Australia in the post-pandemic world
by Alex Harvey
- Part I - Containment
- Patient 31 events
- Can Australia withstand Patient 31 events
- Mandatory testing
- Mandatory surveillance
- Mandatory quarantine
- Part II - Treatment and prophylaxis
- Vitamin D supplementation
- Part III - Containment fails
- Overdispersion and superspreading
- Targeted restrictions
- Against full lockdowns
- Expert opinion in January
- Expert opinion in October
- Empirical evidence for lockdowns
- Household transmission
- Herd immunity in New York City
- Spurious correlations
- Early lockdowns
- Lockdowns have no place
- Part IV - Focused protection
As Australia enters the post-pandemic world, we will be forced to make difficult public health and economic choices. Our policies currently aim to build a robust public health infrastructure for containment of Covid-19 until a vaccine is available. In the event of health capacity breaching, lockdowns are used to suppress the epidemic. Recent experiences in Victoria, however, where over 800 people died, despite one of the strictest lockdowns in the world, show that our current policy framework is optimised for neither public health nor economic outcomes. In response to these inadequacies, our containment strategies must be further tightened to emulate the surveillance and quarantine approach of Taiwan, South Korea, and other successful nations of Asia. Meanwhile, if containment does fail again, we must adopt the targeted restrictions strategies used in Germanic and Scandinavian nations. Further lockdowns should be ruled out of consideration in any circumstance as they are generally not the most efficient way to suppress an epidemic, but they are always most economically impactful.
In February, 2020, scenes of deserted streets in South Korea gave early warning to us in Australia that a probably unstoppable viral pandemic was on its way. The outbreak in the Korean city of Daegu had come a few weeks after the central government of the People’s Republic of China had quarantined the entire city of Wuhan, population 11 million.
A quarantine in a free country like South Korea, however, was never seriously considered. Authorities instead pledged to test more than 200,000 members of the Shincheonji Church of Jesus, whose followers comprised a major cluster of early cases. They also offered tests to everyone in Daegu who had cold-like symptoms, an estimated 28,000 people.
The measures worked. By March 17th, Science Mag was writing of South Korea’s surprising success. The country had contained an outbreak of more than 7,000 cumulative confirmed cases.
In the mean time, however, Europe had become the new epicenter of pandemic. Deaths were soaring in Italy, Spain, France, and Germany, and the countries of the EU had, one by one, adopted the policies, not of South Korea, but of the Chinese Communist Party in Wuhan. They had closed their borders, ordered non-essential businesses to shut, and they had put their citizens in house arrest.
Infectious diseases specialist Kim Woo-Joo at Korea University told Science Mag:
South Korea is a democratic republic, we feel a lockdown is not a reasonable choice.
In this article, I make eight recommendations for change to Australia’s Covid-19 policies. I argue that our policies are inadequate; that failure inside Victoria was predictable; that success outside Victoria has been lucky; and that we must adopt the surveillance and quarantine policies of the successful Asian nations like South Korea. And in the event of containment failing again, we must look to the suppression strategies based on the targeted restrictions in use, for instance, in the Germanic and Scandinavian nations.
I agree with South Korea that lockdowns are not a reasonable choice in a free country. Over our history, more than 100,000 Australians died in wars fighting for the freedom of their children and grandchildren. I do not believe that Covid-19 policies based on perpetual rolling lockdowns is what these Australians fought and died for. But perhaps more importantly than that, the science suggests that lockdowns are not an efficient way to suppress an epidemic anyway.
Part I - Containment
Australia began the pandemic in a lucky position. We have sea borders, a low population density, and a warm, sunny climate. The prime minister’s early decision to ban travel from China bought us valuable time. We purchased test kits and tested widely. Australia was a global leader in the number of tests we did.
But Western nations have struggled with balancing ideals of freedom with the needs of containing a virus. Our attitudes have been inconsistent. We must recognise that to avoid the severe and economically catastrophic losses of freedom in lockdowns, we must be willing to give up some minor freedoms.
Patient 31 events
The western world could learn a lot from the case of South Korea’s Patient 31, one of the biggest superspreading events of the pandemic.
Patient 31 started the outbreak in the Korean city of Daegu. On February 7th, this patient, a woman in her 60s, presented herself to hospital after a minor traffic accident. While in hospital, she developed a fever and sore throat. The doctors advised her to get tested for Covid-19. She refused, twice apparently. She had not travelled internationally, and she believed she had a cold. Instead, she went out with a friend to a buffet.
While staying in the hospital, she attended church services at the Shincheonji Church of Jesus twice, on February 9th, and February 16th. There were 9,000 others attending the services with her, in a huge hall. During these services, she sang hymns loudly with the others.
On February 17th, her symptoms worsened, and finally she agreed to be tested. She tested positive, becoming South Korea’s 31st confirmed case of Covid-19. This diagram shows her movements during the infectious period:
Can Australia withstand Patient 31 events
NSW has been much praised for its effective contact tracing, mostly due to comparisons with Victoria, where control of the epidemic was lost. But if Korea’s Patient 31 were to attend church in Sydney, could we stop her from infecting hundreds here?
One of the mistakes made in the case of Patient 31 was that the woman was not tested. She was not tested, because she refused to be tested, more than once. If Patient 31 had been tested on February 17th, however, it is possible that the Daegu outbreak could have been avoided.
The Chief Health Officer in Australia has the power to compel people to be tested under the Biosecurity Act. The government, however, has stated that they do not wish for testing to be mandatory. Patient 31 shows that making testing voluntary poses an unacceptable risk. Testing must be strongly incentivised or made mandatory.
Recommendation 1: Australian governments must strongly incentivise or compel people suspected of infection to be tested.
Contact tracing in Australia is done manually by teams of contact tracers who speak to infected individuals, and find the contact details of others who may also have been infected. But given a virus that can spread exponentially, these methods do not scale beyond a small number of cases per day. And if contact tracing capacity is exceeded, then governments tend to order economically harmful restrictions and lockdowns.
In the east Asian nations, it is different. In fairness, these countries have experience with the SARS and MERS epidemics. In response to MERS, South Korea rewrote much of its infectious-disease-prevention legislation. To expand contact tracing, South Korean health authorities were given warrantless access to all CCTV footage and the geolocation data from the patients’ cell phones. With access to this data, the South Koreans proved they could contain an epidemic of more than 7,000 cumulative cases after a 1,000 person cluster within a month, without the need to panic or order lockdowns.
In Australia, The federal government provided the COVIDSafe phone app and encouraged people to download it. However, only 6 million or so did download it; the campaign was undermined somewhat by spurious privacy arguments, as well as technical flaws in the app itself.
This pandemic should have taught us by now, especially after Victoria’s disastrous second lockdown, that we must sacrifice a small amount of privacy, in order to respond to the Covid-19 and future viral pandemics.
Recommendation 2: Australia must deploy surveillance to assist contract tracers and enforce quarantine.
Perhaps even more concerning is that Australia as well as other Western nations generally ask infected Covid-19 patients to self-isolate for 14 days. These self-isolation orders are mandatory, and there are significant penalties for failing to obey. In NSW, for example, there is an $11,000 fine or a 6 months prison sentence. Other states are similar, although in the Northern Territory it is only a $1,256 fine.
These are tough penalties, although we know that people can and do breach the self-isolation orders, all the same. Spot checks in NSW had already found 14 cases of people breaching the self-isolation orders, by 1st April. In late August, during Victoria’s second lockdown, it was revealed that 50 people had been fined for breaching self-isolation orders.
With people known to be breaching the self-isolation orders, it should be obvious that it is only a matter of time until a Patient 31 event occurs. Indeed, this has already happened, and more than once. On 28th August it was revealed to the Victorian hotel quarantine inquiry that a security guard had been ordered to self-isolate, but breached the order. Meanwhile, in NSW, it was revealed that a hotel quarantine security guard in NSW was fine twice for breaching self-isolation orders.
This same critical mistake was made in Iceland. Two French tourists arrived for a holiday and tested positive for the virus. They were ordered to self-isolate but instead went out to clubs. This caused Iceland’s second wave.
The Asian nations again do this differently. In South Korea, the elderly and those with serious illness go straight to hospitals. The moderately sick are sent to quarantine camps, where they are monitored by the surveillance. And then only the asymptomatics and contacts of recently diagnosed cases are asked to self-isolate at home, and use separate bathrooms, dishes, and towels from their cohabitants. In Hong Kong, people are given wristbands and they are also monitored.
We must do this too.
Recommendation 3: Self-isolation orders should be monitored by surveillance.
Part II - Treatment and prophylaxis
Another key departure from the Western and Asian responses epidemics and pandemics is the attitude towards therapeutics and prophylaxis. On the 13th February, based on the best advice of Chinese doctors, the Korean COVID-19 Central Clinical Task Force agreed on early treatment principles for patients with COVID-19. They decided: if patients were old or had underlying conditions with serious symptoms, physicians were to consider an antiviral treatment. Treatments were based on the HIV medication Kaletra, as well as the anti-Malarials Chloroquine and Hydroxychloroquine.
The same is true for most of Asia.
Australia and many other western nations including the United States have onerous approvals processes for new therapeutics. Cynics observe that these processes seem to favour the big drug companies and their patented new drugs. Bringing a new therapeutic to market typically involves succeeding in a large scale Phase III clinical trial. This usually takes years.
The ongoing controversy surrounding the anti-Malarial drug Hydroxychloroquine is instructive. At the time of writing, at least 126 studies have looked at the efficacy of Hydroxychloroquine, making it one of the most well-studied drugs in history. A recent meta-analysis published in the journal Clinical Microbiology and Infection found that Hydroxychloroquine lowers mortality risk by a statistically significant 21%.
This is not a large effect, but Hydroxychloroquine has a number of other advantages: its use over 65 years means its safety profile and associated risks are well-understood. It is usually well-tolerated with few, if any, side effects. It is a cheap drug that is administered orally. And many nations including Australia have already stockpiled it.
But Hydroxychloroquine is not only candidate for drug repurposing. To name a few others that also show promise, there is the anti-parasitic Ivermectin, which is backed by credible Australian research; over-the-counter cough suppressants Ambroxol and Bromhexine; and the SSRI anti-depressant Fluvoxamine (Luvox).
Recommendation 4: TGA should make promising, safe medicines available as early antiviral treatment for at-risk populations.
Related to therapeutics are prophylactics. India in particular has deployed chemoprophylaxis to at-risk health care workers, notably Hydroxychloroquine and lately Ivermectin, since early in the pandemic. As with vaccines, proving efficacy is large-scale Phase III trials is extremely difficult, but there is enough evidence of efficacy in a few to be hopeful and to make these drugs available to at-risk health care workers and others who wish to use them.
Recommendation 5: TGA should also make promising, safe drugs available as chemoprophylaxis in at-risk populations.
Vitamin D supplementation
The Covid-19 second wave in the northern hemisphere has made it clear that this virus is highly seasonal. But why are viruses seasonal anyway? While this is not fully understood, it is generally assumed that virus seasonality is caused by changes in sunlight, temperature and humidity.
In this section, however, I show that emerging evidence suggests that sunlight and its impact on vitamin D levels could be the main driver of seasonality. Vitamin D supplementation, therefore, could have a large effect on slowing Covid-19 transmission.
Hope-Simpon’s seasonal stimulus
The idea that seasonality might be linked to sunlight was first proposed by British epidemiologist Edgar Hope-Simpson, in 1981. Back then, his hypothesis was not taken seriously, although, in the years since, evidence has mounted that vitamin D, a hormone that is created when sunlight strikes the skin, plays a key role in our immune system and protects us from viral infections.
A picture is sometimes worth a thousand words, and I think this figure from a recent paper from UC Louvain, Belgium, is such a picture.
The figure shows that the autumnal Covid-19 surge in the northern hemisphere is highly correlated to latitude, and not correlated at all to temperature. This would tend to suggest that sunlight is indeed the driver of seasonality, for Covid-19 anyway.
But Hope-Simpson’s idea was well-supported prior to the pandemic. A literature review in 2006 (Cannell et al, Epidemiol Infect) had concluded that vitamin D deficiency may indeed be the driver of seasonality. Then in 2010, a Japanese randomised control trial demonstrated that vitamin D3 supplements given to children reduced the risk of influenza A infection, by over 75%.
Turning how to Covid-19-specific observations, a randomised control trial from Spain recently showed that early calcifediol (25-hydroxyvitamin D) treatment in hospitalised Covid-19 patients reduced risk of progression to intensive care by 94%. Also notable was a huge observational study in the US that studied 191,779 patients and found Covid-19 infection to be strongly correlated to circulating 25-hydroxyvitamin D levels. Another study from Spain found that 80% of 216 hospitalised Covid-19 patients compared to 197 matched from the general population had vitamin D deficiency.
Calls from experts
In recent weeks, experts have increasingly called for government interventions in the northern hemisphere to address vitamin D deficiency as a means of slowing tranmission. In the UK, where 40% are estimated to suffer vitamin D deficiency in winter, a group of experts are calling for vitamin D fortification in foods to protect against vitamin D deficiency. (Note that Australia already fortifies some foods to a minimum of 220 IU vitamin D per 100g food in some milks, yoghurts, and table confections.) In a letter by two digestive endoscopists to the editors of a journal, it was noted that the north-south gradient in Covid-19 mortality appeared correlated to vitamin D deficiency. They recommended vitamin D prophylaxis without overdosing. In Scotland, a decision has already been made to give free vitamin D supplements to the Scots at most risk of severe Covid-19 disease. And in the Netherlands, media campaigning on the vitamin D issue has led to many in the population taking supplements.
Vitamin D deficiency in Australia
It may come as a surprise to learn that Australia, despite its abundant sunlight, has problems with vitamin D deficiency. This is actually often the case in sunny countries like Australia, as people avoid sunlight due to concerns about sunburn and skin cancer. In fact, almost one quarter (23.5%) of Australians are estimated to have vitamin D deficiency, a figure that rises to 36% in the winter. And, unsurprisingly, the southeastern states of Victoria and Tasmania have the highest rates, 49% and 43% respectively. And Victoria, of course, is where the seasonal second wave hit Australia hardest.
In summary, it may be premature to conclude that vitamin D status is the main driver of Covid-19’s seasonality. But there is more than enough evidence to justify a government campaign to educate people on getting adequate sunlight and considering vitamin D supplementation. There is probably enough evidence to justify a policy similar to Scotland’s, of giving free supplements to the at-risk group.
Recommendation 6: The government should campaign to eradicate vitamin D deficiency in Australia.
Part III - Containment fails
So far, I have looked at measures aimed at preventing and containing future Covid-19 outbreaks. But we must be realistic and accept that, no matter how good are our defences, we may lose control of the epidemic. South Korea, after all, lost control in Daegu. What should we do then?
In this section, I argue that we must give up on the full national lockdown, implement lighter, targeted restrictions aimed at preventing superspreading events, and do what we can to protect those most at risk of severe disease.
Overdispersion and superspreading
South Korea’s Patient 31 caused a superspreading event that exposed over 1,000 individuals and may have infected hundreds. In another superspreading event, 175 executives from around the world came to a Biogen conference in Boston. Around 100 of these went home infected by SARS-CoV-2. In yet another example, 260 children and teenagers were infected at a sleepaway camp in Georgia in July.
The average number of infections, however, that someone infected by SARS-CoV-2 causes - the value known as R0, the basic reproduction number – is thought to be about 2.5.
It turns out that SARS-CoV-2 is highly overdispersed. Overdispersion means a small number of infected individuals cause most of the transmission. It is sometimes expressed as the fraction of infected individuals who cause 80% of transmission. For SARS-CoV-2, the value may be 10% or lower. So, while on average, a group of 10 infected individuals might cause 25 secondary infections, just one of those might infect 20 people, while the remaining nine combine to infect only five.
This is represented in the following diagram:
An important consequence follows from this immediately. If we could remove the 1 person who infects 20 others, the reproduction number is immediately reduced to about 0.5.
Lloyd-smith and backwards contact tracing
The discovery of overdispersion in epidemiology was made by a young post-doc James Lloyd-smith and co-author Wayne Getz in a seminal 2005 paper on the overdispersion of SARS, entitled, Superspreading and the effect of individual variation on disease emergence. Since the time of Typhoid Mary, it had long been known that some people spread disease more than others. By analysing contact tracing data, however, Lloyd-smith and colleagues were for the first time able to model and quantify the effect of individual variations in infectiousness.
Llyod-smith’s paper also mathematically compared two different types of infection control interventions: population-wide controls that aim to reduce the infectiousness of everyone in the population (think of lockdowns), compared to interventions based on contact tracing to trace and isolate superspreaders. “If highly infectious individuals can be identified predictively … then the efficiency of control could be greatly increased … Focusing half of all control effort on the most infectious 20% of cases is up to threefold more effective than random control”.
In the case of a highly overdispersed virus like SARS, whose overdispersion is similar to its cousin SARS-CoV-2, they found that contact tracing had effect of lowering the herd immunity threshold by about half. A huge effect.
This method of contact tracing - now known as backwards contact tracing - became standard in Asia. It is part of the reason South Korea was able to contain the Daegu outbreak.
What is a superspreader
Whether or not someone is a superspreader will depend on the pathogen and the individual’s biology as well as their environment and behaviour. Some infected individuals might shed more virus than others, perhaps because of differences in their immune system. Some people do not feel sick and may continue their daily routines, inadvertently infecting more people. Alternatively, people with weaker immune systems that allow very high amounts of virus replication may be very good at transmitting even if they reduce their contacts with others. Individuals who have more symptoms – for example, coughing or sneezing more – can also be better at spreading the virus to new human hosts.
A person’s behaviours, travel patterns and degree of contact with others can also contribute to superspreading. An infected shopkeeper might come in contact with a large number of people and goods each day. An international business traveler may crisscross the globe in a short period of time. A sick health care worker might come in contact with large numbers of people who are especially susceptible, given the presence of other underlying illnesses.
Gabriela Gomes and individual susceptibility
Another consequence of overdispersion has been pointed out recently by M. Gabriela M. Gomes, an infectious diseases modeller at the University of Strathclyde. “More susceptible and more connected individuals,” her team wrote, “have a higher propensity to be infected and thus are likely to become immune earlier. Due to this selective immunization by natural infection, heterogeneous populations require less infections to cross their herd immunity threshold.” According to her model, the herd immunity threshold is closer to 10-20%, rather than the usual 60% or so.
Gomes’ work has been controversial, as it has been popularised by proponents of Sweden’s so-called herd immunity model. All the same, her key insight is not disputed, namely, that even in the absence of interventions, a highly-overdispersed viral pandemic will tend to burn out quicker and be easier to suppress than it otherwise would, due to the tendency of superspreaders to be infected and naturally immunised early.
Places of superspreading
Superspreading events require an infectious superspreader, but they also require an environment that is favourable to aerosol transmission. In the case of South Korea’s Patient 31, the superspreading during a church service in a huge hall where the infected woman sang hymns loudly.
By now, it is well understood that superspreading is associated with shared rooms, where significant time is spent in the environment, where there may be crowding or overcrowding, poor ventilation, and where people talk, and especially where they sing or shout. So, many outbreaks have occurred in choirs, restaurants, bars, meat packing plants.
Contact tracers already use a method of tracing known as backwards tracing in some states and other states are in the process of adopting the method. This method aims to trace cases specifically backwards to find the superspreader that infected them, remembering there there is a 70% chance that an infected individual has not themselves infected anyone else.
The other way to stop superspreading is to target the types of venues that are associated with high transmission. These venues tend to be almost always indoors, are places where people gather for a long time, are overcrowded, poorly ventilated, and where people may talk loudly, or sing, or shout. So, many outbreaks have occurred in choirs, bars, meat factories, etc.
Iceland is a good example of a country that has adopted targeted restrictions instead of lockdowns. After Iceland lost control of the epidemic in September, they have closed gyms, bars, and nightclubs, and have limited gatherings to 20 people. Prior to this they had already mandated face masks on public transport. Their epidemic curve at the time of writing appears to have peaked, implying that these measures did lower R below 1.
Against full lockdowns
Expert opinion in January
On 23rd January, 2020, the central government of China quarantined the entire city of Wuhan, as well as other cities in Hubei, in an effort to stop Covid-19 spreading to other cities. This was the first time in modern history that a huge city of 11 million people had been quarantined; it became known in the media as the “Wuhan lockdown” (武汉封城).
The attitude of the WHO towards the move was ambiguous. “The lockdown of 11 million people is unprecedented in public health history,” said Gauden Galea, the WHO’s representative in Beijing. The WHO continued to not recommend restrictions on travel or trade. WHO chief Tedros Adhanom Ghebreyesus said, “China has taken measures it believes is appropriate to contain the spread of coronavirus in Wuhan and other cities. We hope they will be both effective and short in their duration”.
Outside of the WHO, other epidemiologists also expressed great skepticism on the lockdown. Wired Magazine wrote that, “Travel bans and quarantines are a centuries-old answer to the spread of disease,” and also “exactly what the World Health Organization asked everyone not to do”. The WHO had advised against travel and trade restrictions, because “almost no one in the game thinks that works”. James Hamblin, M.D. (Yale School of Public Health), writing in the Atlantic, called the lockdown a “radical experiment”, and despite “an especially high mortality rate”, “[t]he moderately virulent nature of the pathogen seems at odds with the fact that the largest quarantine in human history is now taking place in an authoritarian state”. Hamblin’s article is worth reading in full as a statement of the prevailing attitudes of epidemiologists at the time.
Notice here crucially that this professional consensus against travel bans and trade restrictions came at a time when the case fatality rate of the virus was considered to be much higher at 2-4% and the virus was considered to be more infectious.
Expert opinion in October
Of course we know so much more about the virus today, in October, 2020, than we did in January. James Hamblin, for instance, has apparently changed his mind, and is a vocal proponent of lockdowns. Less clear is what the evidence is that caused him to change his mind. Does it mean that everyone changed their mind?
I don’t think so. On October 4th, in Great Barrington, Massachussets, three distinguished epidemilogists from Oxford, Stanford and Harvard, wrote a document opposing lockdowns, known as the Great Barrington Declaration. At the time of writing, it had gathered the signatures of 11,600 “public health experts”, 33,000 doctors, and 604,000 citizens.
A few weeks later, on October 16th, an opposing declaration was published in The Lancet, called the Jon Snow Memorandum. This document opposed the so-called “herd immunity” strategy mentioned in the Great Barrington Declaration, and instead called for containment at any cost. At this time, it is signed by only 6,400 academics. Many of these academics appear to be public health experts, although it appears open to anyone with a Ph.D. I easily found computer scientists, climate scientists etc, in the list.
Let me be very clear about this: I do not support either of these declarations. All the same, the vastly higher number of public health scientists and doctors signing onto the Great Barrington Declaration suggests that support for lockdowns and quarantines continues to be a minority view. Whereas many scientists prefer to keep out of public controversies.
Empirical evidence for lockdowns
Ben Cowling, professor of infectious disease epidemiology at the University of Hong Kong, noted in March that China had paired its lockdown with rigorous containment measures, including widespread testing, isolation of those infected, and quarantining of contacts who might be infected. This was not, however, done in Australia, or anywhere else in the West.
It turns out that the baseline risk of being infected in your home by a housemate or family member is quite low. A large Danish household found that there is only a 17% chance of being infected, all other things being equal. The CEBM says it is 16%.
In a Chinese meta-analysis studying secondary attack rates in households, the authors conclude:
All these challenge the value of home isolation for COVID-19 patients, as it may put household members at high risk of infection, propagating the disease. When the hospital isolation of all cases becomes unfeasible, other sheltering facilities, such as the Fangcang Shelter Hospital used in Wuhan, China, might be a better option.2
Of course, stay-at-home orders during lockdown are likely to increase this risk of household transmission in two ways:
- More time spent indoors with people increases the risk of transmission.
- More time spent indoors weakens the immune system, especially causing vitamin D deficiency.
But this is not merely theoretical. Observational evidence exists from huge seroprevalence studies. A study of 60,000 workers in Spain found that 6.3% of confined workers had been infected compared to only 5.3% of essential workers. These include hospital workers and police and others who should have been a far greater risk of infection - if lockdowns worked as advertised! A bit later, similar results were found in Italy.
Herd immunity in New York City
Serology studies in New York City had already found by May, 23% antibody prevalence. Modeling studies out of Isreal and the United States had gone further and predicted that New York City had already reached herd immunity. And at the time of writing, New York City is one of a handful of regions in the northern hemisphere that is not seeing a ferocious second seasonal wave, as shown in the following chart:
But what about Governor Cuomo’s strict lockdown? On March 20th, Cuomo had, so he said, in his “New York City on PAUSE” executive order. Non-essential workers were confined to their houses. Meanwhile, non-essential businesses were ordered to close.
But if New York City reached herd immunity, despite Cuomo’s strict lockdown, it would appear that herd immunity, not the lockdown, is what actually ended the epidemic.
Government and media communication around the world have always presented lockdowns as the primary cause of suppression of our epidemics. On Italy’s lockdown, for example, Business Insider’s headline read: “Italy and Spain have started to slow the coronavirus with total lockdowns — but France is yet to feel the effects. Here’s how long it’s taking to work”. Or CNN: “Italian doctors hope for a sign the coronavirus lockdown is working, because there’s no plan B”.
Indeed, early in the pandemic, it was widely communicated that epidemics follow an exponential curve. This is really not true however. Epidemics in fact follow a well-known curve in epidemiology called the epidemic curve or epi curve for short. The following chart shows the epidemic curve of Swine Flu in Australia in 2009. Remember that the Australian government gave up early on containing the Swine Flu epidemic, so this helps us to guess what our epidemics would have looked like without any intervention:
We must remember, therefore, that the assumption that our lockdowns flattened the epidemic curves is not proven. We don’t know, for example, if the lockdowns did it, or if widespread testing, tracing, and isolating infections did it. Certainly, South Korea flattened their Daegu curve using only test, trace, isolate, and Australia was a world leader in ttests per day.
A number of states did use lockdowns very early and if used early, there are some common sense reasons to believe that the lockdowns are effective in suppressing an emerging outbreak. Some states, including Queensland and New Zealand, locked down when there were hardly any cases at all. The reason for these early lockdowns was said to be to give public health autorities time to prepare the health infrastructure during the second wave.
But on August 18th, the government of New Zealand ordered a second lockdown after the country recorded a tiny 13 cases after months of no transmission. But lockdowns used as early as New Zealand did in August suggest that that country has no intention to ever use test, trace, isolate to suppress an epidemic. Lockdown would always be the preferred method of containment.
Notice that if NSW had initiated lockdowns each time it recorded 13 cases, the state of NSW would have been in lockdown during 2020 as much as Victoria. The following chart shows cases-per-day in NSW over 6 months (source: covid19data.com.au):
The longer-term economic consequences of New Zealand’s experiments with rolling lockdowns are yet to be felt. The government has so far spent 20% of GDP, its JobSeeker-equivalent has flowed to almost 60% of workers.
We will know in years to come how much of this burden has been shifted onto New Zealand’s children and grandchildren.
Lockdowns have no place
The uncomfortable but inescapable conclusion is that lockdowns have no place in modern pandemic management. We have to choose one:
- Will we be South Korea and use test, trace, isolate contain 7,000 cases in the Daegu outbreak?
- Or will we be New Zealand and shut the entire country to contain 13 cases in the Christchurch Hotel outbreak?
WHO pandemic preparedness documents do not, and have never, recommended lockdowns. The following screenshot shows the WHO’s 2005 pandemic preparedness checklist section on social distancing and quarantine. Lockdowns are not mentioned in any documents:
The WHO coronavirus envoy Dr. David Nabarro has recently urged world leaders not to lockdowns as a primary control method:
The only time we believe a lockdown in justified is to buy you time to reorganise, regroup, rebalance your resources; protect your health workers who are exhausted,” Dr Nabarro said. But by and large, we’d rather not do it.
Similarly, countries like Germany, Norway, Switzerland and others already conceded that their March lockdowns were overreactions.
The case of the New Zealand lockdowns is interesting. According to WHO advice, updated in this pandemic, their first lockdown may have been justifiable, but only if they had used that time to build containment infrastructure. But the fact of a second lockdown at only 13 cases is clearly not justifiable in terms of any pandemic best practice.
Lockdowns do not work. Political leaders must have the courage to hold their nerves, even if cases appear to be increasing exponentially.
Recommendation 7: Rewrite stage 3 restrictions in favour of targeted restrictions.
Recommendation 8: Ban lockdowns.
Part IV - Focused protection
On 17th June, 2009, at the peak of Australia’s Swine Flu epidemic, the Rudd Labor government introduced a new phase in its pandemic response called PROTECT. Containment having failed, the government moved to focus on protecting people with high risk of complications from the disease. In the case of Swine Flu, this included pregnant women, a subpopulation that was particularly susceptible to Swine Flu. Testing at airports was discontinued. And the national stockpile of antiviral drugs was no longer made available to people with the flu unless there were more than mild symptoms or a high risk of dying.
Sweden’s policies have been significantly misrepresented in the media and other public discussions. In reality, Sweden’s policies were a suppression strategy quite similar to what was adopted by the Morrison government in Australia. Indeed, Sweden’s policies essentially implement the flatten-the-curve memes that Australians shared on social media in February and early March, prior to Austalia’s lockdown. For example:
Often missed in these discussions is that Sweden had invested in a huge hospital capacity and therefore could afford to allow the virus wave to spread faster without risking hospital capacity.
Sweden’s policies of focused protection were supposed to protect especially care homes. Sweden’s government however claims that care homes did not implement the guidelines and this led to huge mortality in care homes. The same happened in Victoria, of course. The Royal Commission into aged care recently found that the Morrison government had failed to adequately prepare aged care homes for Covid-19.
The following chart shows excess mortality over a number of years in Sweden, the huge first wave of Covid visible at the end (source: EUROMOMO):
Another factor that also no doubt contributed to Sweden’s high death rate in the first wave is also that excess mortality had been well below average in Sweden over the previous two years, leading to a build-up in people who were at high-risk of death from respiratory infection.
Scenarios of focused protection
I am not a proponent of Sweden’s herd immunity or focused protection. But it is easy to imagine scenarios where we might choose to adopt this:
- The virus could mutate to become less lethal.
- Outbreaks could occur even after vaccination.
- Repeated outbreaks could bring us closer to herd immunity than we are now. Focused protection could be the best way to finish off that process.
As I write this, I am aware of huge danger, in a way I never have been before, both to public health and to Australia economy and future, posed by Covid-19. I have low confidence in the processes that are in place for containing this virus, painfully aware of how easy it is for a single mistake to lead to a superspreading event, and knowledge that Labor states in particular believe in elimination rather than suppression and have response plans based on lockdowns.
In this piece, I have argued for 8 changes to Australia’s Covid-19 response:
- Recommendation 1: Australian governments must strongly incentivise or compel people suspected of infection to be tested.
- Recommendation 2: Australia must deploy surveillance to assist contract tracers and enforce quarantine.
- Recommendation 3: Self-isolation orders should be monitored by surveillance.
- Recommendation 4: TGA should make promising, safe medicines available as early antiviral treatment for at-risk populations.
- Recommendation 5: TGA should also make promising, safe drugs available as chemoprophylaxis in at-risk populations.
- Recommendation 6: The government should campaign on educating immune system health and vitamin D supplementation.
- Recommendation 7: Rewrite stage 3 restrictions in favour of targeted restrictions.
- Recommendation 8: Ban lockdowns.