eHealth: Exploiting our Potential

1
e-Health Exploiting our Potential

• Professor Ronan Lyons
• Centre for Health Information, Research and
  Evaluation (CHIRAL)
• Institute of Life Science
• School of Medicine
• Swansea University

2
Strategic context UK research base

• UK Office for Strategic Coordination of Health
  Research (OSCHR)
• Translational Medicine Research
• Public Health Research
• E-Health Records Research Board
• UK (England) Government Strategy for E-Health
  Research
• 200M Research Capability Programme (CfH)

3
Translational research paradigms

• Traditional perspective
• From Bench to Bedside
• Limited focus on biomedical/clinical issues
• E-Health and Public Health perspective
• From Byte to Bedside and Population Health

4
Health Information Research Unit (HIRU)

• Funded by Wales Office of Research and
  Development
• Part of the strategic infrastructure for research
  in Wales
• Large scale computing and support from IBM as
  part of 50M Institute of Life Science
  development
• Created the Secure Anonymised Information Linkage
  (SAIL) system

5
HIRU/SAIL status in UK context

• A Strategic Framework for Health Informatics in
  Support of Research
• Developed by E-health Funders Group (CRUK, CSO
  (Scotland), EPSRC, ESRC, MRC, NIHR, Welcome
  Trust, WORD)
• UK Health Research Informatics Platform
• Health Research Support Service (England)
• SAIL (Wales)
• SHIS (Scottish Health Information System)
• UK E-Health Dataset Consortium

6
The essence of HIRUs work . .

• Linking anonymised data, at individual and
  ecological levels, across multiple datasets,
  drawn from operational systems in health
  services, national databases, clinical datasets
  and databases in social care, education, housing,
  etc.

7
Steps in utilising health information for research

• Development of anonymisation and linkage
  techniques
• Building partnerships and collaboration
• Quality assessment and appraisal of datasets
• Use of datasets to support research

8
Anonymised Linkage Fields (ALFs)

• Primary ALF encrypted NHS number (HSW)
• Secondary ALF encrypted primary ALF (HIRU)
• No organisation holds both ALFs
• Household (residential) ALF (RALF)

9
Some datasets individual and ecological

• Individual level - NHS
• Population (NHSAR)
• Inpatients -PEDW (HES)
• Births
• Deaths
• Outpatients
• AE
• GP Data
• Laboratory systems
• Out of Hours Services
• Child Health Database Wales
• NHS Direct Wales
• Radiology- Imaging
• Individual level non-NHS
• Social Services
• Educational Attainment

• Clinically rich databases
• Cancer
• Screening (multiple conditions)
• Congenital Anomalies
• Myocardial Infarction
• Diabetes
• Arthropathies
• Etc.
• Ecological datasets (many are GIS)
• Census- small areas
• Ordnance Survey - Mastermap
• Environmental Health
• Government departments and agencies

10
Patient Journey Analysis- Health and Social Care
11
Types of research supported by data linkage

• Health Services Research and Epidemiology
• Clinical and Public Health Trials
• Evaluation of Policy Initiatives
• Clinical and Population Health Studies (cohorts,
  biobanks and data mining)

12
E-Health research success for Wales

• HIRU major contribution to research platform
  success for Welsh universities
• 20M won (majority not in HIRU)
• Most are multi-university collaborations
• HIRL link with IHC/industry/US universities

13
Successful research bids involving HIRU

• DECIPHer - UK Public Health Centre of Excellence
• WISERD ESRC/HEFCW
• FSBI free school breakfast initiative data
  augmentation
• Ankylosing Spondylitis Cohort
• INTEGRIS linking AE and inpatient data in
  Europe
• ALSPAC GP linkage
• MS Register (UK pilot)
• WECC - Wales Electronic Childrens Cohort
• PsyCymru - all Wales psychosis cohort
• Biomarkers - biomarker validation platform
• Caerphilly SAIL enhanced cohort mental health
• Trial support CONSTRUCT, SAFER2

14
HSR exemplar

• Are blood pressure levels taken during a
  secondary care diabetic clinic likely to be
  higher than when measured in primary care?
• Existing studies say yes, replication with
  e-data
• 661 anonymous individuals from GP and OPD in
  SAIL
• SBP 141.9 vs 137.6, plt0.0005
• 50.8 vs 36.8 not optimally controlled
• Big implications for treatment
• Brooks CJ, Tang TS, Bain SC, Ford DF, Lyons RA,
  Price DE, Stephens JW (Submitted Primary Care
  Diabetes)

15
E-Health and Trials

• E-Health can contribute to a variety of aspects
  of trial design, conduct, and efficiency
• Feasibility
• Recruitment
• Outcomes

16
Trial Feasibility

• Two factitious diabetes trials
• Drug X vs Glibenclamide as add on therapy in
  those with suboptimal control with Metformin
• Drug Z vs therapy with Metformin, Acarbose,
  Glibenclamide or Rosiglitazone in drugnaive
  patients with type 2 DM
• 250,086 individuals, including 10,205 with type2
  DM
• 284 and 711 potential participants

17
Trial Feasibility contd

• GP diagnosis only vs GP diagnosis and GP
  pathology vs GP diagnosis and Hospital pathology
• GP Dx/Path has sensitivity of 80-87 and false
  ve rate of 19-25
• Brooks CJ, Stephens JW, Price DE, Ford DV, Lyons
  RA, Prior SL, Bain SC. (Submitted Primary Care
  Diabetes).

18
Trial Recruitment

• Following on from feasibility
• Devise inclusion/exclusion algorithms in SAIL
• Develop recruitment queries in Audit and GPES
  software
• Research practices would access queries
• Test efficacy of Electronic Enhanced Recruitment
  (EER) in RCTs
• Make a standard feature across UK if
  cost-effective

19
More reliable outcome measures

• Two similar large scale (ngt3,000) Vitamin D
  trials in similar populations from residential
  homes, but very different fracture rates/power
• Lyons et al 7.4/100
• Law et al 3.9/100
• Lyons RA, Johansen A, Brophy S, Newcombe RG,
  Phillips CJ, Lervy B et al. Preventing fractures
  among older people living in institutional care
  a pragmatic randomised double blind placebo
  controlled trial of vitamin D supplementation.
  Osteoporos Int 2007 June18(6)811-8.
• Law M, Withers H, Morris J, Anderson F. Vitamin
  D supplementation and the prevention of fractures
  and falls results of a randomised trial in
  elderly people in residential accommodation. Age
  Ageing 2006 September35(5)482-6.

20
Unusual and long term outcomes

• Free School Breakfast Initiative Data
  Augmentation and Analysis - NPRI/MRC funded
• Originally a cluster RCT of breakfast vs not,
  with cognitive outcomes and attitudes towards
  breakfast as outcomes
• Now more, including relationship between
  breakfast and school performance SATS linkage
• Laurence Moore, David Benton, Ronan Lyons, Simon
  Murphy, Katy Tapper

21
Population health studies

• Electronic cohort/longitudinal studies E-Cohort
• Wales Electronic Cohort for Children (WECC)
• Electronically enhanced cohorts
• UK Biobank
• Evaluation of policy initiatives interrupted
  time series analyses
• Warm Wales Initiative
• Studies using integrated health, local government
  and environmental data (GIS)
• Advocacy in Action Study

22
Wales Electronic Cohort for Children (WECC)

• An all-Wales public health and paediatrics
  collaboration
• Ronan Lyons, David Fone, John Gallacher, David
  Ford, Caroline Brooks, Sinead Brophy, Frank
  Dunstan, Mike Gravenor, Kerina Jones, Sailesh
  Kotecha, Gareth Morgan, Shantini Paranjothy,
  Sarah Rodgers, Rhys Williams, Gareth John.

23
Anonymised datasets to be included in WECC

• NHSAR population register
• NCCHD - Community Child Health
• ONS Public Health births and deaths
• AWPS Perinatal Survey
• PEDW inpatients
• OPMD outpatients
• CCD critical care
• CARIS congenital anomalies
• GP data depending on availability
• FSM entitlement from NPD
• Environmental data (RALF and small area derived)

24
Wales Electronic Cohort for Children (WECC)

• Anonymised data from 350,000 children
• Platform for translating information into child
  health population policy answer two questions
  (of many possible)
• What factors determine the future health service
  need for individuals that are vulnerable at
  birth, and inform the development of
  interventions to reduce health inequalities in
  these groups?
• What is the influence of the social and physical
  environment on childhood obesity?

25
Green space

• Neighbourhood greenness and 2-year changes in
  body mass index of children and youth. Bell JF et
  al, Am J Prev Med 200835547-553.
• Higher greenness associates with lower odds of
  increasing BMI
• OR 0.87 (0.79-0.97)

26
Research using housing location data

• Objective Create RALFs for every residence in
  Wales
• Central list held securely within HSW
• Assign each ALF to a RALF,
• Track movements of ALFs between RALFs over time
  (never knowing who or where)
• Conduct research
• Likelihood to walk and take exercise
• Household exposures (e.g. air quality, heating,
  etc)
• Model infection disease spreads through household
  contacts
• Link survey data with environmental exposure and
  health outcomes

27
Environmental metrics and links to health and
social data fictional example
Margaret Williams 8 Main St, Swansea Female age
35 Diabetes Smoker
David Williams 8 Main St, Swansea Male age
8 Asthma -- Free School Meals
Identifiable
Social Database
Anonymous
Same environmental metric e.g. dwelling
density, land use mix, metres to bus stop or GP
surgery
Same RALF cohabiting
GP Database
28
Potential other research questions - WECC

• Wide variety possible
• Impact of health conditions on educational
  outcomes
• Birth anomalies and household infections in early
  pregnancy (antibiotic use as proxy within RALFs)
• Influence of peri-natal environmental and
  biological parameters on asthma/other disorders
• Ability of E-Cohorts to replicate findings from
  traditional cohorts e.g. MCS.

29
WECC, n123,633
30
Research platforms patient cohorts and biobanks

• Exemplars Ankylosing Spondylitis Cohort,
  PsyCymru and Biomarker Discovery
• AS Cohort MRC Stefan Siebert, Sinead Brophy,
  David Ford, Mike Gravenor, Sharon Mary Jones,
  Ronan Lyons, Ceri Phillips, Jane Skerrett
• PsyCymru NISCHR pilot Keith Lloyd, Mike Owen,
  Nick Craddoch, Richard Bentall, David Linden,
  Ronan Lyons, David Ford, John Abbott
• Biomarker Discovery NISCHR pilot John White,
  Gareth Jenkins, Paul Lewis, Shareen Doak, David
  Ford, Paul Griffiths, Christine Davies

31
The Future

• Speed up translation from data capture to
  knowledge production
• SAIL maximises the utility of routine collected
  data but
• More data needed to maximise hypothesis
  responsiveness
• Laboratory reports
• Clinical information systems
• Web enabled patient and participant data
• Infrastructure development
• Capability vs Capacity Scalable
• Training analysts complex data
• Secure remote access for analysis
• Maximise outputs publications from trickle to
  torrent

32
Research Model
33
Exploiting our Potential

• Why fund research in Wales?
• Scientific excellence (local and collaborative)
• Efficiency
• Collaborative advantage replacing competitive
  advantage
• E-Health has a big part to play

34
Key References

• Lyons RA, Jones KH, John G, Brooks CJ, Verplancke
  JP, Ford DV, Brown G, Leake K. The SAIL databank
  linking multiple health and social care datasets.
  BMC Medical Informatics and Decision Making 2009
  93. http//www.biomedcentral.com/1472-6947/9/3
• Rodgers SE, Lyons RA, Dsilva R, Jones KH, Brooks
  CJ, Ford DV, John G, Verplancke JP. Residential
  Anonymous Linking Fields (RALFs) A Novel
  Information Infrastructure to Study the
  Interaction between the Environment and
  Individuals Health. Journal of Public Health
  (doi10.1093/pubmed/fdp041).