Temperature dominates dengue transmission in Thailand: Machine learning reveals critical thresholds and COVID-19 disruption

BackgroundDengue fever remains a critical public health challenge in Thailand, with transmission dynamics driven by complex interactions between environmental and socioeconomic factors. Understanding these drivers is essential for developing robust prediction systems. MethodsWe developed a machine learning framework to classify spatiotemporal dengue risk and identify key drivers of transmission across Thailand. We analyzed 20 years of monthly dengue hemorrhagic fever surveillance data (2003-2022) from 77 provinces, integrating 54 environmental, climatic, and socioeconomic variables. Using eXtreme Gradient Boosting (XGBoost) with SHapley Additive exPlanations (SHAP), we classified provinces as high-risk or low-risk based on the national median incidence. The dataset was stratified into training (2003-2016) and testing periods, with the latter subdivided into pre-COVID-19 (2017-2019), COVID-19 (2020-2021), and post-COVID-19 (2022) phases. ResultsThe model achieved robust performance with an area under the curve (AUC) of 0.94 during training and 0.80 in pre-pandemic testing. Temperature emerged as the dominant predictor, with temperature-related variables comprising seven of the ten most influential features. Critical transmission thresholds were identified at approximately 21{degrees}C for a 1-month lagged minimum temperature and approximately 32{degrees}C for a 3-month lagged maximum temperature. Interestingly, precipitation contributed minimally to model predictions, while a higher Gross Provincial Product was associated with an increased risk of dengue, reflecting urban transmission patterns. Model performance deteriorated significantly during the COVID-19 pandemic (AUC = 0.62 in 2021), with systematic overprediction suggesting that behavioral factors outweighed environmental drivers during the pandemic disruption. ConclusionsTemperature, particularly with lags of 1-3 months, is the primary predictor of dengue risk in Thailand. The pandemic-induced disruption of model accuracy underscores the crucial role of human behavioral factors in influencing dengue transmission dynamics. Our results challenge traditional precipitation-focused models and highlight the importance of temperature-driven approaches for dengue prediction in Thailand. Plain language summaryThis study used machine learning to predict dengue fever outbreaks across Thailands provinces from 2003 to 2022. By analyzing climate data, economic indicators, and satellite imagery, our machine learning model could accurately identify high-risk areas about 80% of the time. We found that temperature is the most important factor determining where and when dengue will spread. Additionally, we identified a critical temperature threshold where dengue transmission essentially stops when minimum temperatures drop below approximately 21{degrees}C. Surprisingly, rainfall patterns, which are often emphasized in dengue predictions, played a much smaller role than expected. Higher economic development was also found to be associated with an increased risk of dengue, probably due to urbanization creating ideal conditions for mosquitoes to breed all year round and higher human population density that facilitates virus transmission. However, the COVID-19 pandemic disrupted the models accuracy, causing it to predict more dengue cases than actually occurred from 2020 to 2022, likely due to the COVID-19 control measures that climate data alone could not capture. This research demonstrates both the power and limitations of using environmental and socioeconomic data to predict dengue outbreaks in Thailand.