Abstract
HPAI outbreaks in Denmark threaten the poultry industry, but the risk of virus spillover from wild birds remains poorly understood. We developed a spatiotemporal quantitative model to assess the HPAI spillover risk driven by wild bird migration and environmental contamination.
Bird abundance data for the selected wild bird species was used to model bird migration, and an epidemic compartmental model was constructed for environmental transmission. The probability of infection was obtained through a dose-response relationship within the wild bird population, and between wild birds and poultry farms, using the median infection doses of HPAI H5Nx viruses assessed in experimental studies. We assumed a homogenous spread of the environmental contamination within each 10 by 10 km grid cell, and categorised farms into six types depending on the species. For wild birds, detection data from 2020/2021 were used to inform the model. Model output consisted of the probability of infection for each individual poultry farm, and this was categorised into at-risk and not-at-risk groups for a spatial scanning analysis. Poultry outbreak data from 2020 – 2024 was used to validate results.
In total, 37 of the total 48 outbreak farms were identified as at-risk farms during the same week the outbreak occurred, indicating a good (77%) model sensitivity. Model specificity, i.e. farms without outbreaks identified as not-at-risk, ranged from 50% to 77% depending on the week number.
The model can support animal health authorities to optimise monitoring and control strategies in space and time to better understand and mitigate future HPAI outbreaks.
Bird abundance data for the selected wild bird species was used to model bird migration, and an epidemic compartmental model was constructed for environmental transmission. The probability of infection was obtained through a dose-response relationship within the wild bird population, and between wild birds and poultry farms, using the median infection doses of HPAI H5Nx viruses assessed in experimental studies. We assumed a homogenous spread of the environmental contamination within each 10 by 10 km grid cell, and categorised farms into six types depending on the species. For wild birds, detection data from 2020/2021 were used to inform the model. Model output consisted of the probability of infection for each individual poultry farm, and this was categorised into at-risk and not-at-risk groups for a spatial scanning analysis. Poultry outbreak data from 2020 – 2024 was used to validate results.
In total, 37 of the total 48 outbreak farms were identified as at-risk farms during the same week the outbreak occurred, indicating a good (77%) model sensitivity. Model specificity, i.e. farms without outbreaks identified as not-at-risk, ranged from 50% to 77% depending on the week number.
The model can support animal health authorities to optimise monitoring and control strategies in space and time to better understand and mitigate future HPAI outbreaks.
Co-Author(s)
Yangfan Liu (1), Lene Jung Kjær (1), Anette Ella Boklund (1), Michael Ward (2), Yuan Liang (1), Carsten Thure Kirkeby (1)
(1): Department of Veterinary and Animal Sciences, Faculty of Health and Medical Sciences, University of Copenhagen, Frederiksberg, Denmark
(2): Faculty of Science, Sydney School of Veterinary Science, The University of Sydney, Australia
Abstract Category
Notable outbreaks, field and molecular epidemiology, and surveillance in poultry