DiFLUsion: A new spatiotemporal early warning system for HPAI

Abstract

Purpose: Highly pathogenic avian influenza (HPAI) represents a global threat due to the devastating economic losses for the poultry industry and a recognized zoonotic potential. The migratory water birds play a critical role for the spread of HPAI between continents. A complex global One Health issue and capturing the disease at the livestock-wildlife-human interface is a major challenge. Addressing this major One Health gap and establishing an early-warning-system requires a multidisciplinary scientific approach combining computer, diverse data sources and decision-making in health sciences. Here we introduce DiFLUsion, a ground breaking early warning detection system for HPAI, that allows early detection of HPAI in free zones connected by movements of wild birds with HPAI affected zones. Methods & Materials: DiFLUsion integrates several data sources and analytical tools using Python, MongoDB, neo4J and ArcGIS that allow modulating alerts according to the location of HPAI outbreaks in Europe, the seasonality of wild bird movements and the temperatures of virus survival. A primary objective was to be able to obtain user-friendly alerts for decision-making, including weekly reports and an interactive map viewer. DiFLUsion´s modular structure offers high flexibility and adaptability allowing adapted it to new geographical areas. To this end, DiFLUsion modules are being transferred to the University of Minnesota for application in a pilot disease surveillance project in the USA, in parallel to its use in Europe. Results: DiFLUsion facilitates the decision-making capacity of livestock health managers in Spain where it is currently implemented, to prepare for and respond in advance to avian influenza epidemics. Here it is shown how cases of the disease originating in wild birds that have occurred recently or in past years are located in areas identified by the alert system as high risk. A validation tool for the system is currently being developed to evaluate its effectiveness in terms of specificity and sensitivity. Conclusion: DiFLUsion represents a useful tool as an early-warning-system for HPAI, once the effectiveness and use of the tool was tested in Europe and in the USA through the current studies, the method can be generalized to be used globally. Funding projects: CON20-171 (Spain); OMB-N°0925d-0001(USA).