Is it time for the Sweden haters to admit maybe, just maybe, the Swedes got it right?
To be sure, your writer is hardly immune from having to eat humble pie, having vociferously pushed a hard lockdown (albeit over a far shorter period than Victoria’s six-month incarceration) only to realise the data now tells us something very different.
Sweden’s chief epidemiologist Anders Tegnell advocated a “light touch” approach, limiting gatherings to less than 50 people and encouraging social distancing, but eschewing mass lockdowns or business closures. Critics of Tegnell’s approach quickly point to Sweden’s relatively high level of deaths as all the evidence needed to suggest that the policy was wrong. And Sweden certainly did have a higher number of fatalities than many countries (currently 5821).
However, Sweden’s death rate is lower than other countries which forced harsh lockdowns, including the UK, Italy, Spain and Peru. Moreover, Sweden’s death rate was inflated by huge errors in its handling of outbreaks in (largely privately-run) aged care facilities (sound familiar?)
Sweden is a powerful case study in herd immunity. Early in the pandemic, it was speculated that COVID-19 would spread wildly until around 60% of the population developed antibodies. Back in March, when northern Italy was being overwhelmed, the World Health Organisation estimated that 3.5% of people who contracted COVID-19 would die.
Applying those two assumptions to Australia’s population, upwards of 525,000 people could have died. This level of death would almost certainly justify locking down the country and slowing the spread.
What’s the actual herd immunity level?
Sweden (as well as other regions which didn’t lockdown hard, but bent the curve, like Florida and Arizona) seemed to show that COVID-19 started to naturally fade away at levels well below the 60% infection rate. Blood sampling taken in May indicated that around 7.3% of Swedes had COVID-19 antibodies.
The reason for this is unclear, but epidemiologists writing in The Conversation suggested that upwards of 20% of the population may already have protective T-cells, possibly due to exposure to other coronaviruses. Interestingly, in cruise ships (like the Diamond Princess) and military boats where COVID-19 spread, it seemed to infect only around 20% of people.
Sebastian Rushworth, an ER doctor who was on the frontline of the Swedish fight against COVID-19 explained:
“The only thing coming into the hospital [in March] was COVID. Practically everyone who was tested had COVID, regardless of their presenting symptoms. People came in with a nosebleed, they had COVID. They came in with a stomach pain, they had COVID.
“Then, after a few months, all the COVID patients disappeared. It is now four months since the start of the pandemic, and I haven’t seen a single COVID patient in over a month.”
“At the peak three months back, a hundred people were dying a day of COVID in Sweden, [a country with] a population of 10 million. We are now down to around five people dying per day in the whole country, and that number continues to drop.”
Rushworth’s anecdotal views are matched by data. Here’s Sweden’s COVID death chart:
The herd immunity level is important. To estimate how lethal COVID-19 will be we need to multiple the infection fatality rate (IFR) by the level to which the population will be infected.
So what about the infection fatality rate?
Here we have another problem: estimating the IFR for COVID-19 is difficult. While the case fatality rate (CFR) is easy to calculate, it’s also largely irrelevant, since the disease presents asymptomatically in the vast majority of cases and it’s very difficult to get a test in Australia without symptoms.
The IFR rate from COVID-19 is highly dependent on the population’s age and quality of healthcare, as well as testing levels. In Singapore, the CFR is 0.05%, so the IFR would be probably around 0.02%. The Diamond Princess (which had an much older age demographic) had an estimated CFR of 2.3% and an IFR of 1.2%.
Australia’s current CFR is 2.3% but our IFR is more likely to be closer to 0.2-0.3%. UNSW blood sampling data indicated that Australia had an IFR as low as 0.04% in April.
Let’s be conservative and assume Australia has an IFR of 0.30% — worse than Singapore and the UNSW blood testing results, but better than Diamond Princess.
Now we have our two key assumptions:
- An infection fatality rate of 0.30%; and
- A herd immunity level of around 10% of the population.
Using those data points, we get to an estimated fatality level of 7500-10,000 for Australia (with a “Sweden level” of lockdown). While that number may be lower than reality, given the terrible management of aged care facilities by the federal government, it’s a fraction of the earlier 525,000 estimate.
The great concern in the community is still that, left unchecked, they are at risk of dying from COVID-19. This thesis is largely based on the scenes in northern Italy and New York, coupled with an unwillingness to take another look at the actual data. Let’s also not forget, of those who die from COVID-19, almost all are above 70 with at least one precondition.
As Rushworth noted, “it is nonsensical to compare COVID to other major pandemics, like the 1918 pandemic that killed tens of millions … COVID will never even come close to this number. And yet many countries have shut down their entire economies, stopped children going to school and made large proportions of their populations unemployed in order to deal with this disease”.