A Continuous Age-Specific Standardized Mortality Ratio for Estimating the Unascertained Rates in the Early Epidemic of COVID-19 in Different Regions

Peipei Du,Peihua Cao,Xiaodong Yan,Daihai He,Xiaotong Zhang,Weixiang Chen,Jiawei Luo,Ziqian Zeng,Yaolong Chen,Lin Yang,Shu Yang,Xixi Feng
DOI: https://doi.org/10.1080/02664763.2021.1947995
IF: 1.416
2021-01-01
Journal of Applied Statistics
Abstract:The difference in age structure and aging population level was an important factor that caused the difference in COVID-19's case fatality rate (CFR) in various regions. To eliminate the age effect on estimating the CFR of COVID-19, our study applied nonlinear logistic model and maximum likelihood method to fit the age-fatality curves of COVID-19 in different countries and regions. We further computed the standardized mortality ratio from the age-fatality curves of COVID-19 in the above regions and found that the risk of COVID-19 death in Wuhan was of a moderate level, while the non-Hubei region was even lower, compared with other regions. Regarding the disparity of CFRs among different regions in the country, we believed that there might be an unascertained phenomenon in high-endemic regions. Based on age-fatality rate curves, we estimated unascertained rates in cities with severe epidemics such as Wuhan and New York, and it was found that the total unascertained rates in Wuhan and New York were 81.6% and 81.2%, respectively. Meanwhile, we also found that the unascertained rates varied greatly with age.
What problem does this paper attempt to address?