Despite the coronavirus pandemic entering its fifth month and having killed at least around 300,000 people, there is still no unified way of counting deaths from the virus – with small differences in methodology having great consequences for how we approach, and ultimately defeat, the virus.
Calculating the exact death rate of COVID-19 is central to policymakers and scientists who are trying to assess the impact of lifting restrictions on movement. If their estimations are too low, they risk exposing people to deaths that could have been prevented by telling them to stay at home. If their estimates are too high, they may prolong lockdowns unnecessarily, causing further economic damage that has a knock-on effect on public health.
Unfortunately for policymakers, there is no unified global method of calculating coronavirus deaths. Instead, governments vary in how they classify deaths as coronavirus, with several key variables.
Coronavirus death toll vs. related death toll
The first is whether they classify a death from coronavirus-related complications, such as a person who died from a heart condition brought on by coronavirus, as death by coronavirus or by the complication itself.
COVID-19 is primarily a respiratory disease, but doctors have also reported a wide range of symptoms including organ failure, blood clots, heart disease and even paralyzed limbs. This range of symptoms, and the fact that coronavirus attacks the immune system, and therefore exacerbates existing illnesses, means it is difficult to define a death from coronavirus.
Different definitions produce wildly ranging death tolls.
Italy and Spain, for example, have included anyone who died with coronavirus in its official toll, even if they had preexisting conditions that may have been the actual cause of death. Both countries’ relatively high tolls therefore may be inflated compared to countries that use other methods.
“The way in which we code deaths in our country is very generous in the sense that all the people who die in hospitals with the coronavirus are deemed to be dying of the coronavirus,” said Professor Walter Ricciardi, scientific adviser to the Italian government on the crisis, to British newspaper The Telegraph.
In contrast, the UK initially used a much less comprehensive method, instead counting only deaths from the virus itself and not preexisting conditions who may also have had the virus.
Whether a country carries out postmortem tests also plays a role: Italy does, while France does not, for example.
Even within countries, the method can vary – US states each decide their own criteria – while others including, China and Iran, have been accused of deliberately underreporting deaths for political goals.
Recorded death toll vs all death toll
A second important variable is whether the government records only the deaths that are confirmed in hospitals, or estimates a larger number including deaths at home based on average death statistics.
Many of the deaths from coronavirus are expected to have happened in care homes for the elderly. Some countries have not reported these as coronavirus cases because many residents die before a doctor has been able to examine them and report them as a virus case.
By mid-April, Case Data England suggested that as many as 7,500 people may have died from coronavirus-related causes in care homes in England alone.
The UK switched from only counting hospital deaths to including deaths at home – but only those diagnosed with coronavirus – a few weeks ago, resulting in a jump in cases.
But as many people who die in care homes or at home are not tested for the virus, scientists have suggested a better method for calculating the real death rate is to measure the overall death toll from all causes during the coronavirus pandemic against an average period.
By accounting for the annual difference, statisticians suggested that there may have been more than 50,000 coronavirus deaths by May 1.
As of May 1, the government’s official count stood at around 28,000. But when looking at data, that figure climbed to 36,000 if death certificates that mention coronavirus are included, leaving a 14,000 average annual increase in deaths still unaccounted for.
Given these variables and a range of other cultural, political, and demographic factors, calculating the death rate of the virus is much harder than it might seem at first glance.