posts ┬╗ COVID-19: personal updates

COVID-19: personal updates

Tonight I've got homemade Chinese food for dinner: delicious (vegan) tofu with black bean sauce. It wasn't that difficult, I just chucked a load of ingredients that I had lying around in the wok together - but this post isn't to provide you with recipes. Instead, I thought I'd give a bit of an update as I've not had so much time to write these past few days, and the world is rapidly getting a crazier place to live. If you previously thought that living under a left-wing dictatorship like you imagine they've got in China was bad, you're gonna have another thing coming when a right wing dictatorship arrives in your beloved western "democracy". Meanwhile, I've been pretty busy recently. So please, a warning: hold your hats on tight, there's a bit of explanation coming up and it's not entirely pleasant reading.

I last wrote on Friday - about Good Things. There are some, but I'm afraid today it's back to a bit of doom and gloom. As I've sat down to write this, the UK has gone into lockdown (finally!) and the Guardian has said the world's largest intensive care unit is currently being created. The United States will be next - well, after South Africa, who're also going into lockdown starting Thursday.

SAMU de Paris

Over the weekend, I managed to get out of the house. I went to Necker Hospital which is where the SAMU de Paris(English description) is based. I have a friend who works there and he'd suggested I come down to visit, to see what their setup is like. I only currently work 30% clinically at the SAMU 93 - plus 100% research time at INSERM doing perinatal epidemiology - so I've got time to do stuff. It also meant I could get out of the house and go for a little bike ride, which enabled me to see the silent streets of Paris. The photo, for example, shows Bastille in full glory, not a person around.

The setup at Necker is pretty crazy. Their "regulation" (where the calls come into) was twice the size of the one in the 93 Department where I work. PLUS they've set up two huge overflow rooms downstairs in what was previously a lecture theatre and a seminar room: these are exclusively to deal with the calls about COVID-19. On a normal day before all this started, they would get about 3000 calls a day to the '15' which is the emergency number for health care (there are separate ones for the police and fire services, although I don't actually know what they are). Towards the end of last week, they were getting up to 12,000 calls a day; it's now settled at a more reasonable number in the range of 7-8000 per day. Everyone of the phone operators and everyone else I met was wearing a face mask and keeping a respectable distance from each other.

As my friend was showing me how it all worked, a call actually came through required attendance so we ran down and jumped the ambulance and headed out. Within a few minutes, we were on the other side of Paris where my friend geared up to go into the apartment to assess the child: someone who had a chronic illness requiring home respiratory support and potentially had COVID-19. As it turned out, the child was well, so my friend was able to come back out within a few minutes, then set about taking off the outer gloves, gown, hat, mask, goggles, and then finally inner set of gloves, using liberal amounts of alcogel at each step to sterilise the gloves and, finally, his hands and face. No one is taking any chances, this disease is serious and it's really killing people.

We headed back to base: it was the end of the shift so we rode back together (although preserving a safe distance of 2m between us) through the silent streets of Paris again. On the way, we got stopped by the police, but as soon as they saw our badges, they waved us through. We also passed a boulangerie with people queueing down the street to get bread - again, respectfully keeping their distance, only entering one by one into the store. Inside, the woman behind the counter was kitted out with mask and gloves so she didn't have to handle the money, although most people are now using contactless payments: another step on the road to all out surveillance.

Queue outside a boulangerie in Paris, March 2020
Boulangerie picture

More modelling

Aside from that, I did a few things over the weekend. First, I tried to do a bit of modelling. The first graph below shows the number of deaths in the France, Germany, Italy and the UK since the first death; the second graph shows the same thing but on a log scale.

Number of deaths by day in France, Germany, Italy and the United Kingdom
Days since the first death

Number of deaths by day using a log scale in France, Germany, Italy and the United Kingdom
Days since the first death, log scale

Using a log scale essentially means that a bigger space is given to earlier events rather than later ones. In this graph, the log scale is on the y-axis, and it means that the curved shape you see in the first graph gets changed into a more-or-less straight line in the second one. We can see that Germany and the UK are following roughly the same track as Italy, although a bit behind. France seems to have done a bit better - it seems they had one death very early, 10 days before the numbers started to increase. However, since then, the line is roughly parallel to Italy.

The problem here, which is nicely illustrated by France, is that the first death doesn't really mean that much - maybe they were "imported" from somewhere else, or it was someone who was particularly fragile because they were very old and/or had lots of other illness and pre-existing conditions as well. This could just be "bad luck" or some kind of statistical fluctuation. Hence, I repeated the graphs, but this time only showing deaths that occurred in each country after they had got to 10 deaths. Those graphs are here:

Number of deaths by day after the first 10 deaths in France, Germany, Italy and the United Kingdom
Days since the 10 deaths

Number of deaths by day after the first 10 deaths using a log scale in France, Germany, Italy and the United Kingdom
Days since 10 deaths, log scale

This is a much more worrying picture, at least for the UK. France and Germany both seem to be following the path of Italy - it's perhaps marginally better in France although there's not much in it. In contrast, the UK is steaming ahead: on the 'normal' (linear) graph, the UK is clearly an outlier to the other countries with more deaths earlier, and on the log graph, the angle seems sharper as if it will go up even quicker than the other countries. At least the country has now gone into lockdown today (there are currently 282 deaths recorded in the data set I am using from Johns Hopkins CSSE, retrieved at 23:40 UTC on 2020-03-23) whereas the Guardian are currently reporting 355. In comparison, France went into lockdown when there were 148 deaths, Italy when there were 463, and China - who managed to get control of the spread - when there were 17. Yes, that's not a typo: seventeen.

I think we in the "free world of the West" are going to be in lockdown for a long time, people.

Some explanations

First, why have I just looked at these four countries? Why did I not look at the USA or Australia or Cuba or Russia or somewhere in Africa, for example? The simple reason is that the UK and France are the two countries that I (currently) have the greatest affiliation to. Italy is ahead of all of us, being one of the first places outside of China to be affected, and thus have lessons to teach us, and Germany have appeared to be doing well - so far.

Furthermore, the populations are roughly similar in size - around about 60-70 million people. There are some pretty important differences in geography, though: France is a much more rural country than the UK, for example. If we wanted to look at the USA, we'd be lookign at something else completely: there are lots of pretty dense urban areas, but also a whole load of space in the middle. The way that disease spreads is therefore very different.

And, why deaths? This is also a relatively simple question to answer: deaths are generally a pretty easy diagnosis to make, and they are also pretty easy to count. There's very little ambiguity, unlike much of the problems that we've seen related to diagnosis of SARS-NCov-2 infection (the coronavirus that causes the COVID-19 disease). Of course, there are debates to be had. Did this person die from the infection, or did they just have the heart attack they were going to have anyway? Indeed, these are the sorts of debates that I'm beginning to see in academic circles; what remains important at the moment is that we need to deal with the clinical problem of treating all the people who are currently getting sick.

Final question for now: how much longer is this going to go on?

Unfortunately, my response to this one is also not very nice: a long time. China locked down when they had 17 deaths - on the 23 January 2020. Two months later, they are just starting to consider lifting some of those restrictions. And this virus is persistant: a recent article in the New England Journal of Medicine described how it can persist for up to several days, depending on where it lands. There was a good explanation about this and why soap may be preferable to more toxic cleaners such as bleach that was published in National Geographic if you want to read more. My current guess for how long lockdown will last is anywhere from 12 to 24 months in one form or another. It may not be quite as stringent all the time, but some form of lockdown is going to be required - at least until we know more and (I hope) we're able to develop a vaccine and scale up production for everyone in the world. That's a lot of people.