The highly pathogenic avian influenza H5N1 epidemic that occurred between 2021 and 2022 in northeastern Italian regions was one of the most severe ever recorded. The epidemic affected a high number of poultry farms (317), with over 14 million affected animals. The spread of the epidemic was very rapid, with peaks of over 50 new outbreaks per week. The speed of spread raised the hypothesis of possible direct contact between infected farms and other poultry companies, or the presence of common sources of infection.
The rapid spread of the epidemic across the territory gave rise to two main hypotheses:
To understand the dynamics of the epidemic, the Epidemiology and Risk Analysis Laboratory in Public Health at the Istituto Zooprofilattico Sperimentale delle Venezie (Veneto Region) used network analysis, a powerful analysis tool that allows the study of characteristics and relationships among objects within a network; the nodes represent entities or objects, while the connections represent existing relationships between these entities.
The study, published in the Journal Pathogens , explains the dynamics of the epidemic by assessing the impact of potential factors in the spread of infection, using data collected during on-site visits to farms and the genetic information obtained from molecular analyses on the isolated viruses from each outbreak.
During the 2020-2021 epidemic, phylogenetic analyses revealed the existence of several viral genetic clusters, supporting the hypothesis of virus spread among domestic farms. Specifically, the complete genomes of 214 viruses were used to construct the phylogenetic network. In this network, each node corresponds to a virus identified in a single outbreak, while the links connect nodes characterized by the highest genetic similarity.
The phylogenetic network serves as the starting database for the study, on which the network method called Exponential Random Graph Model (ERGM) was applied.
ERGM is a statistical model capable of explaining why there is a link between two nodes based on a series of epidemiological variables. When applied to a phylogenetic network, ERGM relates the epidemiological characteristics of the affected farms to the more strictly genetic characteristics of the found viruses. This approach allowed evaluating the impact of these variables on the possibility of infection spread among farms.
The analyses highlighted that certain variables, such as farms belonging to the same poultry supply chain, the duration of exposure to active outbreaks, and the geographic distance between farms, have a significant effect on the network structure and disease transmission. These results suggest important implications for control and prevention strategies of future avian influenza epidemics.
The study showed the effectiveness and innovation of applying network analysis in integrating virological and epidemiological data. This approach could contribute to a better understanding of disease spread dynamics and provide more effective tools for epidemic control and prevention.
In the future, further developments of the network analysis approach could include the integration of temporal information to analyse the evolution of epidemics over time, as well as the analysis of diseases involving different populations through the use of multi-layer networks.
This study highlights the practical potential of network analysis for better understanding and management of epidemics, offering valuable tools to epidemiologists for disease control and prevention in the territory.
The study has been published in the Journal Pathogens
Source: IZSVe