THERE are two important things we need to know about the coronavirus: how fast it spreads, and how many of those it infects will be killed. Thanks to recent data from New York City we now know that peak mortality was reached three weeks faster than it was in the Spanish flu epidemic of 1918, but also that the virus is substantially less lethal, especially among young people but also among the old.
On 25 February, the World Health Organisation (WHO) told its member states that the virus produced few symptomless infections. This implied that it was very dangerous, as most people who were infected would become ill and therefore the death toll would be high. This was the beginning of global panic over Covid-19.
To measure the lethality, we need to work out how many people have been infected. The best way of doing this is through so-called ‘serosurveys’, which test for coronavirus antibodies in blood samples, preferably from a representative sample of the total population. This should give you a direct assessment of the number who have been infected so far, and from this it is straightforward to calculate the infection fatality rate – the key metric which tells you the proportion of those infected who have died. This is your measure of how dangerous the virus is.
It’s fair to say that the world has been relatively slow to perform serosurveys in the wake of the coronavirus pandemic, but the first major one appeared last week, from Santa Clara County, California. The results of the study, performed by a team from Stanford University, were startling. Although they estimated that 50,000-80,000 people in the county had been infected, there had only been around 1,000 confirmed cases. In other words, for every person with symptoms, 50 to 80 were shrugging the infection off, exhibiting mild symptoms or none at all. There had been only 100 deaths from the virus in the area, so this implied that only one or two people were dying per thousand infected. This was strong evidence that the virus was not particularly dangerous.
Although there were some powerful critiques of the Stanford team’s statistical treatment of their results, another serosurvey, this time from Los Angeles County, was published a few days later, and reached similar conclusions. Moreover, there are many other supporting studies, either based on serosurveys or PCR (polymerase chain reaction) analysis of nasal swabs. I hope readers will forgive me if I present a long list here; it is important to show just how much evidence there is.
· A PCR survey of staff at Linköping Hospital in Sweden found that half were infected, but almost none had any symptoms.
· 37 per cent of residents at a shelter for the homeless in Boston, Massachusetts, were infected according to a PCR-survey, but none had any symptoms.
· Of 200 people tested in Chelsea, Massachusetts, a third tested positive in a serosurvey. All were apparently healthy.
· My own review of serosurvey data taken from blood donors in Copenhagen suggested that there were 27 asymptomatic infections for each confirmed case.
· A serosurvey of Finnish blood samples told a similar story.
· Modelling of outpatient survey data in the USA suggested that ‘SARS-CoV-2 had potentially infected millions in the US several weeks ago’.
· A New York obstetric clinic tested all women admitted: 13 per cent PCR-tested positive, but none had any symptoms.
· Another PCR survey of obstetric patients, this time from the Karolinska Hospital in Sweden, found 7 per cent were positive, but without symptoms.
· Rapid tests of blood samples taken from staff working in Swedish care homes found that a third had antibodies.
· A serosurvey of staff, parents and pupils in a French high school found that 25 per cent were positive for antibodies three weeks ago.
Just three days ago, a study from Stockholm suggested that that up to 16 per cent of blood donors have the antibodies; according to the Swedish newspaper Aftonbladet, modelling of the spread of the virus in the city suggests that as many as 25 per cent of residents may have the antibodies. The results of this study are being checked but they will not change materially.
This data all seems to be telling the same story: that we are looking at a virus that spreads quickly but that is not particularly lethal, perhaps only as dangerous as seasonal flu.
The apparently rapid spread of the virus has one huge advantage. It would mean that herd immunity is much closer at hand than previously thought in large cities such as NYC. In fact, one recent report suggests that some places are already nearly there. In the town of Ortisei, in the Italian Dolomites, 50 per cent of those tested had antibodies to the virus. It is generally assumed that there is full herd immunity above 60 per cent.
These results are all suggesting that one, two or three people will die per thousand infected. Where this leaves coronavirus hotspots like New York City and London is still quite hard to assess though. Estimating the number of people infected from the fatalities figure should be a matter of simple arithmetic, but it is complicated by doubt over exactly how many have died from the virus (as opposed to died with it). The new serosurvey data from New York State suggests that at worst eight or nine people could die per thousand infected, if you include both ‘confirmed’ and ‘probable’ fatalities, which may overstate things. Nevertheless, there are good (if rather technical) reasons to believe that the true figure might be substantially lower. And even eight deaths per thousand is much, much lower than the WHO’s original estimate of 20-40 deaths per thousand.
This leaves the WHO with a problem, having told the world that the virus was very dangerous, contrary what we know about nonsymptomatic SARS-CoV-2 infections. Unsurprisingly, they are now in self-defence mode. In an article three days ago in the Guardian, they claimed that only 2-3 per cent of the global population had the antibodies, suggesting that any attempts to reach herd immunity were doomed to failure. This is certainly not true in virus hotspots such as London and New York. Disturbingly, the WHO technical lead on Covid-19 seemed to be trying to avoid the big picture presented by all the new data that has become available, instead trying to pick holes in some of the individual studies.
The world’s economy faces disaster in part because of the WHO’s erroneous claim that most people infected by the virus would develop symptoms. The implication that the world faced another Spanish flu was baseless. We will be living with the consequences for many years to come. It is a very sad moment in the history of the World Health Organisation.