MODELLING ENDEMIC DISEASES IN SPACE AND TIME

Abstract

Infectious diseases pose sizable disease burden, especially in less-developed countries. Control of endemic infectious diseases can be a useful starting point to reduce the burden of infectious diseases. Countries with long-standing surveillance system can readily use the rich data-stream it offers, while many other countries are still in the process of establishing such surveillance systems. In this thesis, we strive to develop endemic infectious diseases models in space and time to forecast future outbreaks, and to evaluate potential control policies. Frequentist and Bayesian statistical methods were implemented to propose novel machine-learning based methods for forecasting endemic infectious diseases including dengue, malaria, chickenpox, and so on, in countries with different climates and to make spatiotemporal predictions, which gives insights to resource planning. One public health control measure—school closure policy for an endemic disease—hand, foot and mouth disease was also evaluated to understand its effectiveness.