Robust Estimation of Infection Fatality Rates during the Early Phase of a Pandemic

  1. Perikles Simon

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

During a pandemic, robust estimation of case fatality rates (CFRs) is essential to plan and control suppression and mitigation strategies. At present, estimates for the CFR of COVID-19 caused by SARS-CoV-2 infection vary considerably. Expert consensus of 0.1–1% covers in practical terms a range from normal seasonable Influenza to Spanish Influenza. In the following, I deduce a formula for an adjusted Infection Fatality Rate (IFR) to assess mortality in a period following a positive test adjusted for selection bias.

Official datasets on cases and deaths were combined with data sets on number of tests. After data curation and quality control, a total of IFR (n=819) was calculated for 21 countries for periods of up to 26 days between registration of a case and death.

Estimates for IRFs increased with length of period, but levelled off at >9days with a median for all 21 countries of 0.11 (95%-CI: 0.073–0.15). An epidemiologically derived IFR of 0.040 % (95%-CI: 0.029%– 0.055%) was determined for Iceland and was very close to the calculated IFR of 0.057% (95%-CI: 0.042– 0.078), but 2.7–6-fold lower than CFRs. IFRs, but not CFRs, were positively associated with increased proportions of elderly in age-cohorts (n=21, spearman’s ρ=.73, p =.02).

Real-time data on molecular and serological testing may further displace classical diagnosis of disease and its related death. I will critically discuss, why, how and under which conditions the IFR, provides a more solid early estimate of the global burden of a pandemic than the CFR.

Article activity feed

  1. ScreenIT

    SciScore for 10.1101/2020.04.08.20057729: (What is this?)

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

    Table 1: Rigor

    NIH rigor criteria are not applicable to paper type.

    Table 2: Resources

    Software and Algorithms
    SentencesResources
    Second, data mining for a final enlarged dataset was done on the pages of the official national health agencies, Wikipedia, and within the data mining community on GitHub using archived webpages if necessary in order to enable a large-scale cross-country assessment and comparison of IFR-values.
    Wikipedia
    suggested: (Wikipedia, RRID:SCR_004897)

    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:
    Even if countries would either go through periods of rapid test rate growths or experience limitations with their testing capacities, the distortion provoked will not lead to huge uncertainty ranges by a substantial unknown error propagation. Correcting CFRs with f is capable of harmonizing differences in CFRs between countries that would otherwise be difficult to explain. Amongst these candidate countries are Japan, South Korea, Iceland, and Norway, which have done meticulous work in dealing with their testing data, protocolling everything transparently and timely, to the public, and moreover, which have strong economies and strong health care systems to cope with the current pandemic. Amongst those, that report their testing data almost in real time and comprehensively, is Pakistan. Pakistan is a country, which seems to fall out of the range of IFRs, with an IFR of 0.007 that is roughly 10-fold lower than the one reported for the so-called developed countries. Since testing was reported transparent and timely, it is important to understand, whether this extremely low IFR figure reported in Table 3 could be possibly realistic, or not. Population statistics of this country compared to any of the developed countries is very informative with this regard. As of 2018, 6.7% of the population in Pakistan were over 60 years old and 45% were younger than 20. In Germany 19 % were younger than then 20 years old, 29% were older than 60 years. The CFR for people aged 60-69 compared to pe...

    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.

  2. Version published to 10.1101/2020.04.08.20057729v1 on medRxiv