Forecasting Dengue Incidence Based on Climatic Factors Using Negative Binomial and Generalized Additive Model in Bandung City, Indonesia

Introduction: Dengue is an endemic disease influenced by climatic factors such as temperature, humidity, and rainfall. Climate-based forecasting of dengue outbreaks can aid disease control, as climatic factors influence the development of the Aedes aegypti vector.. Aim This study aims to explore the influence of climate factors on dengue incidence and compare the accuracy of predictions between negative binomial regression and generalized additive models. Method This study used a correlational design with monthly data on the number of dengue cases from the Bandung City Health Office and climate data from the Meteorology, Climatology, and Geophysics Agency during the period 2014–2023. Two statistical models were applied: negative binomial regression to handle overdispersion and a generalised additive model to capture non-linear relationships. Result The results of the study indicate that temperature, humidity, and rainfall significantly influence the incidence of dengue. In the negative binomial regression model, the coefficients for temperature, humidity, and rainfall show a significant positive influence on the number of cases (AIC = 1456.5, MAE = 98.96, and RMSE = 147.88). The generalised additive model yields the same AIC (1456.5) but is more accurate (MAE = 92.85 and RMSE = 137.37). Conclusion Climate factors, particularly temperature and rainfall, play a significant role in the occurrence of dengue. The generalised additive model is more accurate in predicting dengue outbreaks and can support more effective climate-based control policies.