Is Trump right about America’s fatality rate? It’s all in the maths

The scale of America’s problem is hard to fathom but so too, sometimes, are the maths. Donald Trump was mocked this week for claiming on Twitter:

“The Mortality Rate for the China Virus in the U.S. is just about the LOWEST IN THE WORLD”. Unlike Trump’s relationship with the caps lock key, his claims about the virus might well be correct. It all depends on how you count.

If you look at the numbers at the time of writing, America has suffered 134,888 deaths with an estimated 3,159,671 cases, which works out at a mortality rate of 4.27%. Across the globe, deaths currently stand at 552,796, from an estimate of 12,196,724 cases. Divide the former by the latter and multiply by 100 and you get the frequently cited 4.5% mortality rate.

By that metric, Trump might be right. America’s number is lower. They’re beating the virus. All hail the Chief…

Except, it’s not quite as simple as that. The sum is problematic because it is a matter of dividing a known quantity (deaths) by an estimate (cases). That means the bigger you make your denominator, the lower your mortality rate.

This is why the UK looks so much worse when our death rate is compared against those of other countries. We’ve had 44,517 deaths but just 286,979 cases, which gives us a mortality rate of over 15%. The number is often used to condemn the UK government, but it’s unlikely to be an indicator of poor treatment rather than our lamentable testing.

This is the paradoxical truth that politicians would have been wise to understand early on in the outbreak: underestimating your problem makes your problem look worse. If you find more cases, your mortality rate begins to drop. Trump was right when he recently described testing as a “double-edged sword” and argued that “when you do testing to that extent you’re going to find more people”. Yet perversely he was wrong if he now wants to boast of a lower death rate.

This also explains what’s currently going on in the US. When the President proclaims a mortality rate coming down rapidly (which it is), it’s only because the number of cases in America are rising so quickly. Just yesterday, the nation reported 59,000 more cases. If the UK had reported an extra 59,000 cases yesterday, our mortality rate would have come down by a whole 3%.

The numbers speak to the problem of mathematically illiterate politicians using statistics to grab headlines, yet it’s perhaps understandable. As we’ve all come to learn in recent weeks about the R number, these figures can be quite specialist instruments for measuring outbreaks. We’ve all obsessed over “R” – the rate of transmission – for months, yet it becomes less meaningful as outbreaks get smaller.

If a country has no infections, then one person arrives and gives the virus to 10 people, the R number shoots up from 0 to 10. That “R” of 10 shouldn’t cause panic, though it most likely would.

Similar discrepancies have become obvious in the way that coronavirus data has been presented by newspapers. The Guardian is just one of many publications running an online map of the UK, shaded in different colours to represent the percentage increase in cases. Looking at it over the past months, you’d think some places in the North West have had far worse outbreaks than areas of the South East.

Dig a little deeper, however, and you’d noticed that “red” areas in the North West, indicating a 100% rise in cases, sometimes represent changes from 5 cases one week to 10 cases the next. It pales when compared to areas around London which are “green” because their cases have declined from the high hundreds to the mid-hundreds. Yet, objectively speaking, which area is doing better?

There is, then, much confusion around numbers, but there was some welcome clarity last week after the World Health Organisation held an online conference with leading scientists and concluded that the infection fatality rate (IFR) for Covid-19 is about 0.6%. That is a factor of 10 smaller than early estimates which pointed to a lethality of around 5%.

Using the new estimate to work backwards based on known deaths, we can get a rough back of the envelope approximation of the scale of the infection in each country: possibly over 7 million people in the UK, while in the US the number could well be over 22 million. If those numbers seem bigger than we might expect, it’s down to that oddity in the maths. The lesser the lethality, the higher the estimated cases based on a known number of deaths.

While the IFR is far more encouraging than many quoted over the past few months, it too shouldn’t breed complacency. As we’re seeing in the US, different numbers can be used to drive different political agendas. We have to look at the whole picture in which Covid-19 is an acutely dangerous virus, not because of its lethality but because of its transmissibility, and the subsequent burden that places on a health system.

While Trump might have been correct (in his bumbling way) about the lowering mortality rates, the significant numbers in the US are the staggering numbers of new cases being reported every day, as well as the potential for continued escalation.

The second most chilling words of last week came from Michael Osterholm, the head of the University of Minnesota’s Center for Infectious Disease Research and Policy, who said: “wherever there’s wood to burn, this fire’s going to burn, and right now we have a lot of susceptible people.”

The most chilling words? Those came from the White House, who have a new slogan to help America through the coming months: “Learn to live with it.”