Statistically enriched geospatial datasets of Brazilian municipalities for data-driven modeling

Area of Science:

• Geospatial health sciences
• Environmental epidemiology
• Data science

Background:

• Georeferencing is crucial for accurate scientific studies, especially in health sciences.
• Brazilian health systems, particularly for Notifiable Diseases, often lack georeferenced data, recording only the municipality of occurrence.
• This data gap limits the development of precise disease prediction models requiring socioenvironmental characteristics.

Purpose of the Study:

• To address the lack of georeferenced health data in Brazil.
• To enrich data-driven models by statistically characterizing Brazilian municipalities.
• To provide comprehensive datasets for improved health event analysis.

Main Methods:

• Compiled thematic data layers (642) representing natural and artificial landscape characteristics over time.
• Statistically characterized all 5,570 Brazilian municipalities.
• Generated datasets with 11,556 descriptive statistics attributes per municipality.

Main Results:

• Created a comprehensive collection of datasets for Brazilian municipalities.
• Each dataset offers extensive statistical attributes characterizing municipal landscapes.
• The data facilitates the integration of socioenvironmental factors into health models.

Conclusions:

• The produced datasets enhance the expressiveness of data-driven health models when only municipal-level data is available.
• This resource can improve the accuracy of disease prediction and epidemiological studies in Brazil.
• Enables more robust analysis of health events by incorporating detailed environmental and artificial characteristics.