This is a transcript of my video about coronavirus, kindly put together and lightly edited for rambling by Dr Marianne Baker. Thanks, Marianne!
I made this video in my front room because of course the whole UK (and much of the world) is locked down due to coronavirus.
Before we start, a bit of terminology: COVID-19 is what we’ve called this disease caused by a new virus that’s in the coronavirus family. The virus is called SARS-CoV-2. I’ll refer to it as “coronavirus” throughout even though there are many coronaviruses—hopefully this isn’t too confusing!
Obviously, as you’re probably aware if it’s still lockdown where you are, you’re watching this from your house, you’re only allowed to leave certain times of the day for certain particular reasons, and what I want to do is address things about this situation that the government and media aren’t really communicating very clearly.
What’s not really being made clear is: how this situation ends.
How does lockdown get released? What’s the long game? What’s going to happen to coronavirus over the next few months or even years?
I want to look at these questions – because I’ve been reading an awful lot of research papers and have trouble tearing myself away from the news and all the results coming out – and try to delve into what’s actually going to happen; what the future really holds when it comes to coronavirus.
Part 1: The End
The “endgame” when coronavirus is “defeated” is when we achieve herd immunity.
You might have heard that phrase being thrown around by various different politicians and journalists so here’s what it means. When a new virus like coronavirus enters a population, nobody has ever “seen” that virus before and so there’s no immunity; no natural ability to fight the virus in the population.
That’s why coronavirus has been so contagious and run so rampant through all the different countries it’s got to. However, people can become immune and there are two different ways you can become immune:
1. You become infected and survive the illness.
Your body has then learned to recognise that infection (the virus) and come up with its own defenses, which means that if you’re infected by the virus again the immune system jumps straight into action and hopefully wipes out the virus before you even know you’re sick.
2. Get vaccinated.
If we develop a vaccine for the coronavirus then you could get an injection like when you were a kid or going traveling or for your annual flu shot. What that does is it teaches your body, giving it a sneak preview of what the coronavirus would look like. Then your immune system can build up its defenses and if you actually are infected with a real coronavirus you’re ready to fight it off.
The number of people in a population required to be immune to a disease in order to prevent its spread depends on how contagious that disease is. Epidemiologists often try to quantify this using a number called R0. You might have heard it mentioned in the media too.
R0 is the number of people that one person infected with the virus will go on to infect in a completely naïve population, i.e. one that has never seen this coronavirus before.
We think at the start of the outbreak that R0 was about 2 to4, so every infected person would go on to infect two to four others with coronavirus.
But as immunity starts to spread through the population it might be that, of those three others who would have been infected, if two people are already immune, only one person will actually catch the virus.
So, if we had about two thirds immunity, R0 would go from a value of two or three to maybe a value of one.
R0 falling below one would mean the coronavirus – rather than growing exponentially in this terrifying way that we’ve all been observing – could actually slowly fizzle out.
If we can get immune in the population high enough (either through infection or – preferably – vaccination) then coronavirus will stop spreading and we can finally get this disease under control.
So that is the ultimate “endgame”, how coronavirus will hopefully finally “die”. One way or another, through infection and/or vaccination, there’ll be enough people in the population immune to the disease that nobody else can catch it.
There are two important caveats to the herd immunity idea.
First, it is much more preferable to get herd immunity through vaccination than just through a huge number of infected people and the reason is that although we don’t exactly know the mortality rate (that’s how likely you are to die if you contract coronavirus), we do know that it could be quite high.
Estimates range from 0.1% to 1% and those are not odds that I particularly like to run myself!
So the problem is that if you want to infect two thirds of the British population (that’s over 40 million people) with coronavirus, and there’s a 1% chance of death, then that’s over 400,000 deaths to achieve herd immunity by this method of just infecting the population at large.
This very simple back-of-the-envelope calculation tallies with a more formal modeling exercise. In the middle of March, Imperial College in London worked out that, unchecked, coronavirus would kill about half a million people. We’d rather not get herd immunity through just letting people get infected because this risk of death is just too high (one hopes that’s obvious, anyway).
Even if it turns out the death rate is lower, it could still be tens-hundreds of thousands of people who would die before herd immunity is achieved. So ideally we will develop a vaccine.
The other caveat with any immunity-based strategy, even regarding vaccinations, is that we don’t know how long immunity to the coronavirus will last.
How Long Will Immunity Last?
This coronavirus disease is obviously completely new so we don’t have any actual data from people who are infected one, five or 10 years ago to see how long their immunity lasted, but we do have other examples of coronaviruses (other viruses in the same family) that can help us build a clearer picture.
There’s SARS, a serious disease that was in the news 20 years ago, but there are also much more common coronaviruses. A few causes of “the Common Cold” are coronaviruses (around three in 10 according to some estimates) and what we know is that immunity to these viruses is often short-lived; only a few months or sometimes years.
Given the evidence, it’s possible that if you have been infected or vaccinated against coronavirus, that protection from your immune system might not last very long.
There have also been a few reports of people being reinfected with coronavirus after they’ve already recovered and, unfortunately – so far at least – it isn’t entirely clear whether these are genuine reinfections (people who’ve got better than encountered the virus again and got a sick a second time) or not.
So we need to watch this space and try to learn a more about how long immunity to coronavirus lasts because, if immunity is temporary if it only lasts years or, at worst: months, then we’ll have a lot more trouble getting it under control.
Part 2: The Beginning of the End (Developing a Vaccine)
The main problem with getting herd immunity by vaccination is that we currently do not have a vaccine – there are thousands of scientists, researchers, all around the world working full-time trying to develop new coronavirus vaccines that could be effective.
The problem is that developing a vaccine isn’t a fast process. We think that even in the most optimistic timelines if everything works first time, seamlessly, we won’t have a vaccine ready for mass deployment for at least 12 to 18 months, maybe two years or more.
Unfortunately, there are a lot of things in vaccine development that you just can’t speed up and one example of that is safety testing.
You need to give a decent number of people a dose of the vaccine and then check that they are safe afterwards; watch them for a few months to make sure they have no adverse reactions.
You also need to test whether or not they’re actually going to avoid getting the disease and, given that coronavirus (thankfully, because of our lockdowns) is under relatively good control, someone wandering around is relatively unlikely – vaccinated or not – to come across the coronavirus, which means that this phase of testing can also take a long time.
That’s why it’s so important that we do absolutely everything we can while those human trials are taking place to speed up vaccine development in other ways. One of the key things that we can do is actually being handled not by the world’s governments but by Bill Gates through the Bill and Melinda Gates Foundation.
The software tycoon, founder of Microsoft, formerly the world’s richest man, is thankfully plowing some of his billions into building coronavirus vaccine factories.
You might think that’s premature because we haven’t got a working vaccine yet but it’s actually very sound logic. We could find ourselves in six – 12 months in the situation where we have a vaccine candidate, it passed all the human trials, it’s safe and effective, but suddenly we’ve got the problem of needing to manufacture it at scale; we need to make tens or hundreds of millions – even billions – of doses.
Bill Gates has figured that we should build the factories anyway even if we don’t know the vaccines are going to work because it means that, once we have a working vaccine then we can deploy it straight away, kicking the factory into high gear. Not only will it have been built, but they’ll also have solved any small teething problems in manufacturing; they’ll have made sure the product they’re making is safe and it’ll be ready.
So even though it might seem like a waste of money (say five factories are built and four of them are never needed), a few billion apparently “wasted” now could save trillions of dollars and potentially millions of lives if it speeds up vaccine development even by a week or month.
Human challenge trials
Another slightly more controversial way that we can try to speed up vaccine development is through something called a human challenge trial and the reason that’s controversial is that it’s almost a euphemism; what we mean by “human challenge” is to deliberately infect people with coronavirus. Let’s look at how that might speed things up.
Imagine you’re trying to do a normal vaccine trial: get a bunch of people and give them your experimental vaccine. Get another bunch of people and don’t vaccinate them to act as your control group. Then you send them out into the world.
Normally, you wait a period of time then ask them back and ask them, did you get infected or not? Count them up and if more people got infected in the control group than in the vaccinated group, you know the vaccine worked.
Unfortunately for vaccine developers, much of the world is trying its hardest to suppress coronavirus through lockdowns so the chances of catching it are very low, regardless of vaccine status, meaning it’s very hard to do a conventional trial.
You either need to wait a long time to get enough people in the control group infected to be sure that your vaccine was what was responsible for keeping cases low in the other group (which we don’t have), or you need to have huge numbers of people in your trial, which gets very expensive.
The way that we can speed things up is by doing a human challenge trial because, rather than waiting for enough people to get infected, you can just deliberately infect them yourself. For example, use a nasal spray to squirt a bit of coronavirus up the nose of the people who’ve had the vaccine and see if they develop symptoms.
But what are the ethical challenges of a human challenge trial?
First, you must make absolutely sure that all participants fully understand what they’re getting themselves into because they’ll be infected with a dangerous and possibly deadly disease, testing a vaccine we aren’t even sure works – because that is the point of the trial, right?
However, this isn’t such a terrible idea.
Firstly because you can recruit people into your trial who are the safest possible people to infect with coronavirus.
We already know that it seems to affect younger people less; they’re less likely to develop serious symptoms, or to end up in intensive care, or to die. Say they’re between the ages of 20 and 40 (the lowest risk groups).
We also know that people are more likely to die from coronavirus if they have underlying conditions; things like diabetes and heart disease. Again, you could exclude those groups from your initial trial.
Then we can make sure they have the absolute best care; access to the top intensive care doctors and all the equipment they could possibly need in the event that they do fall sick enough to need it.
This can have a counterintuitive effect because ironically, these people could actually be safer in your trial being deliberately infected than a member of the public just going about their business – if the health service is overwhelmed, people who catch the disease might not otherwise have access to that best possible care.
So you can try and make it as attractive as possible for your participants to the point where it might even turn out better for them than just bumbling around in quarantine society hoping to avoid the virus.
It’s also worth noting that although a human challenge trial does sound quite scary when looking at a dangerous new disease like this coronavirus, actually, we already do these kinds of trials for other diseases like malaria, flu, typhoid and cholera. So it’s not necessarily as controversial as it first appears.
Part 3: The End of the Beginning (trying to lift lockdown)
That’s the long game. We need to develop herd immunity one way or the other – preferably by developing a vaccine.
But even if everything goes to plan – if Bill Gates’ vaccine factories all work and we do an ethical human challenge study to identify a vaccine candidate quickly – it’s still going to potentially be years before we have access to a vaccination that works.
What are we going to do in the intervening time? We can’t just stay locked down.
There are a variety of different approaches to go through. The first is antiviral drugs.
There are now many trials of coronavirus drugs ongoing in hospitals around the world.
Drugs normally used for everything e.g. HIV, malaria, other conditions – the big advantage of antiviral drugs is that, if they work, we could roll them out quickly. Particularly if they’re repurposed existing drugs; we already have a really good idea of when an HIV drug is safe and what the potential side-effects are so we’re able to get them into hospital clinics fast.
The problem with antivirals is they’re never going to make a huge dent in coronavirus.
They might slightly improve the survival of people with the disease or they might slightly shorten the time someone’s sick but the problem is: you can only give them to people who test positive or have symptoms, which means it’s then too late to stop the spread, which is what we really need to do.
If we can slightly improve people’s chances once they’re sick, that’s great, but it’s not a solution to this problem.
Test, Trace, Isolate
The key to getting past lockdown will be testing and contact tracing because through lots and lots of testing we can work out where the disease is in the population and, once you’ve identified infected people, start contact tracing.
That means trying to find everyone a positive person has been in touch with (and therefore might have infected) and, if you can then isolate them, you can make sure that they don’t pass it on.
Currently in most places this just isn’t possible because there isn’t enough testing and the infection’s far too widespread. There’d be potentially millions of people who’d need contact tracing.
So the only option we have is complete lockdown; encourage everybody to stay as physically far apart as possible to deny the virus opportunities for transmission.
But as our lockdowns start to work and slow the spread, if we can ramp up testing at the same time, we can get to a point where there are few enough cases, enough testing and an efficient contact tracing network.
There are a few countries that have shown this approach can work; places like South Korea, Singapore and Taiwan. Through extensive testing, contact tracing, and isolating people, they’ve managed to keep the epidemic under control.
Hopefully by emulating this strategy countries currently in lockdown can start to slowly release it and, with lots of testing and contact tracing, maybe get to the point where the economy could restart.
A central part of this is the very laborious contact tracing. If someone tests positive you literally ask them (as in call them up) who they’ve been in contact with and where they’ve been and try and get in touch with all those people to tell them they might have been infected and they should go to a testing center.
However there are potential ways to automate the process and one getting a lot of coverage is using mobile phones.
China deployed a rather controversial app; although it was technically optional, you were required to show it when you went anywhere like public transport or into public buildings, which meant that people more or less had to have the app installed. By some mysterious not-entirely-understood process it gives people a green or red signal to indicate whether or not they may travel.
However, in many countries this kind of draconian mobile phone data use wouldn’t be tolerated. Luckily there are some ideas in development that would permit contact tracing using mobile phones without the controversial “black box” app aspect.
Google and Apple have come together to try work on this tech and, instead of tracking people using GPS, which would require you to know their every movement and compare it to everyone else to see who was in the same place at the same time (which would be a privacy nightmare), they’re using Bluetooth.
Your phone has something called Bluetooth LE (Bluetooth Low Energy) that’s always on and is transmitting out short randomly generated keys; non-identifiable strings of letters and numbers.
Then if you go onto public transport, your phone pings out its key to people around you, your phone collects the keys from all their phones; each of your phones will know that they’ve been in proximity to one another.
However, because the keys are random and anonymous, you can’t work out who has been in close proximity – just the identity of a phone. The keys can also be changed, say every 15 minutes, which means it’s very hard to piece together any identifiable information.
The whole time you haven’t got coronavirus, the data would stay on your phone. If you test positive and know you’ve been in touch with other people, you can then upload a file from your phone onto a central server. Again that file doesn’t contain anything identifiable, just a bunch of seemingly random keys. The server then sends a message out into the network saying, “Has anyone ever transmitted this key from their phone?”
This means that you can match up to the other people who transmitted you those messages as you wandered around during the day and it’ll then pop up an alert on their phone saying “you’ve been in proximity to someone who’s diagnosed positive for coronavirus. You should go to get tested.”
Even if we combined the automated mobile phone-based method and the old-fashioned bank of people in a call center contacting everyone you’ve been in touch with, it’ll be difficult to get coronavirus under control with testing and contact tracing.
This is a graph from a study about how difficult it would be to get coronavirus under control this way:
The color scheme is simple; red is bad, green is good. Along the bottom of the graph (x) is the percentage success isolating cases; the number of people that we can find successfully who have coronavirus.
Up the side of the graph (y) is percentage success in quarantining their contacts; finding those people and making sure they don’t contact anybody else and potentially spread the disease.
We need to try to reduce R0, reduce the number of people that every case of coronavirus goes on to infect, to below one – that would get us in the top-right green area of the graph. It’s very optimistic, assuming we can do the whole process within a day.
There are a few other scenarios they looked at. As you can see if it takes us more than a day we’re basically doomed, the whole graph is red:
And if we can do it instantaneously, then we do have a lot of green to play with. So if we can do this really rapidly in less than a day then it’s possible that we can use contact tracing and testing to really try and keep a lid on this thing.
Although testing and contact tracing is the smart way to try to keep coronavirus under control, we also need to use every tactic at our disposal.
The fact is that even once we start testing and contact tracing we’re probably going to have to maintain some degree of lockdown at first – gently lift the restrictions – and make sure the virus can’t get out of control as it has done previously.
There are other things we can do that will chip away at R0 (slightly reduce the number of people who catch coronavirus from every infected person). One is by always wearing masks.
In fact, the US has already made this recommendation. If you leave the house to do some “essential activity” like grocery shopping, going to a medical appointment, or your morning run, you’re supposed to wear a homemade mask e.g. a cotton t-shirt you’ve improvised or a bandana or scarf wrapped around.
While this isn’t as effective as “proper masks” (surgical, FFP3 or N95 masks that we hope healthcare workers and other key workers are using because they’re vulnerable dealing with active cases and others) it’s still perhaps worth using these homemade masks in order to try to reduce transmission.
This is important because we know the coronavirus can be spread by people who are asymptomatic – they have no symptoms – so if you’re wearing a mask you’re protecting others. If you’re breathing out little droplets with coronavirus or coughing then your mask will hopefully catch and block some or most of them.
There’s also some inbound protection as well. If you’re breathing air that contains coronavirus-y droplets they’re less likely to make it all the way to your nose and mouth (and eyes) where they can infect you. If you’re wearing a mask properly.
Since these are homemade masks and, as I say they’re not as effective, they’re not going to completely stop transmission; they don’t make you 100% safe outdoors, it’s really important that you don’t start risk compensating by going to the supermarket more often or walking closer to people; we must keep physically distant.
We have to stay at home as much as possible at least for now.
But if masks can reduce R0, even just by 10-20%, that gives us more room to play with. It means we can release lockdown a little when there’s less strain on our testing and contact tracing systems.
So the idea of masks is just to slightly reduce R0 and make it a little bit likely all our other measures will work.
Keep up the good work
As well as wearing masks we can all do our part to try to keep R0 as low as possible by continuing good habits we’ve learned during lockdown, because all of this robs the virus of opportunities to spread so other aspects of the lockdown can be lifted a little earlier.
Washing your hands regularly and thoroughly; staying at home as much as possible, even if you are technically allowed to leave the house; physically distancing by keeping two metres or more away from others when you are out, etc.
The other thing to remember is that R0 is an average across everybody in society and there’ll be some who find it harder to do these things. For example, key workers who have to go to work (doctors, nurses, people who work in shops) find it very hard to reduce their number of contacts during the day.
Also those who find staying at home harder – if you’re a family with children, maybe living in a small flat, you might find those opportunities to go out once a day much more valuable.
So if we can all try to reduce our social contacts as much as possible, even as lockdown lifts, we can give opportunities to others who might be less able to do so.
Another thing we can try and do is increase the capacity of our hospitals – intensive care units (ICUs) in particular. This is a worthwhile and important thing to do because if the virus does get out of control, greater capacity means we are able to treat more patients.
But, like antiviral drugs, this isn’t a way out of the problem. By the time someone is put on a ventilator they’re already very sick. Even if we could mass produce specialist devices, you also need staff; the trained nurses and doctors and support staff to operate them and make sure all the other functions patients need are supplied safely.
You also need the drugs that accompany the use of ventilators and other intensive care equipment and the physical beds; intensive care beds are themselves specialised bits of equipment.
I think it’s fair to say that we can’t ventilate our way out of this crisis.
Although we should ramp up production, because we should be hitting this in every way we can, ventilators alone are not a solution and research is now beginning to suggest they may not be a solution at all, but actually unhelpful in the case of this new disease.
If we can all try to keep R0 nice and low then the government can come in with testing and contact tracing and isolating those who test positive. Then, as the lockdown starts to lift, we could return to a degree of normality.
But the fact is this is a very long game we’re playing with coronavirus; it will be months, maybe years, before we finally get out of the woods into herd immunity.
It’s really important that we give this a go even if it’s not guaranteed to succeed because the cost of not doing so is immense. If we let coronavirus get out of control there may be so many cases that it would take an impossibly long lockdown to get things under control to the point where we could have this option available again.
So, while it’s on the table, we should try it. If it fails, so be it, but if we lose the opportunity it may not come back again if numbers get too high.
If we fail to implement testing and contact tracing then we could be stuck in this situation of continuous lockdowns for a very long time.
This is the result of another modeling study and what we’re looking at here is: time along the bottom of the graph and the slowly moving up and down green line is the number of cases that need to be referred to intensive care.
The grey blocks are lockdowns, initiated when we exceed a threshold number of people in intensive care beds.
This first graph uses our current effective intensive care capacity and when there are more than a thousand patients in intensive care, the lockdowns are almost continuously triggered; some short breaks in between long periods of lockdown and then, as soon as the virus just starts to take off again, back to lockdown.
There are more optimistic scenarios if we could effectively double the number of ICU beds we have in the UK.
What that would mean is that we could finally get to a point where there’s more time not locked down than locked down but the difference is small; per this graph we’re constantly in and out of lockdown well into 2021.
That’s why it’s so important that we try solve this “smartly” with testing and contact tracing and pushing R0 down through everyone’s efforts.
If we don’t we’re going to be hokey-cokeying in and out of lockdowns, which would be very challenging for all of us, long into the future.
The only way that this graph stops is when we finally reach herd immunity. Whether that’s through repeated bouts of infection meaning more and more of the population is slowly infected so we get herd immunity that way, or because we develop a vaccine which, as we’ve said, could be years away.
We’re clearly going to be in this coronavirus crisis for some time. At the time of recording the UK at least will be locked down for quite some time to come but even as it slowly gets lifted it’s going to require a complicated combination of measures to come together in order to allow us to slowly relax those restrictions and bring the economy and broader society back to “normal” life.
Probably it will be a case of trial and error learning because, although we can look to countries like South Korea that seem to have things under control, we haven’t seen how any country copes with trying to lift lockdowns yet.
Our own mistakes and the mistakes of other countries will provide important lessons, lockdown can potentially be reinstated if things start to fail, and we’ll be hoping that everything comes together.
Even if it’s successful, we’re still going to be looking over our shoulders until not only has a safe vaccine been developed but it’s also been rolled out and administered to a large proportion of the population, which may be years into the future.
Part 4: Reasons for Optimism
Coronavirus is clearly here to stay, which might sound like a depressing thought if you’re already suffering cabin fever in your house, but I think there are some reasons for optimism as we continue to grapple with the crisis.
1. Coronavirus could be worse
The first reason is that coronavirus isn’t as deadly as it could have been.
SARS, a disease also caused by a coronavirus (SARS-CoV-1), had a mortality rate of about 10% and there’s no real biological reason why coronavirus couldn’t be that deadly as well.
We’re just quite fortunate that this pandemic – this really contagious disease – doesn’t seem to be at that level.
2. We’ll be better prepared for the next one
We are going to learn so much scientifically, socially, as governments, assessing all the different ways countries are responding to this coronavirus crisis and it means that we’re going to be better equipped if (or when) another pandemic appears.
This is important precisely because the next one could be much more deadly than this.
In fact, it might not even be a natural virus; experts are concerned about the idea of a genetically engineered virus that bio terrorists could release. Hopefully, by becoming more prepared for pandemics through research and understanding now, building resilience and needed response facilities worldwide, we’ll be more ready to deal with deadlier pandemics in future.
As we can see from our response to the current crisis, there have been serious blind spots for governments.3. We’ll be better prepared for other crises
My final reason for optimism is even broader: I’m really hoping as a result of this crisis, people and governments around the world will realise the importance of listening to the grim predictions of experts.
Scientists and doctors have been warning about the risk of a pandemic spreading from animals to humans due to our increasingly interconnected transportation systems for years (indeed it’s how HIV finally spread from its origin) – for decades now – and yet governments have failed to prepare.
What this shows us is how important it is to listen to those predictions, whether it’s about pandemics, which I very much hope will address, or other broader issues; things like climate change or risks of artificial intelligence or even asteroid strikes.
Because it’s only by huge preparation on the part of governments, by international collaboration, by scientific research, that we can mitigate these enormous risks to our civilization.
I think that this coronavirus crisis is going to set back human progress by at least a year because so much is on ice at the moment; whether the economy or scientific research or tackling the complex problems we face as a civilization from poverty to climate change.
But we’ve got to come out of this swinging – we’ve must use this crisis as the paradigmatic example of why it’s so vitally important to invest in science, to work together in international collaborations and to work on preventing disasters like this ever taking off in the first place.
I hope this video/article has given you more of an evidence-based idea of where we are, how things are going to progress over the next few months and maybe the years, and hopefully… things will change for the better.
For now, this is a long haul for all of us. It’s going to be a difficult crisis to get through.
So: stay safe, stay well, stay at home, and thanks for reading.
Stay at home, save lives.
Here are some of the sources I used when researching this video:
- The ‘Imperial model’ that forecast 500,000 deaths in the UK if we did nothing: https://www.imperial.ac.uk/media/imperial-college/medicine/sph/ide/gida-fellowships/Imperial-College-COVID19-NPI-modelling-16-03-2020.pdf
- How long will immunity last? https://twitter.com/TheMenacheryLab/status/1246865058468114437
- Bill Gates ‘wasting’ money: https://www.businessinsider.com/bill-gates-factories-7-different-vaccines-to-fight-coronavirus-2020-4
- Funding for vaccine development globally is scarily low: https://scienceogram.org/blog/2020/03/coronavirus-covid-vaccine-funding-cepi/
- The case for a human challenge trial: https://www.nature.com/articles/d41586-020-00927-3
- Mobile phone contact tracing: https://unherd.com/2020/04/can-your-smartphone-crack-covid/
- Study investigating the feasibility of contact tracing: https://doi.org/10.1126/science.abb6936
- The science behind #masks4all: https://www.fast.ai/2020/04/13/masks-summary/
- US guidelines for mask-wearing: https://www.cdc.gov/coronavirus/2019-ncov/prevent-getting-sick/cloth-face-cover.html
- And article on making and using your own mask that’s better than the official US guidelines: https://www.self.com/story/homemade-coronavirus-mask
- Great video about why ventilator design is really, really hard: https://www.youtube.com/watch?v=7vLPefHYWpY
- Podcast about the logistics of making ventilators: https://www.npr.org/2020/03/31/824886286/episode-987-the-race-to-make-ventilators
- Why we should try physical distancing now just in case (the same arguments apply to trying test, trace and isolate in the hope that it works): https://theincidentaleconomist.com/wordpress/the-option-preserving-value-of-social-distancing/
- The graphs of repeated UK lockdowns comes from this study: https://cmmid.github.io/topics/covid19/control-measures/uk-scenario-modelling.html