Applying Field and Genomic Epidemiology Methods to Investigate Transmission Networks of Highly Pathogenic Avian Influenza A (H5N1) in Domestic Poultry in British Columbia, Canada (2022–2023)

TL;DR

This study combines field and genomic data to track how H5N1 avian flu spread among poultry farms in British Columbia, revealing local transmission patterns and the importance of proximity.

Contribution

The study introduces an integrative approach combining field and genomic epidemiology to better understand and manage HPAI transmission dynamics.

Findings

Five genetically distinct viral clusters were identified in British Columbia.

Local spread was more common in commercial premises, with proximity to infected sites being a strong predictor of infection.

Noncommercial premises with outdoor access were mostly due to independent introductions from wild birds.

Abstract

Highly pathogenic avian influenza (HPAI) virus A (H5N1) was first detected in North America in 2021. Ongoing spillover events from wild to domestic birds and subsequent transmission between domestic birds resulted in Canada's largest recorded epidemic of HPAI. Between December 2021 and April 2024, 422 A (H5N1) outbreaks in domestic birds were confirmed in Canada. Of these, 158 (37%) occurred in British Columbia (BC). This study integrates field and genomic epidemiology (GE) results to investigate the potential for lateral or local spread between infected poultry farms in BC. Five genetically distinct viral clusters were identified during this period. Among the 31 noncommercial premises, 27 (87.1%) were classified as independent introductions, while four premises (12.9%) were attributed to local spread within 10 km of a phylogenetically connected infected premises (IP). No lateral spread events were identified among noncommercial premises. All infected noncommercial premises housed birds with outdoor access, emphasizing their susceptibility to wild bird exposure. Of the 127 infected commercial poultry premises, 21 (16.5%) were classified as independent introductions, 82 (64.6%) as local spread, 18 (14.2%) with potential for lateral spread, five (3.9%) with potential for both local and/or lateral spread, and one (0.8%) for which sequencing was unavailable. Local spread emerged as a prominent feature, with most IP in proximity to one another having genetically similar viruses. Results suggest that proximity (<200 m) to an IP was a more reliable predictor of future infection status than contact with an IP. These findings underscore the critical value of combining field and GE to understand outbreak dynamics comprehensively. This integrative approach improves resource allocation, informs targeted containment strategies, and supports the need for effective biosecurity measures to mitigate future risks, particularly in densely populated poultry production regions. Robust interventions are needed to address both independent introductions and secondary spread pathways.

Figures