Modelling to inform interventions during Highly Pathogenic Avian Influenza outbreaks in Great Britain
Lead Research Organisation:
University of Oxford
Department Name: Mathematical Institute
Abstract
Avian influenza is a highly contagious disease, affecting both wild birds and domestic poultry. The poultry industry suffers substantial economic losses due to this disease, both in Great Britain (GB) and around the world. The 2021-22 outbreak in GB has been the largest in the country's history, and seasonal outbreaks are expected to continue to occur.
Mathematical models are increasingly used during outbreaks of a range of diseases to inform future scenarios and to guide control measures. This project involves developing a national-scale mathematical model of avian influenza transmission around GB, in collaboration with the UK Government's Animal and Plant Health Agency (APHA). It builds on our previous work with APHA during the 2021-22 outbreak, in which we spent six weeks developing a preliminary transmission model to project future spread of the virus and to guide control interventions. In the current project, we will build on our previous rapid-response research, taking advantage of the longer timescale of this project to develop a flexible modelling framework with a high level of rigour. This includes fitting the parameters of the transmission model to a range of data sources, and accounting for both transmission between poultry premises within GB and importations of infection from elsewhere. By working closely with APHA, we will ensure that the model can be run on APHA computer systems, permitting its use by policy advisors to guide interventions in future outbreaks.
Once we have developed the transmission model, we will use it to explore the effectiveness of a range of different control interventions that could be applied during future outbreaks in GB. A benefit of our collaboration with APHA is that we are in a position to use the model to test practical strategies that could be deployed effectively. We will also develop a user-friendly software tool and Graphical User Interface based on the model, allowing the user to change model parameters themselves and test different control interventions without requiring specialist knowledge about the computing code. To encourage others to use the software, we will run an outbreak simulation exercise in which we will provide simulated outbreak data to policy advisors, who will then use the tool to determine optimal interventions. Similar exercises have been conducted before for a range of other livestock diseases. This exercise will be conducted at the start of the third year of this project, enabling any feedback to be incorporated into the model and software tool. We will also run a two-day workshop for other researchers in which we describe the modelling framework and software tool, and demonstrate their use.
In summary, this project will involve the development of a national epidemiological modelling resource that can be used to plan control measures during future avian influenza outbreaks in GB. It complements existing research on avian influenza by providing a tool that will be used by policy advisors for the foreseeable future to determine optimal interventions during outbreaks.
Mathematical models are increasingly used during outbreaks of a range of diseases to inform future scenarios and to guide control measures. This project involves developing a national-scale mathematical model of avian influenza transmission around GB, in collaboration with the UK Government's Animal and Plant Health Agency (APHA). It builds on our previous work with APHA during the 2021-22 outbreak, in which we spent six weeks developing a preliminary transmission model to project future spread of the virus and to guide control interventions. In the current project, we will build on our previous rapid-response research, taking advantage of the longer timescale of this project to develop a flexible modelling framework with a high level of rigour. This includes fitting the parameters of the transmission model to a range of data sources, and accounting for both transmission between poultry premises within GB and importations of infection from elsewhere. By working closely with APHA, we will ensure that the model can be run on APHA computer systems, permitting its use by policy advisors to guide interventions in future outbreaks.
Once we have developed the transmission model, we will use it to explore the effectiveness of a range of different control interventions that could be applied during future outbreaks in GB. A benefit of our collaboration with APHA is that we are in a position to use the model to test practical strategies that could be deployed effectively. We will also develop a user-friendly software tool and Graphical User Interface based on the model, allowing the user to change model parameters themselves and test different control interventions without requiring specialist knowledge about the computing code. To encourage others to use the software, we will run an outbreak simulation exercise in which we will provide simulated outbreak data to policy advisors, who will then use the tool to determine optimal interventions. Similar exercises have been conducted before for a range of other livestock diseases. This exercise will be conducted at the start of the third year of this project, enabling any feedback to be incorporated into the model and software tool. We will also run a two-day workshop for other researchers in which we describe the modelling framework and software tool, and demonstrate their use.
In summary, this project will involve the development of a national epidemiological modelling resource that can be used to plan control measures during future avian influenza outbreaks in GB. It complements existing research on avian influenza by providing a tool that will be used by policy advisors for the foreseeable future to determine optimal interventions during outbreaks.
Technical Summary
We will develop a national-scale model of HPAI transmission in GB for guiding interventions during future outbreaks.
In the model, the risk of virus importations to GB by wild birds will be informed by data on bird migration pathways. We will use multiple bird observation datasets to generate spatio-temporal risk maps describing the arrival of target species in GB.
After quantifying the risk of imported infections from wild birds, we will develop the transmission model including both infection incursions into the poultry industry and premises-to-premises transmission. Guided by APHA, we will incorporate farm-level parameters capturing heterogeneity in farming practices within the poultry industry (including different biosecurity levels), informed by data from the National Poultry Register and APHA HPAI risk assessments.
We will run model simulations of outbreaks in GB and explore scenarios involving different interventions. The inclusion of interventions in the computing code, such as farm-level enhanced biosecurity and culling of hosts on at-risk premises, is essential to allow these measures to be assessed in real-time during future outbreaks.
We will develop a software tool (including a Graphical User Interface) based on the model to allow scientists and policy advisors to run outbreak simulations and test interventions themselves. In the tool, users will be able to upload their own data to inform the model and set the values of parameters governing transmission using sliders. Model simulations will then be adaptable to the exact conditions arising during future outbreaks.
Through our collaboration with APHA, we will run a simulation exercise in which we generate synthetic outbreak data and APHA policy advisors use our tool to guide interventions. This will increase uptake of the software and provide us with feedback to improve the underlying model. The software will be presented to stakeholders and other scientists at an end-of-project workshop.
In the model, the risk of virus importations to GB by wild birds will be informed by data on bird migration pathways. We will use multiple bird observation datasets to generate spatio-temporal risk maps describing the arrival of target species in GB.
After quantifying the risk of imported infections from wild birds, we will develop the transmission model including both infection incursions into the poultry industry and premises-to-premises transmission. Guided by APHA, we will incorporate farm-level parameters capturing heterogeneity in farming practices within the poultry industry (including different biosecurity levels), informed by data from the National Poultry Register and APHA HPAI risk assessments.
We will run model simulations of outbreaks in GB and explore scenarios involving different interventions. The inclusion of interventions in the computing code, such as farm-level enhanced biosecurity and culling of hosts on at-risk premises, is essential to allow these measures to be assessed in real-time during future outbreaks.
We will develop a software tool (including a Graphical User Interface) based on the model to allow scientists and policy advisors to run outbreak simulations and test interventions themselves. In the tool, users will be able to upload their own data to inform the model and set the values of parameters governing transmission using sliders. Model simulations will then be adaptable to the exact conditions arising during future outbreaks.
Through our collaboration with APHA, we will run a simulation exercise in which we generate synthetic outbreak data and APHA policy advisors use our tool to guide interventions. This will increase uptake of the software and provide us with feedback to improve the underlying model. The software will be presented to stakeholders and other scientists at an end-of-project workshop.
