Predictions of Relative Risks for Dengue Disease Mapping in Malaysia Based on Stochastic SIR-SI Vector-Borne Infectious Disease Transmission Model

Abstract

The main aim of this paper is to propose a procedure for predicting relative risk in order to suggest approximate future dengue disease progression patterns for Malaysia. Firstly, we discuss the posterior predictive distributions that are used to generate forecasts of relative risk. This description includes explanations of the simulating process that is required to generate posterior predictive distributions. Then, we apply this procedure to generate predictions for our case studies relating to dengue disease in Malaysia for discrete time-space data. Finally, we present the findings of our predictive analysis, comparing and displaying the forecasts of relative risk in graph, table and map. Results of the numerical analysis that we implemented to generate predictions of relative risk based on discrete time-space stochastic SIR-SI vector-borne infectious disease transmission models using dengue data of Malaysia suggest that the future forecasts of posterior predictive relative risk have strikingly similar patterns as the fitted posterior expected relative risks. This similarity shows the stationary process between the observed and the predicted values under our model assumptions. In practical application, we recommend that shorter-term forecasts should be used. These can then be updated continually as more incidence data become available.