One Health Investigation of a Crow mortality event linked to Potential Circulation of Highly Pathogenic Avian Influenza Virus at Live Bird Markets in Northwestern Bangladesh

Background: Highly pathogenic avian influenza viruses can cause extremely rapid deaths in wild, non-anseriforme birds, and its epidemiology is closely linked with diet, foraging behavior, habitat use, migratory pattern, population size and frequency of aggregation. A One Health, multidisciplinary investigation was initiated by the Government of Bangladesh in response to a crow mortality event near Rajshahi Medical College Hospital (RMCH) between 12-18th January 2017.The aim was to identify the etiologic agent, the extent of the outbreak and identify possible associated human infections. Methods: The wildlife team surveyed crow roosts in the area and observed crows feeding on poultry offal and wastage in neighboring live-bird markets (LBMs) and roosting with other species of wild birds. The investigation included sampling crows, LBMs (environmental samples and offal), poultry feces on farms, and patients as part of existing hospital based influenza surveillance in RMCH, Rajshahi.The team collected cloacal and oropharyngeal swabs from moribund and dead crows (N=58), environmental samples from 12 live bird markets (LBMs) (N=137) and 14 poultry farms (N=65) as well as samples from 28 severe acute respiratory illness (SARI) patients as part of existing hospital-bases influenza surveillance. All samples were tested by real-time RT-PCR for Influenza-A (M gene) and for H5, H7 and H9 and N1. Results: Ninety-one percent of dead or moribund crows (n=58 95% CI: 81-97%) were positive for A/H5N1 and 2% (n=1; 95% CI: 0-9%) for influenza A and were untypable. Among LBMs samples, 16% were positive to H5N1 (n=22; 95% CI: 10-23%), 4% for A/H9 (n=6; 95% CI: 2-9%), 2% had a co-infection with both A/H5 and H9 (n=3; 95% CI: 1-6%) and 12% for influenza A/untypable (n=16; 95% CI: 7-18%). None of the crow samples were positive for H7 or H9 and in LBMs no samples tested positive for H7. Phylogenetic analysis based on partial sequence of the HA gene suggests that it is similar to the Bangladeshi 2.3.2.1a clade that previously circulated in 2011. All human samples and poultry farm samples were negative for influenza A. Conclusions: This study provides evidence of an influenza contaminated LBM environment. Our findings suggest that influenza viruses are circulating in LBMs and crows may have been infected by consuming dead poultry and their waste. Continuation of surveillance in wild birds and live bird market to better understand the evolution of influenza viruses, transmission pathways and potential sources of infection in crows and the risk of spillover highly pathogenic influenza virus to people