Spatial Distribution and Risk Factors of Highly Pathogenic Avian Influenza (HPAI) H5N1 in China

• DOI: 10.1371/journal.ppat.1001308
• Publication Date: 2011-03-03T21:29:05Z
• Authors: Vincent Martin; Dirk U. Pfeiffer; Xiaoyan Zhou; Xiangming Xiao; Diann J. Prosser; Fusheng Guo; Marius Gilbert
• Keywords: Population density; Highly pathogenic

Highly pathogenic avian influenza (HPAI) H5N1 was first encountered in 1996 Guangdong province (China) and started spreading throughout Asia the western Palearctic 2004–2006. Compared to several other countries where HPAI distribution has been studied some detail, little is known about environmental correlates of China. clinical disease outbreaks, virus (HPAIV) isolated from active risk-based surveillance sampling domestic poultry (referred as HPAIV positives this manuscript) were modeled separately using seven risk variables: chicken, waterfowl population density, proportion land covered by rice or surface water, cropping intensity, elevation, human density. We used bootstrapped logistic regression boosted trees (BRT) with cross-validation identify weight each variable, assess predictive power models, map risk. outbreak occurrence mainly associated chicken elevation. In contrast, infection identified water. Both models had a high explanatory (mean AUC ranging 0.864 0.967). The based on data emphasized areas south Yangtze River, while reported extended further North, density humans higher. quantified statistical association between outbreak, post-vaccination levels seropositivity (percentage effective seroconversion vaccinated birds) found that provinces either outbreaks 2007–2009 appeared have lower antibody response vaccination. China appears more limited geographically than previously assessed, offering prospects for better targeted control interventions.