Abstract
AbstractBackground: Dengue fever is one of the world’s most important vector-borne diseases and it is still a major public health problem in the Asia-Pacific region including Indonesia. Makassar is one of the major cities in Indonesia where the incidence of dengue fever is still quite high. Since dengue cases vary between areas and over time, these spatial and temporal components should be taken into consideration. However, unlike many other spatio-temporal contexts, Makassar is comprised of only a small number of areas and data are available over a relatively short timeframe. The aim of this paper is to better understand the spatial and temporal patterns of dengue incidence in Makassar, Indonesia by comparing the performance of six existing spatio-temporal models, taking into account these specific data characteristics (small number of areas and limited small number of time periods) and to select the best model for Makassar dengue dataset.Methods: Six different Bayesian spatio-temporal conditional autoregressive (ST CAR) models were compared in the context of a substantive case study, namely annual dengue fever incidence in 14 geographic areas of Makassar, Indonesia, during 2002–2015. The candidate models included linear, ANOVA, separate spatial, autoregressive (AR), adaptive and localised approaches. The models were implemented using CARBayesST and the goodness of fit was compared using the Deviance Information Criterion (DIC) and Watanabe-Akaike Information Criterion (WAIC).Results: The six models performed differently in the context of this case study. Among the six models, the spatio-temporal conditional autoregressive localised model had a much better fit than other options in terms of DIC, while the conditional autoregressive model with separate spatial and temporal components performed worst. However, the spatio-temporal CAR AR had a much better fit than other models in terms of WAIC. The different performance of the models may have been influenced by the small number of areas.Conclusion: Different spatio-temporal models appeared to have a large impact on results. Careful selection of a range of spatio-temporal models is important for assessing the spatial and temporal patterns of dengue fever, especially in a context characterised by relatively few spatial areas and limited time periods.KeywordsBayesianConditional autoregressive priorsCARBayesSTSpatio-temporal models
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.