Can We Predict the Next Bird Flu Outbreak? A New Model Says Yes

During a recent presentation, Yangfan Liu, postdoctoral researcher at the University of Copenhagen (UCPH), introduced a prediction spatiotemporal model within the PANRISK framework, a pioneering tool designed to forecast pathogenic virus outbreaks. In particular, the presentation focused on the work on PANRISK for highly pathogenic avian influenza (HPAI) in wild birds, poultry, and the environment.

HPAI is a zoonotic disease, meaning it can jump from animals to humans and other mammals, primarily through contact with infected birds. This study aims to predict when and where outbreaks may occur across 43 European countries and regions, covering 227 HPAI-susceptible wild bird species and domestic poultry.

To achieve this goal, the PANRISK model combines two complementary approaches:

– a spatiotemporal mechanistic SIR model – a mathematical framework that tracks how individuals move between susceptible, infected, and recovered states – to capture the infection dynamics
– a local spatial point process model to estimate outbreak probabilities across geographical areas.

Thanks to this dual-structure approach, researchers will be able to anticipate both the timing and geography of potential outbreaks. By integrating historical surveillance data of birds and disease, environmental conditions, and citizen science-based bird population information, along with behavioural and virus characteristics from scientific literature, the model will offer a powerful tool for early warning and prevention.

This also provides valuable insights for the PANPOC detection scheme, the system designed in the PAIR project to identify and monitor RNA virus cases, ultimately strengthening Europe’s preparedness against future epidemics.

With the contribution of Yangfan Liu and Carsten Thure Kirkeby (UCPH)

Photo by Mark Stebnicki on Pexels