Patterns of Avian Influenza Virus detection from active surveillance in wild birds: A systematic review and meta-analysis

Abstract

We summarised active surveillance data to identify patterns and ecological factors correlated with AIV detection in wild birds, through a systematic review and meta-analysis. We screened 2851 articles from the PubMed and Scopus databases, out of which 197 met our eligibility criteria and were selected for further analysis. The dataset encompassed 367 wild bird species from 72 avian families. The pooled prevalence of sampled birds was 4.80% (95% CI: [3.91-5.77%]) across 900,469 samples collected during the period 1971-2023, with substantial heterogeneity (I2 = 99.7%) across studies. Among continents, Central America displayed the highest prevalence at 9.89% (95% CI: [2.25-21.99%]), albeit based on relatively few samples (n=4205). In temperate regions, prevalence peaked in autumn, at 5.82% (95% CI: [3.84-8.15%]), while in tropical regions, prevalence was notably higher during the dry season (2.33%, 95% CI: [0.38-5.54%]) than during the wet season (0.22%, 95% CI: [0.00-0.90%]). Prevalence varied significantly across avian families, with Anatidae, the most extensively sampled family, exhibiting a prevalence of 6.19% (95% CI: [5.10-7.37%]). Migratory species and those associated with freshwater habitats also exhibited higher AIV detection. Moreover, meta-regression analyses revealed that seasonal patterns of AIV detection differed across regions. Despite this, high heterogeneity across studies remained, likely driven by differences in surveillance intensity, diagnostic methods, and unmeasured ecological factors. This meta-analysis highlights key spatial, taxonomic, and temporal patterns in AIV prevalence among sampled birds. The findings underscore the need for harmonised, representative surveillance to better anticipate emerging avian influenza risks.