The potential benefits of data sharing for secondary research purposes are well recognized, but certain ethical barriers must be addressed before these benefits can be fully realized. American Indian and Alaska Native (AI/AN) individuals and tribes that participate in research are often interested in sharing their data when doing so will help answer important questions or contribute to improved population health outcomes, and when appropriate protections are in place. While perspectives on data sharing agreements and processes can vary among AI/AN individuals and tribes, some views are consistent and much can be learned by examining how specific studies have navigated data sharing in ways that both respect tribal sovereignty and meet the overall study goals. What best practices, lessons, and examples exist for IRBs, researchers, and funders who wish to include a data sharing component in research with AI/AN individuals and tribes?
In October 2017, PRIM&R hosted a webinar in collaboration with the National Congress of American Indians Policy Research Center titled Data Sharing in Research with American Indians and Alaska Natives: Informed Practices, Considerations, and Case Studies. Researchers from the Strong Heart Study, the American Indian and Alaska Native Head Start Family and Child Experiences Survey (AI/AN FACES), and the Safe Passage Study comprised the speaker panel for this webinar. Through these three case studies, webinar attendees learned some approaches for operationalizing data sharing arrangements in research with AI/AN populations.
After the webinar, speakers addressed some audience questions that time did not permit them to answer during the live webinar. Lizabeth Malone, PhD and fellow AI/AN FACES study team members Michelle Sarche, PhD, Laura Hoard, PhD, and Meryl Barofsky, PhD, elaborated on the importance of community engagement and how it enriches all types of research. Marcia O’Leary, RN, from the Strong Heart Study, discussed her experience working as manager for a Native-owned research organization, Missouri Breaks, and the specific processes and procedures her organization employs in order to ensure protection of AI/AN research partners and their data. Jyoti Angal, MPH, CIP, described how the Safe Passage Study addressed the specific complexities of genetic data with study subjects. We’re pleased to share the speakers’ responses with the readers of Ampersand.
Q: Are there lessons from your work with tribal populations that may be important for data sharing conversations in research with non-tribal populations? Or for those on the panel that have worked with both tribal and non-tribal populations, are there clear differences in your work across these groups?
AI/AN FACES Team: Prior to conducting research with tribal populations, particularly those within the jurisdiction of federally-recognized tribal governments, researchers must first obtain approval from the relevant tribal authority. In order to obtain research approval, most tribal research review entities require that researchers demonstrate the involvement and support of tribal community stakeholders. While work with non-tribal populations may not carry the same imperative for formal community research review and approval, researchers working with non-tribal populations can still learn from the ways in which researchers working with tribal populations engage in extensive consultation and collaboration with community stakeholders. Some non-tribal communities or settings (such as school districts) may also require local review, but the engagement of the community as a partner and participant in the research likely differs. While a requirement in tribal populations, community engagement is not just a means to an end. Rather, it enriches research— by aligning priorities, by grounding the understanding of the phenomenon of interest within local ways of knowing, by informing study implementation, and most importantly, by increasing the likelihood of community benefit. As it relates to data sharing in particular, researchers should be upfront and clear about the data sharing approach, including plans for how data can be accessed and by whom, the protections that will be in place for ensuring data security, and how the data will be used and findings shared.
Marcia O’Leary (MO): As a Native-owned organization, Missouri Breaks’ primary focus has been research concerning tribal communities. It is important to acknowledge that research benefits everyone, and that these benefits are bilateral. Tribal populations have been beneficiaries of the research that has been done across the nation and the world; in turn, other populations learn and benefit from research conducted in tribal communities. Data is, by definition, quantifiable. Data sharing carries the risk of privacy violation, which in turn can lead to the perpetuation of harmful stereotypes. Tribal communities have seen the damage that can occur when data is used without appropriate oversight. It is very difficult to undo misinformation. Researchers need to work closely with tribal leadership to ensure appropriate data oversight is addressed in their research plan.
Jyoti Angal (JA): Some of the key ethical issues surrounding tribal data sharing are also important for non-tribal populations. For example, the informed consent form should include a provision for data sharing and indicate how it will be protected from misuse. In addition to confidentiality, privacy considerations may be important when collecting sensitive data that may stigmatize certain groups or communities.
Q: With regard to sharing either genetic data or data from children, have the populations you’ve worked with expressed any unique concerns and, if so, how were you able to address them in your data sharing agreements?
AI/AN FACES Team: Tribal communities, like many communities, are especially protective of their children. To this end, we worked to ensure: 1) that the experience of research participation was a positive one for children; 2) that findings would be shared in ways that were respectful of tribal children, families, and communities; 3) that data are shared in ways that do not identify individuals, programs, or communities; and, 4) that access to AI/AN FACES data would only be granted to researchers who demonstrated both a technical and cultural understanding of research with tribal populations.
To maximize the likelihood that the experience of research was positive for children, data collectors participated in an AI/AN cross-cultural understanding training and practiced data collection in tribal settings to ensure cultural sensitivity in engaging with AI/AN children and families. To maximize the likelihood that findings are shared in ways that are respectful of tribal children, families, and communities, we are working closely with tribal Head Start partners to bring a cultural and contextual perspective to bear on the interpretation of findings. To ensure that data are shared in ways that do not identify individuals, programs, or communities, findings are reported at the group level and not for any one individual, program, or community, and the restricted use dataset has been stripped of any individual, program, or community identifiers. To maximize the likelihood that data access is only granted to researchers who demonstrate both a technical and cultural understanding of research with tribal populations, we have, in consultation with our tribal Head Start partners and in accordance with the tribal agreements for the study, developed a rigorous application process for data access.
MO: A commitment to privacy and protection of human subjects is central to any professional relationship in medical care, regardless of the patient or the research procedures. The Strong Heart Study has developed guidelines for both the collection and sharing of data in order to ensure appropriate oversight—both from the participant as well as from tribal partners. Although participants understand in medical research that the testing is intended to inform public health questions, there is still a need to address actionable results. In the Strong Heart Study, all participants are given a copy of their results, and it is their choice if they want a copy of those results sent to their health care provider. It is also the participant’s right to choose not to receive a copy of results, or not to have it shared with medical providers. In the case of children, this is a decision for parents and guardians.
The Strong Heart Study has developed processes for data sharing that ensure data is used only to address research questions that stay within the original scope of the participants’ consent and within the tribally and IRB approved aims of the research. This stringent process requires that individuals work closely with one of the principal investigators to review and agree to a data sharing contract, and that they seek permission of the tribal partners prior to publication.
JA: The tribes participating in the Safe Passage Study were concerned about whether participants would understand the complexity surrounding genetic data collection and sharing. To address this issue, the study developed a ‘genetic informational sheet’ to help communicate the importance of genetic data collection and how the data would be used. Participants were allowed to opt out of genetic data collection. Additionally, the consent form included language to indicate that the future use of genetic data would be restricted to studies related to fetal and infant development and wellbeing.
Q: In some recent conversations about data sharing in research, people have begun to make a distinction between data sharing and data use agreements. In your work is there a distinction between the two?
AI/AN FACES Team: In our work, we think of data sharing agreements broadly to reflect agreements between researchers and tribal research review authorities about the conditions under which researchers can, or are required to, share findings or data files. Alternatively, we consider a data use agreement to be an agreement between two research institutions that covers the conditions under which researchers outside of the original study team may be granted access to the data for the purpose of secondary data analysis. In the case of AI/AN FACES, a data use agreement specifies how researchers will protect the privacy and confidentiality of the data and the individuals and communities (even though not identified in the restricted-use data file). Separately, secondary researchers must acknowledge they will share any findings prior to presentation for review by a study data committee (with tribal representation) and “strongly consider” the committee’s feedback.
MO: Each tribal community has its own processes, and it is important for the researcher to become familiar with the procedures and practices of each tribal nation they work with. Some of our tribal partners prefer to utilize a Memorandum of Understanding (MOU); others have a paper approval process determined by a committee, a designated office, or an individual with designated authority to review and approve publications. Using data outside of the proposed aims of the consent and approvals of the tribal partners is never acceptable. For the Strong Heart Study, data sharing is a process through an application which is reviewed by a paper proposal committee to insure that it is within the intent of the consent. A Strong Heart Study investigator must be a part of the writing team, and the paper must be approved by a designated tribal authority prior to publication.
JA: In our experience, tribes have used data sharing and data use agreements interchangeably. These agreements are very detailed and outline terms and conditions under which data from individual projects would be sharing with the broad research community.
The responses provided here and throughout the webinar highlight how community engagement and respect for tribal sovereignty are essential elements of the data sharing approaches employed by each of the three studies featured on the panel. Data sharing in tribal research requires flexibility and an investment in the time required to identify an approach that both benefits and protects AI/AN individuals and tribes and allows for the primary goals of the study to be met. While these study examples do not provide a step-by-step guide to data sharing in research with AI/AN individuals and tribes, they do describe approaches and lessons that IRB staff, researchers, and funders can build upon to strengthen their work with AI/AN populations.
PRIM&R thanks the National Congress of American Indians Policy Research Center for partnering to produce this webinar. The NCAI Policy Research Center was established by NCAI in 2003, and its mission is to provide tribal leaders with the best available knowledge to make strategically proactive policy decisions in a framework of Native wisdom that positively impact the future of Native peoples. The NCAI, established in 1944, is the oldest, largest and most representative American Indian and Alaska Native organization serving the broad interests of tribal governments and communities.
The recording of this free webinar can be viewed on PRIM&R’s website.
For more information and resources on data sharing, visit the webpage for PRIM&R’s 2017 workshop, Ethics of Data Access, Use, and Sharing for Human Subjects Research.