Spatio-temporal dengue risk modelling in the south of Thailand: a Bayesian approach to dengue vulnerability

PeerJ

article

Fatima Ibrahim Abdulsalam¹  Pablo Antúnez²  Warit Jawjit¹ 

[1] School of Public Health, Walailak University;División de Estudios de Postgrado, Universidad de la Sierra Juárez

关键词: Bayesian inference;    CARBayesST;    Climate variables;    Dengue risk;    Nakhon Si Thammarat;    Spatio-temporal modelling;   

DOI  :  10.7717/peerj.15619

学科分类：社会科学、人文和艺术（综合）

来源: Inra

【 摘 要 】

BackgroundMore than half of the global population is predicted to be living in areas susceptible to dengue transmission with the vast majority in Asia. Dengue fever is of public health concern, particularly in the southern region of Thailand due to favourable environmental factors for its spread. The risk of dengue infection at the population level varies in time and space among sub-populations thus, it is important to study the risk of infection considering spatio-temporal variation.MethodsThis study presents a joint spatio-temporal epidemiological model in a Bayesian setting using Markov chain Monte Carlo (MCMC) simulation with the CARBayesST package of R software. For this purpose, monthly dengue records by district from 2002 to 2018 from the southern region of Thailand provided by the Ministry of Public Health of Thailand and eight environmental variables were used.ResultsResults show that an increasing level of temperature, number of rainy days and sea level pressure are associated with a higher occurrence of dengue fever and consequently higher incidence risk, while an increasing level of wind speed seems to suggest a protective factor. Likewise, we found that the elevated risks of dengue in the immediate future are in the districts of Phipun, Phrom Kili, Lan Saka, Phra Phrom and Chaloem Phakiat. The resulting estimates provide insights into the effects of covariate risk factors, spatio-temporal trends and dengue-related health inequalities at the district level in southern Thailand.ConclusionPossible implications are discussed considering some anthropogenic factors that could inhibit or enhance dengue occurrence. Risk maps indicated which districts are above and below baseline risk, allowing for the identification of local anomalies and high-risk boundaries. In the event of near future, the threat of elevated disease risk needs to be prevented and controlled considering the factors underlying the spread of mosquitoes in the Southeast Asian region.

【 授权许可】

CC BY