Why Are Covid-19 Mortality Rates By Country Or Region So Different?: An Ecologic Study Of Factors Associated With Mortality From Novel Coronavirus Infections By Country

Background In order to find out the factors associated with the large disparities in COVID-19 mortality rates by country, we conducted an ecological study by linking existing statistics. In Japan, a large variation was observed in between geographical areas when assessing mortality. We performed a regional correlation analysis to find factors related to regional mortality.Methods This study design was an ecologic study. A multiple regression analysis was performed with COVID-19 mortality rates of different countries as the dependent variable together with various health care and economic factors. We calculated the cumulative mortality rate as of June 30, 2020. For the regional correlation analysis of Japan, 47 prefectures were divided into nine regions. The factors examined were health care and tourism. Data for 33 Organization for Economic Cooperation and Development (OECD) countries were analyzed. In Japan's regional analysis, the whole country was classified into nine regions.Results Factors related to mortality were the incidence of Kawasaki disease (KD), number of computed tomographies (CTs), and alcohol consumption. Mortality was low in countries with high incidence of KD and high number of CTs, as well as in countries with high alcohol consumption. In European countries, high smoking prevalence and a high Gini coefficient were positively related to high mortality. According to a regional analysis in Japan, mortality was related to proportion of population in the densely inhabited districts, the number of foreign visitors per capita, and the number of Chinese visitors per capita.Conclusion Low mortality in East Asia was associated with specific disease morbidity ( KD), alcohol consumption, and CT numbers. It was suggested that the mortality gap in Japan was related to the number of foreign tourists and the proportion of population in the densely inhabited districts.