Explaining the UK’s official COVID-19 death statistics

Three key death statistics are published on the UK Government’s official COVID-19 dashboard: Daily deaths within 28-days of a positive COVID-19 test by date of death; the same, but by date reported; and daily deaths with COVID-19 on the death certificate by date of death.

The most widely cited daily figure is “deaths within 28-days of a positive COVID-19 death by date reported. This daily figure covers all newly reported deaths in the past 24-hours, whereby the individual who sadly died had, at some point in the past 28-days, tested positive for COVID-19. Those that follow the daily updates will no doubt know of the ‘weekend lag’ that occurs on the reported Sunday and Monday figures (as they cover the prior 24 hours, i.e. Sunday’s figures cover Saturday’s) which make these figures typically lower than those reported in the rest of the week. This is due to a slowdown in reporting done over the weekend, and this effect can also be seen on Bank Holidays and therefore often makes comparisons of daily death figures less clear.

If we look at by date of death however, we do not get this ‘weekend lag’ effect, as obviously people do not stop sadly dying of COVID-19 because it is a weekend or a bank holiday. This is illustrated in the below graph; the blue line showing daily deaths by date reported, and the orange line showing by date of death:

Source: UK Government

The measure of date of death therefore gives a more accurate and comparable measure of COVID-19 deaths. However, this measure is less ‘timely’ as it can take up to a week (or sometimes longer) from a person dying of COVID-19 to be reported in the official statistics. Imagine an individual who sadly died of COVID-19 on the 12th April and whose death was reported on the 18th April. This would show up in the official figure of ‘Date Reported’ as a daily death on the 18th April, whereas it would show as a daily death on the 12th April for the official figure of ‘Date of death’. Therefore, we cannot confidently use the latest 5ish days of data for the ‘Date of Death’ measure; a likely reason why ‘Date Reported’ is the more widely cited measure.

The final measure is ‘Daily deaths where COVID-19 features on the death certificate by date of death’. This measure is where an individuals death certificate mentions COVID-19 as one of the causes of death, regardless of whether or not the individual recently tested positive for COVID-19. This is shown in the graph below:

Source: UK Government

Recently, this has been very similar to the ‘Date of death within 28-days of a positive test’ measure. However, you can clearly see that in the first wave, deaths which featured COVID-19 on the death certificate were significantly higher than by 28-days of a positive test. This is likely due to the fact that there was a significantly lower amount of testing in the first wave than in the second wave. The above graph would also suggest that, for the entirety of the pandemic, there have been more COVID-19 deaths where an individual did not recently test positive than there have been deaths not due to COVID-19 where an individual did recently test positive.

The one downside of this measure is that it there is even more of a time delay; the latest figures for deaths where COVID-19 was listed on the death certificate are for 30th April – around half a month behind the current date.

In conclusion, care should be given in interpreting the COVID-19 death figures by date reported, as they often include weekend and holiday lags. For a more comparable measure, we should look to use COVID-19 deaths by date of death, taking care to ignore the past 5ish days for comparison as figures will be incomplete.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s