Skip to main content
SearchLoginLogin or Signup

On Identifying and Mitigating Bias in the Estimation of the COVID-19 Case Fatality Rate

Published onJul 16, 2020
On Identifying and Mitigating Bias in the Estimation of the COVID-19 Case Fatality Rate

You're viewing an older Release (#2) of this Pub.

  • This Release (#2) was created on Jun 01, 2020 ()
  • The latest Release (#9) was created on May 23, 2022 ().


The relative case fatality ratios (CFRs) between groups and countries are key measures of relative risk that guide policy decisions regarding scarce medical resource allocation during the ongoing COVID-19 pandemic. In the middle of an active outbreak when surveillance data is the primary source of information, estimating these quantities involves compensating for competing biases in time series of deaths, cases and recoveries. These include time- and severity-dependent reporting of cases as well as time-lags in observed patient outcomes. We detail such biases and their significance in the context of COVID-19 CFR estimation. We analyze theoretically the effect of certain biases, like preferential reporting of fatal cases, on naive estimators of CFR. In a simple setting, we correct time-varying reporting rates that may differ among fatal and non-fatal cases. We ultimately suggest that randomized data obtained through contact tracing can help mitigate many of these confounding factors. Our analysis is supplemented by numerical results and a simple and fast open-source codebase.

Just Accepted - Preview

6/1/20: To preview this content, click below for the Just Accepted version of the article. This peer-reviewed version has been accepted for its content and is currently being copyedited to conform with HDSR’s style and formatting requirements.

A Supplement to this Pub
Anastasios Nikolas Angelopoulos:

Contact tracing is usually meant as a disease control strategy. But the point we emphasize in our article is that by collecting data prospectively via contact tracing, we can control many of the biases present in normal surveillance data.

Please go ahead and comment on this paper using this HDSR comment feature! I’ll be monitoring the page every so often and answering questions.