Your browser doesn't support javascript.
loading
Spatiotemporal hierarchical Bayesian analysis to identify factors associated with COVID-19 in suburban areas in Colombia.
Cortes-Ramirez, J; Wilches-Vega, J D; Caicedo-Velasquez, B; Paris-Pineda, O M; Sly, P D.
Afiliación
  • Cortes-Ramirez J; Centre for Data Science. Queensland University of Technology, Australia.
  • Wilches-Vega JD; Faculty of Medical and Health Sciences, University of Santander, Colombia.
  • Caicedo-Velasquez B; Children's Health and Environment Program, Child Health Research Centre, The University of Queensland, Australia.
  • Paris-Pineda OM; Faculty of Medical and Health Sciences, University of Santander, Colombia.
  • Sly PD; Epidemiology Research Group, Faculty of Public Health, University of Antioquia, Colombia.
Heliyon ; 10(9): e30182, 2024 May 15.
Article en En | MEDLINE | ID: mdl-38707376
ABSTRACT

Introduction:

The pandemic had a profound impact on the provision of health services in Cúcuta, Colombia where the neighbourhood-level risk of Covid-19 has not been investigated. Identifying the sociodemographic and environmental risk factors of Covid-19 in large cities is key to better estimate its morbidity risk and support health strategies targeting specific suburban areas. This study aims to identify the risk factors associated with the risk of Covid-19 in Cúcuta considering inter -spatial and temporal variations of the disease in the city's neighbourhoods between 2020 and 2022.

Methods:

Age-adjusted rate of Covid-19 were calculated in each Cúcuta neighbourhood and each quarter between 2020 and 2022. A hierarchical spatial Bayesian model was used to estimate the risk of Covid-19 adjusting for socioenvironmental factors per neighbourhood across the study period. Two spatiotemporal specifications were compared (a nonparametric temporal trend; with and without space-time interaction). The posterior mean of the spatial and spatiotemporal effects was used to map the Covid-19 risk.

Results:

There were 65,949 Covid-19 cases in the study period with a varying standardized Covid-19 rate that peaked in October-December 2020 and April-June 2021. Both models identified an association of the poverty and stringency indexes, education level and PM10 with Covid-19 although the best fit model with a space-time interaction estimated a strong association with the number of high-traffic roads only. The highest risk of Covid-19 was found in neighbourhoods in west, central, and east Cúcuta.

Conclusions:

The number of high-traffic roads is the most important risk factor of Covid-19 infection in Cucuta. This indicator of mobility and connectivity overrules other socioenvironmental factors when Bayesian models include a space-time interaction. Bayesian spatial models are important tools to identify significant determinants of Covid-19 and identifying at-risk neighbourhoods in large cities. Further research is needed to establish causal links between these factors and Covid-19.
Palabras clave
Texto completo
Añadir a Mi BVS
Imprimir
XML
PubMed Links
Buscar en Google

Texto completo: 1 Colección: 01-internacional Base de datos: MEDLINE País/Región como asunto: America do sul / Colombia Idioma: En Revista: Heliyon Año: 2024 Tipo del documento: Article País de afiliación: Australia Pais de publicación: Reino Unido
Texto completo
Añadir a Mi BVS
Imprimir
XML
PubMed Links
Buscar en Google

Texto completo: 1 Colección: 01-internacional Base de datos: MEDLINE País/Región como asunto: America do sul / Colombia Idioma: En Revista: Heliyon Año: 2024 Tipo del documento: Article País de afiliación: Australia Pais de publicación: Reino Unido