Coronavirus: How to understand the death toll
Each day news of more deaths is a huge source of alarm to people across the country - as well as a tragedy for the families involved.
Projections of how bad it could get have prompted strict measures to be taken, with pubs, clubs, theatres and schools closed and people told to stay at home.
A key piece of modelling which has informed government has been done by Imperial College London.
It suggested 500,000 could die if we do nothing, while the government's previous strategy to slow the spread was likely to lead to 250,000 deaths.
Instead, it is hoped the steps which have been taken, which are essentially about suppressing the virus, will limit deaths to 20,000.
In an appearance before the Health Select Committee in March, Sir Patrick Vallance, the government's chief scientific adviser, said that would be "horrible", but still a "good outcome" given where we are.
He likened it to those killed by flu, citing a figure of 8,000 deaths a year.
Could these deaths be prevented?
But what is not clear - because the modellers did not map this - is to what extent the deaths would have happened without coronavirus.
Of course, this will never truly be known until the pandemic is over, which is why modelling is very difficult and needs caveats.
Every year more than 500,000 people die in England and Wales - factor in Scotland and Northern Ireland, and the figure is around 600,000.
The coronavirus deaths will not be in addition to these, as statistician Prof Sir David Spiegelhalter, an expert in public understanding of risk at the University of Cambridge, explains.
"There will be substantial overlap in these two groups — many people who die of Covid [the disease caused by coronavirus] would have died anyway within a short period."
The flu comparison
In contrast, the figures put forward for flu - 8,000 deaths a year - is different.
It is actually the number of deaths over and above what you would expect to happen in any given year.
In fact, it is perhaps a little low. Public Health England uses a figure of 17,000, based on recent winters.
Many more actually die with flu, but this figure gives you an indication of how many more die because of flu.
In comparison, the daily coronavirus death figures and the modelling by Imperial, simply look at those who die with the virus.
They do not tell us is to what extent coronavirus contributed to the death.
How many extra deaths could there be?
A team from University College London (UCL) has attempted to unpick this by looking at the expected number of deaths you would normally see, and then mapped out how many extra deaths coronavirus could cause.
The paper, which has not been peer reviewed yet, shows those from the at-risk groups - the over 70s and people with health conditions - have a 4.4% risk of dying in the next year regardless of coronavirus.
That is to say, one in 20 would not be expected to be alive one year later.
To factor in the impact of coronavirus, there are two variables which are as yet unknown - how much it increases the risk of death by, and how many people become infected.
If it turns out to be as deadly as flu and just 1% of people are infected (the upper limit of what we should be aiming for given the measures in place according to researchers), the number of extra deaths will be under 1,400.
But many believe it will be more virulent than flu, so the researchers mapped different scenarios.
If it was five times as deadly - a "reasonable" estimation, researchers said - there could be 6,900 excess deaths.
If 10% of the population were to be infected with more relaxed measures, the excess death figures would increase 10-fold. But these are just models.
What is needed now that the virus is spreading, is for researchers to get good access to hospital data and in particular intensive care records, so they can more accurately work out the impact the pandemic is having, says Amitava Banerjee, who is leading the UCL research.
Why better intelligence is needed
Prof Robert Dinwall, an expert in sociology from Nottingham Trent University, says it is also important to consider "the collateral damage to society and the economy".
He cites the mental health problems and suicides linked to self-isolation, heart problems from lack of activity and the impact on health from unemployment and reduced living standards.
The economic hit is something University of Bristol researchers have now looked at. Their conclusion? Trashing the economy costs lives.
They found the benefit of a long-term lockdown in reducing premature deaths is outweighed by the cost in terms of lost life expectancy from a prolonged economic dip.
And the tipping point is a 6.4% decline in the size of the economy - on par with what happened following the 2008 financial crash - which leads to a loss of three months of life on average across the population because of factors from declining living standards to poorer health care.
There are, of course, other factors at play here. Left unchecked - or not checked enough - the deaths would come very quickly.
This in itself would overwhelm the health service, putting even more lives at risk, because care may not be available for others whether that is a heart attack victim, a stroke patient or simply someone who has had a fall.
- A SIMPLE GUIDE: What are the symptoms?
- GETTING READY: How prepared is the UK?
- MAPS AND CHARTS: Visual guide to the outbreak
- PUBLIC TRANSPORT: What's the risk?
But, given that suppressing the virus is almost certainly not going to make it go away, and the prospects of a vaccine being available in the near future are considered slim, at some point the government is going to have to weigh up the full benefits and costs, to help it decide on the next step.
Seeing the full picture will be essential in getting those calls right.