This chapter describes the roles and responsibilities of the key stakeholders involved in the sharing of clinical trial data: (1) participants in clinical trials, (2) funders and sponsors of trials, (3) regulatory agencies, (4) investigators, (5) research institutions and universities, (6) journals, and (7) professional societies (see Box 3-1). These parties have differing perspectives on the benefits, risks, and challenges associated with sharing clinical trial data. Further, all stakeholders have a role and responsibility in helping to maximize the benefits and minimize the risks of data sharing for others, as well as themselves.
As outlined in Box 3-1, participants in clinical trials include individual patients and healthy volunteers. Research Ethics Committees (called Institutional Review Boards [IRBs] in the United States), Data Monitoring Committees (DMCs)/Data and Safety Monitoring Boards, and disease advocacy organizations—which oversee the informed consent process, help recruit participants for trials, and monitor the quality and safety of trials during the course of participant recruitment and follow-up—are integral to the participant experience and hence are discussed in this section as well.
Key Stakeholders Involved in Sharing Clinical Trial Data
Participants in clinical trials
- – Individual patients and healthy volunteers*
- – Research Ethics Committees (termed Institutional Review Boards [IRBs] in the United States)
- – Data Monitoring Committees (DMCs), also called Data and Safety Monitoring Boards (DSMBs)
- – Disease advocacy organizations
Funders and sponsors of trials
- – Public and nonprofit funders/sponsors (including disease advocacy organizations in this role)
- – Industry sponsors (including large and small private sponsors of pharmaceutical, device, and biologic clinical trials)
- – European Medicines Agency (EMA)
- – U.S. Food and Drug Administration (FDA)
- – Clinical trialists
- – Secondary users (e.g., reanalysts, meta-analysts)
- Research institutions and universities
- Professional societies
* Individual participants in a clinical trial (who are the initial “providers” of data to researchers) may hold data to the extent that they self-generate the data and transmit them (from self-quantifying devices), retain copies of their data, or receive information from investigators. Participants may, in turn, share their data with organizations that aggregate data from many participants (e.g., disease advocacy groups, research platforms such as PatientsLikeMe, Reg4ALL, or Sage Bionetwork’s Bridge).
Individual Patients and Healthy Volunteers
Clinical trial participants are the initial “providers” of data to investigators in clinical trials; they may be either patients or healthy volunteers, depending on the condition being studied. Without willing participants, sponsors and investigators would be unable to carry out clinical trials to advance science and improve clinical care. Thus, it is vital to the clinical trials enterprise that participants be respected, that trust be maintained, and that data sharing not become a barrier (and ideally that it become an incentive) to broad participation in clinical trials (Terry and Terry, 2011).
Participants’ attitudes toward data sharing are mixed for a variety of reasons. First, participants are not monolithic, and certain individuals and
groups are more or less supportive of data sharing than others. Second, individual participants may have positive attitudes toward some types of data sharing and negative attitudes toward others. Third, individuals’ attitudes may change over time with changes in real or perceived benefits and risks.
In its October 2013 public workshop, the committee heard testimony from Sharon Hesterlee, Vice President of Research for Parent Project Muscular Dystrophy, which focuses on this rare fatal disease that affects boys. Ms. Hesterlee stated that this group of participants and their families have expressed the belief that it would be “morally repugnant to not share that data given the burden of participating in trials” and the urgency to find treatments (Hesterlee, 2013). On the other hand, healthy participants may not feel the same urgency and may therefore place greater weight on the protection of their data (Hesterlee, 2013).
According to Deborah Collyar, founder of the Patients Advocates in Research (PAIR) International Communication Network (primarily a cancer patient advocate network), “it is very clear that people volunteer their time, their effort and the[ir] bodies into clinical trials so that we can get better results. They are hoping better results for themselves, but if not, certainly for other people” (Collyar, 2013). However, many participants and their advocates also have concerns about increased sharing of individual participant data in particular. They are concerned that data sharing will lead to privacy breaches or that their data will be used for a purpose (or by individuals or organizations) that they do not sanction (IOM, 2009). If concerns about data sharing make participants less willing to enroll in clinical trials, the benefits gained by the public from clinical trials will be reduced.
Informed consent is required when individuals volunteer to participate in a clinical trial. It often is the primary vehicle for both informing potential subjects about the risks and benefits of a trial and documenting their agreement to participate (see Box 3-2). The informed consent process entails having research participants sign a document that describes the study, including the potential risks and benefits, and the participants’ rights and responsibilities. It usually involves a conversation with an investigator about the study as well (CIOMS, 2002).
Sharing clinical trial data was not envisioned for many of the trials for which the data have already been collected. Consent forms for these legacy trials may be silent with respect to data sharing, expressly disallow any sharing, limit sharing to specific entities or uses, allow open sharing, or be uninterpretable or internally inconsistent (O’Rourke and Forster,
Overview of Informed Consent Laws
Informed consent is fundamental to the ethical conduct of clinical trials, and regulations governing human clinical trials both in the United States and internationally typically require the informed consent of research participants. This is particularly the case when the trial is testing an intervention that may place the subjects at some risk (EMA, 2014a; FDA, 2014; HHS, 2014b).
Concern has been raised that informed consent forms for research often are filled with legal or technical language and are difficult for subjects to understand (Dawson and Kass, 2005; O’Rourke and Forster, 2014). Informed consent processes also focus on the individual and frequently fail to take community norms into account; in some developing countries, for example, individuals defer to or rely on the views of family members or community leaders (Benatar, 2002; Dawson and Kass, 2005; Molyneux et al., 2005).
When research involves identifiable data on individuals, regulations allow a waiver of informed consent processes if the research poses no more than minimal risk to the data subjects and could not practicably be conducted without the waiver* and, when appropriate, if individuals are provided with additional information after participation (HHS, 2014a). Some cluster randomized trials and some large simple trials may qualify for a waiver of informed consent.
* 45 CFR § 164.512(i)(2)(ii) (2006).
2014). Problems associated with consent for data sharing for legacy trials may be compounded because different sites in a multisite clinical trial may have altered the consent forms in accordance with local values and Research Ethics Committee requirements.
For most prospective trials, however, the informed consent process provides an opportunity to obtain participants’ approval for planned data sharing and to be transparent about potential future data sharing. Although the initial consent process is unlikely to provide full details of future data sharing, investigators and sponsors can explain what data will and will not be shared with the individual participant during and after the trial, as well as under what conditions data might be shared beyond the investigators’ organization or research institution. The consent process also provides a good opportunity to educate participants on the benefits and risks of data sharing so they can factor these into their decision to participate.
Debate exists among researchers and participant representatives about whether consent for data sharing should be a condition of trial participation, or whether consent forms should include a separate provi-
sion allowing participants to opt out of subsequent data sharing while still being able to participate in a trial (called “compound consent”) (Bierer, 2014). On the one hand, the argument for including data sharing as a condition of trial participation (i.e., not allowing choice) is that if data from some participants are not shared, differences between the original and shared data sets will lead to discrepancies between analyses carried out by the primary team and by secondary users and inhibit the ability to reproduce analyses and conduct meta-analyses. This is a concern not only for investigators wishing to perform reanalyses or meta-analyses but also for participants who want their shared data to be of the greatest value. On the other hand, there is an argument to be made for allowing people to participate in a clinical trial without sharing their data. Under compound consent, people need not choose between sharing data they are uncomfortable with releasing and not enrolling. If large numbers of people from a specific demographic, cultural, or ethnic/racial group do not participate in clinical trials, the results will not apply to that group, weakening the evidence base for clinical decisions for the group. Proponents of compound consent argue that at least initially data sharing needs to be approached cautiously (particularly with sensitive conditions and populations that are vulnerable with respect to a particular clinical trial) so as to strengthen trust among participants. This issue requires further consideration in the context of specific clinical trials.
Concerns About Privacy
Clinical trial participants have concerns about their privacy being breached during data sharing (i.e., information about them being made public or released to individuals or organizations that could cause them harm). Participants can be vulnerable to privacy breaches in many ways. Some breaches may cause tangible harms, such as stigma or discrimination directed at persons identified as having sensitive conditions (e.g., mental illness, HIV infection, and other sexually transmitted infections), being at risk for such conditions, or engaging in illegal or stigmatized activities (e.g., use of alcohol or injection drugs, commercial sex work, or certain reproductive practices). Often such vulnerable persons have suffered discrimination in the past (Corbie-Smith et al., 1999). Those who are vulnerable because of these medical conditions often are also vulnerable because they are poor, poorly educated, and politically powerless (Benatar, 2002). The level of stigma and discrimination varies by culture and country, a fact that needs to be kept in mind because clinical trials are increasingly conducted in nations and communities around the world.
The committee examined the current landscape of international privacy protection laws to see how they provide protection against privacy
breaches and whether they would offer sufficient protection in an environment of increased transparency of clinical trial data.
International laws protecting personal data—commonly referred to as data protection laws—regulate the collection and use of personal data and are commonly based on fair information practices (Privacy International, 2014). In general, fair information practices help ensure that data are collected only for specified and legitimate purposes, that individuals are informed about data collection, that data are kept secure and accurate, and that appropriate remedies exist should data be breached (Privacy International, 2014). In the case of personal health data, a variety of approaches to protection are taken across jurisdictions (see Box 3-3).
Data protection laws often apply only to identifiable data. For example, European data protection laws cover only information that relates directly or indirectly to an identified or identifiable individual (Retzer
International Protections for Health Data
In Europe, protections for sensitive data—typically defined as including data on health or medical conditions—commonly require the consent of the subject prior to data access, use, or disclosure; however, there are public policy exceptions (Retzer et al., 2011). When data are used for scientific purposes, for example, most European countries exempt those activities from some or all of the data protection obligations.
Countries often impose various additional requirements on the use of health data. Italy and Austria have security measures specific to the health sector: Italy requires researchers to follow the data protection authority’s guidelines, whereas Austria requires encoding of health data. Belgium requires data anonymization and approval from the data protection authority. Ireland, Poland, Slovakia, and the United Kingdom mandate that use of the data not adversely affect individuals. Germany permits secondary use of data for research purposes “if the scientific interest significantly outweighs the individual’s interests.” Finland and the Netherlands impose no additional safeguards for such secondary use (Retzer et al., 2011).
Failure to comply with these data protection laws can jeopardize a clinical research project. Although not all laws are actively enforced, violations can lead to administrative and criminal penalties, and in some cases even to imprisonment (Retzer et al., 2011).
Although data protection laws are proliferating globally, not all countries have enacted them. In the United States, for example, laws govern research uses of health data in some contexts but not universally (IOM, 2009). Many countries in Asia have considered data protection laws but have failed to implement them or, where they have been implemented, regulators lack adequate powers to enforce them (Privacy International, 2014).
et al., 2011). In the United States, HIPAA and the Common Rule apply only to identifiable data. As discussed in Chapter 5, however, there are no uniform international standards for determining when data have been sufficiently anonymized or de-identified to be exempt from regulatory requirements. Similarly, pseudonymized data—data for which personal identifiers have been replaced with a pseudonym or code—are subject to less onerous restrictions in some European countries (and under HIPAA in the United States), but other countries apply the full range of data protection requirements even to such data (Retzer et al., 2011).
As big data analytics becomes more widespread, even data that have been de-identified may lead to violations of privacy. “Big data” is characterized by the quantity and scope of data that can be analyzed, as well as the large scale of the analysis. Combining rich data from various sources into a data set increases the likelihood of being able to re-identify individuals in the data set or determine whether they belong to a subgroup with certain characteristics (Barocas and Nissenbaum, 2014).
Concerns About Unsanctioned Uses of Data
For participants from vulnerable populations who historically were victims of unethical research, the possibility that if data sharing becomes obligatory their data will be used for purposes or by individuals or organizations they do not sanction (IOM, 2009) may be a particular concern. The history of human subjects research is sullied by several scandals in which vulnerable participants were enrolled in trials without their knowledge or ethically valid consent. In the United States, these scandals included the Tuskegee Syphilis Study, drug clinical trials carried out on prisoners, and research on institutionalized persons with mental disabilities or psychiatric illnesses. Many participants in these unethical trials were poor, poorly educated, or members of racial and cultural groups that suffer discrimination.
In the African American community, the Tuskegee study and other examples of unethical research have led to ongoing mistrust in research generally. Partly because of such mistrust, fewer African Americans than Caucasians enroll in clinical research. In turn, this low participation rate results in a weaker evidence base to guide the clinical care of this population, potentially contributing to health disparities in the United States (Corbie-Smith et al., 1999).
Internationally, violations of informed consent in clinical research involving vulnerable populations have created mistrust toward research sponsored or directed by entities from wealthy developed countries. Recent examples include unethical research on sexually transmitted infec-
tions in Guatemala and alleged violations of consent in HIV prevention trials of pre-exposure prophylaxis with tenofovir (Singh and Mills, 2005).
Although this topic has not received extensive empirical study, this evidence of mistrust suggests that certain vulnerable populations—those that have been the victims of unethical research, are disadvantaged socially or economically, or have been discriminated against—might also mistrust sharing of clinical trial data. This possibility might be addressed, strengthening engagement in clinical trials, if representatives of such vulnerable populations were included in the design and implementation of trials and if effective ways of helping participants understand the benefits of data sharing and protecting the subjects of trials were developed.
Research Ethics Committees
Research Ethics Committees are tasked with reviewing, revising, and approving clinical investigations involving humans, with the goal of protecting research participants and ensuring they are treated ethically as a result of their participation. The committees pay close attention to the informed consent process for a particular study (EUREC, 2014; WHO, 2009).
As noted, in the United States, Research Ethics Committees are called IRBs.1 Testimony provided during the course of this study revealed that currently IRBs generally do not approve informed consent documents that require sharing of individual participant data, citing their charge to protect participants and minimize the harms of research participation (Bierer, 2014). Many countries follow the International Conference on Harmonization’s (ICH’s) Guideline for Good Clinical Practice (GCP) (ICH E6), which is similar to the U.S. Food and Drug Administration’s (FDA’s) regulations on IRBs and consent (O’Rourke and Forster, 2014).
FDA regulation 21 CFR Section 50.25(c), which requires disclosure to prospective trial participants that trial results will be posted on ClinicalTrials.gov, includes mandated consent language specifying that
1 A uniform U.S. federal policy for the protection of human subjects (i.e., the Common Rule) applies to proposed research involving human subjects funded by federal departments and agencies that have adopted the Common Rule. Under this framework, research institutions formally file a commitment (i.e., Federalwide Assurance, or FWA) with the federal government to protect participants in federally funded research at their institution by adhering to the Common Rule requirements. IRBs are responsible for implementing these protections at the institutional level and reviewing each proposed research study protocol and the associated informed consent document to ensure compliance with federal rules, institutional policies, and community expectations. IRB members are not associated with the proposed research they review. One academic institution may have multiple IRBs organized by areas of expertise (e.g., biomedical and nonbiomedical research).
only summary results (i.e., no individual participant data) will be posted publicly on that site. This regulatory provision also requires that informed consent forms include statements about the “extent, if any, to which confidentiality of records identifying the individual will be maintained.” This language might be interpreted in different ways. One interpretation is that data sharing, even of individual participant data, is not prohibited. On the other hand, some might suggest that because ClinicalTrials.gov is the only data sharing platform mentioned in the regulation and only summary data are to be shared there, sharing of individual participant data in any form is precluded or at least discouraged.
Research Ethics Committees play an important role in protecting research participants and giving due consideration to the interests and values of communities. For example, there is always some level of risk that individual participant data, even if de-identified, could be used to re-identify a research participant, particularly if other auxiliary information were linked with the clinical trial data set (Dwork, 2014). In addition, using auxiliary information, it may be possible to infer or learn information about individuals in a research data set—for example, whether they have a sensitive condition such as alcoholism or mental illness—even without specifically re-identifying them (Dwork, 2014). Chapter 5 examines these privacy challenges in greater detail, as well as appropriate protections and controls that reduce these risks.
Research Ethics Committees can establish policies that allow and promote responsible sharing of individual participant data in the clinical trials they review. To this end, they can ensure that investigators discuss with research participants during the informed consent process both the prospective benefits and the risks, including privacy risks, of sharing clinical trial data. When reviewing trials, Research Ethics Committees also can ensure that the risks of data sharing are minimized and that they are acceptable in light of the anticipated benefits.
Data Monitoring Committees
DMCs/DSMBs have broad responsibilities for monitoring data quality and assessing the risk/benefit ratio during the course of participant recruitment and follow-up (Ellenberg et al., 2002). Initially, the DMC reviews the trial protocol to gain familiarity with the details but generally not to approve it. The DMC also must review the draft DMC Charter carefully in order to execute it as requested by the sponsor and/or the investigators, and make suggested modifications as necessary before final approval. During the conduct of the trial, the DMC typically reviews a detailed report on interim data by intervention arm, usually unblinded to intervention assignment. This review encompasses recruitment progress,
baseline data describing the participants’ characteristics, comparison of baseline data, concomitant medications or interventions, adverse event data, serious adverse events, primary outcome data, secondary outcome data, and a small list of prespecified subgroups. In addition, DMCs typically request additional analyses motivated by trends in interim data. Interim DMC reports are strictly confidential until the trial is completed. Throughout the conduct of the trail, DMCs are on the alert for signals regarding inadequacies in data integrity and data quality and may ask for specific follow-up reports to clarify or rectify these inadequacies. DMCs usually complete their work with the last participant’s last visit and may be asked by the sponsor and investigators to share their view or interpretation of the results when the data files are absolutely complete and locked down. However, their role is not to review or approve papers presenting final results, although their input is often solicited.
DMCs indirectly facilitate data sharing because they commonly ask the trial’s data management and biostatistics teams to modify their presentation of data in interim reports so as to make the data clearer and more comprehensible. The DMCs’ directions likely lead to improvements in data organization and presentation that are helpful not only to the clinical trial team but also to other investigators who later analyze shared data sets.
Although DMCs are likely to be strongly in favor of data sharing, they are not in a position to enforce that practice any more than they can enforce registration of trials on ClinicalTrials.gov and completion of the uploading of summary results. To give DMCs any responsibility for enforcing data sharing would substantially increase their responsibilities and require them to remain operational for some time beyond completion of a trial.
Disease Advocacy Organizations
Disease advocacy organizations are groups of individuals with a common condition or disease that share resources and knowledge to support clinical research, patient education, and clinical care. These organizations are active in the United States (e.g., American Cancer Society, American Heart Association, National Kidney Foundation) and are becoming more common around the world, including in developing countries (Landy et al., 2012). Traditionally, disease advocacy organizations have been involved primarily in promoting and facilitating participation in clinical trials and raising money to fund research. More recently, their roles have expanded to encompass
- collaborating with investigators on the study design and review of clinical trial protocols;
- developing and managing clinical trial networks;
- incorporating into clinical trials data collected directly by participants, such as from personal devices or sensors, as well as reports of participant-centered outcomes; and
- creating online platforms for patient engagement, such as PatientsLikeMe, Chronology, and CureTogether (American Cancer Society, 2014; CFF, 2013, 2014; Davidson, 2010; Greenwald, 2013; JDRF, 2014; Marcus, 2012; Olivas, 2014).
Furthermore, disease advocacy organizations, through online networks, have acted as a conduit for the expression of participant frustrations regarding the lack of data sharing by investigators. In many online forums, participants often express frustration about the lack of communication from investigators at the conclusion of a clinical trial (Terry and Terry, 2011). A recent study in the cancer community (Ramers-Verhoeven et al., 2014) reinforces these expressions of participant frustration. According to the authors,
Unfortunately, this feeling of being special in many cases vanished when participation was over. Although all participants were appreciative of the care they received during the trial, there was a very clear sense of feeling that they were no longer a priority when the trial ended. “You are extremely well informed, but once you come off the trial there is not one letter. Nothing. . . . This is the major problem I had with it.” (French respondent/patient)
Many participants also expressed frustration at never being told the results of the clinical trial in which they participated:
The clinical trial experience was similar to how I had imagined it, but I was surprised that I didn’t get more information about it all as it progressed and when I was withdrawn. (UK respondent/participant)
Will the results of the clinical trial be provided? That’s what preoccupied me the most. (Japanese respondent/participant)
Disease advocacy organizations are uniquely positioned to address these frustrations—both as a conduit for the expression of concerns and as a potential partner with investigators to create frameworks for continual engagement with trial participants.
An increasing number of patient groups have responded to this frustration by bypassing traditional investigator- or company-initiated clinical trials and organizing themselves to conduct their own trials of experimental agents. For example amyotrophic lateral sclerosis (ALS) patients have banded together to develop homemade versions of an experimental agent and test it on themselves (Marcus, 2012). When acting in this capacity,
disease advocacy organizations share many of the same concerns, roles, and responsibilities as those of other nonprofit funders and sponsors of clinical trials with regard to data sharing.
Whether through direct financial support (either alone or as part of a funding syndicate) or other forms of assistance (e.g., participant recruitment, clinical research networks), disease advocacy organizations make significant contributions to the development and execution of clinical trials. These efforts give these organizations an opportunity to influence policies and strategies so as to encourage responsible sharing of clinical trial data.
Both funders and sponsors of clinical trials have significant leverage to set standards and to encourage data sharing for the trials they fund. When considering data sharing, however, it is important to consider the context for each clinical trial in terms of the remit of the organization that has funded the work and the type of organization or institution that is the sponsor of the trial.
Public and nonprofit organizations (e.g., the U.S. National Institutes of Health [NIH], The Wellcome Trust, The Bill & Melinda Gates Foundation, the European Clinical Trial Development Platform, and disease advocacy organizations such as the ALS Association) often will support clinical trials by providing a research grant to a university or other research organization, such as a hospital or charity. The recipient organization is required to take on the role of sponsor or may delegate this role to an appropriate organization or group that has the capacity to act as sponsor. The trial sponsor is defined by ICH GCP as the organization that has specific responsibilities for trial conduct, such as ensuring that the trial is scientifically robust, that its conduct and procedures comply with safety and ethical standards, and that participants will be compensated for any harm that may result from their participation (ICH, 1996). Sponsors also are required to ensure that the trial is listed on a recognized clinical trial registry. Alternatively, a private company (e.g., a pharmaceutical or device company) may directly sponsor a clinical trial for one of its products. Contract research organizations also may work with private companies and research institutions/universities to conduct clinical trials.
Public and Nonprofit Funders
U.S. National Institutes of Health
In the United States, NIH is by far the largest public funder of clinical trials. Currently, NIH supports more than 3,000 open trials (10 percent of all open trials registered on ClinicalTrials.gov), and the agency has expressed support for ensuring that data from every trial are made public to improve the reproducibility of research results and ultimately facilitate their use to improve health (Hudson, 2013). Currently, however, the results of only 46 percent of NIH-funded trials are published within 30 months of trial completion (Ross et al., 2012). As the primary funder of translational and clinical science, NIH has been a crucial force behind major innovations designed to transform and restructure research. First, during the Human Genome Project, NIH was a key driver of researchers’ sharing of genome sequencing data soon after they discovered the sequence so that other scientists could benefit from this knowledge to make further discoveries (NIH issues genomic data sharing policy, 2014). The U.S. Department of Health and Human Services’ (HHS’s) regulation on “intangible property” generated through federal awards (which is part of a comprehensive set of HHS administrative rules on awards) states that the federal government has the right to
- Obtain, reproduce, publish, or otherwise use the data first produced under an award;
- Authorize others to receive, reproduce, publish or otherwise use such data for Federal purposes.2
Second, to promote broader dissemination of the results of the trials it sponsors, NIH requires investigators applying for new funding to link
2 34 CFR § 74.36 (c)(1-2).
3 Of interest, patent law may somewhat constrain public release of publicly funded data. Were public funders to mandate data sharing on an extremely compressed time schedule (for example, within 24 hours, as was done with the effort surrounding the NIH-funded Human Genome Project), grantees could argue that such rapid public release interfered with their ability to file patent applications, in violation of the Bayh–Dole Act of 1980, which allows recipients of federal funding broad discretion to patent products of federally funded research. Indeed, although they never brought legal action based on these concerns, certain university technology transfer offices associated with the Human Genome Project did note tensions between immediate data release and patenting.
4 See, e.g., Contreras (2011) and Eisenberg and Rai (2006).
the publications in their biosketch to the ID number in PubMed Central or similar platforms for sharing articles (NIH, 2014c). Most recently, in November 2014, HHS proposed a rule to clarify and extend the requirements of the Food and Drug Administration Amendments Act of 2007 (FDAAA). The FDAAA originally required that summary-level results of trials of FDA-approved products (including demographic and other baseline participant characteristics, primary and secondary outcomes, and adverse events) be shared on the ClinicalTrials.gov registry within 12 months of study completion. The proposed rule extends the requirement to register and share summary-level results so that it covers trials of unapproved products. Also in November 2014, NIH proposed a draft policy to require registration and summary results reporting of all interventional clinical trials (i.e., surgical and behavioral trials, phase 1 trials) funded by NIH (Hudson and Collins, 2014). To facilitate such reporting of results, ClinicalTrials.gov is increasing one-on-one staff support. Failure to comply with these requirements could result in civil penalties of up to $10,000 per day (assessed by the FDA) and withholding of funding for federally funded trials (Hudson and Collins, 2014). NIH also is taking timely reporting of clinical trial results into account during the review of subsequent funding applications.
Third, through the Clinical and Translational Science Awards (CTSA), NIH has provided funding and leadership to promote team-oriented scientific research and collaborative research among different institutions and research teams (NIH, 2014a).
Finally, the recent NIH Genomic Data Sharing Policy sets forth the responsibilities of NIH-funded researchers for sharing genomic data (including clinical trial data) and, notably, “encourages researchers to get consent from participants for future unspecified use of their genomic data” (NIH, 2014b).
NIH could be a driver for the sharing of clinical trial data by making it a requirement in the grant approval process and funding stipulations. Currently, NIH requires grantees to have a plan for data sharing if they request direct costs of $500,000 or more in any budget year, but it does not require data sharing, monitor whether data are shared as planned, or expressly allow a line item for expenses due to data sharing activities (NIH, 2003).
NIH’s experience with the legal and policy justification for requirements to share publicly funded genomic data has implications for the sharing of clinical trial data. In the context of human genome data, NIH has sometimes implemented controlled access to accommodate concerns about participant privacy and informed consent.5 However, the
5 For a recent articulation of the NIH approach to addressing issues of informed consent and privacy for human genome data, see Final NIH Genomic Data Sharing Policy, 78 F.R. 51345, August 27, 2014.
requirement that investigators must deposit data free from claims of trade secrecy/commercially confidential information has not changed. This experience with policies on genomic data could inform policies that NIH and other agencies adopt with respect to publicly funded clinical trial data.
Resolution of the issue of whether and how NIH might enforce data deposition requirements against violators similarly could be informed by the experience with genomic data. For its policies regarding genomic data deposition, NIH generally has refrained from articulating legal enforcement mechanisms. But NIH’s most recent genomic data sharing policy does refer to potential sanctions under 45 CFR Section 74.62,6 which addresses enforcement of the terms and conditions of a grant. Such sanctions include withholding of future research awards and even suspension of an entire institution from receipt of federal funding.
The wellcome Trust
The Wellcome Trust is a charitable foundation based in the United Kingdom that provides funding for research in the United Kingdom and low- and middle-income countries (The Wellcome Trust, 2014a). The foundation supports increased transparency and sharing of clinical trial data from the research it funds by (1) requiring a data management and data sharing plan for all applications for funding, regardless of whether the investigation is taking place in a resource-poor country or the United Kingdom; (2) requiring any research papers published in peer-reviewed journals to be made available through PubMed Central (PMC) and Europe PubMed Central (Europe PMC) within 6 months of publication and providing funding to cover open access charges; and (3) requiring authors and publishers that receive open access payments to license research papers to be freely copied and reused with proper attribution to the original authors, using the Creative Commons Attribution license (CC-BY) (The Wellcome Trust, 2014b).
In addition, the foundation supports a number of major initiatives that promote data sharing, such as MalariaGEN—the Malaria Genomic Epidemiology Network, a community established in 2005 comprising more than 20 countries and 100 researchers. MalariaGEN serves as the main driver of collaborative research and genomic data sharing for malaria research, both of which play a crucial role in researchers’ efforts to work to develop and improve tools for controlling this disease (MalariaGen, 2014a). MalariaGEN receives funding from a variety of nonprofit sources,
including The Wellcome Trust, the Foundation for the National Institutes of Health, and The Bill & Melinda Gates Foundation (MalariaGen, 2014b).
The Bill & Melinda Gates Foundation
The Bill & Melinda Gates Foundation is a global foundation that supports clinical research and other projects in resource-poor countries. It has joined the International Aid Transparency Initiative (IATI), which works to improve the transparency of aid, development, and humanitarian resources by establishing a common standard for the publication of aid information and providing an online repository for all raw data published to the IATI standard. In March 2014, the Gates Foundation began publishing open data on its development activities in accordance with IATI standards (Aid Transparency Index, 2014).
Starting in January 2015, the Gates Foundation will require grantees to make publications, and the underlying data sets, available for free immediately upon publication and with no restrictions on use. The foundation will pay open access fees for publication. The foundation is allowing a phase-in period: Until 2017, open access to publications and to underlying data may be delayed for 12 months (The Bill & Melinda Gates Foundation, 2014; Van Noorden, 2014).
In the past, the culture of clinical research in industry did not include proactively sharing clinical trial data. There are several reasons for this culture, including concerns about incorrect or conflicting analyses generated by secondary users; concerns about participant privacy; and the desire to allow participating researchers and investigators the unique opportunity to publish data from trials in which they participated, which is important for their careers. Moreover, clinical research generated in support of marketing applications generally was considered to be commercially confidential information. Sponsors spend significant time and effort in developing drugs, clinical and regulatory strategies, and clinical research protocols and analysis plans. These plans and documents may include considerations based on confidential interactions with regulatory authorities and internal scientific expertise and on strategic ideas. Data gathered from studies often are used in the development of subsequent studies or products and are part of the institutional knowledge base in research and development (R&D) units within companies. Access to this information could give competitors a significant competitive advantage. Thus, sharing clinical trial data could shorten the time between the marketing of a first-in-class product and the marketing of similar products.
This interval, a key driver of return on investment for R&D, has already decreased significantly over time (Lanthier et al., 2013). Further decreases could discourage future investment in new product development.
Appendix C provides additional legal analysis, conducted by the committee, of industry intellectual property concerns. This legal analysis focuses on small molecules and biologics. Of course, clinical trial data also are generated in medical device trials. With medical devices, however, issues regarding data exclusivity, even in jurisdictions like the United States, are less clear-cut. Indeed, according to a submission to this committee from the trade group AdvaMed, release of data associated with receiving FDA clearance through the expensive premarket approval process could facilitate market entry for competitors using the so-called 510(k) pathway for approval (AdvaMed, 2013).7 Under the 510(k) pathway, an applicant must prove that its device is “substantially equivalent” to a device already on the market. Proving substantial equivalence requires showing that the new device has the same intended use and technological characteristics as the predicate.
In the past, secondary users requesting access to individual participant data or study reports from industry-sponsored studies would approach individual investigators and/or authors of study manuscripts to request access, just as in academia. Many times, if both parties agreed, secondary users requesting access could be asked, for example, to share hypotheses and data analysis plans and/or sign confidentiality agreements, and access would be granted. Some companies had procedures and review committees for external proposals (Rosenblatt, 2014), while others did not.
Independent researchers who have obtained clinical study reports (CSRs)8 and individual participant data from industry-sponsored trials have identified significant problems with underreporting of negative results and serious adverse events and with failure to publish results of negative trials for widely prescribed therapies and a vaccine (Doshi et al., 2013). These claims, however, have been disputed by sponsors. In several
7 In contrast to AdvaMed’s point about data exclusivity, its arguments about the relative weakness of medical device patents or the purported exemption of medical devices from the FDAAA appear to lack foundation.
8 A CSR is an “integrated full report of an individual study of any therapeutic, prophylactic or diagnostic agent . . . conducted in patients. The clinical and statistical description, presentation, and analyses are integrated into a single report incorporating tables and figures into the main text of the report or at the end of the text, with appendices containing such information as the protocol, sample case report forms, investigator-related information, information related to the test drugs/investigational products including active control/comparators, technical statistical documentation, related publications, patient data listings, and technical statistical details such as derivations, computations, analyses, and computer output” (FDA, 1996, p. 1).
cases, the ensuing scientific and public debate has led to changes in labeling and marketing of drugs, legal settlements, and further clinical trials to address contested clinical hypotheses (see Table 3-1). The results of these additional trials have shown a more complex and nuanced picture than did either the original clinical trial results or the first independent analyses. This back-and-forth debate is part of the scientific method, which leads to ongoing clarification of scientific issues. Although complex and sometimes confusing to the public, this process illustrates how scientific knowledge generally progresses through debate, new data and analyses, and further debate.
Cases in which requests for access by independent investigators were repeatedly denied or delayed have sparked calls for a more systematic and transparent approach to the sharing of industry clinical trial data. In addition, as discussed in the section below on regulatory agencies, the European Medicines Agency (EMA) has generated a great deal of discussion on its mandate to share clinical trial data after marketing approval (EMA, 2013). In parallel with these discussions, recent HIPAA guidance has led to increased comfort with sharing de-identified clinical trial data for scientifically important analyses (HHS, 2012). Thus, over the past several years, the culture in industry has been changing, so that, as noted in Chapter 1, industry now is often leading data sharing initiatives. Several new initiatives launched by pharmaceutical companies and two device companies have significantly changed the paradigm for sharing of clinical trial data (see Appendix D) (Krumholz et al., 2014). In addition, in 2013, the Pharmaceutical Research and Manufacturers of America (PhRMA) and the European Federation of Pharmaceutical Industries and Associations (EFPIA) announced a commitment from all of their member companies to develop a process for and commit to sharing clinical trial data (PhRMA and EFPIA, 2013).
Currently, the costs of sharing clinical trial data (see Box 3-4) are borne by trial sponsors that agree to share the data. A substantial portion of this cost is for redacting commercially confidential information and participant identifiers from the data manually—for example, handwritten notes in CSRs that identify participants or reveal a company’s strategy for future research or for interactions with regulators (Shoulson, 2014).
Small companies that account for a significant proportion of new therapeutic discoveries have stated that they currently do not have the revenue to support sharing clinical trial data. In 2012, 42 percent of new drug approvals were for emerging sponsors (FDASIA, 2013). There are precedents for reducing fees for small companies. For example, when small companies seek drug approval from the FDA, they have reduced application fees (FDA, 2011b, p. 9). Sharing clinical trial data may be an
TABLE 3-1 Examples of Effects of Independent Analyses Carried Out on Clinical Trial Data
|Concerns Raised by Independent Analyses of Clinical Trial Data||Effects|
|Concerns Raised by Independent Analyses of Clinical Trial Data||Effects|
|Selective Serotonin Reuptake Inhibitors (SSRIs)||
In 2002, about 900,000 prescriptions (costing $55 million) were written for children with mood disorders in the United States for a drug with potential harms and poor evidence of efficacy (Chan et al., 2014)
|Concerns Raised by Independent Analyses of Clinical Trial Data||Effects|
analogous situation in which small companies should not pay the same costs as large companies.
Submission of clinical trial data to regulatory authorities to gain approval for a new product or indication is an important consideration in the sharing of clinical trial data. Although health authorities exist around the world,9 this section focuses primarily on the EMA and the FDA because many other countries worldwide rely on those agencies’ review of products instead of conducting their own.
The EMA has been a pioneer in the sharing of clinical trial data; its
9 It is important to note that regulatory agencies in different jurisdictions differ in the types of data they hold and in what authority they have to provide access to data submitted to them or to other parties.
Costs of Sharing Clinical Trial Data
Costs associated with current data sharing activities among private sponsors of clinical trials may include the following:
- Protections for privacy, including de-identification of data*
- Redaction of documents
- Setting up databases/participating in data sharing sites (e.g., SAS®)
- Setting up and maintaining websites/portals
- Payments for review panel/steering committee as required
- Payments for third-party administrator
- Solicitation of external experts
- Due diligence assessments—finding data, getting data in the correct format, working with partners to assess data sets, etc.
- Compliance and auditing efforts—registry and publication
- Creation of lay summaries/synopses for posting on external sites (over and above ClinicalTrial.gov requirements)
- Work to create templates for informed consent forms, clinical study reports, etc. to allow for greater data sharing
- Data sharing coordinators—independent roles assigned to manage the intake and fulfillment of requests
* The committee was unable to find estimates of the costs associated with anonymizing or de-identifying health data in the published literature. However, the authors of the commissioned paper on de-identification (see Appendix B), both of whom are external consultants who de-identify health data sets, estimate that costs for de-identifying a particular data set can range from $10,000 to $100,000, depending on the amount of effort required. When the work to de-identify data sets goes from being episodic to more routine or ongoing, it may make sense to develop in-house expertise. A de-identification course offered by Privacy Analytics, for example, costs about $2,000 per person. Developing the capacity to automate the deidentification of data sets could cost between $100,000 and $500,000, but these costs would be spread over a number of data sets and years.
plan for sharing data submitted to it and its engagement with stakeholders and the public on the issue have stimulated international discussions of data sharing. In the United States, the FDA plays a special role in the review of submitted applications. It obtains the entire database of studies conducted on products submitted for approval: protocols and all amendments, data analysis plans, case report form, summary data, and individual participant data (FDA, 2011a). In addition to a comprehensive review, the FDA conducts its own analyses and produces its own summaries. The EMA also conducts its own extensive reviews of submitted applications (although it does not usually obtain individual participant
data), as do agencies in several other countries, most notably the Pharmaceuticals and Medical Devices Agency (Japan) and the China Food and Drug Administration.
It is important to keep in mind that only a small proportion of all clinical trial data is submitted to regulatory authorities. Most academic and publicly funded trials do not have a regulatory goal, and indeed many industry trials will not automatically generate data that are submitted in a regulatory file or for other reasons are ever submitted to an authority. Therefore, although there may be benefit in regulatory agencies’ release of the data that have been submitted to them, which are the basis for regulatory decisions, these data do not represent all the clinical trial data that are being generated internationally.
European Medicines Agency (EMA)
As noted, the EMA has been a key promoter of greater transparency in sharing of clinical trial data. From pharmaceutical and device companies, the EMA receives detailed CSR and selected individual participant data that are used in making decisions to approve or reject products for marketing authorization.
The EMA’s policy10 on sharing the data it receives has evolved over the last 4 years, beginning in November 2010 when it began to release CSRs and other documents on a case-by-case basis in response to a formal freedom-of-information request (EMA, 2014c). Then, in June 2013, the EMA published a draft policy proposing to publish proactively both CSRs and de-identified individual participant data at the time of regulatory decisions (EMA, 2013; Wathion and EMA, 2014). In that document, the EMA noted that CSRs do not contain commercially confidential information and therefore could be released with no redactions. The draft policy received more than 1,000 comments from more than 150 organizations (EMA, 2014d). According to the EMA, the comments were centered on three main issues: (1) protection of patient privacy, (2) whether information contained in CSRs could be considered commercially confidential and be used by competitors for commercial advantage, and (3) the legality and enforceability of the data sharing agreement between the EMA and data users. The draft policy also faced legal challenges from pharmaceutical companies AbbVie and Intermune (Adams, 2014; Mansell, 2014).
In October 2014, after extensive deliberations and discussions with industry sponsors, academic researchers, journals, patient representa-
10 The policy relates to clinical trial data submitted under the centralized procedure after the “effective” date. The policy does not relate to legacy data submitted under arbitration/referral procedures or the centralized procedure.
U.S. Food and Drug Administration (FDA)
The FDA is currently examining the possibility of releasing “masked” nonsummary data—nonsummary data from which information has been removed so that the data will not identify any specific product or application.11 In response, various commentators have urged that “masking” is likely to be ineffective (AdvaMed, 2013; Gaffney, 2013). Absent further detail regarding how masking will be accomplished, it is difficult to parse these arguments. Furthermore, critics have objected that effective masking of products would limit the usefulness of the data for secondary analyses of individual clinical trials and meta-analyses.
The power of U.S. regulatory agencies over data submitted by clinical trial sponsors is governed by two statutes: the Freedom of Information Act12 (FOIA), which addresses disclosure in response to citizen requests; and the Trade Secrets Act13 (TSA), which addresses the limits of affirmative disclosure by the government. In the case of the FDA in particular, a relevant regulation that essentially mirrors the constraints of these two statutes is 21 CFR 20.61(c). This regulation provides that “data and information submitted or divulged to the Food and Drug Administration which fall within the definitions of a trade secret or confidential commercial or financial information are not available for public disclosure.” Moreover, these statutes, together with case law interpreting them, generally prohibit regulatory agencies such as the FDA from releasing
11 68 Fed. Reg. 3342, June 4, 2013.
12 5 U.S.C. § 552.
13 18 U.S.C. § 1905 (1982).
information that is likely “to cause substantial harm to the competitive position of the person from whom the information was obtained.” Courts have, however, required that those who might allege competitive harm make arguments that go beyond the “conclusory and generalized.”14 For practical purposes, this means that any regulatory steps the FDA might take in the direction of making nonsummary clinical data publicly available would have to take into account (as the EMA has) redaction of information regarding potential new uses, as well as restrictions against copying full data sets for purposes of seeking marketing approval in other jurisdictions.
One important open question is the extent to which the FDA may have the authority to issue regulations that override the ordinary constraints of the TSA. The TSA does allow agencies to disclose trade secrets/commercially confidential information to the extent that such disclosure is “authorized by law.”15 Moreover, under standard principles of administrative law, substantive regulations that the FDA (or any agency) has authority to promulgate constitute “law.” Further, some legal scholars have argued that the FDA potentially has the power to disclose trade secrets for public health reasons, citing the provision in the Hatch-Waxman Act stating that the FDA is supposed to release clinical trial data after Hatch-Waxman data exclusivity expires, absent “extraordinary circumstances.”16
Section 301(j) of the Food, Drug, and Cosmetic Act (FDCA) does specifically prohibit the FDA from releasing to the public information “concerning any method or process which as a trade secret is entitled to protection.”17 Thus, the FDA presumably would not have the authority to issue regulations that disclosed this specific subcategory of trade secrets. However, under section 501(i) of the FDCA,18 the FDA does have expansive authority to impose on regulated parties “other conditions” that “relat[e] to the protection of the public health.” Moreover, in January 2001, the FDA did rely in part on this authority to propose disclosure rules with respect to clinical data concerning human gene therapy.19 The proposed FDA rules invoked section 505(i) in the context of arguing that “several significant public health goals” would be served through greater disclosure of data. Although these proposed rules were never finalized, they
14 5 U.S.C. § 552(b)(4).
15 18 U.S.C. § 1905 (1982) (emphasis added).
16 Drug Price Competition and Patent Term Restoration Act, Public Law 98-417, 98th Cong. (September 24, 1984).
17 21 U.S.C. § 331(j) (1982).
18 21 U.S.C. § 355(i).
19 See FDA “Proposed Rule on Availability for Public Disclosure and Submission to FDA for Public Disclosure of Certain Data and Information Related to Human Gene Therapy or Xenotransplantation,” 66 Fed. Reg. 4692 (2001).
represent an important precedent to consider in thinking through questions of the FDA power in the context of disclosure of clinical trial data.
Two types of scientific researchers are vitally involved in sharing clinical trial data: (1) the researchers who are key figures in the clinical trial design and at the participant interface, and (2) the researchers who analyze data collected by projects and processes—such as clinical trials, disease registries, and clinical care—in which they were not involved. In this report, the term trialist is used to refer specifically to researchers who design and conduct clinical trials, while investigators is used for all researchers. Among investigators, the term secondary users refers to investigators who use clinical trial data for purposes including reanalyses, novel analyses, and meta-analyses but were not involved in generating the primary data.
Traditionally, trialists are people whose careers are built on conducting clinical trials that provide the evidence base needed for the high-quality practice of medicine. Because trialists have expertise in the condition to be studied by a trial, in the science of designing and conducting trials, or both, they are able to frame the research questions, study the key variables, perform state-of-the-art outcome measurements pertinent to the study question, and determine how best to gather that information. They may be based in academia; in the pharmaceutical, biotechnology, or device industries; in clinical practice; in contract research organizations (CROs); or in nonprofit organizations. In recent years, patients and disease advocates also have increasingly been involved in the design and conduct of trials out of personal interest and not as a career.
Trialists play the key roles in trial design, participant recruitment, and data accrual. Often, the success or failure of a clinical trial hinges on specific aspects of the trial design. Knowledge of how to design trials is thus a critical skill; successful trialists have the experience and expertise to create a design that is simple enough to be practical, comprehensive enough to be informative, and rigorous enough for results to be convincing (Anturane Reinfarction Trial Policy Committee, 1982; Anturane Reinfarction Trial Research Group, 1980; Baron et al., 2008; Bresalier et al., 2005; Canner, 1991; The Coronary Drug Project Research Group, 1980; May et al., 1981; Nissen, 2006; Temple and Pledger, 1980; Thackray et al., 2000; Wedel et al., 2001).
Once a trial has been designed, trialists identify eligible participants,
train members of the trial team, organize recruitment, obtain informed consent, and either gather the key outcome data themselves or arrange to have the data gathered by others. For trialists based in CROs or clinical practice, participant recruitment and enrollment may be their only role in the trial. Finally, it is trialists who have the understanding and expertise to interpret the trial data and give them clinical meaning. Thus, without motivated and knowledgeable trialists, the clinical trial process would come to a halt.
Clinical trials are the most rigorous approach to assessing the efficacy and safety of an intervention. Thus, clinical trialists play an essential role in developing the evidence base on which clinical care rests. However, being a successful clinical trialist is a long and arduous career path. To attract talented young people to clinical investigation, a clear career path and rewards for this path are needed. Because most clinical trials are carried out in a time frame of years, if not decades, within one’s career, an investigator may participate in a relatively small number of trials and lead only a handful of them. In contrast, basic science investigators, epidemiologists, health services researchers, and other types of scientists may complete many studies in the time it takes to perform one clinical trial.
Because of the time and skill set needed to design a trial, enroll the participants, and accrue and interpret the data, many trialists view the data gathered in a trial as their intellectual property, even if it technically “belongs” to a third party (Royal Society, 2012). Typically, trialists intend to carry out and publish a number of secondary analyses of the collected data. Often the prospect of such secondary analyses helps academic senior clinical trialists recruit trainees and junior faculty members to their research group. Because academic and industrial success depends on published output, and because only a small fraction of the accrued data is published in the primary report of a trial, many academics have been reticent to share data. A principal concern is that others will use the data to publish findings in a data set—and gain the accompanying recognition and academic prestige—before those who labored to design the trial and gather the data have had the opportunity to fully examine and analyze the data.
Clinical trialists also may be concerned that third parties who do not fully understand the subtleties of the trial design and data accrual may draw erroneous conclusions that could cloud or even vitiate the published findings. Thus, trialists want to protect their data from misuse by others. In addition to the potential career benefits of this more restricted approach, many trialists believe it is important and feel a responsibility to limit data access both to protect the research participants from having their data misused and, more broadly, to protect the public from misinterpretation of the data based on flawed analyses by nonexperts. Clinical
trialists also fear that responding to invalid challenges to their publications will consume large amounts of time and effort, taking them away from their own work (Wallentin et al., 2014).
Because it takes decades to create seasoned clinical trialists and people at the outset of their careers make choices for the future based in part on the possibility of long-term success, data sharing may be a disincentive to choosing a career as a clinical trialist. Indeed, many trialists may view people who use their data for structured reviews and meta-analyses as “parasites” on the system and as antagonists to the medical progress resulting from their work (Reidpath and Allotey, 2001; Share alike, 2014). Thus, there is a cultural gap that needs to be bridged. This bridge will best be built when trialists see the value to them in sharing data. For example, trialists’ examination of a data set accrued by others may help in designing future trials or interpreting data gathered by others. Funders can accelerate this process by providing tangible rewards to trialists for data sharing activities.
Finally, clinical trialists rely on third parties (e.g., industry, public funding entities, private charities) to fund their work. Therefore, they respond to priorities and direction from these organizations. Trialists are concerned that if data sharing becomes an unfunded mandate, the costs of sharing will reduce the funding available for new grants, which in turn would result in fewer new trials. This concern is particularly cogent for trialists working in low-resource settings such as those affected by neglected global diseases, where funding for new clinical trials is already scarce. According to a 2013 World Health Organization (WHO) report (WHO, 2013), unless there is more research, there will be no real improvements in public health in the world’s poorest regions. The report emphasizes that these regions “need to be the generators and not the passive recipients of data” (The Global Health Network, 2014; WHO, 2013). However, data sharing also could potentially increase the long-term return on grants by catalyzing secondary data analyses and helping to avoid future research that is redundant or based on an unpromising approach. Furthermore, making clinical trial data shareable could make future clinical trials more efficient in the long run because new research could build on secondary analyses of the shared data. New models for funding the sharing of clinical trial data may alleviate some concerns if the costs of sharing are spread more equitably among the various stakeholders.
Only a small minority of clinical investigators are trialists conducting interventional trials (Nussenblatt and Meinert, 2010). Other researchers conduct observational studies (e.g., disease registries, cohort studies) or
secondary analyses of existing data from such sources as electronic health records, administrative data, and clinical trials. Meta-analysts are the investigators most interested in sharing clinical trial data, for they analyze and synthesize results from multiple trials to arrive at the overall findings from the evidence base (Stewart and Tierney, 2002). Meta-analysts are critical for advancing summary understanding of a body of research, and meta-analytic results are highly valuable for informing the design of new clinical trials. Currently, relatively few investigators analyze publicly available results of clinical trials for new hypotheses and findings, simply because trial results currently are not widely shared, but this type of investigator is likely to be more prevalent if responsible data sharing becomes more widespread. Trialists, of course, can also be involved in these other types of clinical research.
Clinical investigators who do not conduct trials themselves may generate numerous publications by analyzing preexisting data sets in less time than it takes clinical trialists to generate even their initial results (IOM, 2010). However, just because data are accessible does not mean that they are usable. Data are usable if an investigator can search and retrieve them, can make sense of them, and can analyze them within a single trial or combine them across many trials. Given the large volume of data anticipated to become available from the sharing of clinical trial data, the data will have to be in a computable form amenable to automated methods of search, analysis, and visualization (the committee discusses this challenge further in Chapter 6).
In contrast with the foregoing disincentives for investigators to share data, the incentives for investigators to share data are almost universally regarded as insufficient (EAGDA, 2014; Tenopir et al., 2011). At least four parties—research institutions and universities, research funders, biomedical journals, and professional societies—could provide meaningful incentives for investigators to share data.
Universities can influence data sharing activities through infrastructure support, incentives, training, and scientific review.
Academic centers typically provide infrastructure in support of investigators who conduct clinical trials and generate new data. This infrastructure can consist of data managers, research coordinators, biostatistical and informatics support, and other clinical trial coordinating center expertise. By contrast, academic centers provide comparatively sparse support for
data curation, archiving, and sharing. High-quality data curation and management are required to prepare for data sharing, so that investigators must both recognize this need and have appropriately skilled personnel available to them. However, the level of such support varies widely among institutions, as can investigators’ recognition of how much of this support may be necessary. Academic centers also often provide insufficient recognition of the time, effort, and value of sharing clinical trial data (Bonham, 2014; Dickersin, 2013; IOM, 2013). Further, few universities have transitioned to being “digital enterprises” (Bourne, 2014) that manage their digital assets to full advantage, although the NIH CTSA program’s continuing emphasis on informatics has substantially improved clinical research informatics capacity at CTSA institutions (Kahn and Weng, 2012) and through common tools such as i2b2 (i2b2, 2014). For sharing of clinical trial data in particular, institutions continue to face challenges, including supporting faculty in sharing data or ensuring that trials are registered on ClinicalTrials.gov as required (Hudson and Collins, 2014). Better overall support of the clinical trials enterprise within most institutions is needed to support the kinds of data structuring and documentation that will be needed for data sharing.
In the eyes of performance review and promotion committees, the primary criteria for academic success rest on publications, funding, leadership, and teaching. Data sharing is not an activity that receives attention from promotion committees, and there is insufficient recognition of the intellectual effort involved in designing, accruing, curating, and completing a clinical trial data set. In this way, the lack of incentives for sharing clinical trial data is analogous to the recognized dearth of incentives for team science within university settings (Chan et al., 2014; NRC, 2012). Positive examples of promotion committees’ acknowledging the important contributions of investigators in creating high-quality, widely used data sets and sharing them with others are currently lacking. Appropriate recognition of data sharing activities in the promotion process would provide incentives for sharing data and obtaining maximal value from completed trials. Other promotion-related incentives for data sharing would exist if promotion committees took into account secondary publications by others based on clinical trial data produced and shared by their faculty.
Most of the workforce that would be involved in activities related to the sharing of clinical trial data are trained in universities. Currently, there
is little or no training within traditional clinical research education in the procedures and structures needed to share data. The development of such modules, either online or in classroom settings, could be instrumental in helping to move the field of data sharing forward.
Many academic institutions perform some form of internal scientific review for studies that are not reviewed externally or that fall within certain areas in which internal review is required (e.g., cancer studies in cancer centers designated by the National Cancer Institute). Including an assessment of data sharing plans in such reviews, together with trial registration and results reporting, would both serve as an incentive to create such plans and potentially provide valuable technical guidance in how to do so in that institution.
Biomedical journals serve as an intermediary between investigators who gather data and write research reports and readers, including clinicians, clinician-scientists, scientists, and laypeople. Journals play an essential role in providing peer review, evaluating research claims for scientific accuracy, and preparing reports for publication in a clear and lucid fashion. They provide a measure of assurance that the claims made by authors have validity and value. In this role as an intermediary between author-investigators and readers, journals want to be sure that the ideas reported under their imprimatur are correct. Journals often are perceived as the arbiters of the scientific and societal value of research in that their decision whether to publish submitted research results determines whether that research reaches an audience. The trialists who conducted the research also have a great interest in the publication decisions of journals, which are the gatekeepers for the academic credit and prestige that accompany being a lead author on research published in a high-impact journal.
The core goal of biomedical journals is to connect people (e.g., physicians, other researchers, patients, policy makers) with valid and important scientific research to improve scientific knowledge, patient care, and health outcomes. This goal is aligned with responsible sharing of clinical trial data. Broad sharing of clinical trial data also would advance the interests of journals in helping to ensure that published research findings are reproducible (Collins and Tabak, 2014).
Although responsible sharing of clinical trial data is consistent with the goals of biomedical journals, journals in general cannot take on data sharing responsibilities that are beyond their scope of work and resources.
It is not feasible for journals to assume the responsibilities entailed in serving as the repository of data for the studies they publish, adjudicating who will have access to the data, and negotiating interactions between authors of primary research reports and investigators whose secondary analyses reach different conclusions. Biomedical journals can, however, leverage their role as evaluators and publishers of research results, implementers of academic standards, and gatekeepers of academic credit and prestige to create and enforce policies that require sharing of clinical trial data (IOM, 2013; Laine et al., 2007).
Biomedical journals have an important role to play in advancing the creation of an environment in which sharing of clinical trial data is a standard and an expectation for publication in the scientific literature. Journals could require and enforce data sharing as a condition of submission and publication, with the goal of increasing the transparency and validity of clinical trial results they publish. Biomedical journals previously have played important roles in setting and enforcing important standards for clinical research. In 2004, for example, member journals of the International Committee of Medical Journal Editors (ICMJE) adopted a position that they would publish only clinical trials that had been registered in an appropriate, publicly accessible database prior to the enrollment of the first participant in the study (De Angelis et al., 2004). There was some initial pushback from the research community, but when the date for trial registration arrived on September 13, 2005, there was a spike in new trial registrations at ClinicalTrials.gov, the largest ICMJE-compliant website, as academic and industrial researchers rushed to bring their trials into compliance with this mandate (Zarin, 2013). Similarly, journals could provide leadership in setting minimal requirements for data sharing that would have a great impact on clinical trial investigators and sponsors. The Annals of Internal Medicine, the British Medical Journal, and the family of journals belonging to the Public Library of Science (PLoS) have already promulgated positions that they will publish only articles whose authors indicate willingness to share data (Annals of Internal Medicine) or agree to share data (British Medical Journal, PLoS) (Godlee and Groves, 2012; Laine et al., 2007; Silva, 2014). The challenge with these statements is that the specifics of what data will be shared, with whom, and by what means have not been clearly delineated. As recommended in this report, these details are key to the success of any data sharing program.
In addition, medical journals could help address challenges to responsible sharing of clinical trial data, particularly concerns about the usefulness and validity of secondary analyses of shared data. Journals could require authors of papers using shared clinical trial data sets to agree to
make the analytic data set supporting their findings, tables, and figures available, similar to requirements for authors of primary analyses of clinical trial data. In addition, journals might require authors submitting secondary analyses for publication to explain how their analytic approach differs from that of the primary analysis. Such an explanation would help reviewers and editors of submitted manuscripts and readers of published articles assess the validity of the secondary analyses and compare them with the primary analysis by the clinical trial team. Moreover, this requirement would discourage secondary analyses based on scientifically invalid analysis plans—for example, because they entail carrying out multiple analyses using statistically inappropriate methods.
Physicians’ professional societies can play a role in setting standards and establishing norms for data sharing. Most societies have an official journal (e.g., the American Medical Association’s Journal of the American Medical Association, the American College of Physicians’ Annals of Internal Medicine, and the American Heart Association’s Circulation), and investigators commonly present abstracts of clinical trials at the societies’ annual meetings. “Late-breaking” abstracts of clinical trials often attract considerable press coverage, as occurred with the recent clinical trial of the use of ezetimibe plus simvastatin in patients with acute coronary syndrome (Kolata, 2014; Peck, 2014; Vaczek, 2014). Moreover, professional societies often convene committees to develop evidence-based clinical practice guidelines, consistent with their mission to improve the evidence base upon which members make clinical decisions for patients.
Professional societies could require that authors in their official journals follow the recommendations in this report for responsible sharing of clinical trial data and that investigators submitting and presenting abstracts at their meetings agree to do so when they publish their clinical trial findings. Members of professional societies commonly take the lead in designing and carrying out clinical trials in their specialty. At annual meetings, professional societies could hold workshops on the sharing of clinical trial data. In these ways, professional societies could set expectations for responsible sharing of clinical trial data in their professional codes for members. Finally, professional organizations could help develop common data elements for clinical trials in their specialty and advocate for the use of those data elements in clinical trials.
The committee concluded that although no one stakeholder alone can achieve the benefits of the sharing of clinical trial data and minimize its risks, all stakeholders have a role and responsibility in responsible sharing of clinical trial data.
Recommendation 1: Stakeholders in clinical trials should foster a culture in which data sharing is the expected norm, and should commit to responsible strategies aimed at maximizing the benefits, minimizing the risks, and overcoming the challenges of sharing clinical trial data for all parties.
Funders and sponsors should
- promote the development of a sustainable infrastructure and mechanism by which data can be shared, in accordance with the terms and conditions of grants and contracts;
- provide funding to investigators for sharing of clinical trial data as a line item in grants and contracts;
- include prior data sharing as a measure of impact when deciding about future funding;
- include and enforce requirements in the terms and conditions of grants and contracts that investigators will make clinical trial data available for sharing under the conditions recommended in this report; and
- fund and promote the development and adoption of common data elements.
Disease advocacy organizations should
- require data sharing plans as part of protocol reviews and criteria for funding grants;
- provide guidance and educational programs on data sharing for clinical trial participants;
- require data sharing plans as a condition for promoting clinical trials to their constituents; and
- contribute funding to enable data sharing.
Regulatory and research oversight bodies should
- work with industry and other stakeholders to develop and harmonize new clinical study report (CSR) templates that do not include commercially confidential information or personally identifiable data;
- work with regulatory authorities around the world to harmonize requirements and practices to support the responsible sharing of clinical trial data; and
- issue clear guidance that the sharing of clinical trial data is expected and that the role of Research Ethics Committees or Institutional Review Boards (IRBs) is to encourage and facilitate the responsible and ethical conduct of data sharing through the adoption of protections such as those recommended by this committee and the emerging best practices of clinical trial data sharing initiatives.
Research Ethics Committees or IRBs should
- provide guidance for clinical trialists and templates for informed consent for participants that enable responsible data sharing;
- consider data sharing plans when assessing the benefits and risks of clinical trials; and
- adopt protections for participants as recommended by this committee and the emerging best practices of clinical trial data sharing initiatives.
Investigators and sponsors should
- design clinical trials and manage trial data with the expectation that data will be shared;
- adopt common data elements in new clinical trial protocols unless there is a compelling scientific reason not to do so;
- explain to participants during the informed consent process
- − what data will (and will not) be shared with the individual participants during and after the trial,
- − the potential risks to privacy associated with the collection and sharing of data during and after the trial and a summary of the types of protections employed to mitigate this risk, and
- − under what conditions the trial data may be shared (with regulators, investigators, etc.) beyond the trial team; and
- make clinical trial data available at the times and under the conditions recommended in this report.
Research institutions and universities should
- ensure that investigators from their institutions share data from clinical trials in accordance with the recommendations in this report and the terms and conditions of grants and contracts;
- promote the development of a sustainable infrastructure and mechanisms for data sharing;
- make sharing of clinical trial data a consideration in promotion of faculty members and assessment of programs; and
- provide training for data science and quantitative scientists to facilitate sharing and analysis of clinical trial data.
- require authors of both primary and secondary analyses of clinical trial data to
- − document that they have submitted a data sharing plan at a site that shares data with and meets the data requirements of the World Health Organization’s International Clinical Trials Registry Platform before enrolling participants, and
- − commit to releasing the analytic data set underlying published analyses, tables, figures, and results no later than the times specified in this report;
- require that submitted manuscripts using existing data sets from clinical trials, in whole or in part, cite these data appropriately; and
- require that any published secondary analyses provide the data and metadata at the same level as in the original publication.
Membership and professional societies should
- establish policies that members should participate in sharing clinical trial data as part of their professional responsibilities;
- require as a condition of submitting abstracts to a meeting of the society and manuscripts to the journal of the society that clinical trial data will be shared in accordance with the recommendations in this report; and
- collaborate on and promote the development and use of common data elements relevant to their members.
Adams, B. 2014. Intermune drops EMA lawsuit over data release. http://www.pharmafile.com/news/187640/intermune-drops-ema-lawsuit-over-data-release (accessed October 27, 2014).
AdvaMed. 2013. Re: Docket no. FDA-2013-N-0271: Availability of masked and de-identified nonsummary safety and efficacy data; request for comments; notice. http://advamed.org/res.download/346 (accessed December 8, 2014).
Aid Transparency Index. 2014. The Bill and Melinda Gates Foundation. http://ati.publishwhatyoufund.org/donor/gates (accessed November 25, 2014).
American Cancer Society. 2014. American Cancer Society clinical trials matching service. http://www.cancer.org/treatment/treatmentsandsideeffects/clinicaltrials/app/clinical-trials-matching-service (accessed October 15, 2014).
Anturane Reinfarction Trial Policy Committee. 1982. The Anturane Reinfarction Trial: Reevaluation of outcome. New England Journal of Medicine 306(16):1005-1008.
Anturane Reinfarction Trial Research Group. 1980. Sulfinpyrazone in the prevention of sudden death after myocardial infarction. The Anturane Reinfarction Trial Research Group. New England Journal of Medicine 302(5):250-256.
Barocas, S., and H. Nissenbaum. 2014. Big data’s end run around anonymity and consent. In Privacy, big data, and the public good: Frameworks for engagement, edited by J. Lane, V. Stodden, S. Bender, and H. Nissenbaum. New York: Cambridge University Press. Pp. 44-75.
Baron, J. A., R. S. Sandler, R. S. Bresalier, A. Lanas, D. G. Morton, R. Riddell, E. R. Iverson, and D. L. DeMets. 2008. Cardiovascular events associated with rofecoxib: Final analysis of the approve trial. Lancet 372(9651):1756-1764.
Benatar, S. R. 2002. Reflections and recommendations on research ethics in developing countries. Social Science & Medicine 54(7):1131-1141.
Bierer, B. E. 2014. Guiding principles for clinical trial data sharing. Paper presented at IOM Committee on Strategies for Responsible Sharing of Clinical Trial Data: Meeting Two, February 3-4, Washington, DC.
The Bill & Melinda Gates Foundation. 2014. Open access policy. http://www.gatesfoundation.org/how-we-work/general-information/open-access-policy (accessed December 10, 2014).
Bonham, A. 2014. Cultural and financial incentives for data sharing: Recognition and promotion. Paper presented at IOM Committee on Strategies for Responsible Sharing of Clinical Trial Data: Meeting Two, February 3-4, Washington, DC.
Bourne, P. E. 2014. What big data means to me. Journal of the American Medical Informatics Association 21(2):194.
Brent, D. A. 2004. Antidepressants and pediatric depression—the risk of doing nothing. New England Journal of Medicine 351(16):1598-1601.
Bresalier, R. S., R. S. Sandler, H. Quan, J. A. Bolognese, B. Oxenius, K. Horgan, C. Lines, R. Riddell, D. Morton, A. Lanas, M. A. Konstam, and J. A. Baron. 2005. Cardiovascular events associated with rofecoxib in a colorectal adenoma chemoprevention trial. New England Journal of Medicine 352(11):1092-1102.
Canner, P. L. 1991. Covariate adjustment of treatment effects in clinical trials. Controlled Clinical Trials 12(3):359-366.
CFF (Cystic Fibrosis Foundation). 2013. TDN protocol review process. http://www.cff.org/research/TDN/WorkingWithTDN/ProtocolReview (accessed October 15, 2014).
CFF. 2014. Therapeutics development centers. http://www.cff.org/research/TDN/TherapeuticsDevelopmentCenters (accessed October 15, 2014).
Chan, A. W., F. J. Song, A. Vickers, T. Jefferson, K. Dickersin, P. C. Gotzsche, H. M. Krumholz, D. Ghersi, and H. B. van der Worp. 2014. Increasing value and reducing waste: Addressing inaccessible research. Lancet 383(9913):257-266.
CIOMS (Council for International Organizations of Medical Sciences). 2002. International ethical guidelines for biomedical research involving human subjects. Geneva: CIOMS.
Cochrane Neuraminidase Inhibitors Review Team. 2011. Does oseltamivir really reduce complications of influenza? Clinical Infectious Diseases: An Official Publication of the Infectious Diseases Society of America 53(12):1302-1303.
Collins, F. S., and L. A. Tabak. 2014. Policy: NIH plans to enhance reproducibility. Nature 505(7485):612-613.
Collyar, D. 2013. what patients want from clinical trials—results from data! Paper presented at IOM Committee on Strategies for Responsible Sharing of Clinical Trial Data: Meeting One, October 22-23, Washington, DC.
Contreras, J. 2011. Bermuda’s legacy: Policy, patents, and the design of the genome commons. Minnesota Journal of Law, Science & Technology 12:61-125.
Corbie-Smith, G., S. B. Thomas, M. V. Williams, and S. Moody-Ayers. 1999. Attitudes and beliefs of African Americans toward participation in medical research. Journal of General Internal Medicine 14(9):537-546.
The Coronary Drug Project Research Group. 1980. Influence of adherence to treatment and response of cholesterol on mortality in the coronary drug project. New England Journal of Medicine 303(18):1038-1041.
Coxib and Traditional NSAID Trialists’ (CNT’s) Collaboration. 2013. Vascular and upper gastrointestinal effects of non-steroidal anti-inflammatory drugs: Meta-analyses of individual participant data from randomised trials. Lancet 382(9894):769-779.
Davidson, M. 2010. Help today, help tomorrow is goal of MDA’s Duchenne Clinical Research Network. http://quest.mda.org/article/help-today-help-tomorrow-goal-mdas-duchenne-clinical-research-network (accessed October 15, 2014).
Dawson, L., and N. E. Kass. 2005. Views of US researchers about informed consent in international collaborative research. Social Science & Medicine 61(6):1211-1222.
De Angelis, C., J. M. Drazen, F. A. Frizelle, C. Haug, J. Hoey, R. Horton, S. Kotzin, C. Laine, A. Marusic, A. J. Overbeke, T. V. Schroeder, H. C. Sox, and M. B. Van Der Weyden. 2004. Clinical trial registration: A statement from the International Committee of Medical Journal Editors. Canadian Medical Association Journal 171(6):606-607.
Diamond, G. A., L. Bax, and S. Kaul. 2007. Uncertain effects of rosiglitazone on the risk for myocardial infarction and cardiovascular death. Annals of Internal Medicine 147(8):578-581.
Dickersin, K. 2013. Principles of responsible data sharing. Paper presented at IOM Committee on Strategies for Responsible Sharing of Clinical Trial Data: Meeting One, October 22-23, Washington, DC.
DOJ (U.S. Department of Justice). 2004. warner-Lambert to pay $430 million to resolve criminal and civil health care liability relating to off-label promotion. Washington, DC: DOJ.
DOJ. 2012. GlaxoSmithKline to plead guilty and pay $3 billion to resolve fraud allegations and failure to report safety data. http://www.justice.gov/opa/pr/glaxosmithkline-plead-guilty-and-pay-3-billion-resolve-fraud-allegations-and-failure-report (accessed December 8, 2014).
Doshi, P., K. Dickersin, D. Healy, S. S. Vedula, and T. Jefferson. 2013. Restoring invisible and abandoned trials: A call for people to publish the findings. British Medical Journal 346:f2865.
Dwork, C. 2014. Differential privacy: A cryptographic approach to private data analysis. In Privacy, big data, and the public good, edited by J. Lane, V. Stodden, S. Bender, and H. Nissenbaum. New York: Cambridge University Press. Pp. 296-322.
EAGDA (Expert Advisory Group on Data Access). 2014. Establishing incentives and changing cultures to support data access. http://www.wellcome.ac.uk/stellent/groups/corporatesite/@msh_peda/documents/web_document/wtp056495.pdf (accessed December 16, 2014).
Eisenberg, R., and A. Rai. 2006. Harnessing and sharing the benefits of state sponsored research. Berkeley Technology Law Journal 21:1187-1213.
Ellenberg, S. S., T. R. Fleming, and D. L. DeMets. 2002. Data monitoring committees in clinical trials: A practical perspective. West Sussex, England: John Wiley & Sons.
EMA (European Medicines Agency). 2013. Publication and access to clinical-trial data. http://www.ema.europa.eu/docs/en_GB/document_library/Other/2013/06/WC500144730.pdf (accessed October 15, 2014).
EMA. 2014a. Clinical trials in human medicines. http://www.ema.europa.eu/ema/index.jsp?curl=pages/special_topics/general/general_content_000489.jsp&mid=WC0b01ac058060676f (accessed October 17, 2014).
EMA. 2014b. European Medicines Agency policy on publication of clinical data for medicinal products for human use. http://www.ema.europa.eu/docs/en_GB/document_library/Other/2014/10/WC500174796.pdf (accessed October 27, 2014).
EMA. 2014c. Finalisation of the EMA policy on publication of and access to clinical trial data—targeted consultation with key stakeholders in May 2014. http://www.ema.europa.eu/docs/en_GB/document_library/Report/2014/09/WC500174226.pdf (accessed October 17, 2014).
EMA. 2014d. Publication and access to clinical data: An inclusive development process. http://www.ema.europa.eu/ema/index.jsp?curl=pages/special_topics/general/general_content_000556.jsp (accessed October 24, 2014).
EMA. 2014e. Release of data from clinical trials. http://www.ema.europa.eu/ema/index.jsp?curl=pages/special_topics/general/general_content_000555.jsp&mid=WC0b01ac0580607bfa (accessed October 14, 2014).
EUREC (European Network of Research Ethics Committees). 2014. European Network of Research Ethics Committees—EUREC. http://www.eurecnet.org/index.html (accessed December 8, 2014).
FDA (U.S. Food and Drug Administration). 1996. Guideline for industry: Structure and content of clinical study reports. http://www.fda.gov/downloads/drugs/guidancecomplianceregulatoryinformation/guidances/ucm073113.pdf (accessed October 17, 2014).
FDA. 2011a. Detailed description of clinical trials database. http://www.fda.gov/AboutFDA/CentersOffices/OfficeofMedicalProductsandTobacco/CDRH/CDRHTransparency/ucm208389.htm (accessed October 29, 2014).
FDA. 2011b. Guidance for industry: User fee waivers, reductions, and refunds for drug and biological products. http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/ucm079298.pdf (accessed October 24, 2014).
FDA. 2014. A guide to informed consent—information sheet. http://www.fda.gov/RegulatoryInformation/Guidances/ucm126431.htm (accessed October 17, 2014).
FDASIA (Food and Drug Administration Safety and Innovation Act). 2013. Small business report to Congress. http://www.fda.gov/downloads/RegulatoryInformation/Legislation/FederalFoodDrugandCosmeticActFDCAct/SignificantAmendmentstotheFDCAct/FDASIA/UCM360058.pdf (accessed October 18, 2014).
Gaffney, A. 2013. Device, pharmaceutical industries say FDA proposal to release trial data illegal. http://www.raps.org/focus-online/news/news-article-view/article/4248/device-pharmaceutical-industries-say-fda-proposal-to-release-trial-data-illegal.aspx (accessed October 29, 2014).
Gibbons, R. D., C. H. Brown, K. Hur, J. M. Davis, and J. J. Mann. 2012. Suicidal thoughts and behavior with antidepressant treatment: Reanalysis of the randomized placebo-controlled studies of fluoxetine and venlafaxine. Archives of General Psychiatry 69(6):580-587.
The Global Health Network. 2014. About The Global Health Network. https://tghn.org/about (accessed December 11, 2014).
Godlee, F. 2009. We want raw data, now. British Medical Journal 339:b5405.
Godlee, F., and T. Groves. 2012. The new BMJ policy on sharing data from drug and device trials. British Medical Journal 345:e7888.
Greenwald, T. 2013. Patients take control of their health care online. http://www.technologyreview.com/news/518886/patients-take-control-of-their-health-care-online (accessed October 15, 2014).
Hammad, T. A., T. Laughren, and J. Racoosin. 2006. Suicidality in pediatric patients treated with antidepressant drugs. Archives of General Psychiatry 63(3):332-339.
Hernán, M. A., and M. Lipsitch. 2011a. Oseltamivir and risk of lower respiratory tract complications in patients with flu symptoms: A meta-analysis of eleven randomized clinical trials. Clinical Infectious Diseases 53(3):277-279.
Hernán, M. A., and M. Lipsitch. 2011b. Reply to Cochrane Neuraminidase Inhibitors Review Team. Clinical Infectious Diseases: An Official Publication of the Infectious Diseases Society of America 53(12):1303-1304.
Hesterlee, S. 2013. Research participant perspectives. Paper read at IOM Committee on Strategies for Responsible Sharing of Clinical Trial Data: Meeting One, October 22-23, Washington, DC.
HHS (U.S. Department of Health and Human Services). 2012. Guidance regarding methods for de-identification of protected health information in accordance with the Health Insurance Portability and Accountability Act (HIPAA) Privacy Rule. http://www.hhs.gov/ocr/privacy/hipaa/understanding/coveredentities/De-identification/guidance.html (accessed October 15, 2014).
HHS. 2014a. Federal policy for the protection of human subjects (“Common Rule”). http://www.hhs.gov/ohrp/humansubjects/commonrule (accessed October 17, 2014).
HHS. 2014b. Informed consent. http://www.hhs.gov/ohrp/policy/consent (accessed October 17, 2014).
Hudson, K. 2013. Strategies for responsible sharing of clinical trial data: NIH perspective. Paper presented at IOM Committee on Strategies for Responsible Sharing of Clinical Trial Data: Meeting One, October 22-23, Washington, DC.
Hudson, K. L., and F. S. Collins. 2014. Sharing and reporting the results of clinical trials. Journal of the American Medical Association. http://dx.doi.org/10.1001/jama.2014.10716 (accessed December 10, 2014).
i2b2 (Informatics for Integrating Biology and the Bedside). 2014. About us. https://www.i2b2.org/about/index.html (accessed December 8, 2014).
ICH (International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use). 1996. Guideline for good clinical practice E6(R1). http://www.ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/Efficacy/E6/E6_R1_Guideline.pdf (accessed December 16, 2014).
IOM (Institute of Medicine). 2009. Beyond the HIPAA Privacy Rule: Enhancing privacy, improving health through research. Washington, DC: The National Academies Press.
IOM. 2010. Transforming clinical research in the United States: Challenges and opportunities: workshop summary. Washington, DC: The National Academies Press.
IOM. 2013. Sharing clinical research data: workshop summary. Washington, DC: The National Academies Press.
JDRF (Juvenile Diabetes Research Foundation). 2014. To find a cure, JDRF needs people with type 1 diabetes to consider participation in human clinical trials of experimental new therapies. https://trials.jdrf.org/patient (accessed October 15, 2014).
Kahn, M. G., and C. Weng. 2012. Clinical research informatics: A conceptual perspective. Journal of the American Medical Informatics Association 19(e1):e36-e42.
Kolata, G. 2014. Study finds alternative to anti-cholesterol drug. The New York Times, http://www.nytimes.com/2014/11/18/health/study-finds-alternative-to-statins-inpreventing-heart-attacks-and-strokes.html?_r=1 (accessed December 8, 2014).
Krumholz, H. M., C. P. Gross, K. L. Blount, J. D. Ritchie, B. Hodshon, R. Lehman, and J. S. Ross. 2014. Sea change in open science and data sharing: Leadership by industry. Circulation: Cardiovascular Quality and Outcomes 7(4):499-504.
Lagakos, S. W. 2006. Time-to-event analyses for long-term treatments—the APPROVe trial. New England Journal of Medicine 355(2):113-117.
Laine, C., S. N. Goodman, M. E. Griswold, and H. C. Sox. 2007. Reproducible research: Moving toward research the public can really trust. Annals of Internal Medicine 146(6):450-453.
Landy, D. C., M. A. Brinich, M. E. Colten, E. J. Horn, S. F. Terry, and R. R. Sharp. 2012. How disease advocacy organizations participate in clinical research: A survey of genetic organizations. Genetics in Medicine 14(2):223-228.
Lanthier, M., K. L. Miller, C. Nardinelli, and J. Woodcock. 2013. An improved approach to measuring drug innovation finds steady rates of first-in-class pharmaceuticals, 1987-2011. Health Affairs (Millwood) 32(8):1433-1439.
Lu, C. Y., F. Zhang, M. D. Lakoma, J. M. Madden, D. Rusinak, R. B. Penfold, G. Simon, B. K. Ahmedani, G. Clarke, E. M. Hunkeler, B. Waitzfelder, A. Owen-Smith, M. A. Raebel, R. Rossom, K. J. Coleman, L. A. Copeland, and S. B. Soumerai. 2014. Changes in antidepressant use by young people and suicidal behavior after FDA warnings and media coverage: Quasi-experimental study. British Medical Journal 348:g3596.
MalariaGen. 2014a. About the malaria genomic epidemiology network. http://www.malariagen.net/about (accessed December 10, 2014).
MalariaGen. 2014b. Funding. http://www.malariagen.net/about/funding (accessed November 25, 2014).
Mansell, P. 2014. Abbvie strikes compromise with EMA on data disclosure. http://www.pharmatimes.com/article/14-04-08/AbbVie_strikes_compromise_with_EMA_on_data_disclosure.aspx (accessed May 22, 2014).
Marcus, A. D. 2012. Frustrated ALS patients concoct their own drug. http://online.wsj.com/news/articles/SB10001424052702304818404577345953943484054 (accessed October 15, 2014).
May, G. S., D. L. DeMets, L. M. Friedman, C. Furberg, and E. Passamani. 1981. The randomized clinical trial: Bias in analysis. Circulation 64(4):669-673.
MHRA (Medicines and Healthcare Products Regulatory Agency). 2003. Safety review of antidepressants used by children completed. http://www.mhra.gov.uk/PrintPreview/PressReleaseSP/CON002045 (accessed December 8, 2013).
Michiels, B., K. Van Puyenbroeck, V. Verhoeven, E. Vermeire, and S. Coenen. 2013. The value of neuraminidase inhibitors for the prevention and treatment of seasonal influenza: A systematic review of systematic reviews. PLoS ONE 8(4):e60348.
Molyneux, C. S., N. Peshu, and K. Marsh. 2005. Trust and informed consent: Insights from community members on the Kenyan coast. Social Science & Medicine 61(7):1463-1473.
Newman, T. B. 2004. A black-box warning for antidepressants in children? New England Journal of Medicine 351(16):1595-1598.
NIH (U.S. National Institutes of Health). 2003. Final NIH statement on sharing research data. http://grants1.nih.gov/grants/guide/notice-files/NOT-OD-03-032.html (accessed October 14, 2014).
NIH. 2014a. Clinical and translational science awards. http://www.ncats.nih.gov/research/cts/ctsa/ctsa.html (accessed October 24, 2014).
NIH. 2014b. NIH issues finalized policy on genomic data sharing. http://www.nih.gov/news/health/aug2014/od-27.htm (accessed October 14, 2014).
NIH. 2014c. Public access policy. http://publicaccess.nih.gov (accessed October 24, 2014).
NIH issues genomic data sharing policy. 2014. Nature Biotechnology 32(10):966-966.
Nissen, S. E. 2006. Adverse cardiovascular effects of rofecoxib. New England Journal of Medicine 355(2):203-205.
Nissen, S. E., and K. Wolski. 2007. Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes. New England Journal of Medicine 356(24):2457-2471.
NRC (National Research Council). 2012. For attribution—developing data attribution and citation practices and standards: Summary of an international workshop. Washington, DC: The National Academies Press.
Nussenblatt, R. B., and C. L. Meinert. 2010. The status of clinical trials: Cause for concern. Journal of Translational Medicine 8:65.
Olivas, R. 2014. International consortium provides $2m funding for phase 1 clinical trial of ReveraGen DMD drug. http://mda.org/media/press-releases/international-consortium-provides-2m-funding-phase-1-clinical-trial-reveragen (accessed October 15, 2014).
O’Rourke, P. P., and D. Forster. 2014. Considerations of informed consent in the discussion of clinical trial data sharing—domestic issues. Paper presented at IOM Committee on Strategies for Responsible Sharing of Clinical Trial Data: Meeting Two, February 3-4, Washington, DC.
Peck, P. 2014. AHA: Improve-it proves ezetimibe benefit. MedPage Today, www.medpagetoday.com/MeetingCoverage/AHA/48637 (accessed December 8, 2014).
PhRMA (Pharmaceutical Research and Manufacturers of America) and EFPIA (European Federation of Pharmaceutical Industries and Associations). 2013. Principles for responsible clinical trial data sharing: Our commitment to patients and researchers. http://www.phrma.org/sites/default/files/pdf/PhRMAPrinciplesForResponsibleClinicalTrialDataSharing.pdf (accessed October 16, 2014).
Privacy International. 2014. The state of privacy. https://www.privacyinternational.org/sites/privacyinternational.org/files/file-downloads/state-of-privacy_2014.pdf (accessed October 17, 2014).
Psaty, B. M., and R. A. Kronmal. 2008. Reporting mortality findings in trials of rofecoxib for Alzheimer disease or cognitive impairment: A case study based on documents from rofecoxib litigation. Journal of the American Medical Association 299(15):1813-1817.
Ramers-Verhoeven, C. W., F. Perrone, and K. Oliver. 2014. Exploratory research into cancer patients’ attitudes to clinical trials. Ecancermedicalscience 8:432.
Reidpath, D. D., and P. A. Allotey. 2001. Data sharing in medical research: An empirical investigation. Bioethics 15(2):125-134.
Retzer, K., J. Lopatowska, and C. Rich. 2011. Navigating the sea of data protection in European clinical research. http://media.mofo.com/files/Uploads/Images/111201-Data-Privacy-SRC.pdf (accessed October 31, 2014).
Rosenblatt, M. 2014. Cultural and Financial Incentives for Data Sharing—Recognition and Promotion. Discussion Panel at IOM Committee on Strategies for Responsible Sharing of Clinical Trial Data: Meeting Two, February 3-4, Washington, DC.
Ross, J. S., T. Tse, D. A. Zarin, H. Xu, L. Zhou, and H. M. Krumholz. 2012. Publication of NIH funded trials registered in ClinicalTrials.gov: Cross sectional analysis. British Medical Journal 344:d7292.
Royal Society. 2012. Science as an open enterprise. London: Royal Society.
Share alike. 2014. Nature 507(7491):140-140.
Shoulson, I. 2014. Cultural and financial incentives for data sharing: Recognition and promotion. Paper presented at IOM Committee on Strategies for Responsible Sharing of Clinical Trial Data: Meeting Two, February 3-4, Washington, DC.
Silva, L. 2014. PLOS’ new data policy: Public access to data. The PLoS ONE Community Blog, http://blogs.plos.org/everyone/2014/02/24/plos-new-data-policy-public-access-data-2 (accessed October 29, 2014).
Singh, J. A., and E. J. Mills. 2005. The abandoned trials of pre-exposure prophylaxis for HIV: What went wrong? PLoS Medicine 2(9):e234.
Stewart, L. A., and J. F. Tierney. 2002. To IPD or not to IPD?: Advantages and disadvantages of systematic reviews using individual patient data. Evaluation & the Health Professions 25(1):76-97.
Temple, R., and G. W. Pledger. 1980. The FDA’s critique of the anturane reinfarction trial. New England Journal of Medicine 303(25):1488-1492.
Tenopir, C., S. Allard, K. Douglass, A. U. Aydinoglu, L. Wu, E. Read, M. Manoff, and M. Frame. 2011. Data sharing by scientists: Practices and perceptions. PLoS ONE 6(6):e21101.
Terry, S. F., and P. F. Terry. 2011. Power to the people: Participant ownership of clinical trial data. Science Translational Medicine 3(69):69cm63.
Thackray, S., K. Witte, A. L. Clark, and J. G. Cleland. 2000. Clinical trials update: OPTIME-CHF, PRAISE-2, ALL-HAT. European Journal of Heart Failure 2(2):209-212.
Tucker, M. E. 2013. FDA panel advises easing restrictions on rosiglitazone. British Medical Journal 346:f3769.
Turner, E. H., A. M. Matthews, E. Linardatos, R. A. Tell, and R. Rosenthal. 2008. Selective publication of antidepressant trials and its influence on apparent efficacy. New England Journal of Medicine 358(3):252-260.
Vaczek, D. 2014. Merck’s IMPROVE-IT trial hits big, supports wider Zetia use. Medical Marketing & Media, http://www.mmm-online.com/mercks-improve-it-trial-hits-big-supports-wider-zetia-use/article/383761 (accessed December 8, 2014).
Van Noorden, R. 2014. Gates Foundation announces world’s strongest policy on open access research. Nature News Blog, http://blogs.nature.com/news/2014/11/gates-foundation-announces-worlds-strongest-policy-on-open-access-research.html (accessed November 25, 2014).
Vedula, S. S., L. Bero, R. W. Scherer, and K. Dickersin. 2009. Outcome reporting in industry-sponsored trials of gabapentin for off-label use. New England Journal of Medicine 361(20):1963-1971.
Wallentin, L., R. C. Becker, C. P. Cannon, C. Held, A. Himmelmann, S. Husted, S. K. James, H. S. Katus, K. W. Mahaffey, K. S. Pieper, R. F. Storey, P. G. Steg, and R. A. Harrington. 2014. Review of the accumulated PLATO documentation supports reliable and consistent superiority of ticagrelor over clopidogrel in patients with acute coronary syndrome. Commentary on: DiNicolantonio JJ, Tomek A, Inactivations, deletions, non-adjudications, and downgrades of clinical endpoints on ticagrelor: Serious concerns over the reliability of the PLATO trial, International Journal of Cardiology, 2013. International Journal of Cardiology 170(3):E59-E62.
Wathion, N., and EMA. 2014. Finalisation of EMA policy on publication of and access to clinical trial data. http://www.eucope.org/en/files/2014/05/2014-05-Finalisation-of-EMA-policy-on-CT-data-Mtgs-with-stakeholders.pdf (accessed December 16, 2014).
Wedel, H., D. DeMets, P. Deedwania, B. Fagerberg, S. Goldstein, S. Gottlieb, A. Hjalmarson, J. Kjekshus, F. Waagstein, and J. Wikstrand. 2001. Challenges of subgroup analyses in multinational clinical trials: Experiences from the MERIT-HF trial. American Heart Journal 142(3):502-511.
The Wellcome Trust. 2014a. Funding. http://www.wellcome.ac.uk/Funding/index.htm (accessed December 10, 2014).
The Wellcome Trust. 2014b. Open access policy. http://www.wellcome.ac.uk/About-us/Policy/Policy-and-position-statements/WTD002766.htm (accessed November 25, 2014).
WHO (World Health Organization). 2009. Casebook on ethical issues in international health research. http://whqlibdoc.who.int/publications/2009/9789241547727_eng.pdf (accessed December 8, 2014).
WHO. 2013. Research for universal health coverage: world health report 2013. Geneva: WHO.
Zarin, D. A. 2013. Trial transparency. Paper presented at IOM Committee on Strategies for Responsible Sharing of Clinical Trial Data: Meeting One, October 22-23, Washington, DC.
This page intentionally left blank.