Abstract: It is unclear to what extent age structure and testing rate explain the differences in mortality rate of COVID19 across countries. Based on the agespecific data of Italy, Spain, China, and South Korea from March 23 to April 27, 2020, this paper uses decomposition method to study the differences and influencing factors for mortality rate of COVID-19 among these countries. Major findings are as follows: (1) Differences in mortality rate and fatality rate vary across countries, and are widen over time. (2) Mortality rates in Italy, Spain, and China are higher than those in South Korea. Age structure effect accounts for 66.62%, 74.82%, and 11.70% of the total difference for Italy, Spain, and China respectively. (3) After introducing relative risk in decomposition, the age structure effect for Italy and Spain, and the agespecific fatality effect for China are reduced gradually.(4) After introducing testing rate in decomposition, the effect of age structure is decreased drastically, and the positive rate of COVID-19 dominates the differences in mortality across countries. (5) The effect of testing rate on mortality for Italy, Spain, and South Korea increases over time.

Keywords: COVID-19, Fatality Rate, Mortality Rate, Age Structure, Testing Rate

摘要： 年龄结构和检测率究竟能在多大程度上解释新冠肺炎死亡率的国际差异尚不清楚。利用2020年3月23日至4月27日意大利、西班牙、中国和韩国的分年龄统计数据，采用死亡率分解方法，研究新冠肺炎死亡率的国际差异及其影响因素。研究发现：(1) 4个国家的病死率和死亡率差异明显，且随时间推移呈扩大趋势；(2)意大利、西班牙和中国较之韩国更高的病死率中，病例年龄结构效应分别占66.62%、74.82%和11.70%；(3)引入相对危险度后，意大利和西班牙的人口年龄结构效应以及中国的年龄别病死率效应有所下降，但仍是影响其与韩国病死率差异的主导因素；(4)引入检测率后，年龄结构对死亡率国际差异的贡献明显下降，确诊率成为影响死亡率国际差异的主导因素；(5)检测率对意大利、西班牙和韩国死亡率的影响逐渐增强。

关键词: 新冠肺炎, 病死率, 死亡率, 年龄结构, 检测率