Risk map for high pathogenicity avian influenza in 2022-2023 winter season in Japan

For high pathogenicity avian influenza (HPAI) outbreaks at poultry farms, both farm and regional characteristics have been reported as risk factors. Because the regional-level component of outbreak risk may reflect the level of virus contamination in the surrounding area, it may also serve as a proxy for exposure risk potentially shared by poultry farms and sympatric wildlife. We examined associations between HPAI outbreaks at chicken farms in Japan during October 2022-April 2023 and both farm factors (flock size, production type, and peripheral water bodies) and regional factors (land cover, meteorological conditions, and topography); we also estimated the regional-level component of outbreak risk and developed an HPAI risk map. We conducted a nationwide cross-sectional study using data from 68 affected chicken farms and 5496 unaffected farms, excluding remote southern islands. Multivariable analyses were performed using stochastic partial differential equation models accounting for spatial autocorrelation. Farm factors were forced into all models, and combinations of regional factors aggregated on 10-km grids were selected using wAIC. Models with ΔwAIC <2 were combined by stacking. Among the 59 models retained for stacking, four contributed to the final stacked prediction. Coefficient estimates from these models suggested that lower elevation, lower forest cover, and high TWI were associated with higher HPAI outbreak risk. Although further investigation is warranted to examine other factors potentially affecting HPAI outbreaks and to validate the regional risk index in wildlife, the regional-level component of outbreak risk may provide a useful basis for mapping HPAI risk potentially shared by poultry farms and sympatric wildlife.