Why are people with blood cancer being told to shield for 12 weeks?

People with blood cancer make up 115,000 of the 1.5 million people in England with underlying health conditions who the NHS says are “extremely vulnerable” to the coronavirus.

Asking people to stay at home for three months is a big ask, and our survey of people with blood cancer has found that over half say it’s already had an impact on their mental health. So how do we know that the risk from coronavirus is so great if you have blood cancer that it’s worth shielding for so long?

Most of the health advice that we get from the NHS is based on lots of research that’s been published over many years. But with just a month between the first reported case of coronavirus in China and the first case in the UK, there hasn’t been anything like enough time for us to get robust evidence on the disease and how it affects different people.

Without solid evidence, the NHS has had to make assumptions based on what they know about the virus, and how people with various health conditions have been affected by similar viruses in the past.

One of those conditions is blood cancer. Research has shown that people with blood cancer are often more severely affected by viruses. For example, figures from the Intensive Care National Audit & Research Centre show that people with blood cancer make up almost one in 20 of those who become seriously ill from pneumonia.

This is because if you have blood cancer, your blood cells don’t fight off infections as well as they should and so you’re more vulnerable to them. Added to that, lots of blood cancer treatments – things like chemotherapy, immunotherapy and stem cell transplantation – compromise the immune system further.

Given this, there’s good reason to think that everybody with blood cancer should take extra precautions to avoid the virus. In the absence of any evidence to the contrary and, given the potential consequences of getting it wrong, we should exercise caution. We would much rather look back in a year’s time and realise we asked too many people to shield, than look back and realise we asked too few. That was the basis for the NHS’s decision to ask people to shield.

Now we're over four months on from the first reported case, more evidence is starting to emerge.

Firstly, the Intensive Care National Audit & Research Centre has been publishing weekly reports about people in England, Wales and Northern Ireland who have coronavirus and are treated in intensive care.

Of the 8,250 people included in its reports so far (up to May 7), 121 have had blood cancer. That works out as one in 68 of them. To give a comparison, the 115,000 people in England who are being asked to shield because they have blood cancer make up one in 500 people in the population.

Looking crudely at the numbers, this means people with blood cancer are over-represented in the people in intensive care by roughly seven times. This means that if you’ve got blood cancer you’re seven times more likely to end up needing intensive care as a result of coronavirus than an average person. But it’s not as straightforward as this because many people with blood cancer are over 70, which in itself puts people at greater risk from the coronavirus. So perhaps the “seven times more likely” is overstating things.

But there’s another factor that might mean the “seven times more likely” is an underestimate. People with blood cancer have been shielding for a month now, and even before that many of them were avoiding crowds and being extra-vigilant with handwashing. If they hadn’t been taking these measures, then more of them would have come into contact with the virus and so the number needing intensive care would have been higher.

Given these two factors are pulling in different direction, we can’t read too much into this data. But on the face of it, the high proportion of people in intensive care who have blood cancer does seem to bear out the NHS’s fears about the coronavirus’s likely impact.

As well as this, there have been scientific studies published. The first few were from other parts of the world and they added to the evidence that the coronavirus is particularly dangerous for people with blood cancer. But because of the small number of people included in each study, it wasn’t possible to be confident that their findings weren’t just down to chance.

But then last week a much bigger study was published, and this has given us the clearest picture yet of the risks of coronavirus for people with blood cancer.

The researchers from the OpenSAFELY study looked at people in England who have died with the coronavirus, and for each person looked at whether they had had any other health conditions and recorded their age, gender and ethnicity. They were then able to work out which of these seem to increase a person’s risk of dying with the coronavirus, and by how much.

The study hasn’t been peer-reviewed, meaning independent scientists haven’t been able to scrutinise its methods. But it’s been done by a respected group of researchers, meaning it would be surprising if there were significant issues with it.

The findings suggest that someone who was diagnosed with blood cancer in the last year is about 3.5 times more likely to die if they get the coronavirus than someone who doesn’t have blood cancer. This risk seems to reduce the longer ago someone was diagnosed, but that the risk of dying with coronavirus is still double even five and more years after diagnosis.

It’s important to emphasise that these aren’t exact figures. While this study looked at many more people than previous coronavirus studies have done, and includes around 200 people with blood cancer. Our understanding of the increase to risk in people with blood cancer will become more precise as the study continues. It may be that the real increase to risk is considerably higher or lower. But, crucially, the study did include enough people to be very confident that the increase to risk isn’t down to chance.

It looks likely that blood cancer increases your risk of dying with the coronavirus more than any other health condition with the exception of having had an organ transplant. And even if the very best-case scenario were to prove true, the coronavirus increases risk of dying in people with blood cancer by as much as it does for people with health conditions such as diabetes and kidney disease.

It’s also important to understand that this analysis lumps together the data from everyone who’s had a blood cancer diagnosis. This means we can’t compare the risk of people with different types of blood cancer, or the risk for people who are on watch and wait compared to those who’ve been successfully treated.

So while the “3.5 times more likely to die” is an average for people diagnosed in the last year, an individual person’s risk may be significantly higher or lower than this. As this evidence starts to emerge, it may be that the NHS is able to change its advice so it’s better tailored to individuals.

OpenSAFELY and other analyses have continued to highlight other COVID-19 risk factors that are just part of who we are as individuals. We know that being from an African Caribbean or Asian background, being male, and, most importantly of all, being over 70 are all risk factors for the coronavirus. Much more data and detailed analysis will be needed to show exactly how having more than one risk factor interact.

But we know that having more than one risk factor does add up to a higher risk.

These kinds of questions will only become clearer as more research is published, and we’ll keep on reporting back to you as this happens. If you want to make sure you get this, sign up to our regular e-newsletter.

But what we can say now is that the evidence is increasingly clear that, as tough as shielding is, it’s really important for people with blood cancer to do it for at least 12 weeks.

In the meantime, if you’re struggling with shielding and want to talk to someone, call our support line, which is open seven days a week, on 0808 2080 888. And if you have any thoughts about what we can do as a charity to support you through this difficult time, please do let us know.

Please note, this article was updated on 13 May 2020 to include newly published research.