Testing Modeling Assumptions in the West Africa Ebola Outbreak

Area of Science:

• Epidemiology
• Mathematical Modeling
• Public Health

Background:

• The 2014 West Africa Ebola Virus Disease (EVD) epidemic caused significant mortality.
• Previous EVD models often relied on assumptions of uniform transmissibility and homogeneous population mixing.
• These simplifying assumptions may limit the accuracy of disease spread predictions.

Purpose of the Study:

• To challenge the assumptions of uniform transmissibility and homogeneous mixing in EVD models.
• To investigate factors influencing EVD spread patterns in West Africa.
• To propose improvements for EVD disease modeling accuracy.

Main Methods:

• Utilized West African migration data and modeling to analyze EVD spread patterns.
• Estimated initial EVD growth rates within administrative divisions.
• Developed and applied a novel statistical test to assess EVD strain transmissibility.

Main Results:

• EVD spread was found to be non-homogeneous, following predictable patterns influenced by migration.
• EVD initial growth rates were inversely correlated with population density.
• A statistical test indicated non-uniform transmissibility among EVD strains, with some strains appearing more frequently than expected.

Conclusions:

• EVD transmission dynamics are complex and influenced by population movement and density.
• Standard modeling assumptions of homogeneous mixing and uniform transmissibility are not fully supported by data.
• Incorporating migration, population density, and strain-specific transmissibility can enhance EVD epidemiological models.