Dark Mode
Multi-Level Mortality Causality Relations
Data Science and Analytics
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
Aggregates all known cause of deaths, including diseases and accidents, focusing on complex biological and mechanistic relationships. This open data product provides stratified death counts based on sex and specific age groups. A key innovation is the entity relations utilizing weighted biological causations, moving beyond simple statistical correlations. It details multi-level causal pathways, such as root, indirect, and direct causes of death events. The curated data for France is ready for immediate use. The expected update frequency for this product is weekly.
Columns
The data product includes 36 columns covering concept identification, causality structures, death statistics, and source quality indicators:
- concept_id: A unique identifier for the specific cause or concept.
- level_of_cause: Indicates the cause's position in the causality chain (e.g., root or indirect).
- predecessors_array/successor_array: JSON arrays that map causality relations by listing the related concept ID and the death impact ratio attributed to that relationship.
- total_yr_deaths_FRANCE: Total annual deaths recorded for France.
- per_100_000_inhabitants_/yr_FRANCE: Annual death rate per 100,000 persons in France.
- female_yr_deaths_ (and male variants): Annual death counts for France, stratified by age groups: those aged 4 years and under, those aged 5 to 64 years, and those aged 65 years and over.
- deaths_quality_FRANCE / causality_quality: Numerical quality indicators ranging from 0 (to review) up to 3 (cross-checked, verified, context defined).
- is accident: A Boolean field indicating if the concept relates to an accident type.
- CIM10: The optional ICD10/CIM10 classification code.
- concept_id_french: The optional French translation of the concept ID.
- (Note: Columns for the USA deaths are present but currently report 100% missing values, as that data is actively under development.)
Distribution
The data is delivered in CSV format, employing JSON arrays within certain fields to articulate the multi-level causality structure. The decimal separator used is the point. The data currently catalogues 102 unique concepts (rows). While the data includes structural placeholders for USA statistics, the currently populated statistics are specific to France.
Usage
This data product is intended for specialized health statistical reporting and advanced epidemiological analysis.
- Developing predictive or descriptive models that rely on multi-level causal relationships.
- In-depth research into the biological mechanisms that underlie different mortality events.
- Performing entity relation analysis, particularly by parsing the available JSON arrays (
predecessor_array). - Statistical studies requiring fine-grained mortality stratification by age and sex groups.
Coverage
Geographic Scope: The primary curated dataset is available for France. Data collection and curation for the USA is a noted task currently underway.
Demographic Scope: Mortality data is meticulously stratified by sex and three distinct age groups (0-4, 5-64, and 65+ years).
Time Range: The source years for the total death figures in France documented within the concepts range from 1992 to 2018.
License
Attribution 4.0 International (CC BY 4.0)
Who Can Use It
This compilation of data is specifically intended for data scientists.
- Public Health Officials: For tracking and analysing mortality trends using causal, rather than correlational, mechanisms.
- Epidemiologists and Health Researchers: To validate and explore complex biological causality chains across different demographics.
- Statistical Modellers: For building advanced health models incorporating detailed age/sex stratification.
Dataset Name Suggestions
- Multi-Level Mortality Causality Relations
- Age and Sex Stratified Cause of Death Causality
- Biological Mechanism Mortality Data (France)
Attributes
Original Data Source: Multi-Level Mortality Causality Relations
QUALITYLoading...