1) The document discusses reflections on cohorts and longitudinal population studies, focusing on their strengths and weaknesses. It summarizes a survey of 77 cohort studies across 32 low and middle-income countries. 2) Key recommendations include improving data linkage, coordination between studies, use of emerging technologies, capacity building, data sharing, standardization, and translation of research outputs. 3) Barriers to effective data sharing are discussed, as well as initiatives by the Wellcome Trust to address priorities like data discoverability, incentives for data sharing, and ensuring ethical standards.

1. Reflections on Cohorts and
Longitudinal Studies
Jimmy Whitworth
Head of Population Health
Wellcome Trust
Florence
October 2014

2. TThhee WWeellllccoommee TTrruusstt
• an independent biomedical research-funding
charity
• established 1936
• funded from private endowment
• managed for long-term stability and
growth
• interests range from science to history
of medicine

3. WWeellllccoommee TTrruusstt
VViissiioonn aanndd ssttrraatteeggiicc aaiimmss
Our vision is to support research leading to extraordinary
improvements in human and animal health
Our five major challenges are:
1. Maximising the health benefits of genetics and genomics
2. Understanding the brain
3. Combating infectious disease
4. Investigating development, ageing and chronic disease
5. Connecting environment, nutrition and health.

4. The Wellcome Trust:
What do we fund?
• Health research - including public health,
health services research, clinical and
fundamental research
• Not-for-profit technology transfer activities
• Medical humanities - including biomedical
ethics and the history of medicine
• Public engagement and communication
See www.wellcome.ac.uk

5. International funding:
Henry Wellcome’s legacy
• Support over 3000 researchers in
more than 50 different countries
• Total spend approx £720 million p.a
• £103.2 million for international activities
“We will take forwards major initiatives to build research
capacity in India and Africa”
“We help to create world-class research environments in the
locations in which we fund”

6. Cohort and longitudinal studies
• Population and Public health 1990-2011,
Portfolio review published April 2013
www.wellcome.ac.uk/populationhealth
• More than a quarter of the funding allocated to cohort
studies and prospective longitudinal research
• Wellcome Trust’s key influences include ‘its committed
support to longitudinal studies, both in the UK and in low-and
middle-income countries’.

7. Longitudinal population studies
• Include:
• Cohorts
• Demographic Health Surveillance Sites
• Biobanks (those with a longitudinal component)
Long-term commitment
Strengths and weaknesses
Niche within landscape
Core resource ± Scientific studies

8. Landscape analyses
• MRC Strategic Review of the Largest UK Population
Cohort Studies. 2014
Maximising the value of UK population cohorts
Covers 33 UK cohorts
• We have conducted a similar survey of cohort studies
in LMICs

9. Survey of LMIC cohorts
• Covers 77 studies, 107 sites in 32 countries.
• Over 50 different funders: national agencies, NGOs,
private companies, charities, academic institutions
• Mainly Africa and Asia, but generally good coverage
• Central Asia and Middle East poorly represented
• Large numbers in India, South Africa, Brazil, China
• Few urban
• Few mental health or injuries
• Few linked to other sources of data, other studies, little
harmonisation or standardisation, few provide open
access to data

10. Distribution of LMIC LPS sites by Research Area

11. Geographical spread and research focus of
longitudinal population studies

12. Recommendations (1)
Data Linkage: to routine datasets,
microbiology and hospital records
Co-ordination: networking between
studies
Emerging Technologies: standardised
techniques, development of best
practice guidelines
Capacity-strengthening: interrogating
large and complex datasets,
infomatics, data management

13. Recommendations (2)
Data-sharing and discoverability:
searchable directories, data sharing
policies, access to samples.
Standardization: common data
standards, improving meta-data quality
Cost-efficiency: new digital technologies
and remote data and sample capture
Translation: stakeholder engagement to
increase the impact of research
outputs and potential for translation

14. As health research funders, we want…
• To support high quality health
research that is:
– timely and widely used
– transparent and available to others
– quickly translated into better policies and
better health
• To minimise duplication
• To get maximum “bang for buck”
– data useful both locally and more widely
– data answer several questions
– data endure for (now unforeseen) future
uses

15. But the current situation for public
health research is often that…
• data analysis and reporting are slow
and incomplete
• lack of access to data limits the
capacity for comparative analysis or
even checking for accuracy
• only immediate research interests
are addressed; broader questions
are neglected
• lots of duplication in data collection
• data are not archived effectively for
long-term preservation and use

16. There are significant barriers…
• Infrastructural - building & sustaining
infrastructures for long-term data
storage and curation
• Cultural – incentives and recognition
for researchers who share their data
• Technical - developing data
standards, metadata, platforms for
inter-operability
• Professional – providing training and
career paths for data managers
• Ethical - protecting the confidentiality
of research participants

17. • researchers doing crucial work to
generate public health research
datasets in often difficult
circumstances will lose out to better-resourced
researchers elsewhere,
who have the skills and tools to
rapidly analyze the data
• increased data sharing will create
unacceptable risks for research
participants
• data sharing carries a substantial
cost in terms of money and time
17
…And particular concerns over health
data in resource poor-settings

18. The funder position – a growing
consensus
• major challenges associated with
increasingly large & complex datasets,
but also tremendous opportunities
• policy convergence in promoting
sharing of research data
- expectation that data outputs be preserved
and shared in a way that maximises value
- requirement for data management plans as
integral part of the application process
• but building the resources, culture and
governance to enable data sharing will
require us to work in partnership

19. Working in partnership – some of our
current initiatives
• Expert Advisory Group on Data
Access – advice on emerging issues
relating to data access across genetics,
epidemiology and social sciences
• Clinical Trial Data – funded programme
of work with IOM and seeking to build an
international consortium to support
access to clinical trial data
• Public Heath Research Data Forum –
global cross-funder initiative to increase
access to research data generated by
public health and epidemiology research

20. Key priorities 1: Data discoverability
• Report to examine potential models
for enabling data discoverability
published July 2014
• Key conclusion: this is a solvable
challenge – with exemplars in other
fields that could be applied
• To address this it is critical that we:
– Link up existing pockets of activity
– Embed good practice in metadata
generation and use
– Address related issues in relation to
incentives
• Funders considering next steps –
welcome feedback from community
http://www.wellcome.ac.uk/About-us/Policy/Spotlight-issues/Data-sharing/Public-health-and-epidemiology/WTP054675.htm

21. Key priorities 2: Incentives
• Report by Expert Advisory Group
on Data Access (May 2014)
• Research culture and environment
not perceived to provide sufficient
support or rewards for data sharing
• Funders, institutions, publishers
and researchers have key roles
• Need to ensure that data sharing is:
– planned and resourced appropriately
– tracked and cited routinely
– more explicitly recognised
– supported by specialist expertise and
resources
http://www.wellcome.ac.uk/About-us/Policy/Spotlight-issues/Data-sharing/EAGDA/WTP056496.htm

22. Key priorities 3: the ‘ethical imperative’
• Major ongoing research project to
gather the views and expectations of
research participants and
stakeholders on data sharing
• Work is being conducted in
partnership with research teams in
five countries (South Africa, Kenya,
Vietnam, Thailand & India)
• Aim to identify good practice and
inform appropriate policy
development
• Report to be launched in March 2015