Identifying areas for infectious animal disease surveillance in the absence of population data: Highly pathogenic avian influenza in wild bird populations of Europe

Abstract

A large number (n = 591) of H5N1 highly pathogenic avian influenza virus (HPAIV) outbreaks have been reported in wild birds of Europe from October 2005 through January 2009. Consequently, prevention and control strategies have been implemented in response to the outbreaks and considerable discussion has taken place regarding the need for implementing surveillance programs in high-risk areas with the objective of early detecting and preventing HPAIV epidemics. However countries ability to define the temporal and spatial extension of the high risk areas has been impaired by the lack of information on the distribution of susceptible wild bird populations in the region. Here, a technique for the detection of time–space disease clustering that does not require information on the distribution of susceptible populations and that has been referred to as the time–space permutation model of the scan statistic was used to identify areas and times of the year in which epidemics of H5N1 HPAIV were most likely to occur in wild bird populations of Europe from October, 2005, through December, 2008. The scan statistic was parameterized considering pre-existing knowledge on the epidemiological and ecological characteristics of the disease in the region. Robustness of the results was assessed using a generalized linear regression model to compare the outcomes of 36 alternative parameterizations of the scan statistic. Ten significant time–space clusters of H5N1 HPAI outbreaks were detected in six European countries. Results were sensitive (P < 0.05) to the definition of the maximum spatial size defined for the clusters. Results presented here will help to identify high risk areas for HPAIV surveillance in the European Union. Assumptions, results, and implications of the analytical model are extensively presented and discussed in order to facilitate the use of this approach for the identification of high risk areas for infectious animal disease surveillance in the absence of population data.