SEEDNet: Covariate-free multi-country settlement-level epidemiological estimates datasets for network analysis

Area of Science:

• Network science applications in public health
• Epidemiological data analysis for global health

Background:

• Population health network science is promising but lacks LMIC datasets.
• Existing small-area estimation (SAE) methods introduce uncertainty, hindering cross-country comparisons.

Purpose of the Study:

• Introduce SEEDNet (Settlement-level Epidemiological Estimates Datasets for Network Analysis), a multi-country data library.
• Provide a covariate-free method for SAE of health indicators in LMICs.
• Facilitate network-based population health analysis in LMICs.

Main Methods:

• Utilized georeferenced national surveys for covariate-free SAE.
• Developed an automated estimation process with harmonized data inputs.
• Mapped all population settlement sizes comprehensively.

Main Results:

• Created a multi-country data library of population health indicators.
• Enabled complete mapping of population settlements.
• Established settlements as functional units for epidemiological network analysis.

Conclusions:

• SEEDNet provides open-access, reliable population health data for LMICs.
• The covariate-free approach enhances data comparability across countries and time.
• Facilitates robust network representations of population health in LMICs.