Sharing Data – Methods to increase data sharing and improve human
health
NIH Collaboratory Grand Rounds
Manesh Patel, MD and Michael Gibson, MD
Disclosures• Manesh Patel – Research Grants: Jansen, Bayer, AstraZeneca,
NHLBI, PCORI, HeartFlow, Procyrion, Advisory Board: Jansen, Bayer, AstraZeneca, Genzyme
• Michael Gibson - Research Grants: Angel Medical Corporation, Bayer Corp., CSL Behring, Ikaria, Inc., Janssen Pharmaceuticals, Johnson & Johnson Corporation, Portola Pharmaceuticals, Stealth Peptides, Inc., St. Jude Medical, Peer to Peer Communications: The Medicines Company, Consultant: Bayer Corporation, Boston Clinical Research Institute, Cardiovascular Research Foundation, Eli Lilly and Company, Gilead Sciences, Inc., Janssen Pharmaceuticals, Johnson & Johnson Corporation, The Medicines Company, Novo Nordisk, Ortho McNeil, Pfizer, St. Jude Medical, Web MD
Underlying mission at all academic organizations and researcher
• Culture of Scientific Curiosity and Excellence
• Environment of Collaboration and teamwork
• Passion for scientific discovery of new knowledge
and sharing that knowledge to improve Human
Health
Sharing of Data in Medicine – Open Science
• Human Genome Project
• Collaborative DNA and exploration – discovery science
• GSK – Welcome Trust – Clinical Study Data Request . com
• Yale Open Data Access (YODA)
• Supporting Open Access for Researchers (SOAR) program,
Risks of Sharing Data
1. Patient Privacy
2. Multiple testing
Benefits of Sharing Data
1. Confirmation of Findings
2. Secondary questions – analysis and discoveries
3. Aggregation of data for increased power to answer multiple possible questions
YODA - NEJM
Data Sharing Efforts
How we have done so far…..
IOM recommendations
IOM Recommendations
ICMJE recommendations
Fairness in Data Sharing Viewpoint - NEJM
Independent statistician confirm primary results
People who want to access the data should provide funds to compensate original investigators for efforts and investments in the trial and for costs of making data available
ACCESS CV Development
• Growing a consortium
• Building and extending on relationships
• Academic Research Organizations as places that can participate, coordinate, curate and lead in data sharing
ACCESS PROPOSAL
Academic Research Organization Consortium for Continuing Evaluation of Scientific Studies—CardioVascular: ACCESS CVImproving health through generating and sharing knowledge
Initial Consortium Partners (authors/representatives):Manesh R. Patel, MD, Paul W. Armstrong, MD, Deepak L. Bhatt, MD, MPH, Eugene Braunwald, MD, A. John Camm, MD, Keith A. A. Fox, MBChB, Robert A. Harrington, MD, William R. Hiatt, MD, Stefan K. James, MD, PhD, Ajay J. Kirtane, MD, SM, Martin B. Leon, MD, A. Michael Lincoff, MD, Kenneth W. Mahaffey, MD, Laura Mauri, MD, MSc, Roxana Mehran, MD, Shamir R. Mehta, MD, MSc, Gilles Montalescot, MD, Stephen J. Nicholls, MBBS, PhD, Vlado Perkovic, MBBS, PhD, Eric D. Peterson, MD, MPH, Stuart J. Pocock, PhD, Matthew T. Roe, MD, MHS, Marc S. Sabatine, MD, MPH, MikkaelSekeres, MD, MS, Scott D. Solomon, MD, Ph. Gabriel Steg, MD, Gregg W. Stone, MD, Frans Van de Werf, MD, PhD, Lars Wallentin, MD, PhD, Harvey D. White, DSc, and C. Michael, Gibson, MS, MD on Behalf of the ACCESS CV Group*
ACCESS CV2
Table S1. Barriers to Data Sharing and Responses b y ACCESS CV Proposal
Key Barrier ACCESS CV Response / Proposal
Rationale Potential Benefit
Evaluation of requests for data
Requests to publications committee after primary publication Review at 24 months—through learned intermediary group
Evaluate conflicts Identify duplicate proposals Evaluate ability to carry out analysis and ensure plan for publications
Increase data use Increase engagement with requesting investigators and collaboration with trial team
Control of data access and use
Secure web link with audit trail Non-trial investigators inform ACCESS CV as soon as a paper or abstract has been accepted for publication or presentation Violation of the ACCESS CV data use agreement by a non-trial investigator may affect future availability of clinical trial data to that individual The non-trial investigator must identify all collaborators who will see the confidential data (confidential until published)
Reduce patient privacy risk Ensure structured analysis
Ensure high-quality science with reduced risk of misuse of data
2
Table S1. Barriers to Data Sharing and Responses b y ACCESS CV Proposal
Key Barrier ACCESS CV Response / Proposal
Rationale Potential Benefit
Evaluation of requests for data
Requests to publications committee after primary publication Review at 24 months—through learned intermediary group
Evaluate conflicts Identify duplicate proposals Evaluate ability to carry out analysis and ensure plan for publications
Increase data use Increase engagement with requesting investigators and collaboration with trial team
Control of data access and use
Secure web link with audit trail Non-trial investigators inform ACCESS CV as soon as a paper or abstract has been accepted for publication or presentation Violation of the ACCESS CV data use agreement by a non-trial investigator may affect future availability of clinical trial data to that individual The non-trial investigator must identify all collaborators who will see the confidential data (confidential until published)
Reduce patient privacy risk Ensure structured analysis
Ensure high-quality science with reduced risk of misuse of data
ACCESS CV proposal
ACCESS CV – Study Completion
ACCESS CV
4
primary publication data before research team can perform them
no later than 12 months post-publication
Components of current proposals
• Data sharing through different mechanisms
• Some review process
– Independent “Learned Intermediary”
– Learned Review Panel – ACCESS group proposes PI, Stats, other researchers from other studies
• Time points of sharing
• Mechanisms of sharing (secure access)
Governance and Key Takeaways
• Data Sharing – Open Science evolving
• Many Groups working engaged with ICMJE recommendations
– Compliance
– Trial Registration
• Work needed
– Governance
– Measures of Sharing Plans
Questions the Groups will need to consider
• Consolidation vs. Multiple opportunities
• Funding
• Academic Credit
• Work flow
• Patient engagement