Abstract
Understanding the spatial characteristics of dengue fever (DF) incidences is crucial for governmental agencies to implement effective disease control strategies. We investigated the associations between environmental and socioeconomic factors and DF geographic distribution, are proposed a probabilistic risk assessment approach that uses threshold-based quantile regression to identify the significant risk factors for DF transmission and estimate the spatial distribution of DF risk regarding full probability distributions. To interpret risk, return period was also included to characterize the frequency pattern of DF geographic occurrences. The study area included old Kaohsiung City and Fongshan District, two areas in Taiwan that have been affected by severe DF infections in recent decades. Results indicated that water-related facilities, including canals and ditches, and various types of residential area, as well as the interactions between them, were significant factors that elevated DF risk. By contrast, the increase of per capita income and its associated interactions with residential areas mitigated the DF risk in the study area. Nonlinear associations between these factors and DF risk were present in various quantiles, implying that water-related factors characterized the underlying spatial patterns of DF, and high-density residential areas indicated the potential for high DF incidence (e.g., clustered infections). The spatial distributions of DF risks were assessed in terms of three distinct map presentations: expected incidence rates, incidence rates in various return periods, and return periods at distinct incidence rates. These probability-based spatial risk maps exhibited distinct DF risks associated with environmental factors, expressed as various DF magnitudes and occurrence probabilities across Kaohsiung, and can serve as a reference for local governmental agencies.
Highlights
Dengue fever (DF) is among the most severe vector-borne infectious diseases spread by mosquitoes in tropical and subtropical regions; 2.5 billion people in over 100 countries are at risk of contracting the dengue virus [1,2,3,4,5]
Numerous DF studies have focused on determining the DF etiology to facilitate space–time prediction, a risk assessment framework to account for DF risk factors and provide risk measures across space must be established to control and manage the disease
This paper proposes a threshold-based quantile regression approach to investigate the functional relationships between the spatial distributions of environmental risk factors, including land use and socioeconomic factors, and DF incidence
Summary
Dengue fever (DF) is among the most severe vector-borne infectious diseases spread by mosquitoes in tropical and subtropical regions; 2.5 billion people in over 100 countries are at risk of contracting the dengue virus [1,2,3,4,5]. Numerous DF studies have focused on determining the DF etiology to facilitate space–time prediction, a risk assessment framework to account for DF risk factors and provide risk measures across space (e.g., the probability and magnitude of DF incidences) must be established to control and manage the disease. Populated areas close to vector-preferred areas exhibit increased DF occurrence risk [15,19,20,21]. Previous studies have indicated that high population densities, residential areas, and areas with low income family exhibit increased risk of DF transmission [22,23,24,25]. The statistical relationships among DF incidence, land use, and socioeconomic factors frequently change across the study area because of the distinct environmental and climatic conditions. We used quantile regression to characterize the relationships between risk factors and DF incidence across the quantiles of the DF incidence probability distribution
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
