Clinical data transparency solutions for sharing participant level - - PowerPoint PPT Presentation

• 1 -

Click to Edit Master Title Style

Clinical data transparency solutions for sharing participant level data

Barbara E. Bierer, MD

• 2 -

To improve the integrity, safety, and rigor of global clinical trials

Collaborating to Improve Multi-Regional Clinical Trials

Engage diverse stakeholders to define emerging issues in global clinical trials and to create ethical, actionable, and practical solutions.

• 3 -

MRCT Center Project Status

Training (PI competency, Protocol Ethics, Causality, Pharmacovigilence, DMC, Drug Development) Return of Results to Participants Global Regulatory Engagement Post Trial Access Data Sharing and Transparency

• 4 -

MRCT Data Sharing and Transparency Efforts

Clinical Trial Data

Study Participants

Researchers Public

Return of Plain Language summary results to participants: Guidance & Toolkit

Return of:

• Incidental findings
• Individual research results
• Study arm results
• MRCT has spearheaded two major initiatives related to sharing of clinical

trials data sharing

• Return of summary results to participants
• Data sharing and transparency
• Registration and results reporting
• Participant-level data sharing

• 5 -

The MRCT Center Deliverables

MRCT Center Return of Results Guidance for groups wishing to return results to participants, and applicable to posting of “lay summaries,” including:

• Content (essential components, source documentation, health literacy

considerations)

• Logistics and detailed processes for results sharing
• Timing
• Special considerations

MRCT Center Return of Results Toolkit including:

• Templates for Phase I, Phase II/III, studies ending early
• Neutral language guide
• Endpoints language guide
• Useful Checklists

http://mrct.globalhealth.harvard.edu/file/377016 http://mrct.globalhealth.harvard.edu/file/377001

• 6 -

MRCT Data Sharing and Transparency Efforts

Clinical Trial Data

Study Participants

Researchers Public

• MRCT has spearheaded two major initiatives related to sharing of clinical

trials data sharing

• Return of summary results to participants
• Data sharing and transparency
• Registration and results reporting
• Participant-level data sharing

• 7 -

MRCT Center fully supports trial registration and results reporting. And we are currently doing a dismal job Of ~13,300 likely applicable clinical trials ending 2008, only 13.4% of trials reported summary results within 12 months after trial completion, and 38.3% reported results at any time. Nothing we are about to describe should detract from the obligation to register and report results.

MRCT commitment to registration and results reporting

• 8 -

MRCT Center’s Focus in Data Sharing

Timeframe Milestones February 2013 Working Group launched with 18 stakeholder organizations March 2013 Convened 4 sub-groups on key issues:

• Rationales for clinical trial data sharing
• Safeguarding patient privacy, consent principles
• Balancing intellectual property interests
• Implications of patient-level data shared in public domain

May 2013 Co-hosted a conference “Issues and Case Studies in Clinical Trial Data Sharing: Lessons and Solutions”

• 9 -

Data Sharing: Common ICF and DUA

TEMPLATE ICF LANGUAGE FOR DATA SHARING DATA USE AGREEMENT TEMPLATE

In addition to efficiencies, of particular importance during public health emergencies

• 10 -

Industry efforts

• Inauguration of multiple sponsor clinical data request site at

https://www.clinicalstudydatarequest.com

Astellas, Bayer, Boehringer Ingelheim, Eisai, GSK, Lilly, Novartis, Roche, Sanofi, Takeda, UCB, ViiVHealthcare

• Johnson &Johnson and

Medtronic engage Yale University Open Access Data Project (YODA)

• Bristol-Myers Squibb aligns with Duke University (DCRI) to

expand access to its clinical trial data

• Joint Statement of EFPIA and PHRMA Principles for Responsible

Clinical Trial Data Sharing

Multiple efforts developed dedicated to sharing clinical data

• 11 -

Other efforts to support data sharing Multiple academic and private-public partnerships

Examples: Project DataSphere, All Trials Initiative (Clinical Trials)

Journal and sponsor requirements

Examples: BMJ Open Trial Campaign, Bill and Melinda Gates Foundation, NIH (e.g. BIOLincc)

Cochrane Collaboration statement on access to clinical trial data Institute of Medicine report on data sharing (2015) EU finalized regulations requiring posting of lay summary results (2016) and intention to post participant level clinical trial data for trials on medicinal products for human use

Data held in silos hinder interoperability and utility

• 12 -

Background: Data Sharing Conference: March 30-31, 2015

70 representatives of pharma, biotech, patient/patient advocates, foundations, academics, journal editors and others: Consensus on future strategic vision:

• Expectations and practices of registration and results reporting
• f all clinical trials would be regularized among industry and

academia;

• Greater access to participants-level clinical trial data could be

facilitated;

• Researchers would be able to access and combine data

across various platforms and sponsors, to multiply

• pportunities for data analysis; and
• Research participant privacy can be safeguarded
• Sponsored by the MRCT Center, LJAF and Wellcome Trust

• 13 -

Data Sharing: Future Vision

Partnering with Wellcome Trust, IOM and Deloitte Consulting to ensure collaborative, sustainable, unified approach to common platform and portal

• Organizational structure and Governance - A coordinating, centralized,

international, not-for-profit organization with accountability;

• A centralized and single portal - A central user interface with a robust search engine

functionality, including information on trials around the world;

• Data requirements – Data standards, definition, data ontology and metadata to allow

for and enable the integration of differing datasets for analysis;

• Shared or common services – Efficient shared or common services across data

generators/sponsors (e.g. policy setting, data de-identification, criteria for independent review panel decisions or reliance, and statistical services); and

• Flexibility – Data platform accommodating differing expectations and research

needs, including ability to host data for those data generators that do not wish to or cannot do so themselves, access data that is hosted elsewhere, or download data if freely available. Ability to utilize middleware to perform analyses.

• 14 -

Data Sharing: Proposed Model Platform - Draft

Researcher

Central multi- stakeholder governance

• rganization

Central repository for academics (or

• thers) who do

not wish to “host” data Provides shared services for:

• Administer

researcher requests

• Review process
• De-identification
• Setting policies
• Define standards

PORTAL: Central user interface portal with search engine building upon existing search engines (e.g. ClinicalTrials.gov and ICTRP) to pull information from registries / provide complete and robust “denominator” of existing data PLATFORM: federated platform model with optional central component enabling access to data, combining datasets and allowing downloading as appropriate

A

Sponsor A Data sets Perform feasibility checks

B

Repository B (other data sets)

Researcher

• 15 -

Data Sharing: Strategy Phase

Mission:

To develop and maintain an international non-profit entity to promote, coordinate and oversee clinical research data sharing through the creation and implementation of a sustainable global data-sharing platform that will:

• Protect study participants privacy and confidentiality
• Encourage harmonization, best practices, and regulatory convergence
• Encompass the breadth of clinical trials conducted by academia,

government and industry

• Respect and leverage existing platforms, with ability to search and interact
• Provide the ability to access or to host and to analyze data
• Ensure sustainability and efficiency

Vision:

To maximize the contribution of clinical trial participants to advance science and patient care through the sharing of participant data for further research.

Data Sharing

• 16 -

Data Sharing: Next Steps

To create a blueprint for the creation of a new, not-for-profit

• rganization whose goal is to create, direct, implement and oversee a

sustainable data-sharing platform

Partnering with Wellcome Trust, IOM, Laura and John Arnold Foundation and Deloitte Consulting to ensure collaborative, sustainable, unified approach to common platform and portal

Strategy Construction Implementation

3 Phases of Realization:

• 17 -

Data Sharing: Strategy Phase

In Strategy Phase: We have launched 3 integrated working groups to develop

• rganizational blueprint for the suggested not-for-profit entity:

MRCT Governance Working Group

Information Technology (IT) Working Group Business Models Working Group

• Define purpose, plan and scope of new entity
• Establish new steering committee
• Develop governing principles
• Transparency
• Accountability
• Data Quality and Security
• Privacy, Data Protection and Confidentiality
• Risk Benefit Analysis
• Recognition and Attribution
• Sustainability
• Education and Training
• Accessibility and Dissemination

• 18 -

Data Sharing: Strategy Phase

In Strategy Phase: We have launched 3 integrated working groups to develop

• rganizational blueprint for the suggested not-for-profit entity:

MRCT Governance Working Group

Information Technology (IT) Working Group Business Models Working Group

• Define purpose, plan and scope of new entity
• Establish new steering committee
• Develop governing principles
• Report on existing IT infrastructure
• Determine required IT specifications
• Develop global-level IT platform blueprint
• Develop sustainable business model, informed

by IT Working Group, evaluated by Governance Working Group

• Provide expertise on financial sustainability

• 19 -

Data Sharing Working Group Members

Governance Work Group IT Work Group Business Models Work Group

Co-Chairs: MRCT Center Wellcome Trust Arnold Foundation Co-Chairs: Ida Sim (UCSF) Barbara Bierer (MRCT) Co-Chairs: Wellcome Trust MRCT Center Team Members: Mark Barnes (Ropes & Gray/MRCT) Barbara Bierer (BWH/MRCT) Stuart Buck (Arnold Foundation) Nina Hill (Pfizer) Rebecca Li (MRCT) Nick Lingler (Deloitte Consulting) Justin McCarthy (Pfizer) Heather Marino (MRCT) Sandra Morris (Johnson & Johnson) Jennifer O’Callaghan (Wellcome Trust) Nicola Perrin (Wellcome Trust) Jessica Scott (GlaxoSmithKlein) Caroline Stockwell (Pfizer) Catrin Tudur Smith (University of Liverpool) Team Members: George Alter (U of Michigan) Munther Baara (Pfizer) Vivien Bonazzi (NIH) Brian Bot (Sage Bionetworks) Anne Claiborne (IOM) Khaled El Emam (U of Ottawa) Michael Khan (U of Colorado) Ghassan Karam (WHO) Rebecca Kush (CDISC) Rebecca Li (MRCT) Gene Lichtman (HCRI) Michelle Mancher (IOM) Heather Marino (MRCT) Chris Mavergames (Cochrane) Eric Perakslis (Takeda) Team Members: Barbara Bierer (MRCT) Rebecca Li (MRCT) Peter Lyons (Deloitte Consulting) Heather Marino (MRCT) Nicola Perrin (Wellcome Trust) Rohin Rajan (Deloitte Consulting)

• 20 -

Data Sharing: Endorsement and Implementation

• Conference in March 2016 in London or Boston to discuss the

blueprint for the new governance structure and IT platform for all stakeholders

• New not-for-profit entity will commence operation of directing,

implementing and overseeing broad data-sharing platform.

• Sustainable, cost-sharing business model for use of data

sharing platform will be implemented to support new entity longitudinally. This is the “how” -- not the “whether” or “what”

Other efforts should be ongoing while platform is being created.

• 21 -

Global Health Data

The Future: We Need to Look Beyond Clinical Trials

Big Data: EMR, Health Care Systems. Genomics/Omics/Imaging

Beyond Clinical Trials Data In scope are capacities to look beyond clinical trial data

Clinical Trial Data

• Interoperability
• Security, Data Protection
• Dynamic risk
• Diversity of data streams

• 22 -

Patient level data are of particular value:

• In small populations that are difficult to recruit
• Development of sufficient numbers to identify subgroups of interest
• Appreciation of natural history of disease (and perhaps using raw data to serve

as historical controls)

• Sophistication of statistical models (timing, covariates) and priors for Bayesian

analyses But the more data and detailed data one has, the greater the risk of re-identification

• And rare populations are of particular concern

Must:

• Ensure participant privacy (and IC concordance) and data protections
• Ensure quality of secondary analysis
• Protect reasonable investigator and sponsor interests

We are stewards, not owners, of the data

Risks and challenges inherent in benefits

• 23 -

Data Sharing: Work Ahead

• Clarify that and how standards apply to all research, to all data generators

and all sponsors, in all instances

• Implementation of plain language participant summaries and engagement

with participants throughout process

• Develop standards and frameworks to integrate data platforms
• Develop informed consent procedures appropriate for all data (trials,

genomics, health care data), for all trials, in all countries

• Develop privacy protections and dynamic anonymization methodologies
• Ensure enforceable legal protections in data use agreements
• Promote fair, balanced and similar standards for data users
• Develop metrics and assessment tools to answer key questions:
• Are (all) stakeholders upholding their recommended responsibilities?
• How is infrastructure, workforce, technology and sustainability working?
• Are we impacting on science and building trust?

• 24 -

Data Sharing: unscripted thoughts

• What does global health engagement mean in this context? In emergencies?
• Creation of platform(s) for data sharing before next emergencies

(and problems with PHEIC)

• Creation of agreements before health emergencies
• DUA and DCA (data contributor agreement)
• IC (including community approval of IC and of approach)

Need to engage local governments and get to “yes”

• Potential to create scientific approach before emergency
• Ring Vaccine design
• Adaptive trial design
• Empowerment of appropriate convener(s)
• Socialize concept of data stewardship
• (and does this apply to ”ownership” by country of origin)
• Need to be rigorous in differentiation between research and care (allowing for

implementation science) and between health care infrastructure, systems improvement, and research capacity

• 25 -

Click to Edit Master Title Style

Questions and Discussion