Record linkage in Birth cohort Biobanks

Record linkage in Birth cohort Biobanks

Carel Thijs, Monique Mommers

Maastricht University

Biolink WP4 3 June, 2013

EU Birth cohorts

77 birth cohorts in 21 EU countries

Opportunities• meta-epidemiological studies: power & variation• case-control studies within cohorts with biobanking• record linkage: enrichment and future follow-up• systems biology

Biolink (BBMRI.nl) Challenges• maintenance of biobanks and cohorts• data management and quality control• harmonisation and interoperability• ethical, legal and societal implications (ELSI) • funding, governance, and publication policy

Collaboration

biobanking: total and specific IgEharmonised definitions of

eczema, airway allergy, wheeze, asthma

blood sampled at age

PIAMA

n=

PREVASC

n=

KOALA

n=

total

n=

1 y 370 230 700 1.300

2 y - 360 700 1.060

6-8 y 750 220 600 1.470

Gene-environment interactions (ALLERGENIC)

Aims for WP4 – birth cohorts

To establish opportunities for • long-term follow-up (on health related outcomes)• keep follow-up rate high • substitute missing data (selective missings!) • harmonize methods and definitions• combine medical data from different source • establish validity of medical data• assess utility• guidelines and access

Case study WP4 – birth cohorts

• validity and utility of medical data• case study ‘asthma’: harmonized definition

in last 12 months: asthma complaints (ISAAC) OR inhalant medication use

AND

ever: doctor’s diagnosis of asthma OR inhalant medication prescription

Case study WP4 – birth cohorts

Linkage with medication prescriptions (SFK)

•validation of linkage (WP3)•WP4: validation of medical data WP4

cross-validation1.parent reported: doctor’s diagnosis and medication

2.SFK: medication prescriptions

3.GP (and specialists): diagnosis and medication

Case study WP4 – birth cohorts

Linkage with medication prescriptions (SFK)

•validation of linkage (WP3)•cross-validation of medical data (WP4)

1. parent reported: doctor’s diagnosis and medication

2. SFK: medication prescriptions

3. GP (and specialists): diagnosis and medication

Case study WP4 – birth cohorts

Linkage with medication prescriptions (SFK)

•validation of linkage (WP3)•cross-validation of medical data (WP4)

1. parent reported: doctor’s diagnosis and medication

2. SFK: medication prescriptions

3. GP (and specialists): diagnosis and medication

•utility: what do SFK data add? (WP4)

Case study WP4 – birth cohortsPriority list of medicines and medical data

•related to asthma as outcome– asthma inhalation medication (also used for wheeze..)– antihistamine medication (airway allergy)– other drugs for airway allergy– topical corticosteroids (eczema) and other topicals

•related to etiology, comorbidity– antibiotics– ADHD medication (common underlying exposures)

in last 12 months: asthma complaints (ISAAC) OR inhalant medication use

AND

ever: doctor’s diagnosis of asthma OR inhalant medication prescription

Challenges

Time & space issues; e.g. medication history

–repeated parental questionnaires - cover calendar periods

–biosample collection - time points

–contacts with care - date of prescription vs start of use and continuation

- different pharmacies

–start of registrations - SFK gradually building up in time

–national coverage and regions

–linkage - delays (cohort data; registries)

Challenges

Person issues; e.g. doctor’s diagnosis

–identification of child – confusion sibs..

–repeated parental questionnaires – father or mother?

–contacts with care - which doctor? GP, specialist, other..- start of symptoms, time of diagnosis- end of disease episode?

–linkage - delays in processing of cohort data, registration, data retrieval

Workplan WP4 – birth cohorts• June 2013

– identity check GBA (started)– data retrieval GPs (started for selected asthma cases)– preparation for pilot (pharmacies, IC; completed)

– proposal submission SFK (in preparation)

• Autumn 2013 – pilot (WP3)– adapted proposal for WP4

– asthma inhalation medication (also used for wheeze..)– antihistamine medication (airway allergy)– other drugs for airway allergy– topical corticosteroids (eczema) and other topicals

• related to etiology, comorbidity– antibiotics– ADHD medication (common underlying exposures)

in last 12 months: asthma complaints (ISAAC) OR inhalant medication use

AND

ever: doctor’s diagnosis of asthma OR inhalant medication prescription

Workplan WP4 – birth cohorts

• 2014 – final linkage with SFK, check by WP3– data retrieval from GPs and pharmacies when needed,

e.g. for discordant medication histories or diagnoses– cross-validation and utility– reporting: concept white paper; article

• Other birth cohorts– LucKi has obtained informed consent (limited region)– NTR birth cohort – PIAMA, PREVASC have no informed consent (WP5)

in last 12 months: asthma complaints (ISAAC) OR inhalant medication use

AND

ever: doctor’s diagnosis of asthma OR inhalant medication prescription