Analysis of the time course of COVID-19 cases and deaths from countries with extensive testing allows accurate early estimates of the age specific symptomatic CFR values

  1. Jessica E. Rothman
  2. David Eidelberg
  3. Samantha L. Rothman
  4. Theodore R. Holford
  5. Douglas L. Rothman

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

Knowing the true infected and symptomatic case fatality ratios (IFR and CFR) for COVID-19 is of high importance for epidemiological model projections. Early in the pandemic many locations had limited testing and reporting, so that standard methods for determining IFR and CFR required large adjustments for missed cases. We present an alternate approach, based on results from the countries at the time that had a high test to positive case ratio to estimate symptomatic CFR.

Methods

We calculated age specific (0–69, 70–79, 80+ years old) time corrected crude symptomatic CFR values from 7 countries using two independent time to fatality correction methods. Data was obtained through May 7, 2020. We applied linear regression to determine whether the mean of these coefficients had converged to the true symptomatic CFR values. We then tested these coefficients against values derived in later studies as well as a large random serological study in NYC at that time.

Results

The age dependent symptomatic CFR values accurately predicted the percentage of the population infected as reported by two random testing studies in NYC. They also were in good agreement with later studies that estimated age specific IFR and CFR values from serological studies and more extensive data sets available later in the pandemic.

Conclusions

We found that for regions with extensive testing it is possible to get early accurate symptomatic CFR coefficients. These values, in combination with an estimate of the age dependence of infection, allows symptomatic CFR values and percentage of the population that is infected to be determined in similar regions with limited testing.

Article activity feed

  1. Version published to 10.1371/journal.pone.0253843
  2. ScreenIT

    SciScore for 10.1101/2020.05.13.20101022: (What is this?)

    Please note, not all rigor criteria are appropriate for all manuscripts.

    Table 1: Rigor

    Institutional Review Board Statementnot detected.
    Randomizationnot detected.
    Blindingnot detected.
    Power Analysisnot detected.
    Sex as a biological variablenot detected.

    Table 2: Resources

    No key resources detected.

    Results from OddPub: We did not detect open data. We also did not detect open code. Researchers are encouraged to share open data when possible (see Nature blog).

    Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:
    We discuss below the potential sources of difference between our study and previous CFR and IFR calculations, and the relative strengths and limitations of our approach. Our calculated corrected CFR for China was 2.19% with a 95% CI of the mean: 1.54%-2.85%, which was considerably lower than for New York City due to the lower fraction of cases between 70 and 79 and 80 and over years old, but higher than the majority but not all of the previously reported values (Table 1), which ranged from 0.9 to 5.6%. Almost all of the studies that calculated an IFR/CFR using data from China performed a time correction similar to that applied here, so that it is unlikely that the time delay to fatality significantly contributed to their lower corrected CFR values.(3,5–7,20–23) A more likely factor explaining the differences between values is the correction for missed cases, which ranged from 1.0 to approximately 4.(20,23) In addition to uncertainties in determining missed cases, a significant amount of the difference can be accounted for when previous values are scaled up by 1.5 fold to match the recent report of the Chinese government of undercounted fatalities.(2) The reported nCFR(t) versus day data shows clearly that the nCFR(t) values have all risen at least several fold from the early values used to justify low estimates of the IFR for COVID-19 (see Figures 5A, 5B, and Appendix 3).(12,14,24–27,30) We therefore calculated the corrected CFR for each country using both the earlier converg...

    Results from TrialIdentifier: No clinical trial numbers were referenced.

    Results from Barzooka: We did not find any issues relating to the usage of bar graphs.

    Results from JetFighter: We did not find any issues relating to colormaps.

    Results from rtransparent:
    • Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
    • Thank you for including a funding statement. Authors are encouraged to include this statement when submitting to a journal.
    • No protocol registration statement was detected.

    About SciScore

    SciScore is an automated tool that is designed to assist expert reviewers by finding and presenting formulaic information scattered throughout a paper in a standard, easy to digest format. SciScore checks for the presence and correctness of RRIDs (research resource identifiers), and for rigor criteria such as sex and investigator blinding. For details on the theoretical underpinning of rigor criteria and the tools shown here, including references cited, please follow this link.

  3. Version published to 10.1101/2020.05.13.20101022v3 on medRxiv
  4. Version published to 10.1101/2020.05.13.20101022v2 on medRxiv
  5. Version published to 10.1101/2020.05.13.20101022v1 on medRxiv