Census data
Provide population counts, demographics, and
socio-economic indicators
Vital statistics
Capture births, deaths, and causes of death
Registries
Track specific diseases, exposures, or health events over time
Health care utilization /
administrative health
datasets
Offer information on hospitalizations, physician visits,
prescriptions, and other system-level metrics
purpose of vital statistics
Track population changes
Monitor health trends (e.g., infant mortality rates)
Support planning and evaluation of health programs
Provide denominators for epidemiologic measures
purpose of registries
Monitor disease trends
Support research and prevention
programs
Improve patient care and outcomes
Health surveys
Provide self-reported data on health status, behaviours, and
determinants not captured in administrative systems
Uses of Administrative Health Data for Epidemiological
Research
Disease surveillance: Monitor population health and detect emerging trends.
* Integration of multiple data sources: Combine information from hospitals, clinics,
and other registries.
* Identification of first-time (incident) cases: Track new occurrences of disease.
* Etiologic research: Study causes and risk factors of disease.
* Longitudinal cohort follow-up: Monitor individuals over time.
* Health care utilization: Measure services used for a specific disease or population
group.
Canadian Community Health Survey (CCHS)
Ongoing, annual survey on
health status, healthcare access, and health determinants
Canadian Health Measures Survey (CHMS):
Combines interviews with direct
physical measurements and biological samples
Canadian Health Survey on Children and Youth (CHSCY)
Focused on health
and well-being of Canadians aged 1–17 years
Canadian Tobacco, Alcohol and Drugs Survey (CTADS):
Monitors substance
use behaviours and trends.
Health and demographic surveillance systems
can monitor vital statistics in a given region
○ Often used for places with incomplete/non-existent civil registrations
§ Ex. Sub-Saharan Africa, some Asian countries
- Also collect information on locally relevant health issues (i.e. vaccinations)
- Follows a group over time
Prevalence Surveys
- Measuring prevalence of a disease not captured by routine statistics or another aspect of health
○ Obesity, health-related behaviours (smoking, sun exposure, diet), use of health services
Migrant Studies
- A challenge we face in trying to interpret disease rates between countries is separating nature and nurture
○ Migrant studies can make a ‘natural experiment’
Ecological or correlation studies
Correlation: Can hint at what may play a role in some conditions or disease
○ But not causational, we still do have to consider other factors that may play a role
Ecological
○ Compare the prevalence of exposure and occurrence of disease in populations or groups of people, not individuals
○ There may be confounding variables
ecological fallacy
○ Ecological fallacy
§ Ascribing characteristics to members of a group that they might not possess as individuals
big data includes
Volume: refers to the amount of data available, either in terms of the number of variables collected on an individual, or the number of individual records for a given variable
○ Epidemiologists traditionally work with large volumes of data
Velocity: the speed of which data becomes available and its variety
○ Has increased lots with the digital age
The simultaneous collection and collation of different types of information
drawbacks to using routine data
Data is based on population level so it can be hard to draw conclusions about individuals
○ Look out for ecological fallacy/cross-level bias
Reports of incidence of disease are not very accurate
