Showcasing ALSPAC as a resource for social and health researchers Andy Boyd The Royal Statistical...
-
Upload
kimberly-short -
Category
Documents
-
view
215 -
download
0
Transcript of Showcasing ALSPAC as a resource for social and health researchers Andy Boyd The Royal Statistical...
Showcasing ALSPAC as a resource for social and health
researchers
Andy Boyd
The Royal Statistical Society
4th April 2011
ALSPAC
The Avon Longitudinal Study of
Parents and Children
A brief history & introduction
2
ALSPAC: Study History• Established by Professor Jean Golding• Initial funding from the Department for the Environment• Need identified at WHO Europe conference in Moscow
for a series of pan-European cohorts with comparable design and data collection tools - ELSPAC
• Known to its participants as ‘Children of the 90s’
“To determine ways in which the individual’s genotype combines with environmental pressures
to influence health and development”Golding 2001
3
ALSPAC: British Birth Cohorts
• 1946 ‘National Birth Cohort’ (NBC)
• 1958 ‘National Child Development Survey’ (NCDS)
• 1970 ‘British Birth Cohort’ (BCS70)
• ALSPAC 1991-1992
• ‘Millennium Cohort Study’ (MCS)
• 2012 Birth Cohort Facility
4
ALSPAC: defining characteristics
• Regional catchment area– Permanent study center in Bristol
• Multi-generational• Health and genetics ‘reputation’
– Extensive social measures
5
ALSPAC: defining characteristics
• Extensive biobank• Intensive and frequent follow-up• Used as a sampling frame• Expanding record linkage arrangements
6
ALSPAC: Eligibility
• Pregnant women resident in Avon (excluding the city of Bath) with an expected date of delivery between 1st April 1991 and 31st December 1992
7
ALSPAC: Sample size
• Core participating sample of 14,541 pregnancies resulting in 14,062 live born children
• ~ 8500 young adults participated between ages 16-18 • 10 000 children attended at least one clinic• Outreach clinics are helping to boost numbers and
target young people from socially deprived areas
8
ALSPAC: The resource
• Questionnaires• Hands-on Clinical Assessments• Biological Samples• DNA & Genotyping• Record Linkage to routine information• Future Collections
9
ALSPAC: Questionnaires• 17 questionnaires about the mother• 23 mother completed about the child• 22 questionnaires completed by the child• 15 questionnaires completed by the partner• Many other single topic or sub-sample questionnaires
– 10 Puberty questionnaires between ages 8 to 18– Web based data collection– Questionnaires administered in schools
10
ALSPAC: Clinical Assessments
• ‘Children in Focus’– 10% sub sample– 10 clinics between 4 – 61 months
• Focus Clinics– Open to all eligible study children– 9 clinics from age 7 – 17
11
ALSPAC: Clinical Assessments
• Mothers clinics– Opportunistic data collection at child focus clinics– First mothers clinic running from 2009-2011– Funding secured for two further waves
• Fathers clinics– Some opportunistic data collection– Funding secured for first fathers clinic
12
ALSPAC: Biological Samples
• Collected since pregnancy– Blood– Urine– Hair– Nails– Teeth– Saliva– Placenta
13
ALSPAC: DNA & Genotyping
• DNA extracted for over 10,000 children and mothers• Cell lines produced for ~7,000 children and 6,000
mothers• GWAS, expression data and shortly whole genome
scans on a sub sample of 1,000 – 3,000 cases
14
ALSPAC: Record Linkage • Linkage to health and administrative routinely collected
records
– Primary Care Maternity and Birth records
– NHS/ONS Flagging and Tracing service
• Death notification & Cancer registration
– Education records
• National Pupil Database (NPD) census and attainment records
– GIS & Environmental Measures
– ALSPAC Friendship Matrix
15
ALSPAC: Future Collections
• Continuation of data collection from the ALSPAC index children and their mothers
• 3rd Generation ‘Offspring’ pilot• Fathers recruitment & first clinic• Sibling recruitment• Development of the biobank and genotype resource• Expanded data linkage to routine records
16
ALSPAC: Principal Findings• Verified the safety of the ‘Back to Sleep’ campaign that led to the
reduction in rates of cot death• Findings led to a reformulation of topical creams to remove
peanut oil• Policy Impact
– Changed US government advice re eating fish during pregnancy
– Evidence of lack of change in social mobility• Genetic & Epigenetic research
– Helped identify common genetic variants that relate to traits such as obesity
– Data used in exploration of new field of Epigenetics
17
ALSPAC: Data Access
• ALSPAC committed to moving towards ‘open access’ solution– UK Data Archive - pilot data sets– MRC DSS - ALPSAC meta data
• Current access arrangements detailed on ALSPAC web site:
www.bristol.ac.uk/alspac/sci-com/collab-policy/
18
ALSPAC: further information
• www.bristol.ac.uk/alspac– Detailed summaries of the resource– Data access policy– Links to further information
19
20
PEARL
Project to Enhance ALSPAC through Record Linkage
21
PEARL: The PEARL Team• A Wellcome Trust funded project
– Part of the Electronic Patient Record (EPR) series of grants
• PI: John Macleod• Other team members:
– Andy Boyd– Kerry Humphries– Kate Angel– Lindsey Brown
22
PEARL: Benefits of Linkage
• Cost effective• Comprehensive source of data• Obtain data that is less subject to self-report or
participation bias• Inform strategies for dealing with missing
observations• May help avoid study fatigue
23
PEARL: Project Goals
• Obtain consent for and establish mechanisms of linkage
• Investigate challenges and develop generalisable solutions
• Demonstrate the value of linkage-based research through exemplar projects
• Establish a training programme to share these methods and insights with other researchers
24
PEARL: Data Sources• Health data
– Patient level primary care records– General Practice Research Database (GPRD)– Hospital admissions data (HES)
• Work and Pensions Longitudinal Study (WPLS)– Employer, benefits and income data
• Ministry of Justice– Criminal convictions and cautions records
• GIS (Geographic Information System)– Data to inform spatial analyses
• Education data– National Pupil Database (NPD), Further and Higher Education
25
PEARL: Governance Structures
26
ALSPAC NHS Other Data bodies owners
Executive
NIGBSection 251
REC
AL&EC
ICO ONS
DWP
DfE & BIS
HMRC
MoJ
PEARL: Research Governance
• Concerns:– Privacy & Cohort acceptability
• Trust, duty of care• maintaining the long term relationship
– Research ethics– Legislation
• Data Protection Act • Data owners
• Balance:– Right of privacy against right of public goods
27
PEARL: Pseudonymisation
• Anonymisation– Is anonymisation possible in this context?
• Pseudonymisation– Removal of personal identifiers
• Restrict precision of geographical scale, date of birth
– Assign new unique key number• To each participant• To schools/employers/health facilities etc• Suppress/Transform small cell counts
28
PEARL: Pseudonymisation in practice
• Balance between privacy and utility• Deductive disclosure still possible
• Does pseudonymisation meet the requirements of the
DPA / data owners?
29
PEARL: ALSPAC Linkage Protocol
• ALSPAC will continue to Pseudonymise data• ALSPAC will continue to mitigate risk through data usage
agreements with data users• ALSPAC will seek consent & meet fair processing requirements • Governance & infrastructure to control for security risk
– ISO27001 & HMG Security Policy Framework– Staff training
• Investigating governance frameworks & accreditation– NHS Information Governance Framework and ‘Data Safe
Havens’• Investigate technological solutions
– DataSHIELD, SAIL
30
EUCCONET Data Linkage Workshop
• Bergen 15-17th June
• Focus on linkage in Child Cohorts– Linkage theory– Governance and disclosure control– Consent– Exemplar projects