Covid-19 kills far more than 0.1% of the people who catch it

14th Sep 2020

Claim

Covid-19 kills no more than 0.1% of the people who catch it

Conclusion

This is incorrect. The real death rate is probably about five or ten times higher.

“Lockdown happened because we were told that Covid could kill one percent. But Covid was never going to kill more than about 0.1 percent – max.”

RT.com, 6 September 2020

An opinion article on RT.com, a Russian state news site in English, claims that no more than about 0.1% of people who catch Covid-19 in the UK will die from it. Several readers have asked us to look into this.

The article argues that lockdown policies were based on a mistake, confusing the “Case Fatality Rate” (the proportion of people known to have the disease who die from it) with the “Infection Fatality Rate” (the proportion of all infected people who die from it, including people who may never have shown symptoms or been identified as cases). It claims that this made the novel coronavirus seem ten times deadlier than it actually is, suggesting it would kill around 1% of people it infected, rather than the “true” figure of 0.1%.

It is difficult to know exactly how many people have caught Covid-19, or died from it, which are the numbers we would need in order to calculate a precise death rate. Nevertheless, even a rough calculation shows that the true figure in the UK is much higher than 0.1%.

How deadly is Covid-19?

The chances of dying if you catch Covid-19 vary a great deal according to your age and health, with older people and those with underlying conditions facing a much higher risk. This means that different countries are likely to have different death rates, depending partly on the age and health of their populations, the quality of their healthcare, and other factors that we may not understand.

However, we do know that up to 30 June, 44,440 deaths occurred in England alone with Covid-19 listed by a doctor as the underlying cause of death, meaning it was “the disease or injury which initiated the train of morbid events leading directly to death”. (The author of the RT.com piece gives a figure of “around 40,000 Covid deaths” for the UK as a whole, which is slightly too low, but roughly the right level.)

Even 44,440 deaths might underestimate the true number in England, if some deaths were registered late, or if there were other deaths from Covid-19 that could not be confirmed by a doctor, perhaps when tests were less readily available. (We also have data for the UK’s other nations, but focusing on England will make it simpler to calculate a death rate.)

So, if 44,440 people had died of Covid-19 in England up to the end of June, and if they accounted for 0.1% of all infections—which is one in a thousand—then that would mean 44.4 million people must have caught the disease by early June, or thereabouts. (It takes an average of around two or three weeks for someone to die of Covid-19, so people who caught it later will generally not appear in June’s death figures.) This would mean it had infected 79% of the total population of England.

Had 79% of England caught Covid by early June?

It is hard to say precisely how many people have caught the disease, because many of them may never report symptoms or get tested. Even so, no reasonable estimate for England would get close to 79%.

The most direct method involves testing blood from a sample of the population, to see how many people have Covid-19 antibodies, which are a sign that they were infected in the past.

Using this approach, the REACT-2 survey estimated that about 6% of the adult community in England had antibodies when they were tested between 20 June and 13 July.  (And an antibody survey in Spain found roughly the same level.)

If only 6% of England had caught Covid by this point, that would imply a death rate of roughly 1.3%. However, an antibody survey is not a perfect way to count how many people were infected, because a negative antibody test might not prove that someone has not had Covid-19.

Some people infected in early June might not have made antibodies quickly enough to test positive. Others may have made antibodies, then lost them again by the time they were tested. It is also possible that some infected people did not produce enough antibodies for the tests to detect. Scientists are still trying to understand these aspects of the disease.

Silent Covid can’t be everywhere

But even if only a fraction of infected people actually tested positive for antibodies, it still does not seem possible that 79% of England had been infected, because of the regional variation that the REACT-2 survey also revealed.

According to the survey data, about 13% of adults in London had antibodies, and no other region had more than about 6.6% (the North West). Assuming that people’s antibody response to Covid-19 does not vary according to where they live, this means that the share of people who caught it in London is at least twice as large as it is in any other English region.

Therefore, even if 100% of the population of London had been infected with Covid-19 by the early summer—almost all without noticing—this does not change the fact that only about half as many in the North West were. And only about a quarter in the South West, which was the least affected region.

Altogether, if we assume that the whole of London somehow did catch Covid-19 (and that children everywhere were infected at about the same rate as adults) then we can calculate the number of people who would have caught it in every other region, and add up the total.

This comes to about 26 million people, which is a theoretical (but very far-fetched) maximum of 46% of England infected by the time of the REACT-2 survey. If this were true, it would mean that 44,440 people died out of 26 million infected, giving a death rate of about 0.17%. So even the most extreme and unrealistic assumptions about the level of invisible infection can’t get the death rate down to the claimed 0.1%.

In reality, of course, a 0.17% death rate is also implausibly low, because there is no evidence that all of London was infected. Indeed there is strong evidence that they weren’t, since reinfection seems to be rare, and people in London are still catching Covid-19. Even if half of the population of London had caught Covid-19 by the summer, that would suggest an English death rate of about 0.34%.

So what is the real death rate?

Many scientists have attempted to estimate the proportion of people who die after catching Covid-19. So far they tend to produce rates between 0.5% and 1%, which would make it between five and ten times more deadly than this article claimed.

By looking at English data, it is clear that the death rate in this country must be much higher than 0.1%. The researchers who conducted the REACT-2 survey produced a more detailed analysis, which estimated an overall death rate that is nine times higher, at about 0.9%. The death rate from the disease may be lower now, and in the future, than it has been previously, because of improvements in medical treatment.