A bootstrap method based on linear regression to estimate COVID-19 Ecological Risk in Catalonia

Abstract

Background: SARS-CoV-2 is a new type of coronavirus that causes COVID-19. It is affecting the entire planet. Despite the widespread use of ecologic analysis in epidemiologic research and health planning, health scientists and practitioners have given little attention to the methodological aspects of this approach. The study of risk factors linked to the COVID-19 pandemic is one of the most current and exciting topics for epidemiologists. These risks in many cases are unknown. This research covers the study of risk factors in the case of COVID-19 and proposes the use of an ecologic method known to epidemiologists in the case of aggregated data. The present study aims to compute a model that allows to easily calculate the risk of infection in different types of populations, using aggregated data to approximate the individual risk of COVID-19 transmission by a person. Methods: The case of Catalonia, in Spain, is presented as an example, as it is one of the areas where the incidence of the virus among the population is being higher. The proposed method is known as an ecological study and is based on the statistical regression model between the incidence (or variable that represents it) and the risk factors but using aggregated data and obtaining a risk ratio (RR). Results: The results obtained have made it possible to find the risk of contracting COVID-19 concerning risk factors for high family income (RR=1.157491), more mobility (RR=1.065475), and high density of population (1.00002). Conclusions: This method could be used to design an app that predicts how the risk will evolve and calculate the risk of contagion in one area or another to take the proper action. The calculated RR can help us to understand how the variables become risks or protective factors at an ecological level (understanding aggregate data).