Association Between Number of Wild Birds Sampled for Identification of H5N1 Avian Influenza Virus and Incidence of the Disease in the European Union

Abstract

Fourteen European Union (EU) countries have been affected by highly pathogenic avian influenza (HPAI) of the subtype H5N1 since 2005. Many of the H5N1 HPAI cases have been identified in wild bird populations from aquatic habitats. The objective of this contribution was to assess the association between the H5N1 HPAI incidence, estimated by a co-kriging approach, and the number of wild birds sampled for H5N1 HPAI surveillance in EU countries. The probability that a sample for H5N1 HPAI surveillance collected in the EU was obtained from a given country was associated (P<0.01) with the probability that a H5N1 HPAI case reported in the EU was located in the same country. The relation between probability of sampling and probability of infection in the Netherlands and in Spain was significantly higher (P<0.01) than that estimated in other EU countries, suggesting that political or social factors may have selectively influenced the number of samples for H5N1 HPAI surveillance collected in those two countries. The approach presented here will be useful for post hoc comparison of the relation between probability or risk for disease and the intensity of surveillance activities among administrative divisions of a region.