Linking administrative records into MCS Millennium Cohort Study - - PowerPoint PPT Presentation

Linking administrative records into MCS

Millennium Cohort Study

MCS 1 MCS2 MCS 3 MCS4 MCS5

|

2001/2 2003/4 2005/6 2008/9 2012/13 9 months AGE 3 AGE 5 AGE 7 AGE11

|

Mother Mother Mother Mother Mother

|

Father Father Father Father Father

|

Child Child Child Child Older Siblings Older Siblings Teachers Teachers Education records Education records Birth records Medical records Medical records Medical records Economic records  18,552 15,590 15,246 13,857 13287

MCS6

Launching 2015

?

Requested consent from participants to link with their administrative:

• Health
• Economic
• Education

records

During data collections

• In 2008 (MCS4)
• 14042 participants : 13169 parents consented
• Until age 16

Education

Education: The data

England Wales Scotland

• N. Ireland

No of consents 8448 1898 1536 1288 No of cases sent for linkage 8445 1890 1536 n/a

• No. of cases

successfully linked 7476 1890 (1735) 1407 n/a % of cases successfully linked 88.5% 100% (92%) 92% n/a

• Test and exam results, prior attainment and progression at different key stages

for pupils in the state sector

• Attainment data for students in non-maintained special schools, sixth-form and

further education colleges

• Information on pupils in independent schools, where available

The database also includes information about pupils’ characteristics, such as:

• Gender
• Ethnicity
• First language
• Eligibility for free school meals
• Special educational needs (SEN)
• Pupil absence and exclusions
• KS1 and KS2 data are

deposited

• School census data to

become available longitudinally

England: The national pupil database (NPD)

The database includes information about pupils’ characteristics, such as:

• Gender
• Ethnicity
• First language
• Eligibility for free school meals
• Special educational needs status

Pupils is Scotland Census

Welsh Pupil Level Annual Schools Census and Pupil Attainment dataset.

• Test and exam results, prior attainment and progression at different key stages

for pupils in the state sector

• Attainment data for students in non-maintained special schools, sixth-form and

further education colleges

• Information on pupils in independent schools, where available

The database also includes information about pupils’ characteristics, such as:

• Gender
• Ethnicity
• First language
• Eligibility for free school meals
• Special educational needs status

• MCS 2008
• 14042 participants 13047 parents consented
• Until age 14

Health

Using health record linkage in the UK MCS to investigate childhood obesity, asthma and infections.

• enhance the cohort data by linking longitudinal survey

data with routine hospital admission records.

• examine the quality, completeness and, consistency of

key child health data by comparing information from these sources with parent/carer report.

Wellcome Trust funded project

Hospital Episode Statistics (linked to ONS mortality data).

• Inpatient data
• Outpatient data
• A&E
• Augmented care
• Maternity records

In negotiations

• On a study by study basis
• Piloting with MCS

Health England: HSCIC

APPROVED IN PRICIPLE!

• Inpatient and day cases
• Outpatients
• Maternity
• Mental Health
• Cancer Registration
• Neonatal
• Child Health Surveillance Programme (pre-

school/school) and Scottish Immunisation & Recall System

• Prescribing Information

Health Scotland: NHS National Services Scotland

APPROVED for MCS. Data available soon.

• Patient Episode Database for Wales (PEDW)
• Inpatient data
• Outpatient data
• National Community Child Health Database
• Birth
• Health examinations
• Immunisations

Health Wales:

• 2008 MCS
• 11063 parents consented
• 9906 NI numbers given

Economic

• HMRC
• Employment
• Pensions
• Earnings
• Tax Credits

HMRC

• Criminal justice
• MCS (2018)
• Ministry of Justice: Police National Computer

Future plans

• More complete picture of participants life history.
• External validation
• Self-reporting
• No recollection bias

The good

• Data errors.
• Linkage errors.
• Missing records.
• Consent rates.
• Administrative data are not collected for research

purposes.

• Differences in data entry requirements.
• Differences in data quality.

The ugly