Understanding the NIH Genomic Data Sharing (GDS) Policy
The NIH GDS policy requires the submission of large-scale genomic data, as well as relevant associated data, to an NIH-designated data repository. The policy became effective January 2015 and expands the scope of the original 2007 policy that was specific to genome-wide association studies (GWAS) studies. The GDS policy applies to all NIH-funded research that generates large-scale genomic data, regardless of the funding level, as well as the use of these data for subsequent research. Examples of large-scale genomic data include genome-wide association studies (GWAS), single nucleotide polymorphisms (SNP) arrays, and genome sequence, transcriptomic, metagenomic, epigenomic, and gene expression data.
Institutions are responsible for assuring, through an institutional certification, that plans for the submission of large-scale human genomic data to the NIH meet the expectations of the GDS policy. An Institutional Certification must accompany the submission of all large-scale human data to the NIH repository. Investigators are responsible for requesting this institutional certification from the IRB prior to the sharing of any data with the federal repository.
How do I know if the NIH GDS policy applies to my study?
If a study is NIH-funded and involves the creation of large-scale genomic data, the GDS policy applies. The policy requires that investigators provide a basic genomic data sharing plan in the Resource Sharing Plan section of grant applications at the time of submission. A more detailed plan should be provided prior to award. The Institutional Certification also should be provided prior to award, when possible. Guidance on genomic data sharing plans is available on the NIH GDS website.
If researchers are unclear whether the GDS policy applies to their study, they are encouraged to consult with their NIH program officer. If a project has already been submitted for IRB review and it is still not clear whether the policy applies, the IRB will work with study teams during the pre-review process to identify the need to comply with the policy.
What do I do if the GDS policy applies to my study?
Requests for institutional certification may be submitted at the time of initial review, or, for already approved studies, as a full change of protocol. In either case, the investigator must provide certain information in order for the IRB to evaluate the request, along with a template certification letter. The NIH provides single-site and multicenter template institutional certification forms, which can be found here.
Study teams are responsible for identifying the correct certification template letter and providing the letter with applicable fields filled out to the IRB with the request for certification.
Study teams should also address the following in the specific pages of ARROW:
- Research Design General page, question 1.4: Include a description of any specifically planned or potential genetic testing, including the whether any will result in large-scale genomic data.
- Risk/Benefit Ratio page: Address the risks of the genomic data sharing and how they will be minimized. If there may be particular risks to groups (e.g., stigmatization of an ethnic or cultural group) from genomic data sharing, these should be identified in this section as well.
- Privacy/Confidentiality Continued page: Answer “yes” to 2.2, to indicate genomic data will be released outside the UW.
- Data Release page: Describe:
- any specific federal repository(ies), if known, to which genomic data will be submitted. If the specific repository is not yet known, indicate this.
- the phenotypic data fields that will be submitted to the federal repository, including whether any dates or other potential identifiers will be associated with the data
- types of genotypic data that will be deposited (e.g., whole genome or whole exome vs. particular SNPs)
- whether the genomic data will be created prospectively or are data that already exist (e.g., from prior genetic testing)
- whether the samples from which the genomic data have been or will be derived were collected before January 25, 2015
- whether genomic data will be shared for all study participants or only those that gave consent for such data sharing
- whether the genomic data to be shared with the repository will be for general research use (which could include the study of population origins or ancestry), health/medical/biomedical use (does NOT include the study of population origins or ancestry), disease-specific (e.g., subjects signed consent forms stipulating use of their data or specimens only in particular diseases of research such as hypertension, diabetes, cancer, neurodegenerative diseases), or other research
- whether the genomic data will be limited to use by not-for-profit organizations
- description of the method(s) to be used to code data for transmission to the repository and how the code key(s) will be maintained by the study team (note specifically that the code will never be shared with the repository)
- Consent Overview page: Provide copies of consent documents used to prospectively collect specimens for genetic testing and ensure they include relevant language regarding the genetic testing, potential for deposition of the genomic data in a federal repository, risks of broad sharing of the data as well as any risks to certain individuals/populations, whether the data could be used for commercial purposes, and scope of the potential uses of the genomic data. If study teams will use specimens already collected for the genetic testing, see the Supplemental Information section for handling these situations.
- Supplemental Information section: If relying upon consent forms already signed by subjects and those documents were not approved as part of the current study, the relevant documents should be uploaded on this page and the study team should identify their source (i.e., provide either the UW IRB number under which the documents were approved or, if approved by an external IRB, documentation of other IRB approval of the documents).
Upload the template certification form with applicable fields filled out.
What do I do if I know my study falls under the GDS policy, but do not have all the information I need at this point to request institutional certification from the IRB?
In some cases, a study includes plans to conduct genetic analysis that would fall under the GDS policy at some point in the future, but does not currently have the details necessary to issue a certification, such as the specific repository to which data would be submitted, the details of the phenotypic data that would be included, and/or the details of the data transfer. As institutional certification cannot be granted until these details are available, a full change of protocol should be submitted to request institutional certification at the time data is ready to be submitted to the federal repository.
In cases such as these, where the genomic analysis and/or data sharing is not immediate, the initial application should include as many details as possible about the type of genetic analysis that is planned in the future, as well as the possibility of sharing with a federal repository. Additionally, the consent form should include template language about the sharing; this will facilitate the certification process in the future.
What language should I include in the consent form?
If you know that your study falls under the scope of the GDS policy, whether sharing is expected to occur immediately or at some point in the future, or if you think it is likely that you may obtain NIH funding in the future and/or collaborate with someone who has NIH funding and your study otherwise appears to fall under the scope of the policy, language regarding the sharing should be included in any applicable consent forms. Suggested language is available at Informed Consent: Additional Model Language.
What happens after I submit the request for institutional certification?
The IRB will review the letter, IRB application, protocol, and/or consent document(s) to confirm that the certification requirements have been met. If the IRB confirms that certification requirements have been met and the letter to the repository is complete and accurate, the IRB staff will ask the Principal Investigator (PI) to sign the proposed letter then the IRB will forward the signed document to the University of Wisconsin-Madison Institutional Official for Human Subjects Protection (IO). The IO then countersigns the letter and the finalized letter will be provided to the PI, who can then provide this document to the relevant NIH repository.
If sharing genomic data is required for a study, is it mandatory for all subjects or can subjects opt out of the sharing?
The GDS policy allows for participants to make choices about whether their genomic data can be shared with a federal repository. Thus, research teams should consider how they will address data deposition in their consent documents. A consent form with checkboxes that allow participants to indicate whether or not they are willing to share their data with federal repositories is strongly encouraged.
My study is multi-site and the UW is not the coordinating center. Do I still need to request institutional certification?
If a multi-site study falls under the scope of the GDS policy, study teams should consult with the coordinating center to determine who/what institution is responsible for sharing study-wide genomic data and how they expect institutional certification to be handled. In most cases, if the UW is not responsible for the sharing and transfer of data, the UW would not provide institutional certification. Rather, this would fall to site that is overseeing the sharing for the multi-site study. However, in some cases, the coordinating center may rely on individual sites to provide their own certification. The IRB can work with study teams to answer questions related to certifications for multi-site studies.
Can the UW provide institutional certification for non-UW research teams?
Unless the UW is serving as the Reviewing IRB for other institutions, generally the UW cannot provide GDS certification for other research teams.This is because the UW IRB would need to be able to assess the consent documents and other information related to the GDS in order to support the certification. Having each institution make its own certification is generally a more streamlined process.