A new method for accurate calculation of case fatality rates during a pandemic: Mathematical deduction based on population-level big data

Abstract

Background

During the course of an epidemic of a potentially fatal disease, it is difficult to accurately estimate the case fatality rate (CFR) because many calculation methods do not account for the delay between case confirmation and disease outcome. Taking the coronavirus disease-2019 (COVID-19) as an example, this study aimed to develop a new method for CFR calculation while the pandemic was ongoing.

Methods

We developed a new method for CFR calculation based on the following formula: number of deaths divided by the number of cases T days before, where T is the average delay between case confirmation and disease outcome. An objective law was found using simulated data that states if the hypothesized T is equal to the true T, the calculated real-time CFR remains constant; whereas if the hypothesized T is greater (or smaller) than the true T, the real-time CFR will gradually decrease (or increase) as the days progress until it approaches the true CFR.

Results

Based on the discovered law, it was estimated that the true CFR of COVID-19 at the initial stage of the pandemic in China, excluding Hubei Province, was 0.8%; and in Hubei Province, it was 6.6%. The calculated CFRs predicted the death count with almost complete accuracy.

Conclusions

The method could be used for the accurate calculation of the true CFR during a pandemic, instead of waiting until the end of the pandemic, whether the pandemic is under control or not. It could provide those involved in outbreak control a clear view of the timeliness of case confirmations.