A Meta-Analysis of the Prevalence of Influenza A H5N1 and H7N9 Infection in Birds

Despite a much higher rate of human influenza A (H7N9) infection compared to influenza A (H5N1), and the assumption that birds are the source of human infection, detection rates of H7N9 in birds are lower than those of H5N1. This raises a question about the role of birds in the spread and transmission of H7N9 to humans. We conducted a meta-analysis of overall prevalence of H5N1 and H7N9 in different bird populations (domestic poultry, wild birds) and different environments (live bird markets, commercial poultry farms, wild habitats). The electronic database, Scopus, was searched for published papers, and Google was searched for country surveillance reports. A random effect meta-analysis model was used to produce pooled estimates of the prevalence of H5N1 and H7N9 for various subcategories. A random effects logistic regression model was used to compare prevalence rates between H5N1 and H7N9. Both viruses have low prevalence across all bird populations. Significant differences in prevalence rates were observed in domestic birds, farm settings, for pathogen and antibody testing, and during routine surveillance. Random effects logistic regression analyses show that among domestic birds, the prevalence of H5N1 is 47.48 (95% CI: 17.15-133.13, P<0.001) times higher than H7N9. In routine surveillance (where surveillance was not conducted in response to human infections or bird outbreaks), the prevalence of H5N1 is still higher than H7N9 with an OR of 43.02 (95% CI: 16.60-111.53, P<0.001). H7N9 in humans has occurred at a rate approximately four times higher than H5N1, and for both infections, birds are postulated to be the source. Much lower rates of H7N9 in birds compared to H5N1 raise doubts about birds as the sole source of high rates of human H7N9 infection. Other sources of transmission of H7N9 need to be considered and explored.