Estimation of COVID-19 case fatality ratio based on a bi-directional correction method

Abstract

The case fatality rate (CFR) can be used to predict the number of potential 19 deaths in the epidemic and thus can reflect the appropriateness and quality of medical 20 measures developed by public health. When a new disease breaks out, it is 21 particularly important to accurately estimate the CFR. However, while the epidemic is 22 still developing, the crude CFR is often lower than the true value and the hospital 23 CFR is often higher than the true value due to differences in occurrence time, patient 24 number, and treatment plans. Therefore, this study proposes a bi-directional correction 25 method to estimate the CFR. COVID-19 data from China were used to evaluate this 26 method. The results show that this method provides more accurate results than both 27 the crude CFR and hospital CFR. Additionally, this method was used to estimate the 28 CFR of COVID-19 in other countries, with an aim to provide a reference for 29 prevention and control decisions for the COVID-19 epidemic and for the evaluation 30 of medical efforts. 31