ABSTRACT
Background: The Coronavirus Disease 2019 (COVID-19) pandemic impacts each country and region differently as countries and regions vary in terms of socioeconomic capacity, healthcare systems capacity, and pre-existing population characteristics. In this study, we analysed the associations of country and region-level attributes with four COVID-19-related outcomes after controlling for testing response by delineated classes of countries and regions. Methods: Country and region-level data of the exposures and COVID-19-related outcomes were obtained from several sources. Two hundred eighteen countries and regions were delineated into three classes using latent class analysis (LCA) based on the following variables: total tests performed per million of the population, current health expenditure per capita, and International Health Regulations (IHR) core capacity index. Bivariate analysis and multivariable linear regression were conducted to detect the associations of country and region-level attributes with COVID-19-related outcomes after controlling for testing response by class. Findings: LCA revealed three classes of countries which we labelled post-hoc as low testing, moderate testing, and high testing. Multivariable linear regression revealed that measures of globalisation, such as the international migrant stock and the number of air travel passengers reported, were positively associated with both total COVID-19 cases and deaths. We also found that democracy, measured through the Freedom in the World index, was negatively associated with total COVID-19 cases per million of the population and positively associated with total COVID-19 deaths. Interpretation: Our findings show that countries and regions with a large migrant stock and air travel, as well as democratic governance structures, may have poorer COVID-19 outcomes, and there is a need for the re-evaluation of their pandemic response. Funding: No funding to declare.Declaration of Interests: None.