Spatial-temporal clustering and high-risk area identification of dengue in Thailand: a 20-year national analysis using space-time scan statistics

Abstract Dengue remains a major public health challenge in Thailand, with substantial geographic variation. This study analyzed national dengue surveillance data (2005–2024; 1,640,368 cases) across five consecutive 4-year intervals to identify spatial clustering patterns, detect space-time clusters, and characterize the evolution of high-risk areas. Province-level pooled incidence rates were assessed using Global Moran’s I, Local Indicators of Spatial Association (LISA), and SaTScan space-time scan statistics with discrete Poisson models. Global Moran’s I showed significant positive spatial autocorrelation in four periods (I = 0.211–0.473, p < 0.01) but not during 2017–2020 (I = 0.065, p = 0.190), indicating spatial fragmentation. LISA revealed a progressive decline in High-High cluster provinces from 11 to 2 and a geographic shift from Central-Eastern to Northern Thailand. SaTScan detected 13 statistically significant clusters; the most likely cluster occurred in Northern Thailand (2013–2016, relative risk = 2.33, p < 0.001). The cluster with the highest cumulative burden encompassed the Central-Western region (2009–2016, 149,010 cases). Dengue exhibits dynamically evolving geographic patterns, with Northern Thailand emerging as the primary high-risk area requiring priority intervention. These findings support geographically stratified control strategies.