Estimating the time of Highly Pathogenic Avian Influenza virus introduction into United States poultry flocks during the 2022/24 epizootic

Following confirmation of the first case of the ongoing U.S. HPAI H5N1 epizootic in commercial poultry on February 8, 2022, the virus has continued to devastate the U.S. poultry sector and the pathogen has since managed to cross over to livestock and a few human cases have also been reported. Efficient outbreak management benefits greatly from timely detection and proper identification of the pathways of virus introduction and spread. In this study, we used changes in mortality rates as a proxy for HPAI incidence in a layer, broiler and turkey flock together with diagnostic test results to infer within-flock HPAI transmission dynamics. Mathematical modeling techniques, specifically the Approximate Bayesian Computation algorithm in conjunction with a stochastic within-flock HPAI transmission model were used in the analysis. The time window of HPAI virus introduction into the flock (TOI) and the adequate contact rate (ACR) were estimated. Then, using the estimated TOI together with the day when the first HPAI positive sample was collected from the flock, we calculated the most likely time to first positive sample (MTFPS) which reflects the time to HPAI detection. The estimated joint (i.e., all species combined) median of the MTFPS for different flocks was six days, the joint median most likely ACR was 6.8 newly infected birds per infectious bird per day, the joint median R0 was 13 and the joint median number of test days per flock was two. These results were also grouped by species and by epidemic phase and discussed accordingly. We conclude that this findings from this and other related studies are beneficial for the different stakeholders in outbreak management. We recommend that combining TOI analysis with complementary approaches such as phylogenetic analyses is critically important for improved understanding of disease transmission pathways. The estimated parameters can also be used to parametrize mathematical models that can guide the design of surveillance protocols, risk analyses of HPAI spread, and emergency preparedness for HPAI outbreaks.