Emma McBryde, James Cook UniversityRecently released modelling from the Doherty Institute, which the federal government used to back its roadmap out of the pandemic, misses one critical point — the importance of vaccinating children.
The Doherty modelling instead focuses on vaccinating 70-80% of the adult population as thresholds for easing various restrictions, such as lockdowns. It says vaccinating younger adults, in particular, is important to reach these thresholds.
However, our modelling shows vaccinating children is vital if we are to reach herd immunity, which would allow us to ease restrictions and safely open up.
This would mean potentially vaccinating children as young as 5 years old.
However, we are still waiting to see if this is safe and effective, with trials under way in the United States. So we need a plan that assumes we may never achieve herd immunity.
Here’s what our modelling shows and how it differs from the modelling used to advise the federal government.
Here’s what we did
Our modelling, which we’ve uploaded as a pre-print and has yet to be peer-reviewed, considers different vaccine strategies for Australia to achieve herd immunity. That’s when we can expect no sustained transmission of the virus in the community.
We take into account the Delta variant, which is twice as infectious as the original Wuhan strain of the virus, and has a reproduction number estimated between 5 and 10. In other words, this is when one person infected with Delta is estimated to infect 5-10 others.
We also consider different contact patterns across various age groups. This is because some age groups are more mobile and have many contacts. If infected, these people are more likely to infect many others, particularly of similar age, which can lead to reservoirs of transmission.
We combine this information with possible vaccine effects. These include the possibility of having the vaccine then becoming infected, having symptoms, and if infected, how serious the illness is and how infectious people are.
This allows us to model what’s likely, given we’re focused on the Delta variant for now, and allows us to assess the impact of strategies across different age groups, types of vaccines and percentage vaccinated.
Our interactive tool also allows rapid response to changing information, such as new variants, or new evidence about vaccine impact.
Delta is more infectious
The Wuhan strain had a basic reproduction number of 2.5. This means, at the start of the pandemic, one person infected with it was expected to infect 2.5 others.
If the Delta variant is twice as infectious, this means its basic reproduction number may be over 5 (at the lower range of international estimates). So this changes the number (and type) of people we need to vaccinate to reach herd immunity considerably.
The simplest form of the herd immunity equation would suggest we needed to fully immunise 60% of the population to achieve herd immunity for the Wuhan strain but as much as 80% for the Delta variant.
If we take into account how different age groups mingle or are in contact with others, the situation is worse.
For the Wuhan strain, children were not as infectious or susceptible to infection and we predict that if we vaccinate 65% of the adults, transmission would not continue among children.
However, with the Delta variant, we predict children will continue to infect other children, even when most adults are vaccinated.
We also know both the AstraZeneca and Pfizer vaccines are less able to protect against the Delta variant, with a reduced efficacy after one dose and slightly reduced efficacy after two doses.
All this makes achieving herd immunity a great challenge.
We estimate if the reproduction number is 5, then vaccinating 85% of the population, including children down to age 5, will be necessary to achieve herd immunity.
If the reproduction number is as low as 3, then vaccinating children will not be necessary to achieve herd immunity and we will only need to vaccinate 60% of the population.
The Doherty modelling uses an effective reproduction number of 3.6. This explains why its modelling does not see vaccinating children as critical to reaching herd immunity. This is the major difference between our model and theirs.
What happens next?
Of course, new variants may arise pushing Delta aside, and the world post-COVID is unpredictable.
The lesson from Delta is if we don’t vaccinate children, we may need to continue some form of public health action to prevent large-scale circulation of the virus.
This would not require stringent lockdown, but may require ongoing mask use and physical distancing, including in children. The alternative is to reduce the focus on case numbers, expect transmission and focus on protecting the most vulnerable.
Do we need to reach herd immunity?
Herd immunity is not the only possible target. Even if we don’t reach full herd immunity, we may achieve “herd protection”. This provides some reduced risk to people who can’t or won’t be vaccinated, and it will make outbreaks smaller and easier to control.
And without full herd immunity, individuals still benefit from vaccination as they are dramatically less likely to die from COVID.
Do we need to change our vaccination strategy?
We predict Australia’s strategy of vaccinating the elderly and vulnerable first is the best strategy for reducing deaths under most circumstances, particularly when there is insufficient vaccine available.
But once the most vulnerable groups have been covered, we should turn our attention to the highest transmitters to achieve herd protection. In Australia, this group is the late teens and young adults.
Whether we next focus on vaccinating children is controversial and many people have voiced their concerns about going down this path. This is because COVID is generally a very mild illness for most children — although long COVID and life-threatening complications can arise.
So we need to balance the risks with benefits. But included in the benefits should be the potential benefit of herd protection and the freedoms that may bring.
Julie Leask, University of Sydney and James Wood, UNSWAs we try to control COVID-19, many people are keen to know what proportion of the population will need to be vaccinated in order to reach “herd immunity”.
It’s a reasonable question. People are asking because they want to know when we’ll see an end to lockdowns; when they’ll be able to reunite with loved ones overseas; when their businesses will have more security; when headlines will no longer be dominated by COVID-19.
Right now expert modellers are plugging in numbers and looking at various scenarios to estimate the scope of protection different levels of vaccination coverage will give us. We’re expecting to see the results of this modelling from the Doherty Institute as early as this week.
But it’s important to acknowledge it’s difficult to pin down a single magic number for herd immunity.
What is herd immunity again?
To understand why experts often avoid pinpointing a single vaccination figure needed to reach herd immunity for COVID-19, let’s first recap the concept.
Herd immunity is when immunity in a population is high enough to block the pathway for the ongoing transmission of the disease.
While vaccination provides each of us with direct protection against disease, with herd immunity, even people who are unvaccinated benefit from that blocked transmission pathway.
Different diseases have different thresholds for herd immunity. For measles, for example, the herd immunity threshold is 92%-94%. Estimates for COVID-19 have varied, with some putting it at 85% or higher.
However, many hesitate to give a single number. Here are three reasons why.
1. Variations in the vaccines, and the disease itself
A single herd immunity figure is difficult to estimate when the infectiousness of SARS-CoV-2 (the virus that causes COVID-19) remains so variable.
We understand the infectiousness of a disease by looking at the R0, or reproduction number — the average number of people infected by one case where no control measures are in place. The ancestral strains of SARS-CoV-2 have an R0 of 2-3, but Delta is estimated to be twice as infectious, with an R0 around 4-6.
The type of vaccine, doses given (whether one or both), and how well the vaccines cover the different variants all factor in.
Estimates from the United Kingdom show two doses of the Pfizer vaccine are between 85% and 95% effective against symptomatic disease with the Alpha variant, while two doses of AstraZeneca are 70% to 85% effective. Overall vaccine effectiveness appears to drop about ten percentage points with the Delta variant.
The lower the vaccine effectiveness, the higher the level of coverage we’ll need to control COVID well.
2. We cannot cover the entire population yet
The Pfizer vaccine has now been provisionally approved for 12-15-year-olds in Australia. If it becomes routinely recommended for this age group, it will still take time to vaccinate them. Even once that occurs, there will remain a gap in our population protection among younger children.
Children should benefit somewhat from adult vaccination. In England, where overall vaccine uptake is 48.5% for two doses, there was initially a decline in infections for children aged under ten years. This is partly due to indirect protection offered by adults being vaccinated.
3. Population protection will vary in time and space
There is rarely a neat threshold after which everything changes for good. Vaccine protection in individuals is likely to wane over time. With that and new variants appearing, we will almost certainly need boosters to sustain population protection against COVID-19.
With influenza vaccination, we rarely even talk about herd immunity, because the duration of protection is so short. By the next flu season, immunity from the current season’s vaccine will be much less effective against the newest viral strain.
Spatially, protection can vary across localities and demographics. Even in a country that has reached a herd immunity threshold for vaccination coverage against measles, you can see small outbreaks in pockets of lower coverage in kids, or where a cohort of teens and adults weren’t adequately vaccinated as children.
The capacity to achieve herd immunity is also affected by population density and how much people in a population mix with a variety of others — what’s called heterogeneity of mixing.
Life will gradually change as more people are vaccinated
Given these factors, it’s understandable experts often avoid giving a single figure for herd immunity.
With the infectiousness of Delta, we will need very high vaccination rates. Then, life will look different, particularly once this happens globally. Australia will be able to relax its border restrictions. We will likely see modified forms of quarantine, such as home quarantine, for those who are fully vaccinated.
COVID outbreaks will happen, but they will be less risky, with fewer people susceptible to serious illness. City or state-wide outbreaks will be replaced by more localised ones.
We will still require good public health measures like rapid contact tracing and isolation. Rapid tests may be used more often. New treatments may be found.
All the while, we need to be as concerned about global vaccine coverage as we are about national coverage. Because all people, regardless of means, have a right to the freedoms and security that come from COVID-19 protection.
And as we’ve heard from global leaders, “None of us will be safe until everyone is safe”.
Tony Blakely, The University of Melbourne and Vijaya Sundararajan, La Trobe UniversityThe first phase of Prime Minister Scott Morrison’s four-stage “pathway out of the COVID-19 pandemic”, announced on Friday, focuses on vaccinating as many Australians as possible, while halving the cap on international arrivals.
Morrison expects phase one, which we’re currently in, to be in place until 2022. But he said it’s hard to give a definitive answer on when we’ll get to phase two because vaccination targets have not yet been set. He’s waiting for the modelling.
Phase two of Morrison’s plan is a move back to the current levels of international arrivals, and a separate cap for vaccinated travellers. Phase three sees Australians able to travel abroad and no cap on returnees.
We are undertaking our own detailed modelling of border openings and while we’re yet to release our models, we can still arrive at fairly solid conclusions now. We’re basing these on the theory that Australia could substantially open its borders in the second quarter of 2022 with quarantine-free travel from many (but not all) countries.
The key trick, though, is to not think of vaccination as the only intervention. It is vaccination — together with three other measures: ongoing aggressive contact tracing, mask-wearing in high-risk settings and some physical distancing — that will make it safe to open.
Put another way, even when we immunise all Australians who want to be protected against COVID-19, we’re unlikely to achieve herd immunity through vaccination alone.
Modelling performed by the University of Sydney, the Burnett Institute and previously by us all point to the reality that opening the borders before about two-thirds of the population is vaccinated (in the absence of strong additional methods) could cause considerable illness and death.
So it seems unethical to substantially open the borders until everyone has had a fair opportunity to get vaccinated. All going well, this should occur sometime in the second quarter of 2022.
This would, however, be dependent on children also having the opportunity to be vaccinated. While other countries have started vaccinating children, Australia has yet to approve this use.
Vaccination alone won’t get us herd immunity
Some vaccination programs alone can achieve herd immunity, or resilience, meaning the virus won’t spread easily and exponentially, in the absence of masks, contact tracing and the other measures we have used during the pandemic.
But given the Delta variant means an average infected person infects five others without any other measures in place, and given vaccines are not perfect, Australia would need 90% of adults and children vaccinated to achieve herd immunity (through vaccination alone). This is unlikely.
There will be some waning in vaccine immunity over time, and new variants for which Pfizer and AstraZeneca are less effective. However, even with 100% of the population vaccinated, herd immunity may not be achieved by vaccination alone until booster vaccines become available.
How did we get to 90%?
The Delta variant has an R0 (the number of people one infected person on average infects) of about 5.0 under pre-COVID-19 ways of living. This is twice that of the original Wuhan virus which had an R0 about 2.5.
For an R0 of 5.0, theoretically, 80% of the population have to be immune — not just vaccinated — for virus transmission not to take off.
But the actual vaccine coverage required is higher, as the vaccines aren’t 100% effective at stopping any infection. And even though a person has been vaccinated, it doesn’t mean they’re always immune from the virus, as the person may not develop a strong immune response.
Vaccination with two doses of Pfizer among adults aged 16-60 years old is likely about 80% successful at stopping any Delta infection. For those 60 or older, the effectiveness of AstraZeneca at reducing the risk of any infection is less: about 60% against the Delta virus.
Over half, 58%, of the population are aged 16-59, and 23% of the population are aged 60-plus. So for 80% vaccination coverage of adults, the estimated percentage of the population who are immune is 48% — well short of the 80% herd immunity threshold.
For those still at risk of getting infected after either Pfizer or AstraZeneca vaccination, they are 50% less likely to transmit it. So the 16% of the population who were vaccinated but can still get infected are half as likely to pass it on.
The remaining 36% of the population are fully susceptible (20% of adults, and all children).
Extending this maths, we need about 90% of everyone (children and adults) to be vaccinated to achieve herd immunity. This is unlikely to happen.
What’s our alternative?
Using both the above theory, and what we have seen so far in our modelling, we can outline scenarios going forward:
From now until about April 2022, we propose we vaccinate as many people as possible and retain the goal of zero community transmission. Eliminating the virus after each outbreak will get easier with increasing vaccine coverage. And the probability each month of a lockdown somewhere in Australia will diminish.
This phase in our modelling is similar to the government’s first phase.
Back to our scenario: in early-to-mid 2022, we propose some modest opening of borders, through travel bubbles to other countries with virtually no community transmission, such as China and Singapore.
This is a divergence from the government’s second phase.
By mid-2022, we could allow many more countries to have quarantine-free travel to Australia, and arrivals might jump to something like half the volume pre-COVID-19.
However, people from high-risk countries — with a peak in infection, or a new variant of concern — would still have to go through 14-day (or modified) quarantine, in Howard Springs or purpose-built facilities, rather than hotel quarantine.
Then we have two choices.
If we open up to pre-COVID settings — with no contact tracing, no masks, no physical distancing — there will be repeated and serious outbreaks requiring lockdowns.
We might see enough natural infection to top-off the vaccination-induced immunity to get us to something like herd immunity. But the numbers are frightening: we may need about a fifth of us to acquire natural infection. That is five million Australians infected with substantial illness and death.
Thankfully, we have another choice: accept that some restrictions and public health measures need to continue beyond mid-2022 to complement vaccination.
As noted in the graph above, there is still a lot of uncertainty about how we open up the borders. We, and other groups, are actively modelling these options to more fully articulate and quantify pathways out.
The good news is even though herd immunity through vaccination alone is no longer a realistic option, there is an alternative path out through high vaccine coverage augmented by ongoing contact tracing, mask-wearing and physical distancing.
Put another way, 80% vaccination of adults pulls the effective reproductive rate down from 5.0 to 2.2, which is not enough to stop exponential spread. But a mix of mask-wearing, contact tracing and physical distancing will be enough to pull the effective reproductive rate to less than 1.0 — low enough to halt the spread of COVID-19.
Tony Blakely, Professor of Epidemiology, Population Interventions Unit, Centre for Epidemiology and Biostatistics, Melbourne School of Population and Global Health, The University of Melbourne and Vijaya Sundararajan, Associate Professor of Epidemiology and Biostatistics, Department of Medicine, University of Melbourne; Professor, Public Health, La Trobe University
Maria De Jesus, American University School of International ServiceIn the race between infection and injection, injection has lost.
Public health experts estimate that approximately 70% of the world’s 7.9 billion people must be fully vaccinated to end the COVID-19 pandemic. As of June 21, 2021, 10.04% of the global population had been fully vaccinated, nearly all of them in rich countries.
Only 0.9% of people in low-income countries have received at least one dose.
I am a scholar of global health who specializes in health care inequities. Using a data set on vaccine distribution compiled by the Global Health Innovation Center’s Launch and Scale Speedometer at Duke University in the United States, I analyzed what the global vaccine access gap means for the world.
A global health crisis
Supply is not the main reason some countries are able to vaccinate their populations while others experience severe disease outbreaks – distribution is.
Many rich countries pursued a strategy of overbuying COVID-19 vaccine doses in advance. My analyses demonstrate that the U.S., for example, has procured 1.2 billion COVID-19 vaccine doses, or 3.7 doses per person. Canada has ordered 381 million doses; every Canadian could be vaccinated five times over with the two doses needed.
Overall, countries representing just one-seventh of the world’s population had reserved more than half of all vaccines available by June 2021. That has made it very difficult for the remaining countries to procure doses, either directly or through COVAX, the global initiative created to enable low- to middle-income countries equitable access to COVID-19 vaccines.
Benin, for example, has obtained about 203,000 doses of China’s Sinovac vaccine – enough to fully vaccinate 1% of its population. Honduras, relying mainly on AstraZeneca, has procured approximately 1.4 million doses. That will fully vaccinate 7% of its population. In these “vaccine deserts,” even front-line health workers aren’t yet inoculated.
Haiti has received about 461,500 COVID-19 vaccine doses by donations and is grappling with a serious outbreak.
Even COVAX’s goal – for lower-income countries to “receive enough doses to vaccinate up to 20% of their population” – would not get COVID-19 transmission under control in those places.
The cost of not cooperating
Last year, researchers at Northeastern University modeled two vaccine rollout strategies. Their numerical simulations found that 61% of deaths worldwide would have been averted if countries cooperated to implement an equitable global vaccine distribution plan, compared with only 33% if high-income countries got the vaccines first.
Put briefly, when countries cooperate, COVID-19 deaths drop by approximately in half.
Vaccine access is inequitable within countries, too – especially in countries where severe inequality already exists.
In Latin America, for example, a disproportionate number of the tiny minority of people who’ve been vaccinated are elites: political leaders, business tycoons and those with the means to travel abroad to get vaccinated. This entrenches wider health and social inequities.
The result, for now, is two separate and unequal societies in which only the wealthy are protected from a devastating disease that continues to ravage those who are not able to access the vaccine.
A repeat of AIDS missteps?
This is a familiar story from the HIV era.
In the 1990s, the development of effective antiretroviral drugs for HIV/AIDS saved millions of lives in high-income countries. However, about 90% of the global poor who were living with HIV had no access to these lifesaving drugs.
Concerned about undercutting their markets in high-income countries, the pharmaceutical companies that produced antiretrovirals, such as Burroughs Wellcome, adopted internationally consistent prices. Azidothymidine, the first drug to fight HIV, cost about US$8,000 a year – over $19,000 in today’s dollars.
That effectively placed effective HIV/AIDS drugs out of reach for people in poor nations – including countries in sub-Saharan Africa, the epidemic’s epicenter. By the year 2000, 22 million people in sub-Saharan Africa were living with HIV, and AIDS was the region’s leading cause of death.
The crisis over inequitable access to AIDS treatment began dominating international news headlines, and the rich world’s obligation to respond became too great to ignore.
“History will surely judge us harshly if we do not respond with all the energy and resources that we can bring to bear in the fight against HIV/AIDS,” said South African President Nelson Mandela in 2004.
Pharmaceutical companies began donating antiretrovirals to countries in need and allowing local businesses to manufacture generic versions, providing bulk, low-cost access for highly affected poor countries. New global institutions like the Global Fund to Fight AIDS, Tuberculosis, and Malaria were created to finance health programs in poor countries.
Pressured by grassroots activism, the United States and other high-income countries also spent billions of dollars to research, develop and distribute affordable HIV treatments worldwide.
A dose of global cooperation
It took over a decade after the development of antiretrovirals, and millions of unnecessary deaths, for rich countries to make those lifesaving medicines universally available.
Fifteen months into the current pandemic, wealthy, highly vaccinated countries are starting to assume some responsibility for boosting global vaccination rates.
Leaders of the United States, Canada, United Kingdom, European Union and Japan recently pledged to donate a total of 1 billion COVID-19 vaccine doses to poorer countries.
It is not yet clear how their plan to “vaccinate the world” by the end of 2022 will be implemented and whether recipient countries will receive enough doses to fully vaccinate enough people to control viral spread. And the late 2022 goal will not save people in the developing world who are dying of COVID-19 in record numbers now, from Brazil to India.
The HIV/AIDS epidemic shows that ending the coronavirus pandemic will require, first, prioritizing access to COVID-19 vaccines on the global political agenda. Then wealthy nations will need to work with other countries to build their vaccine manufacturing infrastructure, scaling up production worldwide.
Finally, poorer countries need more money to fund their public health systems and purchase vaccines. Wealthy countries and groups like the G-7 can provide that funding.
These actions benefit rich countries, too. As long as the world has unvaccinated populations, COVID-19 will continue to spread and mutate. Additional variants will emerge.
As a May 2021 UNICEF statement put it: “In our interdependent world no one is safe until everyone is safe.”
Maria De Jesus, Associate Professor and Research Fellow at the Center on Health, Risk, and Society, American University School of International Service
At that stage, many commentators were arguing we should let COVID-19 rip through populations so we could get enough people immune to the virus that it would stop spreading. As I argued at the time, this was a terrible idea that would overwhelm hospitals and gravely sicken and kill many people.
Now we have safe and effective vaccines, we can aim to reach herd immunity in a much safer way. It’s certainly possible we’ll be able to reach and maintain local herd immunity in certain regions, states and countries. However the pandemic ends, it will involve this immunity to some extent.
But it’s still very uncertain whether long-term, global herd immunity is achievable. It’s quite likely the coronavirus could continue to spread even in places with high proportions of their populations vaccinated. It will probably never be eliminated.
However, if we’re all vaccinated, we’ll be largely safe from the worst ravages of the infection even if it does break out.
What is herd immunity again? And what does it mean for us long-term?
There are a few different definitions of herd immunity. Nevertheless, they all deal with the “reproductive number” of a disease, known as the R number. This is the average number of people an infected person will pass a disease on to, at a certain point in time.
The R number depends on how infectious a disease is. Measles is often used as an example, because it’s one of the most infectious diseases. In a group of people among whom no one is immune to the disease, on average one person will pass measles on to around 15 others.
But as more people in the community become immune, either through vaccination or getting the disease and recovering, each infected person will pass on the infection to fewer and fewer others. Eventually, we reach a point at which the R number is below 1, and the disease starts to die out. The R number falling below 1 here is in a population where there are no social restrictions, so the disease starts to die out because of immunity and not because of measures like lockdowns. This is one definition of herd immunity.
However, another potential definition is that herd immunity is a state where enough people are immune in a population that a disease won’t spread at all. One of the more confusing parts of the pandemic is we scientists haven’t always used the same definition across the board.
For example, when we say “reached the herd immunity threshold”, we could be talking about a transient state where we’re likely to see another epidemic in the near future, or a situation where the vast majority of a population is immune and thus the disease won’t spread at all. Both are technically “herd immunity”, but they’re very different ideas.
How’s herd immunity calculated?
COVID-19 has an R number somewhere between 2 and 4 in groups of people where no one is immune. Using a simple mathematical formula, 50-75% of people need to be immune to COVID-19 for the R number to fall below 1 so it starts to die out, in a population with no social restrictions. Some researchers have done more complex versions of this calculation throughout the pandemic, but that’s the basic idea behind them all.
However, herd immunity is a moving target. For example, if everyone in your local population is taking great care to socially distance, COVID-19 won’t spread as much. Therefore, in practice, different cultures spread diseases to different extents, so the R number varies in both place and time.
Vaccines are the ultimate path to long-term immunity
Our COVID-19 vaccines are safe and effective. Without going too much into the debate over which one is better, they are all capable of getting us to a point at which the disease would no longer spread through the community. For some vaccines, the percentage of people who we need to immunise is higher. But it’s the same basic idea regardless, and we need to vaccinate as many people as we can to have a shot at herd immunity.
We can already see this happening in some places. For example, in the United Kingdom and Israel, enough people have been vaccinated that even though restrictions are being relaxed, infection rates are staying low or continuing to drop. This is a beautiful sight.
The coronavirus will probably never be eliminated
Even with great vaccines, the problem is complex. There are almost always communities who aren’t immunised, for various reasons, even in countries with large proportions of the total population vaccinated. These small communities can continue to get sick and spread the disease long after the general population has passed the herd immunity line, which means there may always be some risk of COVID-19 outbreaks.
On top of this, new variants of the virus have emerged. Our current vaccines are probably enough to provide most people with immunity to the original strain in the long term. But several variants may substantially reduce our vaccines’ effectiveness as time goes by, so we may need boosters at some point.
What’s more, the global situation isn’t rosy. India and Brazil are currently experiencing horrifying COVID-19 outbreaks. The global case count continues to rise, partially because developed nations have hoarded vaccine doses jealously, despite this being a terrible approach to a pandemic. Rising case numbers anywhere increase the chances even more variants pop up, thereby impacting us all.
Even if we overcome vaccine hesitancy and global inaction, and we immunise most of the world, we may not be protected against the virus forever. Even higher-income nations may never get rid of COVID-19.
It’s quite likely this virus will never be eradicated (eliminated from every country across the globe). There may be places where the disease is gone, where local campaigns are successful, but there’ll also be places where the disease is still spreading.
What does this mean for Australia?
This presents a challenge for Australia. We have virtually no local COVID-19 transmission, so there’s no real risk from the virus as long as our border controls hold steady.
However, we probably can’t maintain this level of vigilance forever. And even with our very effective vaccines, we may not have long-term herd immunity — of any definition — to COVID-19.
At some point in the future, it’s likely we will see some cases of COVID-19 spreading in even the safest places in the world, including Australia.
Even so, getting vaccinated enormously reduces your risk of severe outcomes like hospitalisation and death. We should aim to vaccinate as many people as possible, while acknowledging that the future is inherently uncertain, and herd immunity is a challenging goal.
White House advisers have made the case recently for a “natural” approach to herd immunity as a way to reduce the need for public health measures to control the SARS-CoV-2 pandemic while still keeping people safe. This idea is summed up in something called the Great Barrington Declaration, a proposal put out by the American Institute for Economic Research, a libertarian think tank.
The basic idea behind this proposal is to let low-risk people in the U.S. socialize and naturally become infected with the coronavirus, while vulnerable people would maintain social distancing and continue to shelter in place. Proponents of this strategy claim so-called “natural herd immunity” will emerge and minimize harm from SARS-CoV-2 while protecting the economy.
Another way to get to herd immunity is through mass vaccinations, as we have done with measles, smallpox and largely with polio.
A population has achieved herd immunity when a large enough percentage of individuals become immune to a disease. When this happens, infected people are no longer able to transmit the disease, and the epidemic will burn out.
As a professor of behavioral and community health sciences, I am acutely aware that mental, social and economic health are important for a person to thrive, and that public health measures such as social distancing have imposed severe restrictions on daily life. But based on all the research and science available, the leadership at the University of Pittsburgh Graduate School of Public Health and I believe this infection-based approach would almost certainly fail.
Dropping social distancing and mask wearing, reopening restaurants and allowing large gatherings will result in overwhelmed hospital systems and skyrocketing mortality. Furthermore, according to recent research, this reckless approach is unlikely to even produce the herd immunity that’s the whole point of such a plan.
Vaccination, in comparison, offers a much safer and likely more effective approach.
An uncertain path to herd immunity
Herd immunity is an effective way to limit a deadly epidemic, but it requires a huge number of people to be immune.
The proportion of the population required for herd immunity depends on how infectious a virus is. This is measured by the basic reproduction number, R0, how many people a single contagious person would infect in a susceptible population. For SARS-CoV-2, R0 is between 2 and 3.2. At that level of infectiousness, between 50% and 67% of the population would need to develop immunity through exposure or vaccination to contain the pandemic.
The Great Barrington Declaration suggests the U.S. should aim for this immune threshold through infection rather than vaccination.
To get to 60% immunity in the U.S., about 198 million individuals would need to be infected, survive and develop resistance to the coronavirus. The demand on hospital care from infections would be overwhelming. And according to the WHO estimated infection fatality rate of 0.5%, that would mean nearly a million deaths if the country were to open up fully.
The Great Barrington Declaration hinges on the idea that you can effectively keep healthy, infected people away from those who are at higher risk. According to this plan, if only healthy people are exposed to the virus, then the U.S. could get to herd immunity and avoid mass deaths. This may sound reasonable, but in the real world with this particular virus, such a plan is simply not possible and ignores the risks to vulnerable people, young and old.
You can’t fully isolate high-risk populations
The Great Barrington Declaration calls for “allowing those who are at minimal risk of death to live their lives normally … while protecting those who are at highest risk.” Yet healthy people can get sick, and asymptomatic transmission, inadequate testing and difficulty isolating vulnerable people pose severe challenges to a neat separation based on risk.
First, the plan wrongly assumes that all healthy people can survive a coronavirus infection. Though at-risk groups do worse, young healthy people are also dying and facing long-term issues from the illness.
Second, not all high-risk people can self-isolate. In some areas, as much as 22% of the population have two or more chronic conditions that put them at higher risk for severe COVID-19. They might live with someone in the low-risk group and they still must shop, work and do the other activities necessary for life. High-risk individuals will come in contact with the low-risk group.
So can you simply guarantee that the low-risk people who interact with the high-risk group are uninfected? People who are infected but not showing symptoms may account for more than 30% of transmission. This asymptomatic spread is hard to detect.
Asymptomatic spread is compounded by shortcomings in the quality of testing. Currently available tests are fairly good, but do not reliably detect the coronavirus during the early phase of infection when viral concentrations can be low.
Accordingly, identifying infection in the low-risk population would be difficult. These people could go on to infect high-risk populations because it is impossible to prevent contact between them.
Sweden’s herd immunity failure
Without sharp isolation of these two populations, uncontrolled transmission in younger, healthier people risks significant illness and death across vulnerable populations. Both computer models and one real-world experiment back up these fears.
A recent U.K. modeling effort assessed a range of relaxed suppression strategies and showed that none achieved herd immunity while also keeping cases below hospital capacity. This study estimated a fourfold increase in mortality among older people if only older people practice social distancing and the remainder of the population does not.
But epidemiologists don’t have to rely on computer models alone. Sweden tried this approach to infection-based herd immunity. It did not go well. Sweden’s mortality rate is on par with Italy’s and substantially higher than its neighbors. Despite this risky approach, Sweden’s economy still suffered, and on top of that, nowhere near enough Swedes have been infected to get to herd immunity. As of August 2020, only about 7.1% of the country had contracted the virus, with the highest rate of 11.4% in Stockholm. This is far short of the estimated 50%-67% required to achieve herd immunity to the coronavirus.
Exposure versus vaccination
There is one final reason to doubt the efficacy of infection-based herd immunity: Contracting and recovering from the coronavirus might not even give immunity for very long. One CDC report suggests that “people appear to become susceptible to reinfection around 90 days after onset of infection.” The potentially short duration of immunity in some recovered patients would certainly throw a wrench in such a plan. When combined with the fact that the highest estimates for antibody prevalence suggest that less than 10% of the U.S. population has been infected, it would be a long, dangerous and potentially impassable road to infection-based herd immunity.
But there is another way, one that has been done before: mass vaccination. Vaccine-induced herd immunity can end this pandemic the same way it has mostly ended measles, eradicated smallpox and nearly eradicated polio across the globe. Vaccines work.
Until mass SARS-CoV-2 vaccination, social distancing and use of face coverings, with comprehensive case finding, testing, tracing and isolation, are the safest approach. These tried-and-true public health measures will keep viral transmission low enough for people to work and attend school while managing smaller outbreaks as they arise. It isn’t a return to a totally normal life, but these approaches can balance social and economic needs with health. And then, once a vaccine is widely available, the country can move to herd immunity.
People infected with SARS-CoV-2, the virus that causes COVID-19, can spread the virus when they speak, sing, cough, sneeze or even just breathe. Scientists think face masks help limit virus spread by protecting everyone else from the infected wearer. As a result, face maks are now mandatory in many cities, states and countries to limit the spread of COVID-19.
People typically wear surgical, cloth or other face coverings that don’t completely prevent the virus from infecting the wearer, though medical grade surgical masks do appear to offer more protection. Nonetheless, these don’t have the same level of protection as N95 or P2 “respirator” masks worn by many health-care workers. Additionally, how we wear the mask matters, as touching it often and not completely covering the nose and mouth renders it ineffective.
While these face coverings may not completely prevent us from getting infected with COVID-19, they probably reduce the number of virus particles we inhale — the “viral dose”. Scientists think a lower viral dose can reduce the severity of the disease we get. Indeed, where universal face masking is implemented, a much higher proportion of new infections with COVID-19 are asymptomatic.
Could this lower viral dose help us build some immunity to the disease? Two researchers from the University of California have raised this possibility, writing in the prestigious New England Journal of Medicine. Although the theory hasn’t been proven yet.
The dose makes the poison
How much virus we are initially infected with is a key determinant of how sick we get, according to evidence from other viruses and animal studies. We also know this is true in hamsters that have been experimentally infected with SARS-CoV-2.
Imagine if you touch a door handle that happens to have one virus particle on it, and then touch your nose and breathe that particle in. You will be infected with that one virus particle. One estimate, published in the Lancet, suggested one SARS-CoV-2 virus particle will have replicated to make nearly 30 new virus particles in 24 hours. Those 30 new particles can then go on to infect 30 more cells, giving rise to 900 new particles in the next 24 hours or so.
Now imagine someone sneezes right in your face and you inhale 1,000 virus particles. After one round of replication you could have 30,000 particles, and then 900,000 in the round after. In the same period of time your body could be dealing with 1,000 times more virus, compared to the first scenario.
Once the immune system detects the virus, it has to race to get it under control and stop it replicating. It does this in three main ways:
telling our cells how to disrupt viral replication
making antibodies that recognise and neutralise the virus to stop it infecting more cells
making T cells that specifically kill virus-infected cells.
While the first step is relatively quick, creating specific antibodies and T cells takes days or even weeks. Meanwhile, the virus is replicating over and over again. So the initial dose of virus really determines how much of the body the virus has infected before the immune system kicks fully into gear.
What about for long-term immunity?
The more virus there is, the bigger the immune response has to be to control it. And it’s the immune response that actually causes the symptoms, like fever. In an asymptomatic infection, we think the immune system has probably managed to get the virus under control early on, so the immune response itself is possibly smaller, and so we won’t see any symptoms.
We also think many cases of very severe COVID-19 might really be a result of the immune system overreacting. This is why the steroid treatment dexamethasone, which suppresses the immune response, shows promise in treating severe cases (but not mild ones).
After we clear an infection, we keep some immune cells around in case we get infected again. These are B cells, which produce antibodies specific to SARS-CoV-2, and T cells, which kill virus-infected cells. This is also the premise behind vaccination: we can trick the immune system into making those SARS-CoV-2 specific cells without having been infected.
Because face masks might allow a small number of virus particles through, wearers might be more likely to get asymptomatic infections. This might be enough to protect them from future infection with SARS-CoV-2. So if we are in a situation where there is high community transmission, and we can’t always maintain physical distance, wearing a face mask might be a factor that helps us in the long run.
It’s another argument in favour of masks
While this sounds promising, there’s still a lot we don’t understand. We don’t know yet whether an asymptomatic infection would generate enough immunity to guard against future infection — or if this is even measurable.
Viral dose is likely to be just one factor among many that determines how sick someone gets with COVID-19. Other factors include age, sex, and other underlying conditions. Finally, even with asymptomatic infections, we don’t know yet what the long term effects of COVID-19 are. It’s best to avoid getting COVID-19 altogether if possible.
Nevertheless, this is yet another reason to keep wearing face masks. As many cases of COVID-19 are asymptomatic, we could still be transmitting the virus even without symptoms. That’s why wearing a mask is a responsible thing to do, even if we feel fine.
How is the world going to go back to the days when we could grab a coffee, see a movie, or attend a concert or footy game with anyone?
Opinion suggests there are two options: an effective vaccine, or herd immunity via at least 60-80% of people becoming infected. Either one of these options requires that people become immune to SARS-CoV-2, the coronavirus that causes COVID-19.
An important new study released online this week could have a large bearing on how our future looks in 2021 and beyond.
It suggests our immunity to SARS-CoV-2 does not last very long at all — as little as two months for some people. If this is the case, it means a potential vaccine might require regular boosters, and herd immunity might not be viable at all.
Immunity dwindles quickly
Antibodies are an important part of our immune system that mainly work by physically binding to virus particles and stopping them infecting cells. They can attach to infected cells to induce cell death in some cases.
We also have T cells, another part of the immune system that is much better at recognising and killing virus-infected cells. But for COVID-19, antibodies are important in the lungs because T cells aren’t good at getting to airways where the virus first invades.
The newly released research, from Katie Doores and her team at Kings College London, looked at how long the antibody response lasted in people who had COVID-19. It has been submitted to a journal but hasn’t been peer-reviewed, so it must be treated with some caution.
Of the 65 patients studied, 63 produced antibody responses. The important measurements in the study relate to how good the response is. This is measured in the lab by putting patients’ blood serum together with infectious SARS-CoV-2 virus and seeing whether the virus can infect cells in a lab dish. This is called a “neutralisation assay”, and here the results were good.
Around 60% of people produced a very potent neutralisation response that stopped virus growing in the lab cells.
Finally, the researchers measured how long the antibody response lasted. This is the most important data. Unfortunately, antibodies levels began falling after day 20 and only 17% of patients retained a potent level at day 57. Some patients completely lost their antibodies after two months.
This suggests our immune response to SARS-CoV-2 may be lost much faster than we might have hoped, and people might thereafter be susceptible to reinfection with the virus.
One vaccine might not be enough
It therefore follows that COVID-19 vaccines may not be as effective as we hope. The fact antibody levels reduce over time is normal, but this typically happens much more slowly. Antibody responses against the mumps, measles and chickenpox viruses last for more than 50 years. A tetanus vaccination wanes more quickly but still lasts 5-10 years before a booster is needed.
So why is this happening? It comes down to the nature of the SARS-CoV-2 coronavirus itself. The four normal strains of coronaviruses that cause common colds in humans also fail to prompt a long-lasting immune response, with most people losing antibodies completely after 6-12 months. Coronaviruses in general seems to be particularly good at not being well recognised by our immune system. Indeed, a feature of common cold coronaviruses is that people get reinfected by them all the time.
SARS, another coronavirus which caused a pandemic in 2003, seems to produce a slightly longer antibody response, lasting up to three years. It’s still a long way short of a lifetime, but it perhaps helps explain why the virus disappeared in 2003.
Herd immunity might be in trouble
So herd immunity may not be the solution some think. This is because if immunity is short-lived, we will be in an ongoing cycle of endless reinfection. For herd immunity to be effective we need a high percentage (perhaps more than 60%) of people to be immune at any one time to disrupt chains of transmission. This can’t happen if a lot of reinfection is occurring.
The hope is vaccines will give much stronger and longer lasting immune responses to the virus than getting and recovering from COVID-19 itself. Indeed, the first vaccine candidates from Pfizer and Moderna, reported in early July, show very strong immune responses.
However, these studies only reported out to 14 and 57 days, respectively, after vaccinations were completed. They don’t tell us whether there is a long-lived response that we would need for a vaccine to be truly protective. Phase 3 trials designed to measure this are due to report in December 2020, so watch this space.
While we wait, we should reflect on the fact that although the results of the Kings College study are in one sense disappointing news, this knowledge adds to the truly remarkable scientific progress we have made in understanding a virus that only emerged in December 2019.
This article is supported by the Judith Neilson Institute for Journalism and Ideas.
As Australia begins to relax its COVID-19 restrictions there is understandable debate about how quickly that should proceed, and whether lockdowns even made sense in Australia in the first place.
The sceptics arguing for more rapid relaxation of containment measures point to the economic costs of lockdowns and appeal to the cold calculus of cost-benefit analysis to conclude that the lives saved by lockdowns don’t justify the economic costs incurred to do so.
Their numbers don’t stack up.
To be able to weigh the value of a life against the economic costs of forgone output from lost jobs and business closures, requires placing a dollar value on one person’s life. This number is called the value of a statistical life.
What are the benefits of the shutdown? This is the value of lives saved plus any indirect economic or health benefits. Lives saved are those excess lives that would be lost if government relied on a strategy that allowed enough people to get infected to result in so-called herd immunity.
How many extra lives would be lost under this second strategy?
To answer this, we need assumptions about the virus.
The lives lost if we let it rip
The initial reproduction rate of the virus, R0, was thought to be about 2.5. This means that every 2 people infected were likely to infect another 5; producing an average infection rate per person of 2.5.
Herd immunity for COVID-19 is estimated to require roughly 60% of the population be infected before the curve begins to flatten and the peak infections fall.
This happens when the reproduction rate, R0, falls below one. Because of subsequent new infections, the total number infected over the course of the pandemic is closer to 90%.
Given a population of 25 million people and assuming a fatality rate of 1%, this would produce 225,000 deaths.
An assumption of a 1% fatality rate is low from the perspective of those making decisions at the onset of the pandemic, at a time when crucial and reliable data were missing.
Those lives are valued at $1.1 trillion
Converting those fatalities to dollars using the Australian value of a statistical life of A$4.9 million per life yields a cost of A$1.1 trillion.
In rough terms, that’s the amount we have gained by shutting down the economy, provided deaths do not skyrocket when lockdown measures are relaxed and borders re-open.
It is about three fifths of one year’s gross domestic product, which is about A$1.9 trillion.
What are the costs of the shutdown?
These are the direct economic costs from reduced economic activity plus the indirect social, medical, and economic costs, all measured in terms of national income.
A starting point is to take the lost income that occurs from the recession that has probably already begun.
What will the shutdown cost?
Let’s assume that the downturn results in a 10% drop in gross domestic product over 2020 and 2021 – about $180 billion – consistent with IMF forecasts of a fall in GDP of 6.7% in 2020 and a sharp rebound of 6.1% growth in 2021, and comparable to the Reserve Bank of Australia’s forecasts in the latest Statement on Monetary Policy.
Comparing this cost from shutting down – about $180 billion – to the benefit of $1,103 billion – makes the case for shutdown clear.
But this calculation grossly overestimates the costs of the shutdown.
The recession is a consequence of both the shutdown and the pandemic.
We need to attribute costs to each.
Most of the economic costs of the recession are likely to be due to the pandemic itself rather the shutdown.
Many costs would have been borne anyway
Even before the shutdown, economic activity was in decline.
Both in Australia and internationally air travel, restaurant bookings and a range of other activities had fallen sharply.
They were the result of a “private shutdown” that commenced before the mandated government shutdown.
Even in a country such as Sweden, where a shutdown has not been mandatory, there has been a more than 75% reduction in movement in central Stockholm and a more than 90% reduction in travel to some domestic holiday destinations.
To be generous, let’s assume the costs attributable to the government-mandated part of the shutdown are half of the total costs, making their cost A$90 billion.
In reality, they are likely to be less, one important study suggests much less.
It is hard to imagine a much bigger private shutdown not taking place had the government decided to simply let the disease rip until its spread was slowed by herd immunity.
Support is not a cost
It is also important to note that the government’s spending of A$214 billion to support the economy during the shutdown is a transfer of resources from one part of society to another rather than a cost.
It creates neither direct costs nor benefits for society as a whole, other than the economic distortions coming from raising the revenue to service the spending.
With long-term government bond rates near 1% (less than inflation), the total cost of distortions is likely to be tiny.
Of course, this discussion simplifies what are incredibly complex social, health and economic questions. There are clearly further costs, from both relaxing restrictions and keeping them in place.
Other costs are not that big
These costs are worthy of serious study and should rightly be part of a comprehensive public policy discussion. But looked at through the lens of a cost-benefit analysis these combined effects are likely to be small relative to the value of preventing mass death.
Among them are the incidence of mental health problems and domestic violence under lockdowns. They are important concerns that should be addressed by targeted and well-designed programs.
Weighing against that is evidence that economic crises are associated with declines in overall mortality rates.
While suicides rise, total mortality, including deaths from heart attacks and workplace and traffic accidents, falls.
In the specific case of this pandemic there is survey evidence based on respondents from 58 countries suggesting that strong government responses to the pandemic have been reducing worry and depression.
Also, we have to acknowledge that recessions and educational disruption have health and economic costs that are unequally spread.
The shutdown disproportionately impacts more-disadvantaged people including short-term casual workers, migrant workers, those with disabilities and the homeless.
The most-disadvantaged suffer, either way
This skewing will also be present in the herd immunity option. As New York City makes clear, a rapid spread of the disease also disproportionately impacts disadvantaged communities. One can only speculate about the disease burden should some of our remote indigenous communities get exposed.
To this we should add further achievements of the shutdown:
elimination of mental trauma and grief from losing our loved ones
avoiding the costs of possible longer-term implications of the disease, which we still know little about
avoiding a collapse in the capacity of the health system to deal with other emergencies through the sheer numbers of COVID-19 infected combined with staff shortages due to illness
Those advocating cost-benefit analysis of this kind have to apply the principle systematically. It is difficult to see how the total of these sorts of considerations on each side of the ledger could compare to the benefit of lives saved. They will be an order of magnitude, if not two, smaller.
$90 billion, versus $1.1 trillion
In the cold calculus of cost-benefit analysis, a highly pessimistic view of the economic costs of Australia’s shutdown comes to around $90 billion.
It is a small price to pay compared to the statistical value of lives the shutdown should save, around A$1.1 trillion.
It produces a simple message. The shutdown wins.
The question we now face is how quickly to relax restrictions. Here, too, there are costs and benefits, and we need to be mindful of the economic cost of a second-wave outbreak, plus mortality costs of disease spread before effective treatments or vaccine become available.
And in all of this bean counting, we should remember that putting a price tag on human life is sometimes unavoidable – such as when a doctor with access to only one ventilator has to choose between two patients.
But we shouldn’t mistake necessity for desirability. We should seek to avoid needing to make such wrenching choices whenever possible.
Dr Jen Schaefer of the Royal Children’s Hospital Melbourne assisted with the preparation of this piece.