Applications of GIS in Health Research (II)

Applications of GIS in Health Research

(II)

West Nile virus

Outline

• Part 1. Applications of GIS in Health research or spatial epidemiology– Disease Mapping– Cluster Detection– Spatial Exposure Assessment– Assessment of Risk of Disease

• Part 2. Exposure Assessment

What is Epidemiology?

“The study of the distribution and determinants of health-related states in populations, and the application of this study to control health problems”. 

SPATIAL

Spatial Epidemiology/Health Geography: the ‘Where”

• There’s always a ‘where’.• The importance of the ‘where’ varies depending

on what you’re examining.• ‘Where’ is a major way statistics are organized -

and can be further organized, processed and analyzed.

Detailed Land Use

Transit Network Data

Network Buffers

Travel Activity

Census Data

Road Networks

Zonal Level Land Use

Sensor Locations

Detailed Land Use

Transit Network Data

Network Buffers

Travel Activity

Census Data

Road Networks

Zonal Level Land Use

Sensor Locations

Network Buffers

Travel Activity

Census Data

Road Networks

Zonal Level Land Use

Sensor Locations

Detailed Land Use

Transit Network Data

Network Buffers

Travel Activity

Census Data

Road Networks

Zonal Level Land Use

Sensor Locations

Detailed Land Use

Transit Network Data

Network Buffers

Travel Activity

Census Data

Road Networks

Zonal Level Land Use

Sensor Locations

Network Buffers

Travel Activity

Census Data

Road Networks

Zonal Level Land Use

Sensor Locations

Components of Spatial Epidemiology

A. Disease Mapping

B. Cluster Detection

C. Spatial Exposure Assessment

D. Assessment of Risk of Disease

Disease Mapping

Statistics Canada

Disease Mapping

The British Columbia Wellness Atlas - http://www.geog.uvic.ca/wellness/

Determinant of Disease Mapping

Disease (and determinant) mapping allows us to:

• Describe patterns of disease.• Explore and analyze spatial patterns.• Portray messages and disseminate information on

health and health determinants.• Hypothesize about possible causal relationships• However, the following needs to be addressed:

• The selection of areal unit (e.g. CTs vs. DAs vs. Health Regions)

• Small number problems: Rates calculated from small populations-at-risk tend to be unstable, because small changes in the count or in the population lead to wide variance in the rate.

• The interpretation of patterns without information on other confounding factors.

B. Cluster Detection

• Clusters are geographically and/or temporally bounded groups of occurrences of sufficient size and concentration unlikely to have occurred by chance.

• Clusters are either related to each other through some social or biological mechanism or they have a common relationship with some other event or circumstance.

Cluster analysis• Incidence of Childhood Leukaemia in Northern England 1968-85.

• Openshaw (1987) Geographical Analysis Machine (GAM).

• Cluster detection generally driven by public/media concern

S Openshaw et al. 1987. International Journal of Geographical Information Science, 1987 Vol 1, No.4, 335-358

Space and Place

• Place and space are often used interchangeably, is there a difference?–Space: location (latitude/longitude) and the physical features (built and natural in that area).

–Place: political, social, cultural, historical meanings attached to a space.

• Rarely, if ever, is space independent of place.

Importance of Space/Place

http://epi.grants.cancer.gov/LIBCSP/reports/NCI_6_06.html

Why are there spatial patterns of health?

• Composition: Similar people tend to aggregate and their shared characteristics explain in part the health and place association.

• Context: Ecological attributes (e.g. physical, social, economical, political) influence whole groups and affect health over and above aggregate individual characteristics.

Need for a Geographic Study of Health

• Spatial epidemiology provides a framework to examine the influences of space and place on health.

• Space and place can often be used as surrogates for influences on health, for example:– Access to primary healthcare.– Exposure to environmental pollutants.– Social networks and capital.– Physical activity opportunities.

Examples of Spatial Epidemiology Studies

Vinceti et al. International Journal of Health Geographics 2009 8:8 doi:10.1186/1476-072X-8-8

-Exposure through air -Living conditions-Exposure through water

Example 1: Risk of Congenital Anomalies Around a Municipal Solid

Waste Incinerator: A GIS Based Case-Control Study

• Examined exposure to emissions from a municipal solid waste incinerator and risk of birth defects in a northern Italian community.

• Modelled incinerator emissions for three dioxin exposure levels.

• Mapped congenital anomalies observed over 1998–2006 and matched controls at postal code level.

Vinceti et al. International Journal of Health Geographics 2009 8:8 doi:10.1186/1476-072X-8-8

Vinceti et al. International Journal of Health Geographics 2009 8:8 doi:10.1186/1476-072X-8-8

Results• No significant increase in total congenital

anomalies, or in specific types of anomalies with exposure.

• Odds ratio for congenital anomalies did not decrease during a prolonged shut-down period of the plant (important sensitivity test).

Example 2: Obesity Relationships with Community Design, Physical Activity, and Time Spent in Cars

• Evaluated the relationship between the built environment around each participant's place of residence and self-reported travel patterns (walking and time in a car), body mass index (BMI)

Frank et al., (2004) Obesity relationships with community design, physical activity, and time spent in cars . American Journal of Preventative Medicine. 27(2): 87-96.

Derived Exposures

Frank et al., (2004) Obesity relationships with community design, physical activity, and time spent in cars . American Journal of Preventative Medicine. 27(2): 87-96.

Frank et al., (2004) Obesity relationships with community design, physical activity, and time spent in cars . American Journal of Preventative Medicine. 27(2): 87-96.

Results

• Land-use mix had the strongest association with obesity (BMI≥30 kg/m2), with each quartile increase being associated with a 12.2% reduction in the likelihood of obesity.

• Strong relationship with time spent in car

Example 3: Relation of Trihalomethane (TTHM) Concentrations in Public Water

Supplies to Stillbirth and Birth Weight in Three Water Regions in England

• TTHM is a chemical used for cooling, it is also a disinfection by-product (result from the reaction of chlorine with organic matter present in the water being treated)

• Modeled estimates of quarterly TTHM concentrations in water zones

• Linked to one million birth and stillbirth records using maternal residence at time of birth.

Relation of Trihalomethane Concentrations in Public Water Supplies to Stillbirth and Birth Weight in Three Water Regions in England Environmental Health Perspectives • VOLUME 113 | NUMBER 2 | February 2005

Water company supply regions.Water supply-zone-level TTHM exposure categories for each

quarter, Severn Trent Water, 1997: (A) January–March; (B) April–June; (C) July–September; (D) October–December.

TTHM Exposure Assessment

Relation of Trihalomethane Concentrations in Public Water Supplies to Stillbirth and Birth Weight in Three Water Regions in England Environmental Health Perspectives • VOLUME 113 | NUMBER 2 | February 2005

Results• Found a

significant association of stillbirths with maternal residence in areas with high TTHM exposure.

Relation of Trihalomethane Concentrations in Public Water Supplies to Stillbirth and Birth Weight in Three Water Regions in England Environmental Health Perspectives • VOLUME 113 | NUMBER 2 | February 2005

Part 1 - Summary• What is spatial epidemiology?

• The study of the spatial distribution of health-related states and health determinants in populations.

• Applications of GIS in Health Research?• Disease mapping• Cluster detection• Spatial exposure assessment and the assessment of risk

• Why are space and place important health determinants?• Similar individuals tend to cluster together.• Health determinants (e.g. behaviors, environmental risks, health

resources, socio-economic conditions) vary across space and place.

Part 2 - Exposure Assessment• Exposure assessment is the study of

human contact with chemical, physical, biological or social agents occurring in their environments.

• Exposure assessment examines the mechanisms and dynamics of events and how they relate to health outcomes.

• Direct Methods– Biomonitoring (blood, urine samples)– Personal exposure monitoring (personal

monitors)– Self-reporting

• Indirect Methods– Spatial modeling (GIS)

How Can We Assess Exposures?

Exposure Assessment in GIS

• A. Proximity (the simplest case)• B. Interpolation (creating a surface out of

a set of known points)• C. Landuse Regression Spatial Modeling

(Prediction based on geographic factors)• D. Satellite data

A. Proximity • Is the simplest form of spatial exposure

assessment.• Assumes all individuals within a specific distance

to a source have the same exposure (GIS buffers).

• Commonly used for measuring access to resources and exposure to environmental hazards (e.g. air pollution).

Example: Toronto Neighbourhood Environments – A Focus on

Diabetes• Examined how access to health foods are

related to diabetes in Toronto.• Identified proximity to health food and

density of health food outlets by dissemination areas.

• Caution! Simple proximity is often a poor surrogate for resource access.

• Enhanced proximity by modeling walking times to health food outlets.

Toronto Neighbourhood Environments – A Focus on

Diabetes

http://www.ices.on.ca/file/TDA_Chp8_PartA_press.pdf

Toronto Neighbourhood Environments – A Focus on

Diabetes

http://www.ices.on.ca/file/TDA_Chp8_PartA_press.pdf

B. Interpolation• Spatial interpolation is used to estimate a value of

a variable at an un-sampled location from measurements made at other sites.

• Spatial interpolation is based on the notion that points which are close together in space tend to have similar attributes.

• Many different methods available:

• Exact or approximate• Deterministic or stochastic• Local or global• Gradual or abrupt

West Nile virus

http://www.ij-healthgeographics.com/content/5/1/21

C. Landuse Regression Spatial Modeling

• Predicts measured values based on geographic proximal variables.

• Used extensively to predict air pollution levels based on land use variables.• Land use• Traffic data• Population density

• Method has been expanded to examine neighborhood factors, such as:• Walkabilty• Resource access

Physical Air Pollution Measurements NOx (nitrogen dioxide)

http://www.cher.ubc.ca/UBCBAQS/welcome.htm

Deriving Predictive Variables

Jerrett et al. (2005) A review and evaluation of intraurban air pollution exposure models. Exposure Assessment and Environmental Epidemiology. 15, 185–204

Final Predictive Surface of NOx Pollution

http://www.cher.ubc.ca/UBCBAQS/welcome.htm

D. Satellite Data: Global Ozone Mapping

Part 2 - Summary• Methods available for exposure

calculation.

• Proximity• Interpolation• Spatial modeling (Land Use Regression)

• Satellite data

Thanks for providing material for the slides:

• Alejandro Cervantes-Larios (PhD student, Geography Department)

• Perry Hystad (PhD student in epidemiology)