These were the predictions of the Swedish health authorities based on their mathematical models. Yet official survey results released this week revealed that only 7.3 per cent of Stockholmers had developed antibodies in late April. Tom Britton, a maths professor who helped develop the government model, said: ‘It means either the calculations made by the [public health] agency and myself are quite wrong, which is possible, but if that’s the case it’s surprising they are so wrong. Or more people have been infected than developed antibodies.’
In mid-April Sweden’s state epidemiologist, Anders Tegnell, claimed the epidemic in the capital was slowing as herd immunity was being approached: ‘According to our modellers, we are starting to see so many immune people in the population in Stockholm that it is starting to have an effect on the spread of the infection.’ He predicted herd immunity to be achieved ‘some time in May’.
Now, though, critics are arguing that herd immunity is a long way off and will never come. Björn Olsen, a professor of infectious medicine at Uppsala University, said herd immunity was a ‘dangerous and unrealistic’ approach and that it ‘is a long way off, if we ever reach it’.
Tegnell has dug in, claiming that the 7.3 per cent relates to a test done around May 3 and hence with people who were infected three or four weeks earlier in early to mid-April. ‘We are somewhere around 20 per cent plus in Stockholm now,’ he said.
Looking at the death curve in Stockholm, though, it peaks on April 6, meaning infections would have peaked around 16 days earlier on March 21, so that early to mid-April is several weeks past the peak. This makes it all but impossible for antibody prevalence to have trebled since mid-April, when the epidemic was well into decline. It is also unclear why Tegnell suggests antibodies typically take 3-4 weeks to appear when it is more commonly stated they appear after 1-2 weeks, which would move the 7.3 per cent prevalence level to even later in the epidemic.
But how reliable is the antibody survey anyway? One mystery still to be explained is how it marries with an earlier antibody survey in Stockholm that found 10 per cent of the city had developed antibodies by mid-April, a result the researchers identified with the state of play in late March: ’10 per cent of the tested individuals were infected during or prior to late March.’
How can 10 per cent of Stockholmers have had antibodies at the end of March, then 7.3 per cent in mid-April? On the face of it this makes no sense and needs some explanation. Either way, 10 per cent or 7.3 per cent, both figures correspond to well post-peak, meaning Stockholm is unlikely to get above 20 per cent in the end.
Results from an antibody survey in Spain were also released last week, showing 5 per cent prevalence across the country, rising to 11.3 per cent in the worst affected region, Madrid. The president of the Spanish Epidemiological Society, Pere Godoy, was quick to point out the dire implications as he saw them for herd immunity: ‘Herd immunity is not going to be possible until there is a vaccine, unless we are prepared to accept an enormous human toll. The virus has proved to be very dangerous, and any policy that allows it to circulate comprises a huge risk.’
Estimates for the herd immunity threshold vary, but it is often assumed to be around two thirds of the population. ‘Around 65 per cent of the population should be immune if we want to control the pandemic by the sole means of immunity,’ a recent report from the Pasteur Institute in France said.
A model from the Liverpool School of Tropical Medicine last month challenged that assumption by showing how collective immunity could emerge at a much lower rate of 10-20 per cent once variations in susceptibility and interaction were properly taken into account. However, that model does not as yet appear to have made much impact on policy and strategy around the world, even though it seems to explain the data much better.
Last week the UK government announced results showing 17 per cent or one in six Londoners have antibodies, compared with an average of 5 per cent across the country. The date of this test was not given, but an earlier announcement indicated up to 10 per cent of Londoners carrying antibodies on April 6, a date which corresponds to around the peak of the infection curve in London two weeks earlier. This figure fits well with 17 per cent at a more recent date and also with the 20 per cent antibody rate found in New York City at the end of the epidemic there.
What are the lessons to take from this? I suggest two. First, more consideration needs to be given to the possibility that collective immunity emerges much lower than has been assumed. Anders Tegnell had earlier said that herd immunity was causing the epidemic to slow in Stockholm in mid-April. Now we know that the antibody prevalence at that point was 7.3 per cent. Does he still think collective immunity was causing it? If not, what was causing it? If so, what does that say about the herd immunity threshold? Why is no one asking these questions?
Second, everyone needs to stop making assumptions about this virus and start looking hard at the data that is emerging and fit their ideas to that rather than vice versa. If lockdowns are needed to prevent the spread, why has Sweden not seen tens of thousands of casualties? If social distancing is necessary, why is Belarus, where there is relatively little social distancing, not seeing tens of thousands of deaths? Why is there no consistency in how social distancing affects epidemiological curves around the world? Is it because the spread is largely concentrated in hospitals and care homes, where it is also more deadly? Why has it fallen to an investment firm, J P Morgan, to demonstrate that lifting lockdowns has almost everywhere been accompanied by a drop in R and ongoing decline in the epidemic, as the graphs below show?
These are not side questions, they’re central questions, and there is nowhere near enough interest in them right now, from anyone except a brave few, from journalists to scientists to policy-makers. Until that changes, we will never get a proper understanding of this virus and its real level of threat to us and what we should be doing to protect ourselves from it without completely trashing our economy, healthcare, children’s education and way of life.