Speaker
Joseph S. Ross, MD, MHS
Professor of Medicine and Public Health
Yale University
Keywords
YODA, Open Data Access, Data Sharing
Key Points
- Data sharing and open science are important because underreporting research through selective publication and reporting are common. Fifty percent of clinical trials are never published. Even when published, many trial publications are delayed more than 2 years, many are underreported, statistically significant findings are most likely to be reported, and nearly two-thirds of studies had a primary outcome that was changed, introduced, or omitted. Yet patients and physicians frequently make treatment decisions based on only a portion of the potentially available clinical data.
- The 1997 FDA Modernization Act (section 113) provided public access to information about ongoing clinical trials, which led to the creation of ClinicalTrials.gov. The International Committee of Medical Journal Editors (ICMJE) realized that ClinicalTrials.gov had slow growth, and it decided not to accept papers that had not been registered.
- In 2007 the FDA Amendments Act (FDAAA) broadened the scope, requiring expanded registry, trial results uploaded within 12 months of study completion, sharing of basic results and adverse events. Other funders started requiring similarly.
- Since 2007 there has been a sea change in thinking about data sharing. ICMJE said effective July 2018 manuscripts must contain a data sharing statement, and trials that begin enrolling participants on or after January 2019 must include a data sharing plan in the trial’s registration. Effective January 2023, all research funded or conducted by NIH must include in the proposal plans for management and sharing of all data necessary to validate and replicate research findings.
- NIH is now implementing these large-scale efforts for data sharing. Researchers have to submit a data management and sharing plan including data types and amount as well as metadata; related tools, software, and/or code that will be generated; standards (formats, documentation, dictionaries); data preservation, access, associated timelines; access, distribution or reuse considerations; oversight of data management and sharing; and budgets for allowable costs.
- The YODA Project platform launched to find ways to make clinical trial data more available for investigators to use. YODA started with core principles to answer what would the ideal clinical trial data sharing platform would look like.
- YODA was launched in partnership with Johnson & Johnson in 2014 after a proof-of-concept effort with Medtronic. J&J started sharing trial data from all pharmaceutical products (including legacy trials), device and diagnostic products as of 2015, and consumer products as of 2017. YODA established data access policy and procedures with input from a Steering Committee, experts, stakeholders and public comment.
- YODA has 459 trials currently available with about 90% that have been thus far requested. Of 385 requests submitted, 368 have been approved, 4 remain under review, 11 were withdrawn/closed, 2 were rejected. Nearly all of the requests that come in require some administrative revision, but one-quarter required scientific revisions after review for clarity. 157 manuscripts and 93 abstracts have been submitted and 119 and 89 of which have been published or presented, respectively.
- There is valuable strengthening of science that happens in these data sharing efforts. Numerous studies that might not otherwise been feasible to pursue. Data sharing has facilitated direct collaborations and developed efficiencies. Replication studies have supported the original study. There have been no instances of patient privacy breaches, no publications of spurious safety findings that received unwarranted attention or disrupted patient care, and no data have been used for commercial or litigious purposes.
Learn More
Visit the YODA Project website.
Discussion Themes
-What were some surprises compared to expectations over the 10 years of the YODA project? It is firmly in our mind that we should be living in an open science data world, and part of that is sharing data and trying to help improve patient care. How do we get the resources on the academic side to help investigators who want to do this work? On the industry side, partners have stepped up and made their data available. Across the field, how do we get more prior data available?
Tags
#pctGR, @Collaboratory1