Skip to main content
NCBI home page
Neurology logoLink to Neurology
. 2022 Sep 20;99(12):e1314–e1325. doi: 10.1212/WNL.0000000000200886

Survey of Investigators About Sharing Human Research Data in the Neurosciences

Saskia Hendriks 1,✉, Khara M Ramos 1, Christine Grady 1
PMCID: PMC9576293  PMID: 35981894

Abstract

Background and Objectives

In the neurosciences, significant opportunities for sharing individual-level data are underexploited. Commentators suggest various barriers to data sharing, which may need to be addressed. Investigators' perspectives on the main barriers are unclear. Furthermore, bioethicists have raised concerns about the potential misuse of neuroscience data, although discussions are hampered by uncertainty about the potential risks. It is unclear how common sensitive data are obtained and whether investigators judge them as sensitive.

Methods

An online survey was disseminated among 1,190 principal investigators (PIs) of active National Institute of Neurological Disorders and Stroke, National Institute of Mental Health, or NIH Brain Research Through Advancing Innovative Neurotechnologies Initiative grants involving human subject research.

Results

A total of 397 investigators responded to the survey (response rate 33%). Most investigators (84%) support efforts to increase sharing of deidentified individual-level data. However, investigators perceive many barriers to data sharing. The largest barriers were costs and time; limited interpretation of the data without understanding the context of data collection; lack of incentives; limited standardization and norms for data acquisition, formatting, and description; and heterogeneity of data types. Several types of data described as sensitive in the literature are common among neuroscience studies, for example, neural correlates of behavior, emotions, or decision making (71%) and/or predictive data (54%). Although most investigators consider it unlikely or extremely unlikely for their research data to be misused to harm individual research participants (82%), the majority were at least slightly concerned about potential harm to individuals if their research data were misused (65%). Investigators with more easily reidentifiable data, data from vulnerable groups, and neural data were more concerned about the likelihood of misuse and/or magnitude of harm of misuse of their research data.

Discussion

We hope these data help prioritize the development of tools and strategies to overcome the main barriers to data sharing. Furthermore, these data provide input on what may be sensitive data for which additional safeguards should be considered.

Neuroscience is in an exciting period of large societal investment and rapid progress.1,2 However, despite the amount of data collected, most public neuroscience data repositories are minimally populated.3,4 Increased sharing of neuroscience data is widely advocated for.1,2,4-8 Although the benefits of data sharing are yet to be fully quantified, various neuroscience data sharing initiatives or projects have resulted in hundreds of peer-reviewed publications.3,9,10 Data sharing may increase the rate of scientific discovery and improve the quality of the science and resulting clinical care (e.g., by increasing the power of studies).1,3-7,9,11-13 So, what is needed to get more investigators to share their individual-level research data?

Commentators describe many factors that may limit data sharing in neuroscience.1,3-7,9,11,14 However, it is unclear which of these potential barriers have the largest influence on data sharing and should be prioritized to develop solutions to facilitate increased data sharing. As funders and journals are increasingly pushing data sharing,1,4,11,12,15,16 taking stock of neuroscience investigators' views about data sharing and barriers to it may inform the development of tools and support structures to overcome the main barriers to data sharing.

Several commentators have recently raised concerns about potential misuse of neuroscience data, which may harm research participants (e.g., discrimination) or harm the research population at a group level.11,14,17-20 Specific concern is expressed about data with sensitive or risky features. For example, neural data (defined here as direct CNS measurements) are arguably sensitive, given the connection to identity and personhood.17,20-25 Furthermore, the potential to reidentify deidentified neuroscience data (e.g., from deidentified cranial MRI scans26) has raised concerns.19,22,27,28 However, how commonly neuroscience data have sensitive features is unclear. Furthermore, there is uncertainty about the likelihood and magnitude of any potential harms of data sharing. To minimize potential harms from data sharing, safeguards may be appropriate, ranging from improving informed consent19,25,29,30 to limiting what data may be reused for1,2,24,31 or even exempting the most sensitive data from data sharing requirements.2 However, it is difficult to judge when such safeguards are appropriate given uncertainty about the harms and how to weigh them against the benefits of data sharing.19 Engaging stakeholders, including investigators, has been recommended.17 As data sharing requirements are increasing, designing evidence-based policies on appropriate safeguards for sensitive data may contribute to the important balance between protecting participants' interests and the benefits of data sharing.

This study aims to learn about neuroscience investigators' views on data sharing and the main barriers to it, how commonly neuroscience data have features described as sensitive, and whether investigators' views vary based on whether they collect sensitive data. We hope that these data will inform more evidence-based data sharing policies.

Methods

Survey Design

The survey was designed based on a literature review. The survey was revised after a review by an expert panel of 20 National Institute of Neurological Disorders and Stroke (NINDS) and National Institute of Mental Health (NIMH) extramural staff members, including those involved in the Brain Research Through Advancing Innovative Neurotechnologies (BRAIN) Initiative, and a review at an Empirical Research Laboratory meeting of the NIH Department of Bioethics. The survey was pilot tested among a convenience sample of 14 neuroscientists not eligible for the survey.32

The survey contains questions about (1) features of neurotechnologies and collected data that may increase their sensitivity, (2) potential strategies to protect data, (3) potential secondary uses of data, (4) the value of data sharing, (5) barriers to data sharing, (6) trade-offs between privacy and data sharing, and (7) the need for additional guidance (eAppendix 1, links.lww.com/WNL/C238).

Data Collection

Principal investigators (PIs) listed as point of contact on active NIH grants involving human subject research (as defined by the US Code of Federal Regulations 45CFR46) that were funded through the NIH BRAIN Initiative, NINDS, or NIMH were eligible to participate.

Only publicly available information was used to identify eligible participants. First, a full list of active grants funded by the NIH BRAIN Initiative,33 NINDS, and NIMH34 (n = 10,633) was created. To reduce the number of abstracts to review, a shortlist was created of NINDS and NIMH grants that were likely to entail human subject research. This shortlist was created by looking up the PubMed publication history of the PIs of the active grants. If 80% or more of a PIs' publications involved human subjects (PubMed classification), their grant was shortlisted (n = 3,077 PIs). Content analysis of the publicly available abstracts of the shortlisted NINDS and NIMH grants was performed to identify human subject grants. As many concerns about data sharing involve cutting-edge neurotechnology, all eligible BRAIN Initiative grants were invited to participate. A random selection of eligible NINDS and NIMH grants was identified, based on the required sample size of n = 293 respondents (based on a precision of 0.05, sample proportion of 0.30, and estimated population size of n = 3,083) and an anticipated response rate of ∼30%. If multiple human subject grants had the same contact PI, only one of these grants was eligible (selected at random). For multisite grants, only 1 contact PI was eligible (randomly selected). Each participant thus represents a unique grant and a unique PI. In total, 1,190 PIs were invited.

Eligible individuals were invited by email to participate in the online survey in April or May 2021. Participants received up to 2 reminders.

Statistical Analysis

Data were analyzed with the Statistical Package for the Social Sciences 28 (IBM Corp., Armonk, NY), SAS, and statxact version 10. Surveys with items missing were included as part of the survey data. The existence of missing data is indicated in the summary statistics.

Because BRAIN Initiative-funded investigators were oversampled, we first analyzed, using chi-square tests, whether simply reporting on the summary statistics of the entire sample would skew the outcome data (survey section “access to and possible uses of the data” and beyond). Most outcome variables were not different for BRAIN-funded investigators than for other investigators; only 2 were significantly associated (p ≤ 0.001, threshold adjusted for multiple testing). We therefore report on summary statistics for the entire respondent group.

Based on the literature, neural data, increased identifiability, data from vulnerable groups, increased numbers of data types, many hours of neural recordings (>10 hours), and/or neural data recorded outside of the laboratory or clinic each may increase risks of sharing. To test investigators' perceptions of risk, we evaluated associations between having data with these features and the perceived risks of misuse and of reidentifiability, as well as the types of secondary users who may be interested in the data (e.g., legal systems). Furthermore, we examined whether investigators with these data types implement different strategies to reduce data sharing risks. We used χ2 tests (or the Fisher exact test where appropriate) for associations between binary variables, Armitage trend exact tests to look for trends between binary and ordinal variables, and Cochran-Mantel-Haenszel nonzero direct tests for trends between ordinal variables (using Monte Carlo if needed because of the required computational power to run the exact test). To compensate for multiple testing, we report the adjusted false discovery rate p values, with values below 0.05 significant.

Standard Protocol Approvals, Registrations, and Patient Consents

This survey was deemed exempt from institutional review board (IRB) review by the NIH IRB (P204935). The survey invitation noted that no identifying information would be collected and that only Drs. Grady, Hendriks, and Wakim (statistician), who are not involved in funding decisions (and are not extramural NIH staff), would have access to the individual-level survey data and that survey responses would have no bearing on NIH funding. Consent was presumed for participants who filled out the survey.

Data Availability

Individual-level data will not be shared because although no identifying data were collected, it may be possible for individuals working in this field to identify respondents by the characteristics of their grant. To facilitate truthful responses, the informed consent listed the specific persons who would have access to individual-level data.

Results

A total of 397 investigators responded to the survey (response rate 33%). Sixty-five percent of investigators were funded by the NIMH (response rate 36%), 35% by the NINDS (response rate 31%), and 11% by the BRAIN Initiative (response rate 37%). Observational studies were most common (47%), followed by experimental (37%) or mixed (17%) study designs.

The Sensitivity of Collected Data

Features of the Studies and Data

Table 1 presents features about the studies. Almost all investigators collect data from adults who can consent for themselves (89%), a minority from minors (33%), adults who cannot consent for themselves (11%), or other groups who may be particularly vulnerable (e.g., with stigmatized conditions; 19%). Twenty-six percent had studies collecting data on a small (<20) number of participants. After deidentifying their research data (although we use the term “their research data” in this article to refer to individual-level research data collected by investigators, we do not mean to take a stance on the debate around who has ownership over research data), 29% thought it would definitely not be possible for someone to reidentify their participants within the next 10 years (deidentification defined as removing information that can be used to distinguish or trace an individual's identity, either alone or when combined with other information).

Table 1.

Features of the Studies and the Collected Data

graphic file with name WNL-2022-200853t1.jpg

Open in a new tab

The most commonly collected types of data were medical data (73%), survey or interview data (70%), behavioral data (67%), and/or neural data (64%). Almost all investigators collected more than 1 different data type from participants (89%).

fMRI (44%), structural brain imaging (36%), and EEG (22%) were the most commonly used technologies to collect neural data. Features of neural data that are described as sensitive in the literature were common, including neural correlates of behavior, emotions, or decision making (71%); real-time data (59%); and/or predictive data (54%). A minority of investigators recorded neural data in naturalistic settings (16%) and/or collected more than 10 hours of neural recordings per participant (15%). Sensitive features of neurotechnologies were common, including the ability to upload data to the internet or wirelessly transmit data (49%) and/or to record data without active participation of research participants (44%).

Potential Users, Potential Misuse, and Future Use of Individual-Level Data

Almost all (97%) investigators thought that other scientists, and a majority thought clinicians (62%) and industry (55%), would be interested in their research data (Table 2). A minority but over 10% thought that, within the next 10 years, insurers, public health surveillance agencies, other government entities/regulators, educational systems, media, marketing companies, and/or legal systems might be interested in their research data.

Table 2.

Access to and Possible Uses of the Individual-Level Data

graphic file with name WNL-2022-200853t2.jpg

Open in a new tab

Although most investigators consider it unlikely or extremely unlikely for their research data to be misused to cause harm to individual research participants (82%) or groups (67%; Figure 1), the majority were at least slightly concerned about the potential harm that could be done to either individual research participants (65%) or groups (63%) if their research data were misused.

Figure 1. Perceived Likelihood and Degree of Potential Harm From Misuse If the Investigators' Individual-Level Deidentified Data Were Publicly Available.

Figure 1

Open in a new tab

Total number of responses: n = 339–357 per item.

Many (32%–39%) investigators were unsure whether, within the next 10 years, neural data could be reliably used to extract several types of private information. More investigators predicted that neural data could reliably reveal unexpressed emotions (37%) than thought it would not be possible (31%). By contrast, more investigators thought it would not be possible for neural data to reliably show someone is lying (42%), reveal preferences and beliefs (43%), or reveal the content of unexpressed thought (52%) than thought these would be possible (19%, 20%, and 13%, respectively).

Data Sharing Practices for Their NIH Grant

Most investigators thought deidentifying data protects privacy to a great extent (82%), and almost all reported always deidentifying their research data when sharing (94%; Figure 2). Most (over 80%) thought that encrypting data, ensuring proper informed consent, and a data access and governance process would protect privacy at least somewhat. Although 79% of investigators shared data only if participants had properly consented to sharing, roughly half would only share encrypted data or data subject to a data access and governance process. Finally, although recognizing that not sharing their research data would protect participants' privacy, most (69%) still shared their research data.

Figure 2. Protections for Sharing Individual-Level Research Data Outside of the Investigators' Research Team.

Figure 2

Open in a new tab

Total number of responses: n = 312 -327 per item. Other protections listed by investigators were a hospital firewall (n = 1) and not sharing variables with small cell counts (n = 1).

When considering direct requests for individual-level deidentified data by other researchers, almost all check whether the proposed reuse is consistent with the study's consent forms (94%) and/or may cause harms (86%; Figure 3). Most also check the requestor's credentials (81%), the rigor and social value of the proposed reuse (80%), and the requestor's information technology security systems (66%). About 34% of those who will share their research data on archives did not know whether the factors listed above would be assessed before users could access their research data.

Figure 3. Factors That Are Assessed to Determine Whether Individual-Level Deidentified Data Will Be Shared After the Study Has Ended.

Figure 3

Open in a new tab

Total number of responses: n = 311-329 per item. *N = 85 -91 respondents indicated that their data would not be shared on archives; this table represents the proportion of the remaining respondents. Under “other” respondents listed: whether the proposed use overlaps with what their own research team is still planning to analyze themselves (concern about scooping their work; n = 2) and whether they would be offered authorship on any publications resulting from their data (n=1). Others require institutional review board and/or institutional approval (n = 3), a data sharing agreement (n = 5), or consideration of what the authorities consider appropriate factors (n = 2). Finally, some said their willingness to share data depends on the technical feasibility of sharing (e.g., given the size of data files, n = 1), whether they have a clear understanding of how the data may be reused (n = 1), and the requestor's commitment to not share the data further or use it for purposes other than the proposed use (n = 1).

Beyond sharing with other researchers, most investigators indicated that NIH (77%) and committees that review or monitor their study, for example, Data and Safety Monitoring Board or IRB (75%), had access to their deidentified data or would be granted access on request (Table 2).

The Perceived Value of Data Sharing

The majority either strongly support (49%) or somewhat support (35%) efforts in neuroscience and mental health to increase sharing of deidentified individual-level data (Table 3). Fifty-seven percent of investigators sometimes used data collected and shared by others to conduct research, whereas 20% did this often. Nine percent had never shared their own data with other researchers.

Table 3.

Perspectives on the Value of Data Sharing and the Need for Additional Guidance

graphic file with name WNL-2022-200853t3.jpg

Open in a new tab

Barriers to Data Sharing

Almost 80% of investigators rated all the potential barriers listed in the survey as at least a slight barrier to increased data sharing (Figure 4). The barriers rated as large or huge by most respondents were limited standardization and norms for data acquisition, formatting, and description (67%); heterogeneity of data types (66%); the costs and time involved in data sharing (66%); limited interpretation of the data without understanding the context in which it was collected (60%); and lack of incentives for investigators to share data (57%).

Figure 4. Perceived Current Barriers to Increased Sharing of Deidentified Individual-Level Data in Their Field.

Figure 4

Open in a new tab

Total number of responses: n = 309–323 per item. Other barriers to sharing data noted by investigators included unclear guidelines about data sharing (n = 1), the lack of requirements to share data (n = 1), the need for data sharing agreements or material transfer agreements (n = 2), and international restrictions. (n = 1) Furthermore, an inopportune timing of data sharing may be a barrier (n = 2). Finally, investigators raised concerns about unfairness in who contributes and who benefits from data sharing (n = 4) and specifically warned against placing undue burdens on junior scientists, those who are less politically connected, and scientists of color (n = 2). EU = European Union; GDPR = General Data Protection Regulation.

Need for More Guidance

Almost all investigators thought additional guidance or best practices were needed relating to sharing human research data (97%; Table 3). Best practices for informed consent for data sharing (77%) as well as standards and norms for data acquisition, common data elements, data formatting, data annotation, data deidentification, and/or technical specifications of devices (76%) were most called for.

Additional Concern and/or Altered Practices for Data With Sensitive Characteristics

Investigators who considered it more likely that their research data could be reidentified more commonly thought their research data would be of interest to employers (p = 0.0016), insurers (p = 0.0257), marketing companies (p = 0.0016), educational systems (p = 0.0239), law enforcement or intelligence (p = 0.0016), legal systems (p = 0.0016), public health agencies (p = 0.0198), and/or foreign governments (p = 0.0016). Similarly, those who responded that it would be more likely for their research data to be reidentified considered it more likely that their research data would be misused to cause individual (p < 0.0001) and group harms (p < 0.0001). They were also more concerned about the magnitude of harm from misuse for individuals (p < 0.0001) and groups (p < 0.0001).

Investigators who enrolled vulnerable participants more commonly thought that educational systems would be interested in their research data (p = 0.0488) and were more concerned about the magnitude of harm to individuals (p = 0.0488) and groups (p = 0.0072) if their research data were misused. They also considered misuse leading to group harms more likely (p = 0.0198).

Investigators collecting more data types were more likely to think their research data would be of interest to industry (p = 0.0145), employers (p = 0.0072), insurers (p = 0.0341), marketing companies (p = 0.0403), law enforcement or intelligence (p = 0.0239), legal systems (p = 0.0488), educational systems (p = 0.0072), media (p = 0.0072), and/or foreign governments (p = 0.0016).

Fewer investigators working with neural data thought their research data would be of interest to public health agencies (p = 0.0001) or insurers (p = 0.0488) than investigators without neural data. On requests to share their research data, investigators with neural data less commonly assessed the rigor and social value of the proposed reuse (p = 0.0184), the potential harms of the proposed reuse (p = 0.0381), or the requestor's credentials (p = 0.0341). Investigators with neural data considered group harms more likely (p = 0.0488) and were more concerned about the magnitude of such harms (p = 0.0257).

Investigators who recorded neural data outside of the laboratory or clinic more often indicated that they would not share their research data (p = 0.0303).

The other tested variables were not significantly associated.

Discussion

This study reports on neuroscience investigators' views and concerns about data sharing. Key findings from our study are that (1) most investigators support data sharing but experience many barriers to sharing data; (2) investigators thought misuse of research data was unlikely but were at least slightly concerned about potential harms from misuse; (3) data types described as sensitive in the literature were common in these neuroscience studies; and (4) investigators who thought their research data were more easily reidentifiable, or who had neural data or data from vulnerable groups, were more concerned about possible misuse of data. Some of our findings are reassuring in reducing certain concerns about data sharing raised in the literature, whereas other concerns are supported.

Given these investigators' expressed support for data sharing, the limited degree to which data are being shared in neuroscience3,4 seems more likely related to remaining barriers, as below, than to a general unwillingness of investigators. A previous study found support for data sharing among 23 deep brain stimulation researchers,35 as well as among scientists in other fields,36,37 although some commentators describe ambivalence about data sharing among neuroscientists.10,38

Seventy-five percent of investigators indicated that they would share their research data, a surprisingly high percentage given the relatively small number of published datasets.3,4 Previously self-reported data sharing in subgroups of neuroscientists was more mixed.35,39 However, willingness to share data is high in other domains,40-42 which may represent mostly on-request sharing41 and/or may overrepresent actual data sharing practices (e.g., socially desirable responses).43 Requirements to share data44 may also play a role.

Data sharing practices reported in this study are reassuring, in contrast to a previous study that found data sharing practices quite ad hoc.39 Most investigators carefully consider data sharing requests from other investigators and apply strategies to protect shared data. For example, most obtain informed consent for data sharing, which significantly reduces concerns around data sharing.22 Few investigators report sharing their research data with groups besides scientists, NIH, and/or review committees. A previous study raised concern about potential widespread data sharing with device manufacturers,35 but our data show that this is uncommon. Most investigators thought that misuse would be unlikely, even if their research data would be public. Still, both our data and studies in other fields show that investigators are concerned about some risks of sharing data.37,42,45 This may be interpreted as investigators trying to be good stewards of their research data. Furthermore, before sharing their research data, 94% of PIs check whether the proposed reuse is consistent with the study's informed consent, and 94% deidentify the data before sharing, which may inform best practices moving forward.

Investigators are concerned about the magnitude of harm that could be done with their research data, confirming some commentators' concerns.2,18,24,25,31,46 Many data types that commentators suggest as sensitive are common in this cohort (e.g., predictive data,17,24,47 data from vulnerable groups,22 combinations of data types,22,25 and data that may be reidentified4,19,22,27). Features of neurotechnologies that raise misuse risk are also common (e.g., internet connection25,47 and ability to record data without participants' awareness22,48). Like previous commentators, our investigators are more concerned about misuse when they have more easily reidentifiable data,4,19,22,27,28 data from vulnerable groups,22 and neural data.20-25

Anticipated interest in data from nonbiomedical stakeholders was somewhat common, emphasizing commentators' call to consider who should have access to neuroscience research data and for what uses.2,24,31 For example, commentators have raised concerns about marketers, employers, insurers, law enforcement, and legal systems reusing neuroscience data.2,24,48 Similarly, patients in other fields are concerned about their research data being used by third parties that they would not have supported13,49 and want to control use of sensitive data.49

Obtaining reliable inferences from neural data about private information (e.g., preferences)—something bioethicists speculate and worry about21,24—is not considered inconceivable within the foreseeable future by a significant number of investigators. Clarifying what inferences can legitimately be drawn is important for risk assessments, but our data confirm that even scientists have trouble anticipating scientific developments.

Our data confirm many of the barriers to data sharing described by commentators,1-7,9,11,14,28,50 but, to our knowledge, ours is the first study to comprehensively prioritize them. Efforts to address barriers to data sharing have been ongoing for years2,3,17 and resulted in several successes3,17; our data identify the barriers perceived by investigators as most meaningful currently. It is important that some of the largest barriers identified in our study are barriers to using shared data, not to sharing data itself (e.g., limited standardization and norms for data acquisition, formatting, and description). Without addressing these barriers, increasing data sharing may yield limited value. Challenges with data usability have been highlighted previously.3,4,35 The higher-level FAIR (Findable, Accessible, Interoperable, and Reusable) principles may help define some of the needed standards.51 Policymakers may also consider the timing of data sharing, which may matter to investigators.42,43

Finally, these data confirm the need for additional guidance and best practices described in the literature.2-4,17,29,50 For example, although almost 80% of investigators always get informed consent to share the research data before sharing it, over 90% have concerns that this consent inadequately informs participants, which they view as a barrier to data sharing. Establishing best practices for consent relating to data sharing was thus the most frequently asked for additional guidance.

This study contributes to a limited existing literature on neuroscience researchers' views and practices about data sharing. Our study may inform the debate on data sharing in neuroscience that has thus far been hampered by a dearth of empirical data52 and is responsive to a call to engage stakeholders.17 Our response rate was modest, although much lower response rates (<10%) are common in other surveys on data sharing among scientists.36,37,40,41 Nonresponse bias is impossible to exclude because we do not know the characteristics of the nonresponders or the reasons for nonresponse. Similarly, although a number of respondents did not complete the study, the required sample size was achieved for all questions. We hypothesize that reasons for nonresponse could include a lack of time, the survey topic being not applicable, disagreeing with or a lack of interest in the survey topic, and/or conflict with a survey coming from this funding agency (although the individual-level survey data were carefully and explicitly kept confidential from the NIH involved in funding decisions). We surveyed PIs, and although training grants were eligible, respondents were likely relatively senior. Finally, because we used previous publications to identify investigators likely to conduct human subject research, our sample overrepresents investigators with previous human subject research experience.

Along with policies promoting data sharing,4,11,12,15,16,44 addressing barriers to data sharing may increase data sharing. Our study identified barriers to be prioritized. Addressing these barriers will require a joint effort by funders, journals, professional societies, academic institutions, and individual investigators.28,45,50 Assessing the effectiveness of strategies to increase data sharing may be appropriate as well as evaluating the cost-effectiveness of sharing different types of data.

Accelerating scientific progress and maximizing the value of clinical research present strong ethical reasons to share data, yet the potential harms of data sharing should also be carefully considered. Identifying appropriate safeguards, especially for high-risk data, will help to protect research participants, groups, and public trust.4,8,17-19,25,30 Safeguards may range from technical solutions (e.g., to deidentify data) to procedural improvements (e.g., improved informed consent19,25,29,30), limitations on how data may be used,1,2,24,31 or even what data need to be shared. However, such safeguards may impose burdens on researchers and/or limit the scientific benefits of data sharing.30 Safeguards should therefore be proportional to the risk level of shared data (considering the likelihood and the magnitude of potential harms and ideally anticipating the potential for the sensitivity of data to increase as neurotechnologies improve).2,30 Further ethical guidance2,3,17 and policy17,24,25 should be advanced. This guidance would benefit from further research clarifying the likelihood and magnitude of potential harms from data sharing, as well as how to weigh the potential harms and benefits from data sharing.19 Furthermore, studying research participants' perspectives may be helpful.

Acknowledgment

The authors thank the survey participants and the extramural staff members of the NINDS and NIMH who provided input on the survey design. They thank George Santangelo and Hao Yu from the NIH Office of Portfolio Analysis for assistance with creating a shortlist of the PIs eligible for study participation. They are grateful to Sarah Bell and Melinda Pruss for their help with identifying eligible participants and to Talia Bernhard for assistance in sending the surveys. They thank Paul Wakim for providing statistical support and assistance. They thank NIH leadership for their support of ethics.

Glossary

BRAIN

Brain Research Through Advancing Innovative Neurotechnologies

IRB

institutional review board

NINDS

National Institute of Neurological Disorders and Stroke

NIMH

National Institute of Mental Health

PI

principal investigator

Appendix. Authors

Appendix.

Open in a new tab

Study Funding

This research was supported in part by the Intramural Research Program of the NIH. The views expressed are the authors' own and do not represent those of the National Institutes of Health, the Department of Health and Human Services, or the United States government.

Disclosure

The authors report no relevant disclosures. Go to Neurology.org/N for full disclosures.

References

  • 1.Bargmann C, Newsome W, Anderson D, et al. BRAIN 2025: A Scientific Vision: Brain Research through Advancing Innovative Neurotechnologies (BRAIN) Working Group Report to the Advisory Committee to the Director. NIH, 2014. [Google Scholar]
  • 2. NIH. The BRAIN Initiative 2.0: From Cells to Circuits, Toward Cures NIH. NIH, 2019. [Google Scholar]
  • 3.Ferguson AR, Nielson JL, Cragin MH, Bandrowski AE, Martone ME. Big data from small data: data-sharing in the ‘long tail’ of neuroscience. Nat Neurosci. 2014;17(11):1442-1447. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Poline J-B, Breeze JL, Ghosh S, et al. Data sharing in neuroimaging research. Front Neuroinform. 2012;6:9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Ascoli GA. Mobilizing the base of neuroscience data: the case of neuronal morphologies. Nat Rev Neurosci. 2006;7(4):318-324. [DOI] [PubMed] [Google Scholar]
  • 6.Milham MP. Open neuroscience solutions for the connectome-wide association era. Neuron. 2012;73(2):214-218. [DOI] [PubMed] [Google Scholar]
  • 7.Ascoli GA. Sharing neuron data: carrots, sticks, and digital records. PLoS Biol. 2015;13(10):e1002275. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Eickhoff S, Nichols TE, Van Horn JD, Turner JA. Sharing the wealth: neuroimaging data repositories. Neuroimage. 2016;124(pt B):1065-1068. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Milham MP, Craddock RC, Son JJ, et al. Assessment of the impact of shared brain imaging data on the scientific literature. Nat Commun. 2018;9(1):2818. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Ascoli GA, Maraver P, Nanda S, Polavaram S, Armañanzas R. Win–win data sharing in neuroscience. Nat Methods. 2017;14(2):112-116. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Choudhury S, Fishman JR, McGowan ML, Juengst ET. Big data, open science and the brain: lessons learned from genomics. Front Hum Neurosci. 2014;8:239. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. NIH. DRAFT NIH Policy for Data Management and Sharing. NIH, 2019. [Google Scholar]
  • 13.Mello MM, Lieou V, Goodman SN. Clinical trial participants' views of the risks and benefits of data sharing. New Engl J Med. 2018;378(23):2202-2211. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Spires-Jones TL, Poirazi P, Grubb MS. Opening up: open access publishing, data sharing, and how they can influence your neuroscience career. Eur J Neurosci. 2016;43(11):1413-1419. [DOI] [PubMed] [Google Scholar]
  • 15.Rousselet GA, Hazell G, Cooke A, Dalley JW. Promoting and supporting credibility in neuroscience. Brain Neurosci Adv. 2019;3:2398212819844167. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Taichman DB, Backus J, Baethge C, et al. Sharing clinical trial data: a proposal from the International Committee of Medical Journal Editors. Rev Med Chil. 2016;144(1):11-13. [DOI] [PubMed] [Google Scholar]
  • 17.Eke DO, Bernard A, Bjaalie JG, et al. International data governance for neuroscience. Neuron. 2022;110(4):600-612. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Fothergill BT, Knight W, Stahl BC, Ulnicane I. Responsible data governance of neuroscience big data. Front Neuroinform. 2019;13:28. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Global Neuroethics Summit Delegates, Jeong SJ, Fukushi T, Fukushi T, et al. Neuroethics questions to guide ethical research in the international brain initiatives. Neuron. 2018;100(1):19-36. [DOI] [PubMed] [Google Scholar]
  • 20.Illes J, Racine E. Imaging or imagining? A neuroethics challenge informed by genetics. Am J Bioeth AJOB. 2005;5(2):5-18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Fuchs T. Ethical issues in neuroscience. Curr Opin Psychiatry. 2006;19(6):600-607. [DOI] [PubMed] [Google Scholar]
  • 22.Ienca M, Haselager P, Emanuel EJ. Brain leaks and consumer neurotechnology. Nat Biotechnol. 2018;36(9):805-810. [DOI] [PubMed] [Google Scholar]
  • 23.Finn RL, Wright D, Friedewald M. Seven types of privacy. In: Gutwirth S, Leenes R, de Hert P, Poullet Y, editors. European Data Protection: Coming of Age. Springer Netherlands, 2013:3-32. [Google Scholar]
  • 24.The Committee on Science and Law. Are Your Thoughts Your Own?: “Neuroprivacy” and the Legal Implications of Brain Imaging. The Committee on Science and Law, 2005. [Google Scholar]
  • 25.Yuste R, Goering S, Arcas BAY, et al. Four ethical priorities for neurotechnologies and AI. Nature. 2017;551(7679):159-163. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Schwarz CG, Kremers WK, Therneau TM, et al. Identification of anonymous MRI research participants with face-recognition software. N Engl J Med. 2019;381(17):1684-1686. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Martin P, Ashcroft R. Neuroscience, ethics and society: a review of the field. Background paper prepared for the 2005 Wellcome Trust Summer School on ‘neuroethics’ 2005.
  • 28.Poldrack RA, Gorgolewski KJ. Making big data open: data sharing in neuroimaging. Nat Neurosci. 2014;17(11):1510-1517. [DOI] [PubMed] [Google Scholar]
  • 29.Fukushi T, Sakura O, Koizumi H. Ethical considerations of neuroscience research: the perspectives on neuroethics in Japan. Neurosci Res. 2007;57(1):10-16. [DOI] [PubMed] [Google Scholar]
  • 30.Alter G, Gonzalez R. Responsible practices for data sharing. Am Psychol. 2018;73(2):146-156. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Bonaci T, Calo R, Chizeck HJ. App stores for the brain: privacy & security in brain-computer interfaces. 2014 IEEE International Symposium on Ethics in Science, Technology and Engineering; May 23-24, 2014: 1-7.
  • 32.Willis GB. Cognitive Interviewing: A Tool for Improving Questionnaire Design. Research Triangle Institute, 2004. [Google Scholar]
  • 33.NIH. The Brain Initiative. Accessed December 2020. braininitiative.nih.gov/funding/funded-awards.
  • 34.NIH. RePORTER. Accessed December 2020. reporter.nih.gov/advanced-search.
  • 35.Zuk P, Sanchez CE, Kostick K, et al. Researcher perspectives on data sharing in deep brain stimulation. Front Hum Neurosci. 2020;14:578687-578687. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Zhu Q, Zhu YY, Wang WN. TRUSS inhibition protects against high fat diet (HFD)-stimulated brain injury by alleviation of inflammatory response. Biochem Biophys Res Commun. 2019;511(1):41-48. [DOI] [PubMed] [Google Scholar]
  • 37.Houtkoop BL, Chambers C, Macleod M, Bishop DVM, Nichols TE, Wagenmakers E-J. Data sharing in psychology: a survey on barriers and preconditions. Adv Methods Pract Psychol Sci. 2018;1:70-85. [Google Scholar]
  • 38.Van Horn JD, Ball CA. Domain-specific data sharing in neuroscience: what do we have to learn from each other? Neuroinformatics. 2008;6(2):117-121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Borghi JA, Van Gulick AE. Data management and sharing in neuroimaging: practices and perceptions of MRI researchers. PLoS One. 2018;13(7):e0200562. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Tenopir C, Christian L, Allard S, Borycz J. Research data sharing: practices and attitudes of geophysicists. Earth Space Sci. 2018;5:891-902. [Google Scholar]
  • 41.Tenopir C, Allard S, Douglass K, et al. Data sharing by scientists: practices and perceptions. PLoS One. 2011;6(6):e21101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Tenopir C, Dalton ED, Allard S, et al. Changes in data sharing and data reuse practices and perceptions among scientists worldwide. PLoS One. 2015;10(8):e0134826. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Wallis JC, Rolando E, Borgman CL. If we share data, will anyone use them? Data sharing and reuse in the long tail of science and technology. PLoS One. 2013;8(7):e67332. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Initiative NB. Notice of Data Sharing Policy for the BRAIN Initiative. NOT-MH-19-0102019.
  • 45.Sayogo DS, Pardo TA. Exploring the determinants of scientific data sharing: understanding the motivation to publish research data. Govern Inform Q. 2013;30:S19–S31. [Google Scholar]
  • 46.Nuffield Council. Novel Neurotechnologies: Intervening in the Brain. Nuffield Council, 2013. [Google Scholar]
  • 47.Glannon W. Ethical issues in neuroprosthetics. J Neural Eng. 2016;13(2):021002. [DOI] [PubMed] [Google Scholar]
  • 48.Farah MJ, Wolpe PR. Monitoring and manipulating brain function: new neuroscience technologies and their ethical implications. Hastings Cent Rep. 2004;34(3):35-45. [PubMed] [Google Scholar]
  • 49.Courbier S, Dimond R, Bros-Facer V. Share and protect our health data: an evidence based approach to rare disease patients' perspectives on data sharing and data protection: quantitative survey and recommendations. Orphanet J Rare Dis. 2019;14(1):175. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Nichols TE, Das S, Eickhoff SB, et al. Best practices in data analysis and sharing in neuroimaging using MRI. Nat Neurosci. 2017;20(3):299-303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Wilkinson MD, Dumontier M, Aalbersberg IJJ, et al. The FAIR guiding principles for scientific data management and stewardship. Sci Data. 2016;3:160018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Baker DA, Schweitzer NJ, Risko EF. Perceived access to self-relevant information mediates judgments of privacy violations in neuromonitoring and other monitoring technologies. Neuroethics. 2014;7:43-50. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Individual-level data will not be shared because although no identifying data were collected, it may be possible for individuals working in this field to identify respondents by the characteristics of their grant. To facilitate truthful responses, the informed consent listed the specific persons who would have access to individual-level data.

Articles from Neurology are provided here courtesy of American Academy of Neurology

RESOURCES