The fuzzy system ensembles entomological, epidemiological, demographic and environmental data to unravel the dengue transmission risk in an endemic city

• Published in: BMC public health 2024-09, Vol.24 (1), p.2587-14, Article 2587
• Main Authors: de Souza Leandro, André, de Oliveira, Felipe, Lopes, Renata Defante, Rivas, Açucena Veleh, Martins, Caroline Amaral, Silva, Isaac, Villela, Daniel A M, Teixeira, Marcello Goulart, Xavier, Samanta Cristina das Chagas, Maciel-de-Freitas, Rafael
• Format: Article
• Language: English
• Subjects: Aedes; Aedes - virology; Aedes aegypti; Analysis; Animals; Aquatic insects; Arboviruses; Brazil - epidemiology; Cities - epidemiology; Classification; Data analysis; Decision making; Demographic variables; Demographics; Demography; Dengue; Dengue - epidemiology; Dengue - transmission; Dengue fever; Disease Outbreaks; Disease transmission; Distribution; Effectiveness; Endemic Diseases - statistics & numerical data; Environmental aspects; Epidemiology; Fuzzy algorithms; Fuzzy control; Fuzzy Logic; Fuzzy systems; Heterogeneity; Hotspots; Humans; Indicators; Mosquito Vectors; Mosquitoes; Pest outbreaks; Public health; Risk Assessment; Risk factors; Spatial data; Spatiotemporal data; Surveillance; Surveillance systems; Uncertainty analysis; Vector control; Vector-borne diseases
• ISSN: 1471-2458; 1471-2458
• DOI: 10.1186/s12889-024-19942-4

The effectiveness of dengue control interventions depends on an effective integrated surveillance system that involves analysis of multiple variables associated with the natural history and transmission dynamics of this arbovirus. Entomological indicators associated with other biotic and abiotic parameters can assertively characterize the spatiotemporal trends related to dengue transmission risk. However, the unpredictability of the non-linear nature of the data, as well as the uncertainty and subjectivity inherent in biological data are often neglected in conventional models. As an alternative for analyzing dengue-related data, we devised a fuzzy-logic approach to test ensembles of these indicators across categories, which align with the concept of degrees of truth to characterize the success of dengue transmission by Aedes aegypti mosquitoes in an endemic city in Brazil. We used locally gathered entomological, demographic, environmental and epidemiological data as input sources using freely available data on digital platforms. The outcome variable, risk of transmission, was aggregated into three categories: low, medium, and high. Spatial data was georeferenced and the defuzzified values were interpolated to create a map, translating our findings to local public health managers and decision-makers to direct further vector control interventions. The classification of low, medium, and high transmission risk areas followed a seasonal trend expected for dengue occurrence in the region. The fuzzy approach captured the 2020 outbreak, when only 14.06% of the areas were classified as low risk. The classification of transmission risk based on the fuzzy system revealed effective in predicting an increase in dengue transmission, since more than 75% of high-risk areas had an increase in dengue incidence within the following 15 days. Our study demonstrated the ability of fuzzy logic to characterize the city's spatiotemporal heterogeneity in relation to areas at high risk of dengue transmission, suggesting it can be considered as part of an integrated surveillance system to support timely decision-making.