The Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) has the contractual authority and responsibility to establish policies of National Children’s Study data access. To achieve Study objectives, these policies must maximize availability of data to the research community for the intended scientific purposes. At the same time, the Study must responsibly carry out its mandate by protecting participants and their communities against risk of intentional or inadvertent disclosures, whether directly or through inference, of private, confidential, or sensitive personal information. The Study must meet both objectives while also protecting the reliability, integrity, and validity of the data. To these ends, DACC established the following Data Access Principles. These Principles will are further contextualized in the DACC Policy Manual.
National Children’s Study Guiding Principles of Data Access and Confidentiality
1. Maximize Data Access to the Research Community
The NICHD has the contractual authority and responsibility to establish policies of National Children’s Study data access. Study analytic data and supporting documentation will be made available to the research community for intended scientific purposes in a timely manner regardless of affiliation. (See the 2003 NIH Data-Sharing Policy.) NIH and NICHD data policies are expected to continue to evolve over time, and Study data-sharing policies will be updated accordingly.
This Principle pertains to all forms of National Children’s Study analytic data including, but not limited to, interview data, images, biospecimens, and environmental samples for both the main Study and any adjunct studies. It does not pertain to Vanguard Study data because the purpose of these data is to inform Study methodology and operations. Exceptions to this Principle will be considered on an ad hoc basis, with justification on the basis of established Study data principles.
2. Relate Access to Task
Access to preliminary National Children’s Study data will be guided by project needs and staff roles in the data collection and the cleaning and packaging processes.
Study-affiliated scientists are those who belong to a National Children’s Study entity involved in protocol development, design, and/or collecting data for the Study. Such entities include Coordinating Center, Interagency Coordinating Committee, Program Office and Project Officer, and Study Center teams. Scientists and staff affiliated with the Study will have access to early versions and highly disclosive data files only as required for their responsibility to the development of the Study and review or evaluation of raw and preliminary datasets. Unaffiliated scientists and their staff, who do not share this responsibility, will not have such access. All researchers party to a Study data-use agreement for nonpublic-use data agree not to attempt identification of subjects unless specifically authorized by the National Children’s Study to do so by virtue of their role in the Study.
Some scientists and their staff, such as approved adjunct study Principal Investigators, will be considered partial affiliates. For example, to conduct a particular adjunct study, it may be necessary to access highly disclosive data. In the partial affiliate relationship, the approved investigator accesses highly disclosive data in partnership with a Study-affiliated co-investigator. Consistent with the overall National Children’s Study access policy, access for partial affiliates will be guided by the requirements of the investigator to complete his or her task.
3. Support Equal Access within Project Roles and Levels of Responsibility
Within each role, researchers will have equal access to data regardless of scientific institution or affiliation. This approach maximizes data access for intended scientific purposes and minimizes unmitigated disclosure risk.
National Children’s Study disclosure limitation policies mandate a shared responsibility to limit disclosure risk, whether inadvertent or intentional, direct or indirect, of participant identities and their communities in a manner consistent with federal regulations.
4. Adhere to Equal Standards
All people in contact with personally or socially identifiable data used by or collected by the National Children’s Study are required to adhere to Study data-security guidelines. This group includes, but is not limited to, Study Centers, Program Office, Interagency Coordinating Committee, and Coordinating Center staff and their subcontractors; adjunct study investigators; and members of the general research community.
5. Prepare Standard Datasets on an Ongoing Schedule
The National Children’s Study will make these standard datasets available to the research community on an ongoing, scheduled basis throughout the course of data collection. The schedule of data release will reflect multiple factors including weighting, operational, and user needs. Subsequent Study data releases will be cumulative, featuring additional data and errata, as available. All Study datasets will carry a version number to aid in data and publication management.
Some datasets will require a greater level of detail in terms of personally or socially identifiable data for scientific analysis, for example, geographic information. Some data, by nature, are particularly sensitive, for example, bio-specimens, environmental samples, or tribal affiliations; or are disclosive, for example, facial images. Other datasets will not contain this level of detail in personally or socially identifiable data.
The National Children’s Study will provide standard datasets for different scientific purposes. The purpose and nature of the dataset determines the degree of and combination in which data suppression and coarsening and monitoring should be employed as disclosure-limitation methods. Only data determined to meet the analytic needs of the user will be made available in a given dataset.
In brief, public-use data files will be created to support data use unmonitored by the Study. These data will be less detailed as a result of more comprehensive disclosure limitations applied to the file itself. Nonpublic-use data files will be more detailed, but access will require Study monitoring for appropriate use.
6. Preserve Data Quality
Disclosure risk-limitation techniques will be carefully selected to preserve the quality of data to the extent possible for the purposes of a given dataset. The extent to which staff access and use a given dataset will also be monitored for disclosure. Specific strategies will be proposed once data are collected, cleaned, and reviewed so as to make the most informed decisions. Choosing specific disclosure-limitation strategies once the source and extent of risk is well evaluated, such as after data cleaning, variable construction, and statistical analysis, will maximize the analytic utility of data to be made available after disclosure review by applying coarsening and suppression only when needed for the purpose of the particular dataset.
The methods of data alteration for disclosure-limitation purposes will rely on suppression and/or coarsening of data and will not include data perturbation techniques. If other methods of alteration are recommended to meet disclosure-limitation requirements, DACC will submit the recommendation for approval by the National Children’s Study Director, on the basis of particular datasets.
Study data documentation will inform users of the methods applied to limit disclosure risk on a particular standard data file, as appropriate, and will advise about how these methods may affect particular analyses, as known.
7. Document Transparency of Data-Access Strategy
Disclosure-limitation decisions will be mindful of the integrity of the National Children’s Study and the quality of its data files. DACC will ensure that Study user manuals and other technical guides describe to readers the particular disclosure-limitation methods applied to particular Study standard datasets.
When appropriate, DACC will consult with both respondents and data users to establish an improved understanding of both respondents’ privacy concerns and investigators’ scientific needs. This consultation will inform the balance which will be struck between the risk of disclosure against loss in data utility and access. Further, disclosure-limitation strategies, both relating to microdata and tabular data, should be a documented and verifiable process, as appropriate.
8. Comply with Federal Regulations
National Children’s Study data collection, processing, and access policies will conform to federal regulations, be consistent with federal guidelines regarding data confidentiality and disclosure limitation, and adhere to the commitment to confidentiality espoused in Study informed consent. (See the Privacy Act of 1974 revised 2004, Computer Security Act of 1987, Federal Information Security Management Act (FISMA) of 2002, NIH Data-Sharing Policy revised 2003, Office of Management and Budget (OMB) Guidance on Privacy, and OMB Standards and Guidelines for Statistical Surveys. Note: the Federal Committee on Statistical Methodology (FCSM) Report on Statistical Disclosure Limitation Methodology (Working Paper 22) revised 2005 will be used as a guide for the disclosure review process.)
For inquiries, please send an e-mail to NCSDataAccess@nih.gov.
Report of the Expert Panel on Public Use Data Access and Disclosure Control (April 2008) (PDF 89 KB)
Federal Committee on Statistical Methodology Working Paper 22 (2005)
NIH Data Sharing Policy (February 2003)