Claims, Caution and Comparisons

Flawed comparisons make “an absolutely astonishing statistic”.

On 3rd June, Nicholas Watt (BBC Newsnight) claimed:

The UK now has more daily deaths from [COVID-19] than the rest of the EU put together.

Tens of thousands of social media users shared the Newsnight graph.

International comparisons are challenging. Countries count COVID-19 deaths in different ways and can change over time.

Claim and caution

On BBC Newsnight, Watt said:

The UK now has more daily deaths from COVID than the rest of the EU put together. There’s that 359 UK figure, against 314 for the 27 member states of the EU. Perhaps we could look at this way: the UK with a population of 66 million has more deaths than the EU with a population of 450 million.

The political editor followed those claims with caution:

It is important, Emily, to add a few words of caution. There’s a two-week lag between the UK and France and Italy that had the highest death rates in the EU.

You have to be very careful about comparing deaths between countries which obviously compile them in a different way.

That UK overnight figure (359): that’s for the deaths registered in that period, not deaths that actually took place in that period.

Once made, claims can become detached from caveats. One Twitter post received over 13,000 shares.

Dr Clarke: “Not even [10 Downing Street] can spin that fact away — it is utterly, irredeemably damning.” (Image: Twitter)

The challenges of international comparisons

Whilst the measure sounds simple, countries count COVID-19 deaths in different ways. There is no standardised measure.

Countries can change their counting methods too.

In the Netherlands, RIVM counts “overleden patienten”, or ‘deceased patients’. The measure is deaths in hospital where the patient has a positive test result.

In Belgium, Sciensano counts “both confirmed and suspected COVID-19 deaths” in all settings. Before 30th March, Sciensano only counted confirmed deaths.

Comprising four nations, the UK has changed its headline measure twice:

  • The England count used to come from NHS England. That figure was for hospital deaths. On 29th April, this measure expanded to confirmed deaths elsewhere.
  • From 1st June, Public Health England widened the definition of a positive test result. Now, it include results from commercial testing partners.

There are nuances within the four nations. Public Health Wales requires clinical suspicion that COVID-19 was a causative factor. Confirmed deaths are in hospitals and care homes.

As Watt notes, the date of report is not when the death happened.

The pain in Spain

On May 25th, el Ministerio de Sanidad changed its measure of COVID-19 deaths. That change led to a major reduction in total confirmed deaths, flat-lining afterwards.

Note the change in method. (Image: Our World in Data)

In daily reports, el Ministerio de Sanidad shows total deaths (“Fallecidos”). There is no reported figure for new daily deaths (“Nuevos diarios”).

The footnote highlights international in-comparability. (Image: MdS)

The increased ‘total’ is from [machine-translated]:

Only the cases in which the date of death is the day before the time of writing this report.

The Ministry states an intention to update complete total deaths each week.

Now, the Spanish ministry only reports a very low number of deaths. (Image: John Burn-Murdoch/Twitter)

Imagine NHS England had taken this approach for its reported deaths on 3rd June. Instead of reporting 179 COVID-19 deaths, there would only be 20.

In short

El Ministerio de Sanidad changed the daily COVID-19 measure for Spain. The daily count only includes recorded deaths which occurred in the previous day. That change means major under-counting, due to reporting lags.

Exercise caution in international comparisons. There are no standardised measures for COVID-19 deaths.

Some countries have changed how they count COVID-19 deaths. This example highlights the difficulties in international comparisons over time.

Aggregator sites may not record the change in definition. It makes drawing statistics from these sites more challenging. In addition, Worldometer has errors and is not transparent.

This blog looks at the use of statistics in Britain and beyond. It is written by RSS Statistical Ambassador and Chartered Statistician @anthonybmasters.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store