Xavier THIERRY Eucconet Bergen 15-17 June 2011 An introduction to Elfe Study.

Xavier THIERRYEucconet Bergen 15-17 June 2011

An introduction toElfe Study

CONTENTAn introduction to the Elfe study

Linkage procedures in Health thematic

Research project based on administrative health data

2

First national Birth Study in France• “ELFE” is an acronym for “Étude longitudinale

française depuis l’enfance” (French Longitudinal Study since Childhood)

• The core methodology of the project will be the enrolment and long-term follow-up of a nationally representative cohort of 20,000 children born in 2011

• The objectives of the project cover the fields of Social Sciences and Public Health, with an emphasis on Environmental Health (multidisciplinary approach)

• It is the result of the merging of two major earlier projects 3

Methodological aspects (1)

• A nationally representative sample of children (births)

• All children born in France on 24 specific days of 2011 (April, June, September and December)

• First visit (and recruitment) at maternity• a 20’ questionnaire, information from medical files

• Biological samples (maternal urines, milk, blood, hair and cord blood)

• Second one at 2 months by phone interview• Daily records on nutrition

• Environmental sensors (dust)

4

The sample of children to be included in the Elfe cohort :

A two stage random stratified plan with a comprehensive drawing in the second stage.

- first : selection of a stratified sample of maternities (primary units),

- second : inclusion of all births (secondary units) occurring the predetermined days of the survey.

Methodological aspects (2)

542 maternities identified in metropolitan France (after updating for mergers, deletions or creations in the last list available -2008 -).

Sample design was set in order to investigate around 40 500 births :

- an estimated participation of maternities of 90 % and

- 55% of mothers accepting their children to be included in the cohort = 20 000 children in the cohort

Methodological aspects (3)

Questions to answer… (examples)

In the field of Health • What are the exposure levels to physical (noise…),

chemical (tobacco smoke, pesticides…) and biological (allergens…) factors during pregnancy and early childhood and what are their long term consequences?

• What are the early origins of obesity?

In the field of Social Sciences• What are the effects of Poverty and Social Policies on

families and children?• What are the consequences of Family composition

(blended families, single parent families…), events (divorces and second marriages…) and constraints (employment of the mother, long distance commuting…) on children development and life?

7

Main topics

Social Sciences

• Demographics• Socialization-Education• Economy-

precariousness

Health- environment

• Chemical exposure• Physical exposure• Environmental

contamination

Cross-disciplinary themes

• Perinatal period • Growth

• Respiratory diseases

• Nutrition

• Psychomotor development • Cancer

Health

• Accidents & trauma• Use of health-care• Infectious diseases

8

Methodological aspects (2)

• Others steps• Phone interviews at 1 year, 2 years (with records from

physician interview)

• Face to face interview at home at 3 years with biological samples of the child (urines) and environmental sensors (dust)

• Phone interviews at 4 or 5 years

• Medical examination at 6 years

• Access to other sources:- Those of the Demographic panel (Census data, vital statistics)- Geo-matching with external data on air pollution, quality of water- Family Allowances - Social security system (RNIIAM, PMSI)

9

Elfe cohort data collection schedule

Xavier THIERRY

Linkage projects in Health thematic

Modes of recruitment

Sample directly extracted from social security system:– patients (sick persons)– persons paying social security contributions

=> Social security number is known for each participant at the beginning of the cohort

Sample extracted randomly in total population :– New born in a sample of maternity=> social security number is unknown

Characteristics of health administrative data

• A centralised database for the whole population resident in France, the so-called National Health Insurance Cross-Schemes Information System, Sniiram (for ‘Système national d’information inter-régimes de l’Assurance maladie’)

• Records on reimbursed healthcare :– medical examinations– medicine (pharmaceutical consumption) – biology and radiology– Hospitalization, surgery

• Storage duration : 3 years • Individual key of identification : Social security number

(SSN)• Children do not have their own SSN but they are linked

to those of their parents

Conditions for scientific uses • Up to recently (2004) the national statistical institute was not allowed to get

administrative health individual data • Linkages using SSN with survey was also not practiced (since the 1978

collapse…) • Using the SSN for linkage statistical purpose is now allowed under 3

conditions :– agreement of the Council of state– individual consent– agreement of the ‘Commission nationale informatique et liberté’ *

• A new attempt was made on aggregate data for the 2003 National Health Survey

– 67% adults gave their consent (N=23.000)– in some cases (10%) linkage was technically not possible (some SSN was

erroneous)• Note : health data are never transmitted associated to the SSN but only

through a anonymous number using cryptographic (or “choping”) record linkage methods

* independent authority controlling the application of the French data protection act (7 August 2004)

Linkage project in the Elfe research

• A consent has been asked to parents few days after the delivery for data to be transmitted from Social security system concerning prescriptions and treatments received by their children – 95% agreed (either mother or father or both)

• It has be decided not gathering their SSN• Instead of SSN, names, date and place

(commune) of birth of each parents have been collected in the questionnaires of the survey in maternity

Next steps1. Submission of these information (associated to a specific children

PIN) to a FIRST administration keeping a national register2. Identification of the social security number of the parents

of Elfe children within this register3. Transmission of the SSN by the FIRST administration to

a SECOND one keeping health records (then Elfe will never see the SSN)

4. Extraction of the personal health data :• for mother during pregnancy • for children during follow-up

In practice step 1 and 2 requires :- good quality of information (name, date, place of birth)

In practice step 3 and 4 will retrieve information :- for children : under the file of the father or the mother- for mother : under its own or those of its spouse/partner

Xavier THIERRY

Examples of research projects using health care expenditures

Gestational diabetes and obesity

• Long term consequences on the risk of obesity for children exposed in utero to gestational diabetes (prevalence = 6% in the first recruitment wave)– Data on reimbursements of treatment for diabetes during

pregnancy enable : • To validate self-declared cases of diabetes in maternity • To get information on insulin doses prescribed during pregnancy (as

proxy of the severity of the disease)

– Follow-up of women who had been diagnosed with gestational diabetes

• These women needs regular screening as they are at risk of developing type 2 (permanent) diabetes

• Reimbursements of biological exams and medical examinations during the follow-up could be a marker of the quality of care

• Reimbursements of an eventual treatment for diabetes would confirm the mother’s declaration on diabetes status

Consequences of the depression during pregnancy and post-partum on the child’s development

– Episodes of depression are collected through questionnaires filled-in just after delivery and 2 months latter

– Reimbursement of medical expenses for treatment for depression will prove a clinical diagnosis

– Duration of the treatment and the existence of an hospitalizations could be considered as markers for the severity of the disease

– Date of the beginning of the medical care and regularity of the examinations will be used as indicators of the quality of early diagnostic and of compliance to treatment

Asthma and allergies of children

– Reimbursement of specific drugs will serve as external validation for information self-declared by their parents

– Type of drugs and its duration, hospitalizations will be considered as markers for the severity of the disease

– Quality of the early healthcare access will be studied as a prognostic factor of the long-term evolution

Exposure of the computed tomography scan and incidence of leukaemia or cancer

– Medical exposure to Ionizing Radiation (IR) has dramatically increased over time and represents the main artificial IR exposure

– Increased use of CT scan representing 5-10% of all imaging procedures but 40-70% of the collective dose

– Estimation of the incidence of leukaemia or cancer following CT scan exposure in the cohort (Children : at risk group)

Thank you for your attention