethical issues—scientific capacity and local contexts
When considering an ethical and sustainable approach to external data-sharing, two critical challenges stood out: the need to build scientific capacity and to acknowledge the local context where the research was done.
The first of these challenges is perhaps unsurprising. There is a well-known tension between maximising access to the data—getting it out there as widely and quickly as possible—and the need to be respectful of the legitimate interests of researchers to use their data to pursue their intended scientific goals, ultimately publication.
This trade-off is addressed in the Fort Lauderdale agreement, for example, but at the time that MalariaGEN was considering its stance on data-sharing, the implications of this still-real tension hadn’t been considered deeply on the same scale and in a context where some collaborators were working in low- and middle-income settings.
“It became apparent through our discussions about data-sharing that there might be ways in which you would undermine scientific capacity if data were shared too quickly. One concern would be people from elsewhere would analyse this data and get publications out of it quicker than partners in developing countries, let’s say. This could mean that partners wouldn’t get cutting-edge publications which would mean that they’d either stay at a fairly low-level in terms of analytic ability or they wouldn’t get promotions or grants,” explains Parker.
“And that would be unfair for one thing,” adds Parker. “But apart from being unfair, the future of science depends on researchers in these parts of the world building capacity and having strong scientific skills. There was a need to manage access to data so that scientific capacity was supported and encouraged. This is complementary to the spirit of Fort Lauderdale and open access, which is to promote and support scientific progress.”
The second challenge arose from the fact that MalariaGEN’s partners were working in more than 20 countries, many with vastly different cultural and political contexts. Parker points out this creates a risk when sharing data, “There is the potential for some ethical issue to arise because people using data elsewhere in the world might not be aware of the particular sensitivities in the kinds of settings where the data was collected.” By way of example, he mentions issues related to ethnicity and stigmatisation; the impact on populations and how they’re understood; and non-paternity.
MalariaGEN needed a data-sharing process that balanced these concerns against the risk of excessive protectionism, and facilitated informed judgements about what constitutes appropriate data use.
In response to this need, MalariaGEN adopted a managed access approach when it came to sharing human data with the wider scientific community which, Parker acknowledges, still has its critics. “Some people would prefer that data be available on a website, for example, but attitudes are changing. I think partly they’ve changed because it’s become apparent that groups like MalariaGEN and others that advocate managed access are not against data-sharing,” he says. “They’re just taking seriously the point that what’s behind data-sharing is the idea that it leads to good science. And, good science has an ethical component but it also has a sustainability component.”
MalariaGEN partners in low- and middle-income settings have had the opportunity to publish their work, including high-impact outputs like the 2009 paper ‘Genome-wide and fine-resolution association analysis of malaria in West Africa’ which published in Nature. The first author on the paper was Dr Muminatou Jallow, a paediatrician who works with children with severe malaria at the Royal Victoria Hospital in Banjul, The Gambia, and a former MalariaGEN Data Fellow.
MalariaGEN's experience developing an ethical framework for sharing human data has been distilled into a number of peer-reviewed publications, including the 2009 PLOS Medicine paper, ‘Ethical Data Release in Genome-Wide Association Studies in Developing Countries,’ on which both Parker and Bull are authors. Although it’s difficult to discern the influence of such a paper—for example, it may impact on the development of policies or grey literature that aren’t captured by citation indices—it can be said that the paper was viewed more than 4,000 times and downloaded 665 times (between its publication on 24 Nov. 2009 and 19 Jan. 2014).
A useful model for other scientific networks
By employing an engaged, context-specific approach to addressing the ethical issues, MalariaGEN and the CGGH did more than define a single ethical framework for MalariaGEN—they set an example that has become an important resource for other large-scale data-sharing networks including the Plasmodium Diversity Network Africa (PDNA), the Human Heredity and Health in Africa Initiative (H3 Africa) and the International Severe Acute Respiratory and Emerging Infection Consortium (ISARIC).
“The data sharing policies and governance, and the conceptualisation of that—it’s one of those frustrating things that doesn’t look particularly exciting from the outside but the amount of work and thinking that went into it is tremendous. It’s an innovative model that’s been valued in its own right and the materials have been useful to help others developing policies in this area,” says Bull.
This isn’t to say that these networks have adopted the same policies wholesale; it’s more that these networks are beneficiaries of the lessons learned and the approach pioneered by MalariaGEN and the CGGH. And, in many cases, from the capacity that was built within a cadre of young ethics researchers who have thought deeply on these issues and are now sharing that expertise with other genomic networks.
If there is a clear message to emerge from MalariaGEN’s ethical deep dives, it’s probably that genomics research raises a number of interesting ethical questions and none of them have a single, simple answer. Bull hits at the crux of the challenge, when she says: “There really is no one-size-fits-all solution. There is room for reasonable people to disagree.”
In this complex landscape, having a place to start deliberations around data ownership and sharing can provide newly-formed networks an enormous leg up.
As a frequent participant in such discussions, Dr Victoria Cornelius, CGGH Business Manager, explains, “It’s best to talk about data policies up front—before a single sample is collected—but this creates a paradox: agreeing to a policy generally requires trust that is often still forming in the early days of a collaboration. When I think about the value of MalariaGEN’s policies, it’s not contingent on other people using the policy as is. It’s the fact that it gives other researchers a reference point for engaging in some quite hard discussions. Even if the policy unites a group to say ‘this isn’t what we want,’ it has started a conversation about what they need, what works for them in their context.”
Talking to researchers involved in these discussions, trust appears as a reoccurring theme. In many ways, the approach to developing the ethical framework helped to solidify MalariaGEN as a strong scientific collaboration. “I felt that working around the very practical problem of data release helped to build the MalariaGEN Consortium scientifically as well. It helped people to talk to each other about other issues. Working around specific problems can actually be a way of generating the network and making it live,” says Parker.
The dialogue around ethical frameworks continues today
“Even after you have appropriate policies in place, life isn’t easy. There is an on-going role for ethicists to work with the scientific community around data sharing—articulating and communicating policies,” says Kwiatkowski. “This is an area where the Ethox Centre has been tremendously successful: generating debate and engaging the scientific community in discussion around these issues.”
As a result, there is new a generation of researchers and ethicists primed to dig deeply into developing appropriate solutions as new ethical challenges arise. Inspired by MalariaGEN’s efforts to build scientific capacity—particularly the Data Fellows programme—Parker and his colleagues at the Ethox Centre approached the Wellcome Trust for support to put in place a similar programme around the ethics work.
“The experience with MalariaGEN and the CGGH was very influential in making our case, which resulted in the largest single largest grant by the Wellcome Trust exclusively for ethics,” says Parker.
This Wellcome Trust Strategic Award supported the formation of the Global Health Bioethics Network (GHBN) and funds a five year programme of ethics research and capacity building with Wellcome Trust’s Major Overseas Programmes in Kenya, Malawi, South Africa, Thailand-Laos, and Viet Nam.
In addition to the GBHN's capacity-building activities, which include an annual summer school and a bursary scheme for early career researchers, they are also conducting a programme of influential research on a number of important topics. A good example of this is a project to gather views in five low-incomce countries about what constitutes good data-sharing practice in research conducted in low-income settings.
And, so the debate around the ethical issues inherent in global health research continues today. In many ways this is the strongest outcome of the CGGH collaboration with the Ethox Centre: a lively ethical community that continues to support health research across the developing world.
About Ethox Centre
The Ethox Centre is a multi-disciplinary bioethics research centre based in the University of Oxford’s Nuffield Department of Population Health. The Centre aims to improve ethical standards in healthcare practice and medical research through education, research, and the provision of ethics support to health professionals and medical researchers. The Centre’s research activities are focused on four areas: global health ethics, clinical ethics, public health ethics, and research ethics. In each of these areas, the Centre seeks to engage with ethical issues faced by real-world practitioners in real-world settings; a particular emphasis is placed on approaches that bring together empirical social science research and ethical analysis.
- Bull SJ et al. Clinical Ethics, 2011; 4: 165-168.
- de Vries J, Bull SJ, Ibrahim M et al. BMC Med Ethics, 2011; Mar 18: 12:5.
- Parker M, Bull SJ et al. PLoS Med, 2009; Nov;6(11): e1000143.