Changing Burial Practices Explain Temporal Trends in the 2014 Ebola Outbreak

TL;DR

This study models how changes in burial practices influenced the spread of Ebola during the 2014 outbreak, highlighting the importance of behavior change in controlling transmission.

Contribution

It introduces a multi-stage transmission model combined with a game-theoretic approach to capture burial practice changes during the outbreak.

Findings

Sanitary burial adoption increased between July and October 2014.

Behavior change modeling improved outbreak size predictions.

Late adoption of sanitary burials limited outbreak containment.

Abstract

Background: The 2014 Ebola outbreak in West Africa was the largest on record, resulting in over 25,000 total infections and 15,000 total deaths. Mathematical modeling can be used to investigate the mechanisms driving transmission during this outbreak -- in particular, burial practices appear to have been major source of infections. Methodology/Principal Findings: We developed a multi-stage model of Ebola virus transmission linked to a game-theoretic model of population burial practice selection. We fit our model to cumulative incidence and mortality data from Guinea, Liberia, and Sierra Leone from January 2014 to March, 2016. The inclusion of behavior change substantially improved best fit estimates and final size prediction compared to a reduced model with fixed burials. Best fit trajectories suggest that the majority of sanitary burial adoption occurred between July, 2014 and October, 2014. However, these simulations also indicated that continued sanitary burial practices waned following the resolution of the outbreak. Conclusions/Significance: Surveillance data from the 2014 outbreak appears to have a signal of changes in the dominant burial practices in all three countries. Increased adoption of sanitary burials likely attenuated transmission, but these changes occurred too late to prevent the explosive growth of the outbreak during its early phase. For future outbreaks, explicitly modeling behavior change and collecting data on transmission-related behaviors may improve intervention planning and outbreak response.