EUROCAT: Surveillance of Environmental Impact

1
EUROCAT Surveillance of Environmental Impact

• Dr Martine Vrijheid (Lyon, France) for
• EUROCAT
• Prof Helen Dolk, University of Ulster
• WHO Collaborating Centre for the Epidemiological
  Surveillance of Congenital Anomalies
• Supported by the EU-Commission Public Health
  Directorate Programme of Community Action on Rare
  Diseases

2
EUROCAT Surveillance of environmental impact

• What is EUROCAT?
• Environmental causes of congenital anomalies
• What is surveillance?
• Approaches to surveillance of environmental
  impacts in causation of congenital anomalies
• Routine monitoring
• Detection and response to temporal and spatial
  clusters
• Evaluation of specific hypotheses
• Conclusions

3
What is EUROCAT?

• European network of registries for the
  epidemiologic surveillance of congenital
  anomalies.
• Started in 1979
• More than 1 million births surveyed per year in
  Europe (25 of European birth population)
• 40 registries in 19 countries of Europe
• Standardised database on gt350,000 cases of
  congenital anomaly among livebirths, stillbirths
  and terminations of pregnancy since 1980
  updated annually

4
EUROCAT Registries

• European network of population-based registries
  for the epidemiologic surveillance of congenital
  anomalies.
• Started in 1979
• More than 1 million births surveyed per year in
  Europe
• 40 registries in 19 countries of Europe
• Standardised database on gt350,000 cases of
  congenital anomaly among livebirths, stillbirths
  and terminations of pregnancy since 1980
  updated annually

5
EUROCAT

• Objectives
• To provide essential epidemiologic information on
  congenital anomalies in Europe
• To facilitate the early warning of teratogenic
  exposures
• To act as an information and resource centre for
  the population and health professionals regarding
  clusters or exposures or risk factors of concern
• To provide a ready collaborative network and
  infrastructure for research related to the causes
  and prevention of congenital anomalies

6
EUROCAT

• Registries provide high data quality
• Population-based
• Multiple sources of ascertainment
• Follow-up cases for diagnostic information
• Ascertain anomalies diagnosed after the early
  neonatal period
• Ascertain terminations of pregnancy following
  prenatal diagnosis
• Quality at the expense of completeness of
  geographical coverage

7
Environmental causes Potential adverse
outcomes of embryonic/fetal exposure

• Congenital anomalies (2-4 of births)
• congenital heart disease 25
• limb anomaly 17
• central nervous system anomaly
• (e.g. neural tube defects incl spina bifida)
  9
• cleft palate or cleft lip 6
• chromosomal syndrome (incl Down Syndrome)
  12
• Spontaneous abortion (15 of recognized
  pregnancies) or stillbirth (lt1 of births)
• Low birthweight (6-7 of births) preterm and/or
  growth retarded in utero
• Poor neurodevelopmental outcomes, including
  motor, cognitive and behavioural outcomes
• Effects later in childhood/ adult life (cancer,
  fertility, etc)

8
Environmental causes of congenital anomalies 3
important principles

• Susceptibility to an environmental agent (and
  type of effect) depends on the developmental
  stage at the time of exposure (sensitive
  period)
• Before pregnancy recognized (later for other
  outcomes)
• The effect of an environmental agent increases in
  frequency and degree as dosage increases and
  there may be a practical threshold dose
• Susceptibility to an environmental agent depends
  on the background of other genetic and
  environmental factors

9
Environmental pollution as a cause of congenital
anomalies?

• Disasters/accidental or deliberate high exposure
  incidents
• Radiation Hiroshima microcephaly, mental
  retardation
• Methylmercury Minamata cerebral palsy
• PCB contaminated oil Taiwan, Japan skin, nail,
  teeth, growth, neurodevelopment
• Chronic exposures
• Water disinfection/chlorination byproducts?
• Residence near (hazardous) waste landfill sites??
• Endocrine disrupting chemicals??
• Pesticides?
• Dolk Vrijheid 2004
• EUROCAT Special Report A review of Environmental
  Risk Factors for Congenital Anomalies
  www.eurocat.ulster.ac.uk/pubdata

10
Public Health Surveillance

• The continuous scrutiny of the distribution of
  disease in a population in order to take and
  evaluate control measures.
• As opposed to epidemiologic research, tends to be
  (Thacker and Berkelman 1988)
• Problem detection/hypothesis generating rather
  than hypothesis testing
• Ongoing, using routine data collection systems,
  with minimal often incomplete data as opposed to
  time limited, tailored, extensive and complete
  data collection
• Simple and descriptive analysis with timely and
  targeted communication to agencies involved with
  policy and intervention, as opposed to complex
  analysis with sporadic communication to clinical
  or academic audience

11
EUROCAT Approaches to Surveillance

• Routine monitoring of temporal trends
• Detection and response to temporal and spatial
  clusters
• Evaluation of specific hypotheses

12
1. Routine monitoring

• EUROCAT Reports / web
• Effects of folic acid supplementation on trends
  in neural tube defects throughout Europe
• Reports of rising prevalence of hypospadias
• Increasing prevalence of gastroschisis.

13
Trends in the prevalence of congenital anomalies
per 10,000 births all and cardiac anomalies, all
EUROCAT registries combined, 1980-2002
14
NTD Prevalence per 10,000 births 1980-2001 (LB
SB TOP)
UK and Ireland Continental Europe
15
Hypospadias Prevalence Rates by Year of Birth in
England and Wales, 1964-1996 (NCAS Data)
16
Gastroschisis per 10,000 births, all European
registers combined, 1980-2002
17
2. Detection and response to temporal and spatial
clusters

• Clusters arising through surveillance
• Cluster concerns from local community
• Routine statistical monitoring to detect temporal
  clusters carried out by EUROCAT
• Empirical investigations
• geographic heterogeneity of congenital anomalies
  in the UK
• Geographical clustering of anophthalmia in
  England
• EUROCAT Cluster Advisory Service
  www.eurocat.ulster.ac.uk.

18
Clusters under random pattern
19
Routine Statistical Monitoring

• Statistical tests for detection of clusters Eg.
  Kulldorf techniques, adjusting for multiple
  testing scan moving window
• Would routine detection of clusters overwhelm our
  capacity to respond?
• Main problems are
• not random variation, but ascertainment variation
• Lack of sensitivity in picking up exposure
  related variation

20
Empirical clustering observations

• Anophthalmia in England (Dolk 98)
• 444 cases born 1988-94, prevalence 1.0 per 10,000
  births
• Clustering consistent with random distribution
• Geographical variation in congenital anomalies in
  Britain (Dolk 2003)
• 5 regions over 9 year period
• 6,959 non-chromosomal anomalies 1611 Down
  syndrome cases.
• Clustering consistent with random variation

21
3. Evaluation of specific environmental exposure
hypotheses

• Chernobyl
• Conclusion no detectable effect of Chernobyl on
  overall prevalence of congenital anomalies in
  Europe (Dolk 1999)
• Hazardous Waste Landfill Sites
• Population covered approx 100,000 births within 7
  km of 26 hazardous waste landfill sites in seven
  regions
• 1089 non-chromosomal cases, 270 chromosomal
  cases, 2508 controls
• Residence within 3 km vs 3-7 km
• Non-chromosomal anomalies OR 1.33 (95CI
  1.11-1.59) near sites (Dolk, Vrijheid 1998)
• Chromosomal anomalies OR 1.41 (95CI 1.00-1.99)
  (Vrijheid, Dolk, 2002)

22
Evaluation of specific environmental exposure
hypotheses

• Risk of hypospadias in relation to maternal
  occupational exposures to endocrine disrupting
  chemicals
• Routine data UK-NCAS, 1980-96
• Includes code for maternal occupation at time of
  birth registration
• Job-exposure matrix to classify exposure to
  potential EDCs (van Tongeren 2002) 7 chemical
  groups (phthalates, pesticides, alkyl phenolics,
  heavy metals, etc) 8 of cases classified with
  probable exposure
• Results (Vrijheid, Armstrong et al 2003)
• No increase in risk of hypospadias in categories
  of possible or probably exposure to EDCs
• Increased risk for hair dressers / exposure to
  phthalates in one time period, before adjustment
  for socio-economic status

23
Evaluation of specific environmental exposure
hypotheses

• Will we be overwhelmed with associations with
  environmental pollution beyond our capacity to
  respond?
• Missing what is real Exposure misclassification
  tends to dilute real associations
• Finding what is not real Ascertainment variation
  and confounding can exaggerate or reduce real
  associations or produce spurious associations
  where none exists

24
Advantages of congenital anomaly surveillance

• Short time lapse between exposure and diagnosis
• Continuously updated (spatially referenced)
  denominator statistics (births)
• Network of established registers with harmonised
  data collection

25
Methodological challenges in environmental
surveillance

• Relating to congenital anomaly data
• Classification lumping and splitting
• Adjustment for terminations of pregnancy
• Diagnostic and ascertainment variation
• Relating to other data sources
• Easily accessible systems of georeferencing of
  cases and births across Europe
• Easily accessible information on environmental
  exposures in Europe
• How to defined similar exposures? lumping and
  splitting again
• Exposure modelling/exposure surrogates/biomarkers?
  
• Easily accessible information on socioeconomic
  status across Europe
• Integration of systems, including expertise

26
Conclusions

• Fetal life is an especially sensitive period to
  environmental exposures
• Crucial fetal exposure occurs before the
  pregnancy is recognized
• The proportion of congenital anomalies
  attributable to environmental pollution may not
  be high, but any excess cases would represent a
  failure of our environmental health protection
  system
• EUROCAT, covering a quarter of European births,
  can play an important part in a European
  environmental health surveillance strategy