The numerous stakeholders involved in clinical research (e.g., patients, volunteers, investigators, sponsors, patient advocacy groups) can all agree that data sharing is essential, but it’s hardly routine as a current part of the clinical trial lifecycle. To address the various challenges and barriers related to sharing of clinical trial data, the Institute of Medicine (IOM) was commissioned to assemble a committee to reviewStrategies for Responsible Sharing of Clinical Trial Data and identify guiding principles and framework for data sharing.1 The committee’s report builds upon the draft document released for public comment January 2014 and released in its final format January 2015. The Sponsors of the committee included various organizations, including the National Institutes of Health (NIH), the Food and Drug Administration (FDA), and various pharmaceutical companies and foundations.

Data sharing can take on many forms, from reanalyses of the original hypothesis to check for reproducibility and validity, to meta-analyses across studies, to possibly de novo analyses for which the trial was not explicitly designed to address. Datasets can be shared in a particular format (e.g., SAS, Excel, ASCII) or even made available via a data sharing repository or website. Regardless of the approach or format, data sharing has the potential to advance scientific discovery and improve clinical care by maximizing knowledge from data collected in trials, stimulating new ideas for research, and avoiding unnecessarily duplicative trials. Building on the current infrastructure and available information regarding data sharing best practices, the IOM’s report provides a comprehensive review of ongoing data sharing activities and ultimately urges global participation in data sharing.

Sharing data is in the public interest, but a multi-stakeholder effort is needed to develop a culture, infrastructure, and policies that will foster responsible sharing—now and in the future.”

– IOM Committee on Strategies for Responsible Sharing of Clinical Trial Data

A number of steps on the data sharing pathway have helped to encourage and increase data sharing, but the ability to enforce existing policies is still inadequate. The IOM report chronicles several important dates with respect to data sharing policies, including the FDA Modernization Act of 1997, which mandated registration of federally or privately funded clinical trials conducted under Investigational New Drug applications on the public ClinicalTrials.gov website, maintained by the National Library of Medicine and NIH. In 2003, NIH issued its Final Statement on Sharing Research Data, which states, “All investigator-initiated applications with direct costs greater than $500,000 in any single year will be expected to address data sharing in their application.”

Despite these current data sharing policies, research shows that of 182,330 clinical trials and observational studies listed on ClinicalTrials.gov, only 15,899 have posted results.2 Furthermore, the trials that are sharing information are more likely to be industry-sponsored trials than NIH studies.3  Findings from a recent New England Journal of Medicine study were consistent; within twelve months of study completion, of 13,327 clinical trials that were designated as most likely to report results, 1790 (13.4%) reported results and 5110 (38.3%) reported results at any time during the 5-year period covered by the study. Of the trials reporting results, 17.0% were funded by industry, and 8.1% of trials funded by the NIH.4

Data are generated at every stage of the clinical trial life cycle, from trial design and registration through participant enrollment, publication, and regulatory submission. In the initial development stages of a study, the protocol is created, and the raw data are collected from trial participants. The data collected are then cleaned, coded, and abstracted into a final analyzable dataset, which may or may not proceed through the steps of submitting to a regulatory authority such as the FDA for review.

One of the more controversial aspects of the group’s recommendations is the inclusion of a time frame for release of particular types of data through data “packages,” which includes the following:

  • At trial registration—data sharing plan and registration elements
  • Twelve months after completion—release of summary results and lay summaries (to inform patients/participants)
  • No later than 6 months after publication—analytic data set supporting publication results
  • No later than 18 months after study completion (or 30 days following regulatory submission)—full analyzable data set with metadata—with specified exceptions for trials intended to support a regulatory application

These IOM recommendations offer a contrast to the current reality, where a significant portion of clinical trial evidence remains concealed long after study completion. In one review of NHLBI trials, it took more than 2 years for many results to be published, and 1 third were still not published within 4 years. Additional studies show that 57% of FDA trials have not released a publication within 5 years.4 These data highlight the lack of transparency in current attempts to release clinical trial data more broadly.5

The committee recommends that all stakeholders foster a culture in which data sharing is the expected norm, including commitment to responsible strategies aimed at maximizing benefits, minimizing risks, and overcoming the challenges of sharing clinical trial data. In summary, the IOM defined framework for guiding principles of data sharing encompasses consideration for the following:

  • Benefits and Risks—Maximize the benefits and minimize the risks of data sharing
  • Study Participants—Respect participants whose data are shared by giving highest priority to data confidentiality and security
  • Public Trust—Increase public trust in clinical trials and the sharing of data, resulting in improved public participation and funding
  • Fair Manner—Conduct the data sharing in a fair manner, applying the same rigorous standards to secondary data use as for hypothesis driven clinical research (e.g., research conducted by qualified investigators using prespecified analysis plans and submitted to peer-reviewed publications)

Some of the more innovative concepts in the report include recommendations to implement operational strategies such as an independent data sharing review panel, involving members of the lay public in governance; in academic settings, make clinical trial data sharing a consideration in faculty member promotions and assessment of programs; provide standard training on data sharing activities; and, finally, have the IOM sponsors convene a global multistakeholder committee to address key infrastructure, technological, and workforce challenges associated with the sharing of clinical trial data on an ongoing basis.

Implementing data sharing activities as early or as far “upstream” in the study design as possible builds efficiencies into data sharing processes, but these efficiencies are not realized without costs. Costs are related to the establishment and maintenance of the databases and technology platforms related to storing large datasets or to the technical staff required to prepare the data for sharing, whether by conducting analyses or performing data deidentification. One of the areas identified as a challenge by the commission is the accurate measurement of these various data sharing areas—technology, work force, infrastructure, and sustainability—in order to measure the return on investment resulting from data sharing activities. All costs considered, it is worth continuing the effort to imagine how much potential benefit could be realized for public health as a whole, from new trials to novel therapies and even cures for diseases.

KAI Research, Inc. is a wholly owned subsidiary of the Altarum Institute. KAI offers full service CRO support to government and pharmaceutical clients, and specializes in clinical study management activities including data management, data standardization, and data sharing.


 

  1. Institute of Medicine. (2015, January 14). Strategies for responsible sharing of clinical trial data. Retrieved from http://www.iom.edu/activities/research/sharingclinicaltrialdata.aspx.
  2. Rice, S. (2015, January 17). IOM report on sharing clinical trial data met with skepticism. Retrieved from http://www.modernhealthcare.com/article/20150117/MAGAZINE/301179965.
  3. Prayle, A. P., Hurley, M. N., & Smyth, A. R. (2012). Compliance with mandatory reporting of clinical trial results on ClinicalTrials.gov: Cross-sectional study. British Medical Journey, 344, d7373. PMID: 22214756. doi:10.1136/bmj.d7373
  4. Gordon, D., Taddei-Peters, W., Mascette, A., Antman, M., Kaufmann, P. G., & Lauer, M. S. (2013). Publication of trials funded by the National Heart, Lung, and Blood Institute. New England Journal of Medicine, 369, 1926–1934. PMID: 24224625. doi:10.1056/NEJMsa1300237
  5. Lee, K., Bacchetti, P., & Sim, I. (2008). Publication of clinical trials supporting successful new drug applications: A literature analysis. PLOS Medicine, 5, e191. PMID: 18816163. doi:10.1371/journal.pmed.0050191
  6. Rosenblatt, M., Jain, S., & Cahill, M. (2015). Sharing of clinical trial data: Benefits, risks, and uniform principles. Annals of Internal Medicine, 162(4), 306–307. doi:10.7326/M14-1299.
  7. Goodman, S. (2015). Clinical trial data sharing: What do we do now? Annals of Internal Medicine, 162(4), 308–309. doi:10.7326/M15-0021

All postings to the Health Policy Forum (whether from employees or those outside the Institute) represent the views of the individual authors and/or organizations and do not necessarily represent the position, interests, strategy, or opinions of Altarum Institute. Altarum is a nonprofit, nonpartisan organization. No posting should be considered an endorsement by Altarum of individual candidates, political parties, opinions or policy positions.