It was reported on Tuesday that 100,000 people in the UK have died from Covid-19 since the start of the pandemic. But is this figure actually misleading the public?

In the UK, there are three main ways of counting the number of deaths attributable to Covid-19: deaths within 28 days of a positive test; deaths with Covid-19 on the death certificate; and excess deaths (the number in excess of the average over the last five years). As the BBC noted, these three methods of counting deaths give the figure of 100,000.

Because it’s not always easy to conclude whether someone died of or with Covid-19, counting the excess deaths has become the preferred measure. And in fact, it’s somewhat coincidental that the three measures currently give the same figure. During the first wave, excess deaths gave by far the largest total of the three (due to lack of testing); but since the beginning of the second wave it has been running lower than the other two.

Counting excess deaths is not a useful way to convey the current level of mortality in a country

In any case, it makes sense to focus on excess deaths. The first reason “100,000 excess deaths” is slightly misleading is that it is probably an overestimate. You calculate excess deaths by taking the number of deaths this year and subtracting the average of the last five years. However, this simple average doesn’t account for population ageing. Between 2015 and 2019, the number of people aged 70+ increased by 13 per cent, so – even without a pandemic – you’d expect more deaths in 2020 than the average of the last five years. However, when you use a trend – rather than a simple average – to calculate excess deaths, you still get a very large number. So, “100,000 excess deaths” isn’t a dramatic overestimate.

There’s a more significant reason why “100,000 excess deaths” is slightly misleading. Excess deaths measures the change in mortality. It tells you how many more deaths there were in a given time period than you were expecting based on previous years. In other words, it tells you how much mortality has changed, relative to where it was before. This is useful when trying to assess the pandemic’s overall impact; but it is less helpful when trying to convey the current level of mortality.

Last year may have seen the largest number of excess deaths in England since the 1940s – but it actually wasn’t the deadliest year in terms of mortality rates. In fact, the age-standardised mortality rate (which measures the current level of mortality) was higher in 2008, 2007, 2006 and every year before that. This means that 2020 is the deadliest year since 2008.

As far as I’m aware, nobody claims that excess deaths measures the level of mortality – the BBC has actually reported the age-standardised mortality rates on at least two separate occasions – but headlines like “2020 saw most excess deaths since World War Two” might lead the public to believe that 2020 was the deadliest year since 1940, which isn’t true.

The two following statements are both true, but they give a different sense of Covid-19’s lethality. First: “2020 saw more excess deaths than any year since 1940”. And second: “2020 had a higher age-standardised mortality rate than any year since 2008”.

None of this is meant to trivialise the pandemic or to suggest we shouldn’t be concerned about Covid-19 (which is much more deadly than the flu). My point is that focussing on the total number of deaths attributable to Covid-19 can be misleading, and news outlets would do well to report both this number and the age-standardised mortality rates going forward.