Abstract
African swine fever (ASF), a highly lethal viral disease with no effective vaccines or treatments, poses a significant threat to the global pig industry. Since its first report in China in August 2018, it has spread rapidly, severely impacting China's pig industry. This study developed a Bayesian spatiotemporal model to explore ASF's spatiotemporal patterns, assess relative risk (RR), and identify key factors, aiming to inform targeted prevention strategies. Data (disease-related deaths, pig inventory, GDP, temperature, and 6 other factors) were collected from 31 mainland Chinese provinces from August 2018 to December 2019. The INLA algorithm estimated parameters, with the optimal model selected via DIC and WAIC. Multicollinearity was addressed using VIF and Spearman's correlation coefficient. Univariate and multivariate models quantified factor effects, with risk classified by natural breaks. Significant spatiotemporal patterns emerged: high-risk clusters in Liaoning, Heilongjiang, and Beijing, lower risk in Yunnan and Chongqing. Economic factors and veterinary resources were crucial: GDP per capita correlated positively (RR = 1.8814, 95% CI: 1.1264, 3.1362), while veterinarian numbers correlated inversely (RR = 0.7233, 95% CI: 0.4776, 0.9637). This study clarifies ASF dynamics and influencing factors in China, highlighting the need to strengthen veterinary services and balance economic development with biosecurity, offering a global reference for infectious disease management.