Coronavirus cases, recoveries and deaths: Could outbreak really kill millions?
Monday 9 March 2020 09:20, UK
People from government officials to armchair experts have been attempting to play down the danger from coronavirus.
Claims that COVID-19, the disease caused by coronavirus, will kill millions of people have been accused of being alarmist and "overblown".
But are they right? Is the virus that is currently paralysing China as dangerous as has been made out?
How bad is it?
It is worth saying that the new strain of coronavirus that has hit the headlines, called 2019-nCov by some health experts, is not as deadly or as contagious as many infectious diseases.
Measles, for example, is capable of spreading much more rapidly through a community that has not been vaccinated against it.
As far as its deadliness is concerned, even in the early days of the outbreak in China - the country first affected - the proportion of people dying from the disease was smaller than illnesses that have caused some recent health scares.
Ebola, for example, has a case fatality rate - the percentage of people who are confirmed to have the disease who die - averaging about 50%. But, in the past it has been as high as 90%. That's nine out of 10 people dying.
Novel coronavirus, which causes the disease COVID-19, killed as many as 15% of people in the first 10 days of January but has now decreased to an average of 0.7% in many countries outside of China.
The reason the fatality rate has fallen sharply in China is because of a number of things.
As well as using a series of measures to contain the disease, experts in the health care system have learned a huge amount about the virus and its effects that have allowed them to save lives.
Part of the reason the case fatality rate is much lower outside of China is because doctors are now tackling the disease more effectively.
The biggest impact on death rates though is age.
In an early study of the first 44,000 Chinese cases of the virus, experts found 14.8% of cases aged 80 and over died, but only two out of every thousand 20-29-year-olds and no under 10s.
Whether someone already suffers from another illness also has a major impact on how likely they are to die.
It has been widely publicised that older people with pre-existing conditions are more at risk.
The study found more than 10% of people who already had cardiovascular disease died, compared with an overall case fatality rate of 2.3%.
Scientists have also learned that the disease appears to be more dangerous for men than for women.
It is thought that this is because the study looked at the impact on Chinese men, who may be generally less healthy than Chinese women.
What the death count does indicate, however, is how many cases are likely to be occurring.
Imperial College Professor Neil Ferguson, an adviser of epidemics to several governments, says that during the early phase of an epidemic, for every death, there are probably about 1,000 cases in a community.
:: Listen to the Daily podcast with Dermot Murnaghan on , , ,
How bad is it going to get?
The government's own worst case scenario is that up to 80% of the UK population could get the new coronavirus and up to 500,000 people could die.
A team from universities in Basel and Stockholm predicted that, worldwide, there could be hundreds of millions of cases, with a peak occurring around January next year, about six months after a sub-peak ahead of a small dip over the summer.
The team's modelling, however, varied depending on the value of R0, which itself depended on many things.
The reason why there is uncertainly is because, even if a pandemic occurs, the rate of spread will depend on factors like government intervention.
In an model done exclusively for Sky News, Lancaster University showed how so-called social distancing measures would impact on the spread of the virus in an English county.
It needs stressing at this stage that a pandemic is only one of many possibilities.
At this stage, experts say it is not certain that it will become an epidemic in Britain, especially with the measures that are being employed.
This is because most of the confirmed cases that have been reported so far have come from outside the UK.
Dr Robin Thompson, a University of Oxford junior research fellow in mathematical epidemiology, told Sky News: "There are two plausible scenarios.
"The first is that we're able to isolate imported cases quickly and chase up any contacts and do contact tracing very quickly, and then none of those imported cases lead on to sustained person-to-person transmission and therefore we don't get an outbreak in the UK.
"We just get a series of imported cases without any subsequent person-to-person spread.
"And then scenario two is that in a situation where we missed some of those contacts or we've missed cases coming into the UK and then what we see is sustained person-to-person transmission and then we see large numbers of cases.
"The chance of anything happening in between those two things is pretty small."
Dr Thompson says while many teams around the country are working towards a model which is going to predict how many will be infected and die, it is too early to say what is going to happen because the virus has such a wide variety of effect - a quality called heterogeneity.
He added: "Heterogeneity leads to a huge amount of uncertainty in terms of things of probabilities, but also things like when the outbreak is going to peak, how large is the outbreak going to be. The heterogeneity around this virus isn't fully understood yet. It could be anywhere in an absolutely massive range."
Another reason why it is impossible to say how many will die is because, even in China, there may be many cases that are not yet known about.
An infection spectrum pyramid for coronavirus shows how, while health authorities may know large numbers of people who are showing symptoms, including those who are severe and those who have died, they do not know how many may have experienced very mild symptoms - because they will not have sought medical attention - and those who have shown no symptoms, but are still carriers of the virus.
The true death rate will not be that calculated by the Chinese from the first 44,000 cases - it will be the number who have died divided by the total number of cases (including those who are asymptomatic) - probably a much smaller number.
Why we need to take it seriously
Modellers work on the principle that most epidemics follow a similar pattern - starting off slowly and then gaining speed - at a consistent rate - before peaking after a certain point, as increasing numbers become immune.
The rate at which the number of cases doubled in Hubei before it hit its peak was roughly every six to seven days.
If the UK cannot prevent an epidemic, it raises the possibility that the number of cases would continue to spiral until it reaches tens of thousands of new cases every day.
It has happened before - when flu hit in 1918-19, in 1957-58, in 1968-69 and in 2009 - outbreaks that affected between 10% and 35% of the population and are used in the UK's pandemic modelling.
If coronavirus spreads at that rate, it would put a severe strain on the economy and a health system where bed occupancy rates are already above the 85% threshold, which NICE says can lead to shortages.
It is why several experts have spoken of the need to delay the progress of the disease in the UK until after the worst of the winter - when more hospital beds will be free.
Dr Mike Tildesley, Warwick University associate professor in infectious disease modelling, said: "If we look at Wuhan, there were very draconian measures put in place. The challenge we have is we just don't have the resources. For instance, we don't have 10,000 workers that can come and build a hospital in a few days, which is what happened (in Wuhan).
"If we start see significant numbers of infected people... at some point we're going to have issues in the NHS. Certain hospitals do have isolation facilities that they've put in place. But at some point we're going to have an issue that potentially we could be exceeding those current resources."
Professor Rowland Kao, professor of Veterinary Epidemiology and Data Science at the University of Edinburgh, said government advisers have to model for the worst case scenario because they do not know how bad it will get.
He said: "If you lock down things more tightly, it does give the healthcare system more opportunity to respond.
"One of the crucial things is, how do we make it through to the summer? We don't know what's going to happen, but there's a chance that the spread of the disease will slow down for various reasons - mixing of schools is over; people are home more; they don't have to travel so much; plus the things that they can be confused with, like seasonal flu; there will be fewer or no cases to confuse them with.
"All those things means the longer you can delay the spread of the disease, the more likely there is more time to control it.
"And if we can get over the hump until next year, things like antivirals, vaccines, all those things start hopefully to come into play, maybe better diagnostics as well."
Consequently, he added, it is wise to listen when the government offers advice.
Prof Kao said: "Individuals may find that their individual circumstance makes them think that they're not at risk, or are different. It is almost certainly a bad thing to ignore the advice that's given."
Methodology for the social distancing county spread model:
This is an SEIR model developed by Jessica Bridgen, PhD student at Lancaster University. The model shows an epidemic of an infectious disease with the current best estimated characteristics of Covid-19 in a single county of the UK, with a population size of approximately 1.1 million, and starting with 100 people infected. The model assumes a R0 of 2.2, an average latent period of 5 days (similar to the incubation period) and an average infectious period of 5 days.
Moderate social distancing aims to reduce transmission by up to 30%; extreme social distancing aims to transmission by up to 60%.
Why not plot the y-axis? The model illustrates how social distancing as an intervention could affect the number of cases during an epidemic, but it is not a predictive tool. To avoid misunderstanding, we have decided not to show the total number of cases predicted by the model, but instead how social distancing can shift the dynamics of an epidemic.