for-phone-onlyfor-tablet-portrait-upfor-tablet-landscape-upfor-desktop-upfor-wide-desktop-up

Fact Check-Comparing raw 2020 death figures to the past 50 years gives a misleading conclusion on mortality rates

Social media users have shared raw mortality data from the last 51 years in the United Kingdom to make a misleading conclusion that undermines the severity of the COVID-19 pandemic.

The posts, shared on Facebook, all include a table of data said to be pulled from the Office for National Statistics (ONS) (here , here , here). Two of the posts link to the ONS data source (here). In the accompanying text, one post reads: “STOP BEING FRIGHTENED! Please view the number of all cause deaths in England & Wales over this past 50 years to see that the year 2020 was 35th on the list of total number of deaths. The word ‘pandemic’ has been used to create FEAR from the outset and the outbreak of Covid-19 but on average we are losing NO MORE people than we have done in previous years” (here).

Another adds: “In the last 50 years of registered deaths in the UK. The year of the deadly pandemic of 2020 came in at 35th. 35th...and we have destroyed thousands of small businesses, denied education to millions of children and seen NHS waiting list for critical care explode. And created a National debt that will take decades to repay. Nice one Boris!” (here).

However, this is a misleading interpretation of the data. Reuters spoke to the ONS, along with academics from the University of Oxford, Imperial College London and The Health Foundation Charity, who all said that a comparison of crude data on deaths from the last 50 years will not give an accurate understanding of current mortality patterns.

Firstly, the ONS table says figures for 2020 are still provisional. It also says the “number of deaths” column is not directly comparable over time as it does not account for growth in population. The same is said for “crude mortality rates (per 100,000 population)”, which appears to be the dataset used to decide the rankings in the Facebook table (here). This rate does not consider the changes in the age structure of the population over time.

Secondly, all the figures recorded in the Facebook table for total deaths per year match the ONS data – except for the year 2020. The ONS lists this figure as 608,002, whereas this is 605,000 in the table. The datasets also do not match for population size, crude mortality rate (deaths per 100,000), nor age-standardised mortality rate. Meanwhile, the table on Facebook has an additional column said to show the “% of pop” that died in each year – but this is not reflected in a column in the ONS table.

Despite these differences, Christl Donnelly, a professor of statistical epidemiology at Imperial College London (here) and professor of applied statistics at the University of Oxford (here), told Reuters that 2020’s ranking in 35th place for most deaths in the Facebook table is almost correct.

She said: “If you look at the 50 years 1971-2020 ONS statistics, then I found 2020 to be 39th highest based on the age-standardized mortality (per person) rate or 34th highest based on the crude mortality rate (per person).”

However, this is still a ranking based on raw data – and experts say this isn’t enough to calculate mortality patterns.

Ridhi Kashyap, associate professor of social demography at the University of Oxford (here), said it would be misleading to compare the past 50 years as this “forgets the tremendous progress that has been made in reducing mortality and improving survival, particularly among older age groups.”

He told Reuters: “Age standardised death rates have actually been consistently falling over the past 50 years due to improvements in survival across all ages, and especially among the older (60+) populations in recent decades. So it does not make sense to compare current death rates to those in 1970 or 1980 as they are clearly much lower now — but once we compare 2020 to more recent years (e.g. those after 2010) we see that 2020 does much worse.”

Looking at data from more recent years is “most meaningful” because it would give a better idea of “the number we would have expected under normal circumstances,” said David Finch, a senior fellow at The Health Foundation charity (here). He told Reuters that by doing this, the data reveals a 14% increase in deaths that were expected in England and Wales in 2020 based on the previous five years (here).

Breaking it down further, Donnelly found a similar “14.5% increase in deaths with a 0.65% increase in population size” between 2019-2020. She added: “We expected 534,315 [deaths in 2020] but observed 608,002. That is 73,000 more deaths than expected.”

Meanwhile, ONS calculations from the last five years come close to Donnelly’s figure. It told Reuters: “In 2020, the overall number of deaths registered in 2020 was 75,925 higher than we would expect when looking at the five-year average between 2015-2019.

“Looking recently, we’ve experienced fluctuating but historically low mortality rates. But the provisional age-standardised mortality rate in 2020 was 1043.5 deaths per 100,000 population around 8% higher than the five-year average and is the highest it has been in more than a decade.”

VERDICT

Partly false. ONS data from 1971-2020 shows the year 2020 was 39th in rankings of age-standardised mortality, or 34th for crude mortality rate. However, a comparison of mortality data between 2020 and the last 50 years is misleading as this does not account for changes in the age structure of the population, nor steadily decreasing mortality rates due to societal advances.

This article was produced by the Reuters Fact Check team. Read more about our fact-checking work here.

for-phone-onlyfor-tablet-portrait-upfor-tablet-landscape-upfor-desktop-upfor-wide-desktop-up