Environmental suitability to Highly Pathogenic Avian Influenza in Kazakhstan using data on Eurasian outbreaks 2020 - 2024

Highly pathogenic avian influenza (HPAI) is a highly contagious disease of domestic, synanthropic, and wild birds that has demonstrated a sharp rise globally since 2020. The study here is intended to assess the quality of reporting HPAI outbreaks in Kazakhstan using outbreaks’ location in neighboring countries (2020 – 2024) as training data to build an ecological niche model of suitability to the disease in the country. Afghanistan, China, Hong Kong, Iran, Iraq, Pakistan and Russia were chosen as training countries for fitting a Maximum entropy (Maxent) model, which was then tested using the locations of HPAI outbreaks in Kazakhstan. The results suggest good performance of the model in explaining Kazakhstani outbreaks (test AUC = 0.88 vs AUC=0.93 for training data). HPAI outbreaks reported in Kazakhstan fall mainly in areas treated as highly suitable by the model. The number of outbreaks per each analyzed country is also well explained by a summary area of suitable places per country as predictor (p<0.001, R2=0.75), with Kazakhstan demonstrating no significant deviation from the found dependency. Results suggest that, although sensitivity of the surveillance system in the country is uncertain, it may be able to accomplish the objective of identifying regions affected by the virus incursions.