“'Safe havens' should be developed as an
environment for population-based research
in which the risk of identifying individuals is
minimised… researchers are bound by a
strict code, preventing disclosure of any
personally identifying information, and
providing sanctions in case of breach of
confidentiality.”
(Thomas and Walport, Data Sharing Review
Report 2008)
DaSH: Staff
• All staff employed by UoA
• Posts funded by UoA/NHSG R&D
• All staff have NHSG honorary contracts
• All staff have valid GCP training
• All staff have valid Information Governance
Training
Why Safe Haven
• 2008 - Data Sharing Review Report, Thomas and Walport
• 2011 - Blueprint for Health Records Research in Scotland, Scottish Health
Informatics Programme
• 2012 - Data linkage Framework for Statistics and Research, Scottish
Government
• 2013 - Information: to share or not to share? The Information
Governance Review, Dame Fiona Caldicott
• 1 of 5 Safe Havens
• Joint venture with NHS Grampian
• Federated network
• Facilitate sharing/access of data
Grampian
DaSH
HIC
ISD: eDRIS
National Safe
Haven
GGC
Safe Haven Lothian
Safe Haven
Grampian Data Safe Haven
• Established in May 2012
• Over 175 projects
• ‘Virtual’ access & ‘Farr’ Researcher room
• Team of analysts, research coordinators and
an administrator
• Safe setting
• Safe people
• Safe projects
• Safe outputs
Risk to patient
confidentiality
Opportunities for
linkage research
Managing risk and enabling health benefit
Safe Haven: For unconsented linked data
• Identifiable data is stored on NHS
Servers
Safe storage, linkage and analysis of data
Safe Setting:
• Identifiable data is stored on NHS
Servers
• Only trained ‘approved’ analysts can
access identifiable information
• A different analyst links the data
• No patient level data leaves the safe haven
Safe storage, linkage and analysis of data
• Linked datasets are stored on
separate servers
• Access to linked data restricted
Safe Setting:
• Researchers cannot access patient
identifiers
• No patient level data leaves the
safe haven
Access to linked data for approved researchers only
• Only named and trained ‘approved’
researchers can access linked data sets
• Researchers analyse anonymous
linked data
• Penalties for misconduct
• Investigators sign a Declaration to
adhere to working practices
Safe People:
• All appropriate approvals in place:
• Sponsorship
• Public Benefit and Privacy
Panel for Health and Social
Care
• Research Ethics
• NHS R&D
• Caldicott Guardian
• Data Custodian
All study approvals in place
• Clear data management plan
Safe Projects:
• Research outputs are released
after disclosure checks
No individual patient level data is released
• No individual patient level data
can be printed, copied or removed
from the safe haven
Safe Outputs:
Benefits for researchers
Secure, managed access to data for research
• Support for data linkage research
• Support for Data Management Plans and permissions
DLP and DMP
Transparency
Benefits for researchers
Secure, managed access to data for research
• Support for data linkage research
• Support for Data Management Plans and permissions
• Remote access desktop
• Safe Room ‘Farr Scotland Researcher Lab’
• Physical safe room - Dumb terminal access
• Secure Pods
• Farr Scotland Institute
Benefits for data custodians
Secure, managed access to data for research
• Data stays on site
• Support for data linkage research
• Datasets for multiple use
• Data sharing for collaboration
• Sensitive data
• Large extracts
• Data release policy e.g. small numbers
What have we learntThe Good
• Virtual environment
• Improved governance
• Allowed projects to go ahead
• Got researchers ‘thinking’
• DMPs – better data management
• Removed ‘old school’ methods
• Improved accountability
• Enabled collaboration
• Developed working relationships/partnerships
What have we learntThe Good
• Virtual environment
• Improved governance
• Allowed projects to go ahead
• Got researchers ‘thinking’
• DMPs – better data management
• Removed ‘old school’ methods
• Improved accountability
• Enabled collaboration
• Developed working relationships/partnerships
• Improved Communication
• Improved linkage specifications
• Timely checks
The Bad
• Cultural change
And The Ugly
• Software issues
• Capacity: Compute & People
• Disclosure checking
The future
Accreditation
HPC
Consented data
• Genomics
• Proportionate review
Acknowledgements
Professor Corri Black, Clinical Lead
Dr Sharon Gordon, Research Coordinator
Ms Steph Hall, Research Coordinator
Mr Alasdair Coutts, Data Programmer/Analyst
Mr Alastair Soutar, Data Programmer/Analyst
Mr Artur Wozniak, Data Programmer/Analyst
Mrs Pam Rebecca, Administrator
Mr Derek Turner, Administrator
UOA IT Services
NHSG Health Intelligence
NHSG eHealth
NHSG R&D Office
Any questions?
Transforming the world
with greater knowledge
and learning

Grampian safe haven, research data network

  • 1.
    “'Safe havens' shouldbe developed as an environment for population-based research in which the risk of identifying individuals is minimised… researchers are bound by a strict code, preventing disclosure of any personally identifying information, and providing sanctions in case of breach of confidentiality.” (Thomas and Walport, Data Sharing Review Report 2008)
  • 2.
    DaSH: Staff • Allstaff employed by UoA • Posts funded by UoA/NHSG R&D • All staff have NHSG honorary contracts • All staff have valid GCP training • All staff have valid Information Governance Training Why Safe Haven • 2008 - Data Sharing Review Report, Thomas and Walport • 2011 - Blueprint for Health Records Research in Scotland, Scottish Health Informatics Programme • 2012 - Data linkage Framework for Statistics and Research, Scottish Government • 2013 - Information: to share or not to share? The Information Governance Review, Dame Fiona Caldicott
  • 3.
    • 1 of5 Safe Havens • Joint venture with NHS Grampian • Federated network • Facilitate sharing/access of data Grampian DaSH HIC ISD: eDRIS National Safe Haven GGC Safe Haven Lothian Safe Haven
  • 4.
    Grampian Data SafeHaven • Established in May 2012 • Over 175 projects • ‘Virtual’ access & ‘Farr’ Researcher room • Team of analysts, research coordinators and an administrator
  • 5.
    • Safe setting •Safe people • Safe projects • Safe outputs Risk to patient confidentiality Opportunities for linkage research Managing risk and enabling health benefit Safe Haven: For unconsented linked data
  • 6.
    • Identifiable datais stored on NHS Servers Safe storage, linkage and analysis of data Safe Setting:
  • 8.
    • Identifiable datais stored on NHS Servers • Only trained ‘approved’ analysts can access identifiable information • A different analyst links the data • No patient level data leaves the safe haven Safe storage, linkage and analysis of data • Linked datasets are stored on separate servers • Access to linked data restricted Safe Setting:
  • 9.
    • Researchers cannotaccess patient identifiers • No patient level data leaves the safe haven Access to linked data for approved researchers only • Only named and trained ‘approved’ researchers can access linked data sets • Researchers analyse anonymous linked data • Penalties for misconduct • Investigators sign a Declaration to adhere to working practices Safe People:
  • 10.
    • All appropriateapprovals in place: • Sponsorship • Public Benefit and Privacy Panel for Health and Social Care • Research Ethics • NHS R&D • Caldicott Guardian • Data Custodian All study approvals in place • Clear data management plan Safe Projects:
  • 11.
    • Research outputsare released after disclosure checks No individual patient level data is released • No individual patient level data can be printed, copied or removed from the safe haven Safe Outputs:
  • 12.
    Benefits for researchers Secure,managed access to data for research • Support for data linkage research • Support for Data Management Plans and permissions
  • 13.
  • 14.
    Benefits for researchers Secure,managed access to data for research • Support for data linkage research • Support for Data Management Plans and permissions • Remote access desktop • Safe Room ‘Farr Scotland Researcher Lab’ • Physical safe room - Dumb terminal access • Secure Pods • Farr Scotland Institute
  • 15.
    Benefits for datacustodians Secure, managed access to data for research • Data stays on site • Support for data linkage research • Datasets for multiple use • Data sharing for collaboration • Sensitive data • Large extracts • Data release policy e.g. small numbers
  • 16.
    What have welearntThe Good • Virtual environment • Improved governance • Allowed projects to go ahead • Got researchers ‘thinking’ • DMPs – better data management • Removed ‘old school’ methods • Improved accountability • Enabled collaboration • Developed working relationships/partnerships
  • 18.
    What have welearntThe Good • Virtual environment • Improved governance • Allowed projects to go ahead • Got researchers ‘thinking’ • DMPs – better data management • Removed ‘old school’ methods • Improved accountability • Enabled collaboration • Developed working relationships/partnerships • Improved Communication • Improved linkage specifications • Timely checks
  • 19.
    The Bad • Culturalchange And The Ugly • Software issues • Capacity: Compute & People • Disclosure checking
  • 20.
    The future Accreditation HPC Consented data •Genomics • Proportionate review
  • 21.
    Acknowledgements Professor Corri Black,Clinical Lead Dr Sharon Gordon, Research Coordinator Ms Steph Hall, Research Coordinator Mr Alasdair Coutts, Data Programmer/Analyst Mr Alastair Soutar, Data Programmer/Analyst Mr Artur Wozniak, Data Programmer/Analyst Mrs Pam Rebecca, Administrator Mr Derek Turner, Administrator UOA IT Services NHSG Health Intelligence NHSG eHealth NHSG R&D Office
  • 22.
  • 23.
    Transforming the world withgreater knowledge and learning