Assessing the temporal modelling for prediction of dengue infection in northern and north-eastern, Thailand.

This study aimed at developing a predicting model on the incidence rate of dengue fever in four locations of Thailand--i.e. the northern region, Chiang Rai province, the north-eastern region and Sisaket province--using time series analysis. Seasonal Autoregressive Integrated Moving Average (SARIMA) model was performed using data on monthly incidence rate of dengue fever from 1981 to 2009, and validated using the monthly rate collected for the period January 2010 to October 2011. The results show that the SARIMA(1,0,1)(0,1,1)12 model is the most suitable model in all locations. The model for all locations indicated that the predicted dengue incidence rate and the actual dengue incidence rate matched reasonably well. The model was further validated by the Portmanteau test with no significant autocorrelation between residuals at different lag times. Our findings indicate that SARIMA model is a useful tool for monitoring dengue incidence in Thailand. Furthermore, this model can be applied to surveillance data for early warning systems for control and reduction of dengue transmission.