Application of seasonal autoregressive integrated moving average (SARIMA) model in prediction of dengue cases in Kuantan

A retrospective study was carried out using epidemiological data in Kuantan, Pahang. The confirmed dengue cases from the year 2011 to 2018 was retrieved and analyzed using time series analysis. The time series model could potentially provide useful information that could be further used to facilitate the planning of public health interventions in an effort to minimize dengue outbreaks. The objective of this study was first, to assess the application of Seasonal Autoregressive Integrated Moving Average (SARIMA) model to predict dengue cases for 2019 by analyzing the trend of confirmed dengue cases (2011-2018); secondly, to compare the model’s parameters accuracy to be used in the prediction of monthly dengue cases in Kuantan for the year 2019 and lastly, to perform the forecast of dengue cases using SARIMA with Box Jenkins approach. The model was fitted with monthly confirmed dengue cases (20112018) and validation of the prediction was made using the dengue cases from January to December 2018. The study revealed that SARIMA (0, 1, 0) (3, 0, 2)12 was the best fit model and could be used to extrapolate the cases up to twelve months in advance. The prediction of the cases in 2019 was relatively close to the actual cases within the confidence interval limit. Thus, the model derived from this study has the capability to not only forecast but also anticipate the future dengue cases. This would in turn enhance the current intervention program which is vital in minimizing the burden of the disease in Kuantan specifically.