Risk-benefit analysis of the AstraZeneca COVID-19 vaccine in Australia using a Bayesian network modelling framework

Area of Science:

• Vaccinology
• Epidemiology
• Biostatistics

Background:

• Thrombosis with Thrombocytopenia Syndrome (TTS) is a rare but serious adverse event associated with the AstraZeneca COVID-19 vaccine (Vaxzevria).
• Assessing the risk-benefit profile of vaccination is complex due to evolving data on TTS incidence, COVID-19 severity, and vaccine effectiveness across different demographics and transmission levels.
• Existing risk-benefit analyses may not be rapidly adaptable to new evidence or varying epidemiological contexts.

Purpose of the Study:

• To develop and apply a dynamic Bayesian network model for optimizing the risk-benefit analysis of the AstraZeneca COVID-19 vaccine.
• To estimate and compare the risks of TTS with the benefits of prevented COVID-19 deaths and other severe COVID-19-related thrombotic events.
• To provide a flexible framework for ongoing risk-benefit assessment adaptable to new data and different populations.

Main Methods:

• A Bayesian network model was constructed, integrating diverse data sources including local/international data, government reports, literature, and expert opinion.
• The model estimated probabilities of various outcomes (TTS, COVID-19 deaths, severe blood clots) under scenarios varying by age, sex, transmission intensity, SARS-CoV-2 variant, vaccination status, and vaccine effectiveness.
• Model outputs were used to compare estimated TTS deaths against COVID-19 deaths averted and deaths from COVID-19-related atypical severe blood clots (CVT & PVT).

Main Results:

• For individuals aged ≥70 years, the model estimated fewer than 1 TTS death per million vaccinated (first and second doses).
• Under low transmission, vaccination prevented an estimated 25 COVID-19 deaths per million; under high transmission, over 3000 deaths were prevented.
• The probability of dying from COVID-19-related severe blood clots was significantly higher (58-126 times) than from vaccine-associated TTS, with benefits of vaccination substantially outweighing risks, especially during high transmission periods.

Conclusions:

• The developed Bayesian network model provides a robust and adaptable tool for real-time risk-benefit analysis of COVID-19 vaccines.
• Vaccination with AstraZeneca (Vaxzevria) demonstrates a favorable risk-benefit profile, particularly in older age groups and during periods of high COVID-19 transmission.
• The model's flexibility allows for extension to other vaccines, outcomes, and geographical regions, supporting evidence-based public health decision-making.