A Bayesian network analysis quantifying risks versus benefits of the Pfizer COVID-19 vaccine in Australia

  1. Jane E. Sinclair
  2. Helen J. Mayfield
  3. Kirsty R. Short
  4. Samuel J. Brown
  5. Rajesh Puranik
  6. Kerrie Mengersen
  7. John C. B. Litt
  8. Colleen L. Lau

Reviewed by

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

Log in to save this article

Abstract

The Pfizer COVID-19 vaccine is associated with increased myocarditis incidence. Constantly evolving evidence regarding incidence and case fatality of COVID-19 and myocarditis related to infection or vaccination, creates challenges for risk-benefit analysis of vaccination. Challenges are complicated further by emerging evidence of waning vaccine effectiveness, and variable effectiveness against variants. Here, we build on previous work on the COVID-19 Risk Calculator (CoRiCal) by integrating Australian and international data to inform a Bayesian network that calculates probabilities of outcomes for the delta variant under different scenarios of Pfizer COVID-19 vaccine coverage, age groups (≥12 years), sex, community transmission intensity and vaccine effectiveness. The model estimates that in a population where 5% were unvaccinated, 5% had one dose, 60% had two doses and 30% had three doses, there was a substantially greater probability of developing (239–5847 times) and dying (1430–384,684 times) from COVID-19-related than vaccine-associated myocarditis (depending on age and sex). For one million people with this vaccine coverage, where transmission intensity was equivalent to 10% chance of infection over 2 months, 68,813 symptomatic COVID-19 cases and 981 deaths would be prevented, with 42 and 16 expected cases of vaccine-associated myocarditis in males and females, respectively. These results justify vaccination in all age groups as vaccine-associated myocarditis is generally mild in the young, and there is unequivocal evidence for reduced mortality from COVID-19 in older individuals. The model may be updated to include emerging best evidence, data pertinent to different countries or vaccines and other outcomes such as long COVID.

Article activity feed

  1. Version published to 10.1038/s41541-022-00517-6
  2. ScreenIT

    SciScore for 10.1101/2022.02.07.22270637: (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

    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:
    Furthermore, limitations to the availability of Australian data introduces uncertainty in the model inputs, so results may change as more data become available. For example, at the time of writing no Australian data were available on the incidence of Pfizer vaccine-associated myocarditis after the third dose and international data were deemed inappropriate as a substitute (see Table 1 assumptions), necessitating the use of rates for the second dose as a worst-case scenario. In another example, when calculating the delta variant-specific CFR from COVID-19, ideally CFR for the unvaccinated population would be used, and the 2-3 week lag between diagnosis and death accounted for. This information was not available in Australia, so the assumptions were made that the time-window of a few months for the delta wave was long enough to minimise the effect of time lag from infection to death, and the great majority of deaths during the delta wave was in unvaccinated people. Other limitations arise from the model development process, where the use of expert elicitation may be perceived to introduce bias in the evidence viewed. This was minimised through broad literature searches and frequent meetings with external experts such as cardiologists about the quality of the data sources used in the model assumptions. Despite these limitations, the use of an evidence-based BN to model the risks and benefits of COVID-19 vaccination has many advantages. BNs allow for interactive scenario analysis...

    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.

    Results from scite Reference Check: We found no unreliable references.

    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 10.1101/2022.02.07.22270637 on medRxiv