SlideShare a Scribd company logo
Digital transformation to enable a
FAIR approach for health data science
ConTech Pharma 2022
Varsha Khodiyar, PhD
1st March 2022
What I’ll cover
• Introduction to Health Data Research UK and our work
• Infrastructure for data discovery (Findability)
• Facilitating access to data (Accessibility)
• Improving data linkage (Interoperability)
• Demonstrating data reuse (Reusability)
• Public patient partnerships (considerations beyond FAIR)
HDR UK is the national institute for health data science
Our mission is to unite the UK’s health data to enable
discoveries that improve people’s lives.
Our 20-year vision is for large scale data and advanced
analytics to benefit every patient interaction, clinical trial,
and biomedical discovery.
| 3
Working in partnership with patients, the national health service (NHS),
universities, business & charities to create a world-leading and robust
health data infrastructure
| 4
Gearing up the UK for quality health
care, research and innovation
Enabling data science and innovation
as a catalyst for change
Facilitating standardization of data
access and information governance
| 5
The Alliance: network of
experts committed to define
best practice in the use of
data
Hubs and DSCs: Data
science centres and
research hubs that
provide services of
data curation and
access
Research: Data-
driven research
through our research
sites and promote
open science
Training for
researchers
through our HDR
UK Future
platform
The Gateway:
platform for
discovery of and
access to UK data
and samples
available for
research
Our strategy focuses on uniting,
improving and using health data
Infrastructure for data discovery
Findability
The Innovation Gateway for health data discovery
| 7
“Really impressed with this resource. I think as a gateway to search by data type and indication, it’s a really powerful tool.“
David Leather, GSK
✓ Search datasets by keywords/publisher or by themed
collection.
✓ For some datasets: a Cohort Discovery Tool enables
search population data based on characteristics required.
✓ Find summary and technical metadata for datasets.
✓ Transparency on requirements for accessing data.
✓ Request access directly via the Gateway, user-friendly
end-to-end data access request management system.
✓ Upload your papers and tools and create your own
thematic collection.
The Gateway for Researchers
www.healthdatagateway.org/about/innovators-and-researchers
Searching for data in the Gateway
• Wide-range of health datasets from
cardiovascular, to maternal health,
primary care, mental health to
diabetes and of course COVID-19.
• Biobank physical samples and
tissues also listed.
• Range of disciplines: e.g. emergency
care, cardiovascular, maternal
health, cancer
• Types: primary care, secondary care,
acute care, palliative care, research
cohorts
• Purpose of initial data collection:
clinical care (40%), clinical trials
(38%), administration (34%)
Covering all four nations of
the UK and discoverable via a
shared ‘front door’
49%
19%
17%
9%
7%
England Wales
Scotland NI
International
Cohort Discovery on the Innovation Gateway
Co-vars
Inc. Crit
Exc. Crit
Custodian
controlled
Cohort Query
Agents
Statistical Disclosure Control Policies:
• User validation (e.g. Bona-fide
Researcher)
• Low number suppression (e.g. >50)
• Query Count binning (e.g. 50, 60, 70, -)
• Query Rate limiting
• Researchers can reuse the cohort query to define their research protocol when
submitting their data access request
• Researchers will be able to reuse and compare cohort definitions between similar
protocols
• Cohort definitions will be able to reuse phenotype definitions (asthma, diabetes)
without the need for ICD-10, Read, SNOMED-CT codes
Cohort Discovery enables researchers to discover, assess and request access to
potential datasets that exactly match the research project cohort definition
using standardized inclusion & exclusion criteria and co-variates
Datasets with
female
patients
between 18-35
who have
asthma and
diabetes and
who are not
smokers and
not pregnant
Innovation Gateway
Query Engine
Dataset 1: 20K
Dataset 2: 3K
Dataset 3: 1K
…
Total Patients:
23K
Demo: https://www.youtube.com/watch?v=L50nqIR6k98
Facilitating access to data
Accessibility
The Gateway for Data custodians
✓ Facilitate discoverability and accessibility of your data resources.
✓ Implement best practice for health data governance and sharing.
✓Cohort Discovery tool to improve discovery of your consented
patient cohorts.
✓ Create collections of datasets and samples to increase findability of
your resources.
✓ Provide data access in a trustworthy manner, using the Five Safes
framework.
✓ Fully digital system to manage your data access requests
Gain insight into data reusability prior to requesting access
| 13
Technical details & metadata wheel Data utility
Enabling a common, transparent approach to data access management
based around the Five Safes framework
| 14
1. Safe People
e.g. Approved researchers scheme
2. Safe Projects
Project proposal reviewed by governance board
convened by the data custodian
3. Safe Settings
Data access provided within a Trusted Research
Environment
4. Safe Data
Data supplied as deidentified to minimize identification
of individuals
5. Safe Output
Analysis outputs checked by data custodians prior to
release
Aligning approaches to evaluation of public benefit in data access
request applications
The National Data Guardian guidance on public benefit will be published later in the year
➢ How will you
demonstrate that
use of data
requested will
deliver public
benefit?
➢ Do you anticipate
any risks to
individuals, and if
so, what steps have
you made in your
proposal to mitigate
the risks?
Improving data integration
Interoperability
Towards an interoperable infrastructure: Federation – Metadata Catalogues,
Cohort Discovery & Trusted Research Environments
Query Engine
Federated Data Custodian Environments (No Direct Access)
Cohort
Discovery
Link
Agents
Federated
Analysis
Federated Trusted Research Environments
Common
API
Trusted
Research
Environments
Federated Metadata Catalogues
A TRE is a highly secure environment used for saving, accessing and analysing health and care data. Within any
TRE there are multiple layers of security and safeguards in place, to minimise the risk of data being misused.
| 18
Data Utility Matrix
International links
Data Quality Tool Evaluation
Data Officers Groups / Hubs
Position/consultation papers
Projects (Eg COVID-19)
Metadata completeness
SIGs (FHIR, Synthetic data,etc)
Gateway development and streamlined access management are underpinned by harmonisation of
principles and best practice as driven by the Alliance
Developing and implementing community agreed standards to
maximise interoperability
Improving Data Utility: produce guidance and policy regarding data quality,
standards, dataset publishing, data provenance, ontology / terminology
services and use. Output: data utility framework
Promoting participation and improving access: Ensure consistency and
harmonise principles, definitions, process steps to develop a best practice for
assessing requests to access UK health data. Output: harmonised five safe DAR
Enabling FAIR data
Aligning approaches to trusted research environments: secure analytics
environments. Output: recommendation for a TRE standard
Improving transparency in data use: understand interests and values of people
in health data. Output: recommendations for a data use register standard
Demonstrating data use
Reusability
| 21
| 21
• A data use register is a record showing when and why a dataset has been shared with another organisation.
• Data uses are not always made public. There is also a lack of standardisation for data use registers (content,
functionality and purpose).
Data Use Registers
By establishing a core set of standards for Data Use
Registers, we hope to…
•develop a culture of openness amongst health data custodians
•increase transparency in the use of health data for research and
innovation
• HDR UK Alliance members agree to transparency of governance and
operations
•build public trust by demonstrating the value and benefit of using
health data
• National Data Guardian and Understanding Patient Data highlight that
‘transparency cannot be separated from public benefit’
•generate insights into how health data are used and accessed | 22
- Public contributor-
“Being clear how data is
used, is a vital step to
ensure the whole process is
meaningful and trusted, in
terms of outcomes, cost
effectiveness and public
trust.”
Developing the Data Use Register standard
Community involvement
• Collaborative approach with input from public and lay representatives,
data custodians, researchers, policy makers and funders (more than 100
people and 50 orgs contributed to the standard)
Analysis of 48 health data custodian entities - May 2021 green paper
• Nearly 50% of data custodians do not publish information about data use
Public consultation on green paper - July 2021
• Recommendations supported by 93% of respondents
Recommendations for Data Use Registers - Jan 2022 white paper
• Recommendations presenting a minimum standard for data use registers
Development of a Gateway data use register - ongoing
• Implementation of the Data Use Register recommendations on HDR UK’s
Innovation Gateway Data use register standards white paper
doi.org/10.5281/zenodo.5902743
Inclusion of patient and lay public groups
for fairer health data
Moving beyond ‘FAIR’
| 25
I am glad you're involving me from what
seems to be the beginning so that you can
actually take my concerns and address
them whilst helping the greater good
Patient / Public Voice Rep
“ It is essential that the public is included in
this ground-breaking work.
Margaret Rogers
Member of the HDR UK Public
Advisory Board
“
92 people consulted to inform decisions on
methods and process for clinical trial
recruitment
Consultation on COVID work across 7 UK-
wide patient and public networks with
168 responses.
16,500 contacts with patient & public
contributors across the institute in 2020
Strong Public Advisory Board providing
strategic guidance on all our work
HDR UK exemplifies working with public and
patients on all aspects of health data science
Webinars for the lay public on health data science
| 26
www.hdruk.ac.uk/events/data-access-discovery
Summary
• HDR UK is uniting the UK’s health data access for researchers.
• The Innovation Gateway facilitates discovery and access to health data
• Increasing the FAIRness of data, by increasing the F, A, I of (meta)data, and indicating the
R of data.
• Demonstration of the importance of patient and lay public involvement in all aspects of
health data research.
• HDR UK is laying the foundations for an international alliance to enable a data-centric
response to COVID-19 and other health challenges
| 27
Thank you
Varsha Khodiyar, PhD
@varsha_khodiyar
Varsha.Khodiyar@hdruk.ac.uk
Health Data Research UK
www.hdruk.ac.uk
@HDR_UK
Visit the Gateway
www.healthdatagateway.org
Links and resources
Health Data Research UK: https://www.hdruk.ac.uk/
UK Health Data Research Alliance: https://ukhealthdata.org/
The Health Data Research innovation Gateway: https://www.healthdatagateway.org/
Data use register standards green paper: https://zenodo.org/record/5084761#.YZq_t9DP02w
HDR UK Futures: https://www.hdruk.ac.uk/careers-in-health-data-science/continued-professional-
development/power-up-your-health-data-science-knowledge/
Working in a Trusted Research Environment: https://www.youtube.com/watch?v=8CBy7mGFcEY
The importance of public and patient engagement and involvement:
https://www.youtube.com/watch?v=tustdIQu4ZQ
An introduction to the Innovation Gateway: https://www.youtube.com/watch?v=U7mZ7cNE_KQ

More Related Content

What's hot (20)

PPTX
Lessons from Journal Research Data Policy Registry Pilot
Jisc RDM
 
PPTX
Rscd 2018 Journal policies - natasha simons
ARDC
 
PPTX
Journal Data Sharing Policies rscd2018
SusanMRob
 
PDF
NHS SE presentation
Jisc
 
PPTX
Open access progress and sustainability
Jisc
 
PPTX
Grampian safe haven, research data network
Jisc RDM
 
PPTX
Recognising data sharing
Jisc RDM
 
PPTX
Stop press: should embargo conditions apply to metadata?
Jisc RDM
 
PPTX
Towards a frictionless data future
Jisc RDM
 
PPTX
DAF Survey Results, research data network
Jisc RDM
 
PPTX
State of open research data open con
Amye Kenall
 
PPTX
H2020 open-data-pilot
Sarah Jones
 
PPTX
Why does research data matter to libraries
Jisc RDM
 
PPTX
Open Science Globally: Some Developments/Dr Simon Hodson
African Open Science Platform
 
PDF
Research Data Management, Challenges and Tools - Per Öster
LEARN Project
 
PPTX
How to elaborate a data management plan
Biblioteca de la Universitat Jaume I
 
PPTX
International perspective for sharing publicly funded medical research data
ARDC
 
PPTX
Has anyone seen my data? Incentivising #opendata sharing with altmetrics
Nick Sheppard
 
PDF
E research17 journal data policies - Natasha Simons and Kate LemMay
ARDC
 
PPTX
Methods for measuring citizen-science impact
Luigi Ceccaroni
 
Lessons from Journal Research Data Policy Registry Pilot
Jisc RDM
 
Rscd 2018 Journal policies - natasha simons
ARDC
 
Journal Data Sharing Policies rscd2018
SusanMRob
 
NHS SE presentation
Jisc
 
Open access progress and sustainability
Jisc
 
Grampian safe haven, research data network
Jisc RDM
 
Recognising data sharing
Jisc RDM
 
Stop press: should embargo conditions apply to metadata?
Jisc RDM
 
Towards a frictionless data future
Jisc RDM
 
DAF Survey Results, research data network
Jisc RDM
 
State of open research data open con
Amye Kenall
 
H2020 open-data-pilot
Sarah Jones
 
Why does research data matter to libraries
Jisc RDM
 
Open Science Globally: Some Developments/Dr Simon Hodson
African Open Science Platform
 
Research Data Management, Challenges and Tools - Per Öster
LEARN Project
 
How to elaborate a data management plan
Biblioteca de la Universitat Jaume I
 
International perspective for sharing publicly funded medical research data
ARDC
 
Has anyone seen my data? Incentivising #opendata sharing with altmetrics
Nick Sheppard
 
E research17 journal data policies - Natasha Simons and Kate LemMay
ARDC
 
Methods for measuring citizen-science impact
Luigi Ceccaroni
 

Similar to Digital transformation to enable a FAIR approach for health data science (20)

PPTX
Data sharing: Seeing & Thinking Together
Communication and Media Studies, Carleton University
 
PPTX
dkNET Webinar: Creating and Sustaining a FAIR Biomedical Data Ecosystem 10/09...
dkNET
 
PDF
Rachel Bruce UK research and data management where are we now
Jisc
 
PPTX
The Role of Data Lakes in Healthcare
Perficient, Inc.
 
PDF
Professor Andrew Morris
DHA2015
 
PPTX
Health Data Sharing Scene Setting
ipposi
 
PDF
Barbara Bierer, "Clinical Trial Data Sharing"
The Petrie-Flom Center for Health Law Policy, Biotechnology, and Bioethics
 
PDF
Samantha Robertson - NHMRC Perspectives on Increasing Access to Data from Pub...
Wiley
 
PDF
HRB_Data_Summary_FINAL (2)
Rosalyn Moran
 
PDF
Building a National Data Infrastructure to Advance Patient-Centered Comparati...
Patient-Centered Outcomes Research Institute
 
PPTX
Open FAIR Data and Open Science: Developing Partnerships, Strategies, Policie...
Academy of Science of South Africa (ASSAf)
 
PPTX
BioSHaRE: Operationalizing responsible data sharing and access: GA4GH - Barth...
Lisette Giepmans
 
PPTX
Turning FAIR into Reality: Briefing on the EC’s report on FAIR data
dri_ireland
 
PDF
Trusted! Quest for data-driven and fair health solutions
Sitra / Hyvinvointi
 
PPTX
Data Management, Research Integrity and Ethics
ARDC
 
PPTX
Susanna Sansone at the Knowledge Dialogues/ODHK "Beyond Open"event
GigaScience, BGI Hong Kong
 
PPTX
Safti net kick off 12092010-mrs
Marion Sills
 
PPTX
NCI Cancer Genomics, Open Science and PMI: FAIR
Warren Kibbe
 
PDF
Sharing and standards christopher hart - clinical innovation and partnering...
Christopher Hart
 
Data sharing: Seeing & Thinking Together
Communication and Media Studies, Carleton University
 
dkNET Webinar: Creating and Sustaining a FAIR Biomedical Data Ecosystem 10/09...
dkNET
 
Rachel Bruce UK research and data management where are we now
Jisc
 
The Role of Data Lakes in Healthcare
Perficient, Inc.
 
Professor Andrew Morris
DHA2015
 
Health Data Sharing Scene Setting
ipposi
 
Samantha Robertson - NHMRC Perspectives on Increasing Access to Data from Pub...
Wiley
 
HRB_Data_Summary_FINAL (2)
Rosalyn Moran
 
Building a National Data Infrastructure to Advance Patient-Centered Comparati...
Patient-Centered Outcomes Research Institute
 
Open FAIR Data and Open Science: Developing Partnerships, Strategies, Policie...
Academy of Science of South Africa (ASSAf)
 
BioSHaRE: Operationalizing responsible data sharing and access: GA4GH - Barth...
Lisette Giepmans
 
Turning FAIR into Reality: Briefing on the EC’s report on FAIR data
dri_ireland
 
Trusted! Quest for data-driven and fair health solutions
Sitra / Hyvinvointi
 
Data Management, Research Integrity and Ethics
ARDC
 
Susanna Sansone at the Knowledge Dialogues/ODHK "Beyond Open"event
GigaScience, BGI Hong Kong
 
Safti net kick off 12092010-mrs
Marion Sills
 
NCI Cancer Genomics, Open Science and PMI: FAIR
Warren Kibbe
 
Sharing and standards christopher hart - clinical innovation and partnering...
Christopher Hart
 
Ad

More from Varsha Khodiyar (20)

PDF
COVID-19 variants, vaccines and tests
Varsha Khodiyar
 
PDF
COVID-19 variants and vaccines
Varsha Khodiyar
 
PDF
Data citation and sharing during article publication
Varsha Khodiyar
 
PDF
The importance of research data repositories
Varsha Khodiyar
 
PDF
What role can publishers play in the open data ecosystem?
Varsha Khodiyar
 
PDF
Five essentials factors for unlocking the potential for Open Research Data
Varsha Khodiyar
 
PPTX
The value of data curation as part of the publishing process
Varsha Khodiyar
 
PDF
Preparing your data for sharing and publishing
Varsha Khodiyar
 
PDF
Facilitating good research data management practice as part of scholarly publ...
Varsha Khodiyar
 
PDF
Practical challenges for researchers in data sharing
Varsha Khodiyar
 
PDF
Update from Data policy standardisation and implementation IG
Varsha Khodiyar
 
PPTX
The challenge of sharing data well, how publishers can help
Varsha Khodiyar
 
PPTX
Data peer review workshop
Varsha Khodiyar
 
PDF
Peer Reviewing Data: experiences from a data journal
Varsha Khodiyar
 
PPTX
Data Publishing and Institutional Repositories
Varsha Khodiyar
 
PDF
Gaining credit for sharing research data
Varsha Khodiyar
 
PDF
Data sharing as part of the research workflow
Varsha Khodiyar
 
PDF
Data sharing as part of the research ecosystem
Varsha Khodiyar
 
PPTX
Workflows for Publishing Data; Scientific Data's experience as an early adopter
Varsha Khodiyar
 
PPTX
Clinical Data Publishing at Scientific Data
Varsha Khodiyar
 
COVID-19 variants, vaccines and tests
Varsha Khodiyar
 
COVID-19 variants and vaccines
Varsha Khodiyar
 
Data citation and sharing during article publication
Varsha Khodiyar
 
The importance of research data repositories
Varsha Khodiyar
 
What role can publishers play in the open data ecosystem?
Varsha Khodiyar
 
Five essentials factors for unlocking the potential for Open Research Data
Varsha Khodiyar
 
The value of data curation as part of the publishing process
Varsha Khodiyar
 
Preparing your data for sharing and publishing
Varsha Khodiyar
 
Facilitating good research data management practice as part of scholarly publ...
Varsha Khodiyar
 
Practical challenges for researchers in data sharing
Varsha Khodiyar
 
Update from Data policy standardisation and implementation IG
Varsha Khodiyar
 
The challenge of sharing data well, how publishers can help
Varsha Khodiyar
 
Data peer review workshop
Varsha Khodiyar
 
Peer Reviewing Data: experiences from a data journal
Varsha Khodiyar
 
Data Publishing and Institutional Repositories
Varsha Khodiyar
 
Gaining credit for sharing research data
Varsha Khodiyar
 
Data sharing as part of the research workflow
Varsha Khodiyar
 
Data sharing as part of the research ecosystem
Varsha Khodiyar
 
Workflows for Publishing Data; Scientific Data's experience as an early adopter
Varsha Khodiyar
 
Clinical Data Publishing at Scientific Data
Varsha Khodiyar
 
Ad

Recently uploaded (20)

PPTX
Q1 - W1 - D2 - Models of matter for science.pptx
RyanCudal3
 
PDF
Continuous Model-Based Engineering of Software-Intensive Systems: Approaches,...
Hugo Bruneliere
 
PPTX
Diuretic Medicinal Chemistry II Unit II.pptx
Dhanashri Dupade
 
PDF
Annual report 2024 - Inria - English version.pdf
Inria
 
PDF
Carbon-richDustInjectedintotheInterstellarMediumbyGalacticWCBinaries Survives...
Sérgio Sacani
 
PPTX
formations-of-rock-layers-grade 11_.pptx
GraceSarte
 
PPTX
MICROBIOLOGY PART-1 INTRODUCTION .pptx
Mohit Kumar
 
PDF
The-Origin- of -Metazoa-vertebrates .ppt
S.B.P.G. COLLEGE BARAGAON VARANASI
 
PDF
GK_GS One Liner For Competitive Exam.pdf
abhi01nm
 
PPTX
Diagnostic Features of Common Oral Ulcerative Lesions.pptx
Dr Palak borade
 
PPTX
Structure and uses of DDT, Saccharin..pptx
harsimrankaur204
 
PPTX
PEDIA IDS IN A GIST_6488b6b5-3152-4a4a-a943-20a56efddd43 (2).pptx
tdas83504
 
PDF
Step-by-Step Guide: How mRNA Vaccines Works
TECNIC
 
PDF
Phosphates reveal high pH ocean water on Enceladus
Sérgio Sacani
 
PPTX
Animal Reproductive Behaviors Quiz Presentation in Maroon Brown Flat Graphic ...
LynetteGaniron1
 
PDF
High-speedBouldersandtheDebrisFieldinDARTEjecta
Sérgio Sacani
 
PDF
Adding Geochemistry To Understand Recharge Areas - Kinney County, Texas - Jim...
Texas Alliance of Groundwater Districts
 
PDF
RODENT PEST MANAGEMENT-converted-compressed.pdf
S.B.P.G. COLLEGE BARAGAON VARANASI
 
PDF
A proposed mechanism for the formation of protocell-like structures on Titan
Sérgio Sacani
 
PDF
Pharma Part 1.pdf #pharmacology #pharmacology
hikmatyt01
 
Q1 - W1 - D2 - Models of matter for science.pptx
RyanCudal3
 
Continuous Model-Based Engineering of Software-Intensive Systems: Approaches,...
Hugo Bruneliere
 
Diuretic Medicinal Chemistry II Unit II.pptx
Dhanashri Dupade
 
Annual report 2024 - Inria - English version.pdf
Inria
 
Carbon-richDustInjectedintotheInterstellarMediumbyGalacticWCBinaries Survives...
Sérgio Sacani
 
formations-of-rock-layers-grade 11_.pptx
GraceSarte
 
MICROBIOLOGY PART-1 INTRODUCTION .pptx
Mohit Kumar
 
The-Origin- of -Metazoa-vertebrates .ppt
S.B.P.G. COLLEGE BARAGAON VARANASI
 
GK_GS One Liner For Competitive Exam.pdf
abhi01nm
 
Diagnostic Features of Common Oral Ulcerative Lesions.pptx
Dr Palak borade
 
Structure and uses of DDT, Saccharin..pptx
harsimrankaur204
 
PEDIA IDS IN A GIST_6488b6b5-3152-4a4a-a943-20a56efddd43 (2).pptx
tdas83504
 
Step-by-Step Guide: How mRNA Vaccines Works
TECNIC
 
Phosphates reveal high pH ocean water on Enceladus
Sérgio Sacani
 
Animal Reproductive Behaviors Quiz Presentation in Maroon Brown Flat Graphic ...
LynetteGaniron1
 
High-speedBouldersandtheDebrisFieldinDARTEjecta
Sérgio Sacani
 
Adding Geochemistry To Understand Recharge Areas - Kinney County, Texas - Jim...
Texas Alliance of Groundwater Districts
 
RODENT PEST MANAGEMENT-converted-compressed.pdf
S.B.P.G. COLLEGE BARAGAON VARANASI
 
A proposed mechanism for the formation of protocell-like structures on Titan
Sérgio Sacani
 
Pharma Part 1.pdf #pharmacology #pharmacology
hikmatyt01
 

Digital transformation to enable a FAIR approach for health data science

  • 1. Digital transformation to enable a FAIR approach for health data science ConTech Pharma 2022 Varsha Khodiyar, PhD 1st March 2022
  • 2. What I’ll cover • Introduction to Health Data Research UK and our work • Infrastructure for data discovery (Findability) • Facilitating access to data (Accessibility) • Improving data linkage (Interoperability) • Demonstrating data reuse (Reusability) • Public patient partnerships (considerations beyond FAIR)
  • 3. HDR UK is the national institute for health data science Our mission is to unite the UK’s health data to enable discoveries that improve people’s lives. Our 20-year vision is for large scale data and advanced analytics to benefit every patient interaction, clinical trial, and biomedical discovery. | 3
  • 4. Working in partnership with patients, the national health service (NHS), universities, business & charities to create a world-leading and robust health data infrastructure | 4 Gearing up the UK for quality health care, research and innovation Enabling data science and innovation as a catalyst for change Facilitating standardization of data access and information governance
  • 5. | 5 The Alliance: network of experts committed to define best practice in the use of data Hubs and DSCs: Data science centres and research hubs that provide services of data curation and access Research: Data- driven research through our research sites and promote open science Training for researchers through our HDR UK Future platform The Gateway: platform for discovery of and access to UK data and samples available for research Our strategy focuses on uniting, improving and using health data
  • 6. Infrastructure for data discovery Findability
  • 7. The Innovation Gateway for health data discovery | 7 “Really impressed with this resource. I think as a gateway to search by data type and indication, it’s a really powerful tool.“ David Leather, GSK
  • 8. ✓ Search datasets by keywords/publisher or by themed collection. ✓ For some datasets: a Cohort Discovery Tool enables search population data based on characteristics required. ✓ Find summary and technical metadata for datasets. ✓ Transparency on requirements for accessing data. ✓ Request access directly via the Gateway, user-friendly end-to-end data access request management system. ✓ Upload your papers and tools and create your own thematic collection. The Gateway for Researchers www.healthdatagateway.org/about/innovators-and-researchers
  • 9. Searching for data in the Gateway • Wide-range of health datasets from cardiovascular, to maternal health, primary care, mental health to diabetes and of course COVID-19. • Biobank physical samples and tissues also listed. • Range of disciplines: e.g. emergency care, cardiovascular, maternal health, cancer • Types: primary care, secondary care, acute care, palliative care, research cohorts • Purpose of initial data collection: clinical care (40%), clinical trials (38%), administration (34%) Covering all four nations of the UK and discoverable via a shared ‘front door’ 49% 19% 17% 9% 7% England Wales Scotland NI International
  • 10. Cohort Discovery on the Innovation Gateway Co-vars Inc. Crit Exc. Crit Custodian controlled Cohort Query Agents Statistical Disclosure Control Policies: • User validation (e.g. Bona-fide Researcher) • Low number suppression (e.g. >50) • Query Count binning (e.g. 50, 60, 70, -) • Query Rate limiting • Researchers can reuse the cohort query to define their research protocol when submitting their data access request • Researchers will be able to reuse and compare cohort definitions between similar protocols • Cohort definitions will be able to reuse phenotype definitions (asthma, diabetes) without the need for ICD-10, Read, SNOMED-CT codes Cohort Discovery enables researchers to discover, assess and request access to potential datasets that exactly match the research project cohort definition using standardized inclusion & exclusion criteria and co-variates Datasets with female patients between 18-35 who have asthma and diabetes and who are not smokers and not pregnant Innovation Gateway Query Engine Dataset 1: 20K Dataset 2: 3K Dataset 3: 1K … Total Patients: 23K Demo: https://www.youtube.com/watch?v=L50nqIR6k98
  • 11. Facilitating access to data Accessibility
  • 12. The Gateway for Data custodians ✓ Facilitate discoverability and accessibility of your data resources. ✓ Implement best practice for health data governance and sharing. ✓Cohort Discovery tool to improve discovery of your consented patient cohorts. ✓ Create collections of datasets and samples to increase findability of your resources. ✓ Provide data access in a trustworthy manner, using the Five Safes framework. ✓ Fully digital system to manage your data access requests
  • 13. Gain insight into data reusability prior to requesting access | 13 Technical details & metadata wheel Data utility
  • 14. Enabling a common, transparent approach to data access management based around the Five Safes framework | 14 1. Safe People e.g. Approved researchers scheme 2. Safe Projects Project proposal reviewed by governance board convened by the data custodian 3. Safe Settings Data access provided within a Trusted Research Environment 4. Safe Data Data supplied as deidentified to minimize identification of individuals 5. Safe Output Analysis outputs checked by data custodians prior to release
  • 15. Aligning approaches to evaluation of public benefit in data access request applications The National Data Guardian guidance on public benefit will be published later in the year ➢ How will you demonstrate that use of data requested will deliver public benefit? ➢ Do you anticipate any risks to individuals, and if so, what steps have you made in your proposal to mitigate the risks?
  • 17. Towards an interoperable infrastructure: Federation – Metadata Catalogues, Cohort Discovery & Trusted Research Environments Query Engine Federated Data Custodian Environments (No Direct Access) Cohort Discovery Link Agents Federated Analysis Federated Trusted Research Environments Common API Trusted Research Environments Federated Metadata Catalogues A TRE is a highly secure environment used for saving, accessing and analysing health and care data. Within any TRE there are multiple layers of security and safeguards in place, to minimise the risk of data being misused.
  • 18. | 18 Data Utility Matrix International links Data Quality Tool Evaluation Data Officers Groups / Hubs Position/consultation papers Projects (Eg COVID-19) Metadata completeness SIGs (FHIR, Synthetic data,etc) Gateway development and streamlined access management are underpinned by harmonisation of principles and best practice as driven by the Alliance Developing and implementing community agreed standards to maximise interoperability
  • 19. Improving Data Utility: produce guidance and policy regarding data quality, standards, dataset publishing, data provenance, ontology / terminology services and use. Output: data utility framework Promoting participation and improving access: Ensure consistency and harmonise principles, definitions, process steps to develop a best practice for assessing requests to access UK health data. Output: harmonised five safe DAR Enabling FAIR data Aligning approaches to trusted research environments: secure analytics environments. Output: recommendation for a TRE standard Improving transparency in data use: understand interests and values of people in health data. Output: recommendations for a data use register standard
  • 21. | 21 | 21 • A data use register is a record showing when and why a dataset has been shared with another organisation. • Data uses are not always made public. There is also a lack of standardisation for data use registers (content, functionality and purpose). Data Use Registers
  • 22. By establishing a core set of standards for Data Use Registers, we hope to… •develop a culture of openness amongst health data custodians •increase transparency in the use of health data for research and innovation • HDR UK Alliance members agree to transparency of governance and operations •build public trust by demonstrating the value and benefit of using health data • National Data Guardian and Understanding Patient Data highlight that ‘transparency cannot be separated from public benefit’ •generate insights into how health data are used and accessed | 22 - Public contributor- “Being clear how data is used, is a vital step to ensure the whole process is meaningful and trusted, in terms of outcomes, cost effectiveness and public trust.”
  • 23. Developing the Data Use Register standard Community involvement • Collaborative approach with input from public and lay representatives, data custodians, researchers, policy makers and funders (more than 100 people and 50 orgs contributed to the standard) Analysis of 48 health data custodian entities - May 2021 green paper • Nearly 50% of data custodians do not publish information about data use Public consultation on green paper - July 2021 • Recommendations supported by 93% of respondents Recommendations for Data Use Registers - Jan 2022 white paper • Recommendations presenting a minimum standard for data use registers Development of a Gateway data use register - ongoing • Implementation of the Data Use Register recommendations on HDR UK’s Innovation Gateway Data use register standards white paper doi.org/10.5281/zenodo.5902743
  • 24. Inclusion of patient and lay public groups for fairer health data Moving beyond ‘FAIR’
  • 25. | 25 I am glad you're involving me from what seems to be the beginning so that you can actually take my concerns and address them whilst helping the greater good Patient / Public Voice Rep “ It is essential that the public is included in this ground-breaking work. Margaret Rogers Member of the HDR UK Public Advisory Board “ 92 people consulted to inform decisions on methods and process for clinical trial recruitment Consultation on COVID work across 7 UK- wide patient and public networks with 168 responses. 16,500 contacts with patient & public contributors across the institute in 2020 Strong Public Advisory Board providing strategic guidance on all our work HDR UK exemplifies working with public and patients on all aspects of health data science
  • 26. Webinars for the lay public on health data science | 26 www.hdruk.ac.uk/events/data-access-discovery
  • 27. Summary • HDR UK is uniting the UK’s health data access for researchers. • The Innovation Gateway facilitates discovery and access to health data • Increasing the FAIRness of data, by increasing the F, A, I of (meta)data, and indicating the R of data. • Demonstration of the importance of patient and lay public involvement in all aspects of health data research. • HDR UK is laying the foundations for an international alliance to enable a data-centric response to COVID-19 and other health challenges | 27
  • 28. Thank you Varsha Khodiyar, PhD @varsha_khodiyar [email protected] Health Data Research UK www.hdruk.ac.uk @HDR_UK Visit the Gateway www.healthdatagateway.org
  • 29. Links and resources Health Data Research UK: https://www.hdruk.ac.uk/ UK Health Data Research Alliance: https://ukhealthdata.org/ The Health Data Research innovation Gateway: https://www.healthdatagateway.org/ Data use register standards green paper: https://zenodo.org/record/5084761#.YZq_t9DP02w HDR UK Futures: https://www.hdruk.ac.uk/careers-in-health-data-science/continued-professional- development/power-up-your-health-data-science-knowledge/ Working in a Trusted Research Environment: https://www.youtube.com/watch?v=8CBy7mGFcEY The importance of public and patient engagement and involvement: https://www.youtube.com/watch?v=tustdIQu4ZQ An introduction to the Innovation Gateway: https://www.youtube.com/watch?v=U7mZ7cNE_KQ