Research use and required cautions
The Atlas is a public-data exploration platform and research-support tool. It is not a completed epidemiological study, a causal analysis, a medical system, a diagnostic tool or a proof engine.
Appropriate uses
- Descriptive mapping. Showing what official aggregate records report for an area and period.
- Source discovery. Finding which agency publishes a measure, and reading its methods.
- Hypothesis generation. Forming questions that a properly designed study could test.
- Transparent public-data exploration. Seeing definitions, versions and limitations alongside every figure.
- Reproducible area-level summaries. Exporting a reproducibility record another person can follow.
Not appropriate uses
- Causal conclusions. Nothing here can show that a facility, industry or condition caused an outcome.
- Individual exposure determination. Area records cannot describe what any person breathed, drank or touched.
- Individual cause-of-death inference. The Atlas cannot say why any child died, and never displays individual records.
- Clinical or legal decisions. These figures are not diagnostic, evidentiary or advisory.
- Identifying a health cluster. Cluster identification requires formal investigation by a health authority.
- Estimating personal risk. Aggregate rates do not translate to an individual's risk.
Required cautions
Ecological fallacy
A pattern between area averages does not hold for the individuals in those areas.
Confounding
Poverty, access to care, age structure, smoking, occupation and many other factors differ between areas and are not adjusted for.
Spatial autocorrelation
Neighbouring areas resemble each other, so apparent patterns can be an artefact of geography.
Multiple comparisons
Scanning many areas, periods and outcomes will produce apparently unusual values by chance.
Selective reporting
Choosing the period or geography that shows the largest contrast is not a finding.
Small numbers
Rare outcomes in small populations produce unstable rates that swing widely year to year.
Migration
People move between the exposure period and the outcome period; residence at the time of an event is not residence history.
Reporting variation
Certification, investigation and reporting practice differs between jurisdictions and over time.
Exposure misclassification
Facility presence is not exposure; distance is a poor proxy for what a person actually encountered.
Temporal mismatch
Environmental records and health periods often do not align; unaligned records are labelled and excluded by default.
Missing data
Suppressed, unavailable and restricted cells are common, are not zeros, and are never reconstructed.
Formal research conclusions should be developed with qualified epidemiology, biostatistics, environmental-health, and community partners.
Non-causation note: the Atlas shows aggregate public records in the same place and period. Co-occurrence, proximity or overlap in an area does not establish exposure, individual risk, or causation, and cannot explain why any individual pregnancy, birth or infant outcome occurred.
Research inquiry
Collaboration intake is being set up. Tell us who you are and what you are working on, and the note will be prepared for the Atlas team.
Do not send individual health records, personal details of any child or family, death-certificate or autopsy material. This channel cannot handle urgent health or legal requests — contact your clinician, local health department, or emergency services instead.