Identifying factors of spatial heterogeneity in COVID-19 incidence for Dhaka Division, Bangladesh

Abstract

The modern world is dealing with a novel infectious disease known as COVID-19, which is rapidly spreading from one person to another or from the environment, causing illness in our communities. The application of geospatial and statistical methods allows researchers to gain a better grasp of the nature of this disease's spatial and temporal dissemination. However, there is a paucity of research on employing the Bayesian framework of spatial modelling for COVID-19 and identifying sources of regional heterogeneity in reported coronavirus (COVID-19) disease. Sociodemographic factors, environmental factors, socio-economic factors, and contact patterns are among the postulated mechanisms that generate regional variation in reported coronavirus (COVID-19) infections. Through the Bayesian framework of the spatial conditional model, we evaluate the impact of socio-demographic factors on observed variance in COVID-19 standardized morbidity rates (SMRs) across seventeen districts of Dhaka divisions. The conditional approach uses random effects in a generalized linear mixed model (GLMM) to account for geographic similarity between observations. We created two types of intrinsic conditional autoregressive (ICAR) models and discovered that the first type performed better in explaining variation in COVID-19 SMR rates throughout the Dhaka division in terms of lowest deviation information requirement (DIC). The results of this model revealed that COVID-19 SRMs were higher in Dhaka, Faridpur, and Rajbari districts, and that only transmission mechanisms-population density was a relevant factor in explaining the change in COVID-19 standardized morbidity rates across the study period. The outcomes of this study will aid in gaining a better understanding of the COVID-19 situation and will serve as a guideline for developing an effective action plan to control and prevent disease spread.