Correlation of population mortality of COVID-19 and testing coverage: a comparison among 36 OECD countries

  1. C. Wei
  2. C. C. Lee
  3. T. C. Hsu
  4. W. T. Hsu
  5. C. C. Chan
  6. S. C. Chen
  7. C. J. Chen

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Although testing is widely regarded as critical to fighting the COVID-19 pandemic, what measure and level of testing best reflects successful infection control remains unresolved. Our aim was to compare the sensitivity of two testing metrics – population testing number and testing coverage – to population mortality outcomes and identify a benchmark for testing adequacy. We aggregated publicly available data through 12 April on testing and outcomes related to COVID-19 across 36 OECD (Organization for Economic Development) countries and Taiwan. Spearman correlation coefficients were calculated between the aforementioned metrics and following outcome measures: deaths per 1 million people, case fatality rate and case proportion of critical illness. Fractional polynomials were used to generate scatter plots to model the relationship between the testing metrics and outcomes. We found that testing coverage, but not population testing number, was highly correlated with population mortality ( r s = −0.79, P = 5.975 × 10 −9 vs. r s = −0.3, P = 0.05) and case fatality rate ( r s = −0.67, P = 9.067 × 10 −6 vs. r s = −0.21, P = 0.20). A testing coverage threshold of 15–45 signified adequate testing: below 15, testing coverage was associated with exponentially increasing population mortality; above 45, increased testing did not yield significant incremental mortality benefit. Taken together, testing coverage was better than population testing number in explaining country performance and can serve as an early and sensitive indicator of testing adequacy and disease burden.

Article activity feed

  1. ScreenIT

    SciScore for 10.1101/2020.05.27.20113969: (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:
    Despite these limitations however, our analysis showed that testing coverage was still highly correlated with country performance and testing coverage provides additional benefits of low-cost and efficiency. The results herein should also be interpreted in the context of other limitations. The negative correlation between testing coverage and population mortality does not imply causation, which can only be verified in a prospective interventional study-although there is anecdotal evidence suggesting that early antiviral treatment and/or supportive care may reduce mortality among COVID-19 patients13, this may also be due to increased identification of patients with mild disease. In addition, the infection fatality rate of Covid-19 may vary from country-to-country, as has been seen in Italy.3 Relevant modifiers include prevalence of risk factors, access to healthcare, robustness of healthcare infrastructure, and population density. Rather than be used in monolithic fashion, testing coverage should therefore be applied in context and with adequate judgment. In conclusion, we demonstrate the negative curvilinear relationship between testing coverage and COVID-19 population mortality and case fatality rate. Testing coverage can be used as both an indicator of testing adequacy and potential unidentified disease burden, and is most accurate in the context of high healthcare accessibility, comprehensive contact tracing, and testing sensitivity.

    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 10.1017/s0950268820003076
  3. Version published to 10.1101/2020.05.27.20113969v1 on medRxiv