Government and governance, Health | Australia

10 May 2021

Australia can only achieve herd immunity from COVID-19 if it changes its vaccine strategy, Quentin Grafton, Zoë Hyde, Tom Kompas, and John Parslow write.

The world has now reached the grim milestone of about 160 million total confirmed cases of COVID-19. More than 20 million cases have been reported by India, which has been devastated by a second wave driven by the premature relaxation of public health measures and more transmissible variants of the virus.

Even some countries with high vaccination rates haven’t been spared. Chile is battling a new wave of infections despite more than one third of its population being fully vaccinated, mostly with CoronaVac.

By comparison, Israel is now only carefully easing restrictions, despite having fully vaccinated 56 per cent of its population, and given at least one dose to 60 per cent, with the high-efficacy Pfizer vaccine. So far, Israel’s world-leading vaccination strategy and cautious reopening of its economy has avoided a renewed surge in cases.

The lessons for Australia are obvious. The country won’t be truly safe until we’ve reached herd immunity. Reaching this threshold ensures an outbreak won’t spread rapidly in the community and lead to an epidemic. It also protects the millions of Australians who can’t be vaccinated due to medical reasons, or aren’t currently eligible for vaccination, such as children and adolescents.

To protect Australians, there are four key things policymakers must do.

First, they must strengthen the country’s quarantine system by acting on the latest evidence of the role of airborne transmission in quarantine breaches.

More on this: Containing COVID-19

Second, the country needs to aim for herd immunity by using vaccines with the highest possible efficacy, such as Moderna, Novavax, and Pfizer. These vaccines provide security against even concerning new variants of the virus. Real world experience shows the Pfizer vaccine is 75 per cent effective against infections caused by the South African (B.1.351) variant and 90 per cent effective against the UK (B.1.1.7) variant. In contrast, the AstraZeneca vaccine has 10 per cent efficacy against the South African variant, although it’s probably still reasonably effective against severe disease.

Third, Australia should not relax international border restrictions until everyone has had the opportunity to be vaccinated.

Fourth, the country should establish mass vaccination hubs in major centres to maximise the number of people who can be vaccinated each day, as Victoria is doing.

Australia needs to revise its strategy due to the emergence of more transmissible and more virulent variants, which have substantially changed who is at risk from COVID-19. Variants threaten both young and old, because they advance the age-related risk of hospitalisation by one to two decades. In other words, the risk for a 20 to 39-year-old is now similar to the risk previously faced by a 40 to 59-year-old.

Variants also markedly increase the likelihood of dying from COVID-19; the United Kingdom variant raises this risk by 64 per cent.

It’s been claimed the vaccines in Australia’s portfolio are 100 per cent effective at preventing severe illness and death, but that’s not supported by science. There were too few severe cases in the AstraZeneca trials to accurately assess this endpoint, and the estimated efficacy lies somewhere between 72 and 100 per cent for severe illness. In contrast, real world experience from Israel has shown the Pfizer vaccine is 97.5 per cent effective against severe disease, with a confidence interval of 97.1 to 97.8 per cent.

We’ve learned that death or full recovery aren’t the only outcomes of COVID-19, and this should also change our vaccination strategy. The United Kingdom’s Office for National Statistics reported that more than one in 10 people are left with persistent symptoms lasting at least 12 weeks. A United States study of over 70,000 people who had COVID-19, but weren’t unwell enough to be hospitalised, found that many survivors were left with substantial chronic health problems.

More on this: A wake up call to address faultlines in Asia and the Pacific

This syndrome, now known as long COVID, is estimated to affect over one million people living in the United Kingdom, and more than 400,000 of them experience limitation in their daily activities because of it. The syndrome affects young and old alike, and there is growing evidence of children with long COVID. Australia’s vaccination strategy must, therefore, also aim to prevent long COVID, and we stand the best chance of doing it with high-efficacy vaccines.

Australia has ordered 40 million doses of the Pfizer vaccine and 51 million doses of Novavax, more than enough to vaccinate the entire population with these highly effective vaccines. But Australia is persisting with the rollout of the AstraZeneca vaccine and has, by default, imposed AstraZeneca as the only vaccine for almost all of those 50 years of age and older because those under 50 years have priority over the limited supplies of Pfizer.

This is inequitable because it knowingly gives older people, who face a higher risk of severe disease and death from COVID-19, a less effective vaccine. It also runs the risk of creating a situation where the South African variant could spread in Australia’s older population should border controls be relaxed or if an outbreak should occur.

The South African Government abandoned their rollout of the AstraZeneca vaccine. Denmark stopped using AstraZeneca in April. Seychelles, which has fully vaccinated the highest proportion of its population than any country at 61 per cent – 40 per cent of these with a version of the AstraZeneca vaccine made under license in India – is now facing a surge of cases in which the South African variant appears to be playing a role. New Zealand changed its vaccination rollout in March and is now using Pfizer exclusively because of its higher efficacy over AstraZeneca.

Australia is persisting with the use of AstraZeneca for Australians aged 50 years and over despite the well-documented risk of blood clots, albeit at low rates of around one in 100,000. The continued use of AstraZeneca for older Australians is based on the recommendations in April of the Australian Technical Advisory Group on Immunisation (ATAGI).

In justifying its decision, ATAGI compared the risks of severe blood clots with the risk of intensive care unit admission due to COVID-19 by age group under two scenarios over the next 16 weeks. In scenario one, Australia suffers another outbreak equivalent to its first wave – approximately 7,000 cases and 100 deaths – and COVID-19 risks then exceed blood clot risks for those 50 years and over but not for those under 50. In scenario two, Australia suffers a Victorian-like outbreak, with around 20,000 cases and 800 deaths, and COVID-19 risks exceed blood clot risks for all age groups.

The ATAGI recommendation is sensitive to assumptions about risks of both COVID-19 and blood clots. The European Medicines Agency recently doubled its estimate of the risk of blood clots for older people since ATAGI carried out its original analysis and this probability – one in 100,000 – is similar to the blood clotting risk experienced with AstraZeneca in older Australians to date. Critically, if the ATAGI had used the revised and higher estimates of blood clot risks, and its scenario one COVID-19 outbreak, it would have also recommended against AstraZeneca for those aged between 50 and 60.

Australia last major COVID-19 outbreak ended in October 2020. Small outbreaks from hotel quarantine have necessitated short lockdowns, contact tracing, and self-quarantine measures, which have succeeded so far. But an outbreak of the kind assumed by ATAGI would cause multiple fatalities and require a severe and prolonged lockdown that would cost the domestic economy tens of billions of dollars.

Vaccine hesitancy towards the AstraZeneca vaccine has increased in the 50 and older age group since the new policy was announced last month. It seems that many Australians have done their own risk assessment, and concluded, based on the track record of the last six months, that the risk of infection and hospitalisation is much lower than the risk of blood clots.

Given its inferior efficacy and the risk of blood clots, the continued rollout of AstraZeneca can only be justified as an emergency measure to prevent an imminent third wave before Pfizer and Novavax doses are delivered to all Australians. However, if Australians genuinely face a high risk of a third wave, they should be told because this would help overcome vaccine hesitancy toward AstraZeneca.

Further, if a third wave is imminent, governments should also be doing everything in their power to avoid a major outbreak. This would include: strengthening border controls by constructing or extending purpose-built air-gapped quarantine facilities, such as at Howard Springs; building mass vaccination hubs in all major centres; and promoting continued cost-effective social distancing measures, such as wearing masks on public transport, and not allowing large public events.

If AstraZeneca is being used as an emergency stop-gap measure, as implied by the ATAGI risk-benefit probabilities, then people aged 50 years and over should be prioritised to get Pfizer as they have a higher risk of developing severe COVID-19 than those under 50 years of age. Further, it is almost certain that people will need to be revaccinated periodically as new variants of concern are identified, much like the annual flu jab. This is possible with the Pfizer, Moderna, and Novavax vaccines but can’t be done with AstraZeneca, because people develop immunity to the adenovirus vector, making successive doses less effective.

The Australian Government must reassure the public that the emergency use of AstraZeneca for those 50 years and over is necessary under current border controls and is not being imposed to allow earlier reopening of international borders. Crucially, an AstraZeneca only vaccination strategy for those aged 50 and over will not provide the level of protection necessary to safely relax border controls, nor will it deliver herd immunity.

While there are low-cost border controls available such as rapid testing on arrival, vaccination passports, and self-isolation, there are serious deficiencies with each. Notably, there is a much larger proportion of false negatives with rapid testing, there is an increasing black market for false vaccination certificates, and home self-isolation is not nearly as effective as supervised quarantine at reducing growth in infections. Moreover, those vaccinated with AstraZeneca may still be infectious with the South African variant because of its low efficacy for that variant and some fully vaccinated travellers can still be infectious.

A three-to-six month delay in reopening the international border until everyone can be vaccinated with Pfizer or Novavax would impose a burden on the economy of around $6-12 billion. This cost, however, is much less than a prolonged lockdown needed to control a major outbreak, or the costs associated with an increased frequency of short and long-term city lockdowns should the international border reopen under the current vaccination rollout.

Australia’s vaccination strategy must change as the facts change. Herd immunity is the only way to eliminate community transmission for good and to avoid costly lockdowns. This can be achieved with a Pfizer or Pfizer-Novavax vaccination strategy, but not with the country’s current approach because it will be impossible to achieve herd immunity against one of the world’s most concerning variants using AstraZeneca.

With the emergence of more transmissible new variants, Australia’s current COVID-19 vaccination strategy doesn’t make sense from either a public health or an economic perspective. It must change or Australia risks a future epidemic and/or very large costs from lockdowns when its international border reopens.

The information provided in this article does not substitute for, nor does it replace formal medical advice in relation to individual COVID-19 vaccinations. In all cases, anyone deciding to be vaccinated, when, and with what vaccine, should first consult with their GP.

Back to Top
Join the APP Society

21 Responses

  1. Sabra Goudas says:

    Whole heartedly agree.

  2. Lisa Nicholas says:

    Finally! Thank you!

  3. Laura says:

    Thank you for stepping up and articulating this. On the vaccine roll out in particular, I hope ATAGI updates the risk benefit analysis promptly based on up to date data. At very least, I wish for all who are willing to take an mRNA vaccine to be granted the same ability to stand in queue, based on order of vulnerability. Given prior government statements about safety being foremost as justification for roll out delays, continuing on the current path for over 50s, and particularly under 70s, *without choice* is unconscionable. This would not prevent anyone of any age opting for AZ faster – we are already running two queues.

  4. Laura says:

    A lucid analysis. It is a case of ‘Jack be nimble, Jack be quick’. Not only is our vaccine roll out wallowing in the past instead of pivoting, it is unconscionable in the face of the evolving evidence for the government to be denying over 50s the same vaccine choices as under 50s.

  5. Debbie says:

    Thank you, this needs to become more evident as a fact now. I hope the government really do take note of this, and make the changes needed. It is their duty to protect ALL of their citizens and to not leave the over 50’s so incredibly vulnerable, exposed and afraid.

  6. Kat says:

    When will the government listen. It is stressing the over 50s

  7. Sally says:

    Completely agree! Thank you.

  8. Wynne says:

    Yes. Making AZ the only option for over-50s is truly inequitable. The blood clot risk is not the real issue. At 62, I am no less deserving of an efficacious vaccine than under-50s and politicians. I will wait for another policy change before presenting for vaccination. Pity about herd immunity.

  9. Pauline says:

    An incisive, sensible analysis of Australia’s current situation.
    Thank you for this article

  10. Wendy Holmes says:

    Thank you, this is well argued and I strongly agree. The only thing missing is to emphasize the need for our wealthy country to ensure that less well off countries in our region have sufficient vaccines. This will also contribute to allowing us to open up more safely. There are health care workers not yet fully immunized in countries with severe spread and we should send them our AstraZeneca vaccines on the next flight.

  11. Paula Lancaster says:

    What about those of us who’ve already had the first AZ vaccine but wish we’d had the Pfizer? Why on Earth did the Australian Government give the much less efficacious vaccine to the very group who has the worst outcome in the event of contracting Covid? Obviously we’ll now need our second AZ jab, but thereafter can we ‘switch’ to the Pfizer or Modena jabs for a booster or by then will it be ‘too late’, to use layperson’s language?

  12. Fiona Russell says:

    There are major issues in this article regarding the understanding of vaccine development, clinical trial design and vaccine evaluation to inform vaccine policy. Throwing out the baby (AstraZeneca vaccine) with the bath water is irresponsible and shows a very poor understanding of the global situation and our connection with our region, the Asia- Pacific- we are supplying countries with AZ vaccine and these countries have no alternatives) and the need for all the highly effective vaccines possible. Similarly, health economics, vaccine development and what data goes into making vaccine policy decisions does not seem to be taken into account.

    1. This article does not understand the difference between vaccine efficacy (clinical trials), vaccine effectiveness (real world) and vaccine impact (direct and indirect effects). Clinical trials (efficacy) don’t measure indirect effects and, as a result, don’t measure the full impact of a vaccine in a population. The pneumococcal conjugate vaccine was developed 20 years ago. Clinical trials of the pneumococcal conjugate vaccine in infants found the vaccine to have high efficacy against strains of the bacteria included in the vaccine. However, when the vaccine was included in the national immunisation schedule for infants, it not only prevented disease in infants but stopped transmission to the unvaccinated including the elderly. In some countries, these indirect effects (herd protection) were greater in the unvaccinated population than the direct effects in the infants. In the US, an estimated 9,140 cases were directly prevented by vaccinating children aged under five years and an additional 20,459 cases were prevented through indirect effects of the vaccine across all ages three years after the vaccine was rolled out. This impact was most marked in the elderly. (
    This was not known at the time the vaccine was licensed and included in the national program and as such the health economics did not account for the full impact of the vaccine on the population’s health. Don’t rely on efficacy to calculate cost effectiveness.

    In the real world the vaccine effectiveness of AstraZeneca is similar to Pfizer/BioNTech:

    Vaccine effectiveness of a SINGLE AZ DOSE in Scotland against hospitalisation/death: 94% (73-99)

    Pooled analysis of Pfizer/BioNTech and AstraZeneca vaccines: reduced odds of infection post-second dose: 70% (62-77)
    There was no evidence that these benefits varied between AstraZeneca and Pfizer/BioNTech vaccines

    Effectiveness of Pfizer/BioNTech vaccine against death in England: Single dose: 44% (32-53); Fully vaccinated: 69% (31-86)
    Effectiveness of AstraZeneca vaccine against death in England: 55% (41-66)
    Effectiveness of single dose AstraZeneca vaccine against hospitalisation in those aged ≥80 years in England: 73% (60-81)
    Effectiveness of Pfizer/BioNTech vaccine against hospitalisation in those aged ≥80 years in England: Single dose: 81% (76-85); Fully vaccinated: 93% (89-95)

    Vasileiou E, Simpson CR, Robertson C, et al. Effectiveness of First Dose of COVID-19 Vaccines Against Hospital Admissions in Scotland: National Prospective Cohort Study of 5.4 Million People. SSRN Electron J. 2021. doi:10.2139/ssrn.3789264

    Pritchard E, Matthews PC, Stoesser N, et al. Impact of vaccination on SARS-CoV-2 cases in the community: a population-based study using the UK’s COVID-19 Infection

    Bernal J, Andrews N, Gower C, et al. Effectiveness of BNT162b2 mRNA vaccine and ChAdOx1 adenovirus vector vaccine on mortality following COVID-19. Prepr medRxiv. doi:10.1101/2021.05.14.21257218

    2. Direct comparison of vaccine efficacy results is like comparing apples with oranges. They must be interpreted in the context of study designs (including case definitions, clinical endpoints, access to testing), target populations, and COVID-19 epidemiology. Each study used a different protocol. The only way to directly compare vaccine efficacy results is by doing head to head studies ie using the same protocol and randomising each arm to a different vaccine. This has not been done.

    The importance of this was highlighted at the last WHO SAGE meeting.

    This article says the AstraZeneca vaccine was not effective against the South African variant- that study was designed for mild-moderate disease. The study against severe disease is underway and the results aren’t even available yet. The are no results for the Moderna vaccine either because the study has not been completed. If the results of the AZ South Africa show low efficacy against severe disease, people who have has AstraZeneca as a first or second dose may need a booster with a vaccine that has higher efficacy against this variant. But we do not know that yet. It is highly likely that we may need annual boosters anyway but to change policy now based on this unknown is too premature. This is what we do with the influenza vaccine every year- change the composition and have an annual dose. This is not a policy failure but a normal process.

    4. The article says we will not achieve herd protection with AstraZeneca. It is completely unclear how this was ever determined as there is only one study with empiric data from Israel using Pfizer/BioNTech that showed herd protection in 80%) was needed for herd effects to manifest- our work shows this is a misconception and actually quite low levels provide substantial indirect effects and that this improves over time with the roll out. Each vaccine is different- but my point is this- statements like this have no basis as yet and perpetuate myths (and misinformation).

    • Quentin Grafton says:

      Fiona, you have misread what we wrote. The opinion piece did NOT state we should not be vaccinating with AZ. Rather, we stated that we should ensure that adult Australians (including those 50+) be allowed to and be able to be vaccinated with mRNA vaccines prior to opening the international border. Evidence for benefit of ‘mix and match’, namely, revaccinating with Pfizer after AZ indicates this should be part of our vaccination strategy. Please see

  13. Fiona Russell says:

    This is interpretation is not correct regarding AstraZeneca vaccine.
    AstraZeneca is equally effective as Pfizer/BioNTech.
    See latest UK vaccine effectiveness data for both vaccines

  14. Quentin Grafton says:

    Thank you for the comment Fiona but you ar not correct. Yes, latest UK data shows that one dose AZ and one dose Pfizer are similar in their effectiveness. But two doses Pfizer is definetly more effective than one dose Pfizer or one dose AZ. As you know, we await updated data from the UK for relative efefctiveness of two doses AZ to two doses Pfizer. When we have the data we can make a proper comparison. In addition to the trial data (supported by Israel data) that indicate two doses Pfizer is more effective than two dose AZ, separately (from Qatar) there is evidence that Pfizer is efefctive against some variants of concern This may be true of AZ but we do not have that data as far as I’m aware.

  15. John A. Hackett says:

    Thank you Quentin for this well documented article which agrees with my own ongoing Research.
    The Politics tends to overshadow the Science so often in these situations.
    As a 77 yo Research Scientist in the most vulnerable age group I am currently excluded from access to the Pfizer vaccine.
    I mighty have to move to New Zealand to access the Pfizer vaccine !

    Best Regards
    John A. Hackett

  16. Alexis T Crowford says:

    Well said and comprehensible analysis especially referring to those 50-65 age group. Hopefully ATAGI updates the risk benefit analysis promptly based on up to date data & hopefully our Federal Government take a step to look & compare to others successfully vaccines roll out countries.

  17. Jean Brewster says:

    Totally agree with this article. I cannot understand why the more at risk portion of the population are not allowed access to the safer and more effective Pfizer vaccine.

  18. Michael says:

    I agree with all but one of the comments. The real issue is not the relative effectiveness of the various vaccines, although AstraZeneca in my opinion does seem a sub optimal option relative to say Moderna and Pfizer. Fundamentally it is a travesty that those Australians over 50 are being treated as second class citizens relative to those under 50 in respect to the right to access the same vaccines as those under 50. Is it not time to formalise this process by setting up a petition documenting this opposition since I believe this will be the only way that equality for all can be achieved. Does anybody know if any formal objection mechanisms are already up and running?

  19. z says:

    Australian government should judge the efficiency of AstraZeneca vaccine based on the development of the COVID-19, especially the UK where more than half of the population has been vaccinated with AstraZeneca, and adjust the Australian vaccine rollout strategy in a timely manner. Choose the right vaccine to avoid the same kind of epidemic situation as the UK, namely: high vaccination rate,and high rate of new infections from the Delta virus happend every day at the same time.

Back to Top

Press Ctrl+C to copy