Spatio-temporal Modelling (4.42)

This research Project investigates the best way to represent the spatial disparity of diseases, to enable decision makers the best access to localised spatial health data. The initial focus of the Project is on cancer, with the results then to be applied to other diseases.

Better management of disease will be achieved through the development and application of new ways of

  • small area disease mapping
  • modelling disease and potential risk factors
  • visualising the results of these models linking this new information on the spatial and temporal distribution of disease and associated risk factors with health service utilisation

These research results will be more easily communicated through the visualisation tool developed in Project 4.4.1.

Highlights

Nicole’s work on understanding service provision has led to the development of a model that jointly predicts individual service catchment areas and the rate of use for each service. This represents an extension from previous project work that focussed solely on modelling rates of service use. This new methodology is also able to estimate outcomes at a smaller geographic scale.

Finally, the model developed jointly models the unexplained demand and capacity of each service, and how these outputs are clustered across services. This work represents a novel methodological contribution to the literature.

Susanna Cramb has completed work on the joint spatial modelling of multiple cancers (breast cancer, colorectal cancer and melanoma) by cancer stage (early vs. advanced). The joint modelling of multiple cancers allows for a sharing of mutual information, in this case relative risk of cancer diagnosis, and thus represents a novel approach to improving estimated of cancer risk. A selection of results from the model developed are being presented as a poster at the SSAI Young Statisticians Conference in February 2013.

Read more and comment via the Blog...

The cost to our society of cancer is measured not only by the cost of providing health services, but also by the overall burden on society. To address this huge burden we must have reliable information about where the current and predicted areas of high cancer risk are together with, cancer risk factors and health service utilisation – a scenario that spatial technology and models are built to handle.

Project Participants

Research & Education   -   Curtin University   -   QUT   -   TICHR

Government   -   Dept Health WA   -   Cancer Council Queensland