AAMC Chief of Scientific Officer Ross McKinney, Jr., MD, Jan. 19 submitted responses to the National Institutes of Health (NIH) Nov. 14, 2016, request for information (RFI) that would address questions regarding Strategies for NIH Data Management, Sharing, and Citation.
The impetus for this request was the White House Office of Science and Technology Policy (OSTP)’s February 2013, memorandum that prompted the NIH to develop the National Institutes of Health Plan for Increasing Access to Scientific Publications and Digital Scientific Data. The implementation of this plan requires NIH to request information on two broad topics: “Data Sharing Strategy Development” and “Inclusion of Data and Software Citation in NIH Research Performance Progress Reports (RPPR) and Grant Applications.”
Echoing AAMC comments provided in May 2013, Dr. McKinney provided extensive comments, including the following:
- Investigators and research leaders commonly and emphatically emphasize the need, regardless of the types of data, “to capture the totality of information required to make data useful, including documentation of context, limitations, and other metadata.”
- “Negative data are especially valuable to post in repositories… (T)hey become particularly valuable in meta-analysis.” Negative data represent in part the results of experiments that do not succeed, like a candidate drug that did not demonstrate the intended effect. While such negative findings are rarely published, they are nevertheless instructive to scientists.
- “The length of time for making data available for secondary purposes would be indefinite. It will often exceed the length of time for research projects or grants themselves, and may well exceed time that key personnel remain at an institution.”
- “It is not clear that the public or political leaders, who increasingly support or call for data sharing (and other ‘transparency’) appreciate the additional burden and cost of creating usable shared data resources” including both “generalist data repositories, for a wide variety of data, and specific repositories... Research sponsors and institutions must commit resources.”
- Cybersecurity is a principal concern for data repositories– "not only for clinical data where privacy of human research participants and patients is an issue but also for non-human and other types of data," which may be subject to theft, sabotage, or altercation.
The NIH RFI states specifically that the information provided would be used only for planning purposes and should not be construed as policy, solicitation of applications, or a government obligation of any kind.