Bridging the gap: Using reservoir ecology and human serosurveys to estimate Lassa virus spillover in West Africa

• DOI: 10.1371/journal.pcbi.1008811
• Publication Date: 2021-03-03T19:46:24Z
• Authors: Andrew J. Basinski; Elisabeth Fichet-...; Anna R. Sjodin; Tanner J. Varrelman; Christopher H. Re...; Nathan C. Layman; Brian H. Bird; David J. Wolking; Corina Monagin; Bruno M. Ghersi; Peter A. Barry; Michael A. Jarvis; Paul E. Gessler; Scott L. Nuismer
• Keywords: Spillover effect; Culling

Forecasting the risk of pathogen spillover from reservoir populations wild or domestic animals is essential for effective deployment interventions such as wildlife vaccination culling. Due to sporadic nature events and limited availability data, developing validating robust, spatially explicit, predictions challenging. Recent efforts have begun make progress in this direction by capitalizing on machine learning methodologies. An important weakness existing approaches, however, that they generally rely combining human infection data during training process thus conflate attributable prevalence population with attributed realized rate into population. Because planning requires these components be disentangled, we developed a multi-layer framework separates processes. Our approach begins models predict geographic range primary subset which occurs. The predicted product specific then fit patterns historical result geographically forecast can easily decomposed used guide intervention. Applying our method Lassa virus, zoonotic regularly spills over across West Africa, results model explains modest but statistically significant portion variation spillover. When combined mechanistic mathematical dynamics, predicts 897,700 humans are infected virus each year Nigeria accounting more than half infections.