Autoregressive spatial modeling of possible cases of dengue, chikungunya, and Zika in the capital of Northeastern Brazil

• Published in: Revista da Sociedade Brasileira de Medicina Tropical Vol. 54; p. e0223
• Main Authors: Costa, Silmery da Silva Brito, Branco, Maria dos Remédios Freitas Carvalho, Vasconcelos, Vitor Vieira, Queiroz, Rejane Christine de Sousa, Araujo, Adriana Soraya, Câmara, Ana Patrícia Barros, Fushita, Angela Terumi, Silva, Maria do Socorro da, Silva, Antônio Augusto Moura da, Santos, Alcione Miranda dos
• Format: Journal Article
• Language: English
• Published: Sociedade Brasileira de Medicina Tropical - SBMT 01.01.2021; Sociedade Brasileira de Medicina Tropical (SBMT)
• Subjects: Chikungunya; Dengue; Economic factors; Major; Socio-environmental factors; Spatial analysis; TROPICAL MEDICINE; Zika; Chikungunya; Spatial analysis; Economic factors; Socio-environmental factors; Dengue; Zika
• ISSN: 0037-8682; 1678-9849; 1678-9849
• DOI: 10.1590/0037-8682-0223-2021

INTRODUCTIONDengue, chikungunya, and Zika are a growing global health problem. This study analyzed the spatial distribution of dengue, chikungunya, and Zika cases in São Luís, Maranhão, from 2015 to 2016 and investigated the association between socio-environmental and economic factors and hotspots for mosquito proliferation. METHODSThis was a socio-ecological study using data from the National Information System of Notifiable Diseases. The spatial units of analysis were census tracts. The incidence rates of the combined cases of the three diseases were calculated and smoothed using empirical local Bayes estimates. The spatial autocorrelation of the smoothed incidence rate was measured using Local Moran's I and Global Moran's I. Multiple linear regression and spatial autoregressive models were fitted using the log of the smoothed disease incidence rate as the dependent variable and socio-environmental factors, demographics, and mosquito hotspots as independent variables. RESULTSThe findings showed a significant spatial autocorrelation of the smoothed incidence rate. The model that best fit the data was the spatial lag model, revealing a positive association between disease incidence and the proportion of households with surrounding garbage accumulation. CONCLUSIONSThe distribution of dengue, chikungunya, and Zika cases showed a significant spatial pattern, in which the high-risk areas for the three diseases were explained by the variable "garbage accumulated in the surrounding environment," demonstrating the need for an intersectoral approach for vector control and prevention that goes beyond health actions.