A Multi-model Counterfactual Analysis of Human Brucellosis in China: Spatiotemporal Epidemiology and Heterogeneous...

This study examined long-term spatiotemporal patterns of human brucellosis in China and assessed whether reported incidence in 2020 differed from expected levels during the COVID-19 pandemic.

We analyzed surveillance data from 31 provincial-level regions in mainland China from 2004 to 2020. Joinpoint regression assessed national trends. Seasonal indices, spatial autocorrelation, and space-time scan statistics were used to examine monthly variation and geographic clustering. Province-specific forecasting models estimated expected case counts.

National incidence increased from 2004 to 2014 and declined thereafter, with an overall average annual percent change (AAPC) of 6.03% across the full 2004-2020 study period. Cases regularly peaked from March to July. Farmers and herders accounted for most cases, while adults aged ≥45 years represented an increasing share. Persistent hotspots were concentrated in northern pastoral and agro-pastoral provinces. The largest space-time cluster included Inner Mongolia and Shanxi during 2008-2015. In 2020, observed cases varied from forecast levels across provinces and seasons. Qinghai and Anhui showed excess cases in both peak and non-peak seasons, while Guizhou exhibited more pronounced excess in the non-peak season; by contrast, Heilongjiang reported consistently fewer cases than expected.

Human brucellosis in China remains seasonal, spatially concentrated, and increasingly linked to older livestock workers. Prevention should focus on persistent endemic hotspots, periods of intensified livestock-related exposure, particularly the livestock birthing season, and sustained surveillance capacity during public health emergencies.