PhD Confirmation of Candidature Seminar – Aswi

Bayesian Spatio-Temporal Modelling of Small Areas: Dengue Fever in Makassar Indonesia

Student: ASWI

When: Wednesday, 9th August 2017  10:00 AM-11:00 AM

Where: GP-M Block, Level 3, Room 312


  • Prof Kerrie Mengersen (Principal)
  • A/Prof Wenbiao Hu
  • Dr Gentry White
  • Dr Susanna Cramb

Panel Members:

  • Kerrie Mengersen (Chair)
  • A/Prof Wenbiao Hu
  • Dr Gentry White
  • Dr Susanna Cramb
  • Prof You-Gan Wang
  • Prof John Aaskov


Dengue fever affects more than one hundred million people every year and is one of the most important mosquito-borne diseases in the world. Dengue fever is still a serious health problem in several countries including Indonesia. Makassar, Indonesia which consists of 14 geographic areas had 6882 new cases of dengue registered from 2002 to 2015, and there have been fluctuations over time. Some papers have employed Bayesian spatial and spatio-temporal models using CAR priors to model dengue fever. As the estimation of the risk in any area depends on neighbouring areas, then the number of areas may influence the spatial and spatio-temporal estimation. Therefore, it is important to evaluate the impact of the number of areas on spatial and spatio-temporal estimation. While the choice of spatial model on these estimates has been evaluated for larger numbers of areas, it has not apparently been evaluated for situations in which the number of areas in the region is small. There are specific limitations and challenges in the Makassar data in terms of the number of areas, the different characteristics of data for incidence and time to event data. This motivates the development and expansion of these approaches for the type of problems faced in this project.

The overall aim of this research is to develop Bayesian models that reveal spatial and spatio-temporal trends in disease for regions with a small number of areas, with a focus on understanding dengue fever in Makassar Indonesia and develop a Bayesian model to describe spatial time to event for hospitalisations of dengue. These objectives will be attained by reviewing the relevant literature on dengue fever with a focus on Bayesian spatial and spatio-temporal modelling and multiple simulation studies which consider different forms of modelling spatial and spatio-temporal structure for a range of numbers of areas. Finally, the appropriate models will be applied to dengue fever incidence and time to event data in Makassar Indonesia. Data will be collected from the city health department from the ministry of health of South Sulawesi Province, and from selected hospitals in Makassar.

There are a number of significant aspects of this research. First, to our knowledge this is the first study to investigate the impact of a small number of areas on the estimation of spatial and spatio-temporal models. This study will provide guidance on which Bayesian models are appropriate for a given number of areas in spatial and spatio-temporal analyses. Results from this study are expected to advance the field of Bayesian spatial statistics. Second, this will be the first time that spatial and spatio-temporal dengue incidence data and time to event data have been jointly modelled using Bayesian methods and applied to Makassar data. This research will help to understand the dynamics of dengue in Makassar, and guide targeted interventions to reduce the disease. Third, this research will provide new insights into the variation in duration of hospitalisation for dengue fever using time-to-event models.



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s