Blog

These blogs were written and posted on the Facebook-Page "https://www.facebook.com/tk.corona.updates/" which I created during the early stages of the Covid19 pandemic in order to share some information and analysis on the Corona data (cases and deaths) worldwide and in a few selected countries (US, Italy, Germany, UK, Spain, Canada).


Coronavirus 5: Reopening         April 22, 2020 

Coronavirus 4: Irrational human thinking       March 23, 2020

Coronavirus 3: Fear is killing you     March 15, 2020

Coronavirus 2: Comparison of covid19 spread in different countries       March 9, 2020

Coronavirus 1: A 5-min read that can hopefully save some lives             March 5, 2020

    Florence, April 22, 2020

Coronavirus 5: Reopening   


Why reopening (if not done absolutely right) is so very risky:

What people should really understand about the potential second wave (that you hear so much about now) is that if it comes it will probably be much worse than the first one. The reason is the following: The first one started in one place and slowly spread from there. However, if we are not careful enough the second wave will start in many different places at the same time (everywhere where there are already infected people that start to mingle again with other people). This is exactly what happened with the Spanish flu in 1918 when its second wave at the end of World War I killed more people than had died during the Great War itself.

With Corona, as we know by now, the first wave brought the virus basically everywhere in the world. The containment measures put into place have brought the spread more or less under control (and that’s a big “more or less” actually). But in most places there are still quite a lot of new cases so the virus is certainly alive and active.

So to prevent the second wave after reopening, the reproduction factor R needs to be kept at bay. The difference between R<~1 (one person infects on average one other person or five persons infect five persons) and R~1.2 (five persons affect six persons which means four of them still infect one and just one of them infects two) does not sound like a lot but the effect on the population if this happens everywhere at the same time could be devastating. Instead of the virus slowly fading away, we could end up with a vast number of small epidemic waves and the superposition of all these small waves would be one huge wave that is everywhere at the same time.

That’s why it is so important to have two things in place before reopening:

- Large-scale testing and contact tracing are needed to dampen each individual new wave right at the start. Everyone who had been in contact with an infected person and is thus at risk of an infection needs to be put in quarantine as quickly as possible.

- All of us individually have to keep on doing everything we can to reduce further transmissions. Don’t be that fifth person who (unknowingly) infects two other people. The best way to prevent this from happening is to continue social distancing and to wear a mask whenever you are in public. So even if your country is reopening (hopefully slowly), please stay vigilant and at least initially use the new freedoms sparingly.

The first point can be controlled in some way. This is mostly up to your government (and there are surely some that you can trust in this and some that you can’t). But reading the comments on social media and seeing the “Open up the economy”-protests going on in so many places, I really have my doubt about the willingness of the people to adhere to the second point. And the worrying thing is that we really need everyone to contribute in order for this to work. Otherwise it will again be mostly the elderly, the sick and the poor as well as the health workers that will pay the prize for all of us…

Florence, March 23, 2020


Coronavirus 4: Irrational human thinking



Quite a long post but please bear with me. The relevance to the Corona epidemic will become clear very soon:


In psychology there is a long line of research on biases/heuristics underlying irrational human thinking. In most cases it can be understood where these biases come from (in an evolutionary sense) and that more often than not they still are good for you even in our fast-changing modern times. But sometimes they can severely cloud your judgment and lead to very bad decision-making.

Here is one example, just slightly simplified from the Wikipedia description (*):

“The availability heuristic, also known as availability bias, is a mental shortcut that relies on immediate examples that come to a person's mind when evaluating a specific decision. Under the availability heuristic, people tend to heavily weigh their judgments toward more recent information.”

The main error that is often made here is that we use availability (ease of recall) to estimate frequency. That’s one major reason why the danger from terrorist attacks is typically vastly overestimated, the more so the more recent the last prominent attack. Once time goes by it is often rather quickly forgotten. The same happens for example in the US with calls for gun restriction. They are very loud after each new mass shooting but then often fade away quietly.


Many of these biases were first discussed by Tversky and Kahnemann (for a good account check out Kahnemann’s book “Thinking – fast and slow”). A more recent exploration can be found in the book “The undoing project” by Michael Lewis. There the discussion often goes a step further, Lewis talks a lot about ways to overcome these biases.

For example, a whole chapter is devoted to Daryl Morey, the general manager of the basketball team “Houston Rockets” and his attempts to make better draft decisions by relying on objective and more unbiased data (the “Moneyball” approach, also popularized by Micheal Lewis). This is mostly about finding the right metrics to characterize players and to explicitly avoid bias-driven pitfalls. The people making the decisions actually watched less basketball of players they were interested in instead of more, in order to not let their judgment be clouded too much. For example otherwise they might become too impressed by the latest spectacular dunk of a star player but miss the small positive intangibles of a “glue guy” that could be even more valuable for their team. But if they detach themselves from the highlight reels and look at the right metrics they expect to find these often more relevant differences better represented in the data

The point here is that these biases exist in every one of use but there are things that can be done in order to overcome them. But for this to happen the very first and probably the most essential step is to become aware of these biases.


So while observing the unfolding Corona situation all over the world, I was wondering whether similar “unbiasing” approaches could be applied to the decision making of politicians. It seems quite obvious that apart from a few exceptions (South Korea, Taiwan, …) the same bad things seems to happen in one country/region after another (the word “lemmings” came to my mind, after the well-known misconception about lemmings committing mass suicide). People see what’s going on far away, they might start to worry a bit but they never really take any decisive until it is already too late.


I think there are two main biases that would need to be overcome:

1. It is completely human to vastly underestimate exponential growth. I guess that apart from very few exceptions the ancient world as perceived by people living at the time was mostly linear. Observable exponential growth is probably a rather modern phenomenon (due to the acceleration of life, overpopulation, larger networks with higher connectivity etc.).

In these days this underestimation still happens a lot to me even though I am a physicist and think that I am absolutely aware of the problem. But during this corona pandemic I can’t count anymore the number of predictions for the next week or two that I made openly or in my mind that were then completely obsolete after just a couple of days.

As with the biases discussed above, in most cases the linear worldview is completely fine. However, we seem not to be aware that it becomes a big problem in the situation we are in now when we are dealing with a truly deadly combination: very short delays between the exponential curves of different countries and a disease with a rather long incubation period. The latter basically acts as a negative delay (or anticipation if you will) so it reduces the time that you have considerably. The overall effect is that every country only has a very short time window to act and unfortunately at the moment this window is wasted in one country/region after another.


2.    The second effect is also understandable. We only care about what is close to us and where we feel some kind of personal connection to. Again, this worked very well in the ancient world. All the dangers that you might face were nearby. Probably you didn’t even hear about the catastrophes that happened just a short distance away.

Even in our modern world this mostly makes sense. Given the onslaught of bad news (availability bias!) on television, the internet etc. it would just be too overwhelming. You somehow have to pick your worries (and hopefully causes), otherwise mentally you would not survive for long. Again, this happened to me as well. As far as I’m aware I don’t know any single person living in China personally (I know some Chinese people but they do not live in China). Accordingly, for the first month or so I heard the news about this new virus in China but I was not yet overly concerned. But that changed rather quickly once it made its first appearance in Italy.

Under usual circumstances all quite reasonable but we should be aware (and seemingly we are not) that a fast-spreading epidemic like this one completely redefines our concept of closeness. Due to globalization, increased connectivity and modern means of transport (equivalent to shrinking distances!), in situations like this the whole world becomes close. What happens in China today can have a big influence on your place in the world tomorrow (note that while I do mention China a lot I’m not calling this the Chinese virus).

So here again we need mechanisms that makes us worry (and act!) earlier and not when it is already too late.


As mentioned above, I tried to think of ways to overcome these biases that make everyone act the way they do. Please note that I’m not blaming anyone, as mentioned repeatedly I’m as affected from these biases as anyone else; on the other hand this is also not to say that there aren’t some really bad actors around who are in it for political or personal gain. Anyway,  these would be my conclusions:

People in power just can not continue to make decisions only based on what they observe around them, they have to start acting from the perspective of the leader of a country that is already affected big time. That should be someone who can already observe the very steep parts of exponential growth right in front of his eyes (1) and for whom it has already become very, very personal (2). For an illustration have a look at the latest numbers in Italy (1) and look at some of the pictures in Italy’s northern regions and imagine this to be someone you love, for example a parent or a grandparent (2).

Remember exponentials keep on getting steeper and steeper, at least until the measures that were taken start having an effect, and this can be very late. So the aim has to be to stay ahead of the curve, not always just run behind, and also to act way before it does become personal for you. Otherwise this could result in a deep regret that you would have to carry with you for the rest of your life.


In one of my previous posts I tried to make use of the delay between different countries by providing this advice:

“Every citizen in every country should try to voluntarily adapt the countermeasures of the country that is a week ahead on the epidemiological curve.”

Now I would like to go one step further:

“The decision about which measures should be carried out in a country should be taken (or in more practical terms at least be strongly influenced) by a committee of experts from a country that is as far as possible ahead on the curve.”


The reasoning behind this is simple: If you could give Italian politicians and medical experts the chance to go back in time and decide again about which measures to implement at what time in their own country they would very gladly take it and do everything they possible can to prevent or at least soften the course of the tragic events that are unfolding now - and unfortunately will still unfold in the future. And I’m certain the same is true for its people. Knowing what they know now they all would gladly go back and follow these rules and not complain anymore about media hype, unnecessary restrictions etc.

From Italy it is truly heartbreaking to observe how one country/region after another gets our imaginary second chance as first chance (but once and only once!) and then throws it away.  For me this is very personal, since my father (due to his advanced age certainly among the ones most at risk) lives in Germany and there they are certainly not doing enough.

This is probably the main reason why I keep on writing all these posts. By sheer circumstance I happen to be in the country where the “second epidemic wave” started that has by now turned into a global pandemic. Unfortunately Italy is and will be for a while at the forefront of things to come. It probably will stay “ahead on the curve” for a while (soon to be joined by the US) and what can be learned and experienced here should be put to much more use in order to help other countries where such things could still make a vast difference. So people living in Italy have the chance and maybe even the duty to first of all tell the world about what is going on here (and the world really should listen). We can show how it can be done or at least give the opportunity to learn from our mistakes. So if you live here, please do as much outreach as possible. Small changes that are made here with quite limited effect now can turn into very big changes when applied to other countries that are still early.

Italy itself has done first steps in this direction and invited experts from China to monitor the situation and provide advice. More and more countries should follow this lead and be more willing to accept good advice from countries such as China, Taiwan, as well as Italy and Spain. Preferably they should get most of their expertise from those countries that managed to curb the spread rather well and the lessons learned and (probably most important) the sense of urgency from those countries that didn’t.


* While writing this post I checked the Wikipedia page on cognitive biases and if you need some distraction in these stressful and frightening times there’s plenty of it there. For example, I had no idea that the “Women are wonderful effect” is a thing - but now that I thought about it, I am certainly affected   ;-)

  Florence, March 15, 2020


Coronavirus 3: Fear is killing you


In the US posts like this one seem to make the rounds (a FB-friend of mine just shared this). This is very dangerous because it prevents people from taking the necessary actions. So here I would like to offer a few responses:

- Fear is the mechanism that evolution uses to protect you. Think of fire. Or another example: It is absolutely no coincidence that the animals we are most afraid of (snakes, spiders, etc.) are the ones that can be poisonous. Often the early humans that were not afraid of these animals did not survive. Each single one of us living today is a descendant of an uninterrupted chain of people that were afraid enough to survive at least until they were able to procreate.

- In most of the cases mentioned in this post it was exactly fear that made people act and then prevent these dangers from becoming catastrophes in the first place. For example, for the very reason that ISIS provoked terror other countries built a coalition to stop them. So fear did exactly the job it was designed to do.

- If you compare the numbers, countries such as Taiwan, Hong Kong, Singapore, South Korea (maybe Japan and now China but in these cases there are other considerations as well) vastly outperform other countries in stemming the acceleration of the spread. These are the countries that have already been affected by a previous epidemic (in most cases SARS) and thus were so afraid that this could happen again that they put procedures into place to effectively deal with such situations. They were just better prepared and this was because of fear.

- If one still cannot see that the Corona epidemics is on a different scale than some of the other viruses mentioned in the post then there’s no help anymore. The response should be proportional to the risk and in this case this means that extraordinary measures need to be taken as soon as and in as many places as possible.

[ The fact that evolution cannot keep up with a rapidly changing environment and some rational fears can turn into irrational phobias is an entirely different debate. So I'm not saying that fear is always a good thing (and I'm speaking of personal experience here). There are many situations where Roosevelt's saying "The only thing we have to fear is fear itself" can be perfectly true. Of course there can be overreaction and unnecessary panic in the world. But this current Coronavirus epidemic is not one of those situations! ]

Florence, March 9, 2020


Coronavirus 2: Comparison of covid19 spread in different countries  


(Source: Jean-Baptiste Mouret, INRIA, France; his original Facebook post;

   Data from https://github.com/CSSEGISandData/COVID-19)


A very good graphical representation of what I was trying to convey in words in my previous post. This is what it looks like when the epidemiological curves of different countries are aligned in time.


First note the acceleration of the spread in all these curves (it's getting steeper and steeper with time).

Note also that there are some differences between the curves. It's probably too early to speculate about the underlying reasons but influencing factors could include variances in

- Demographics (e.g. more or less elder people)

- Population density, connectivity within the population (typical housing situation, average social life etc.)

- Number of tests administered (Accuracy of the data)

- Resources and efficiency within the health system

- Time to prepare for the epidemic (!)

- Measures taken and their effectiveness (!)

The countries where the virus arrived later had more time to prepare. And for them the future course of the curve is still to be decided. So let's not waste this opportunity. Taking and adhering to measures that prevent the spread can make all the difference!!!

      Florence, March 5, 2020


Coronavirus 1: A 5-min read that hopefully can save some lives.          

Short summary:

Please all remain calm. No need to worry too much. But also: Please worry just enough so that you take all the necessary precautions against the coronavirus. If you are not overly concerned for yourself because you're young and healthy, that's good. However, please do not forget those with pre-existing conditions, the sick and the elderly, since they are the ones that are most at risk. This includes some children but maybe also your parents or your grandparents. For them the extra years that they may or may not get to live make all the difference in the world.     

   

The following post on the coronavirus and its spread was written from the perspective of someone living in Florence, Italy, on March 5, 2020. However, given the current situation and outlook everywhere many of the underlying ideas should be quite generalisable to other places and times. I just posted this on the local Facebook page “Foreigners in Florence”, mostly in reply to local people (understandably) urging people to not give up on visiting Italy (in particular Tuscany):

     

As a scientist with some knowledge about the spread of epidemics I would like to offer some explanations and thoughts about the current situation regarding the Coronavirus. In particular I would like to explain the reasoning behind some of the measures being taken right now and even more to encourage everyone to follow them as much as possible.

In order to bring my points across I need to be a bit alarmist now and then but I really think this is necessary (“better safe than sorry”). I tried to be as understandable as possible, but if you have questions or comments I would be happy to either address them myself (if I can) or refer you to someone who can (a website or a person). Also please feel free to share this. I would be even happier if someone could translate this to Italian. I do speak the language but it would be much clearer if a native speaker could do the translation. The more people start to act the better for all of us.

     

Epidemics tend to spread exponentially and this is currently true for the coronavirus as well. What this means is that for now every person who has the coronavirus infects on average between 2 and 3 other persons. The current best estimate for this number is 2.2 but for simplicity let’s assume it is 2. This means the first sick person in Italy infected 2 persons, these 2 infected 4, those 8, then 16 and so on.

One can see where this leads, both the number of new cases and the overall number of cases increase with bigger and bigger jumps, so there is not only further and further spread but the spread actually accelerates. That’s exactly what is happening now all over the world. For example, it took Italy about two weeks to go from the first case to 1500 cases but it took just three days to go from 1500 cases to 3000 cases.


For now unfortunately this acceleration will continue. We are clearly on the rising part of the curve (“it will first get worse before it will get better”). In order for the epidemic to die out as quickly as possible we need to press this number mentioned above from 2.2 to below one (which would mean that over time both the number of new cases and the overall number of cases would fade away).

There are various ways to do that, for example large-scale vaccination. This would decrease the number of susceptible people in the population until at some point the virus can’t find any new hosts and slowly fades away. Unfortunately, current estimates expect the availability of a corona-vaccine to be about a year away (first the development itself, then many tests for side effects etc.).

So the next best thing and at the moment our only viable solution is to slow down the spread. For this people need to finally take the epidemic seriously and behave accordingly (i.e. follow medical and governmental advice, avoid large crowds, wash hands etc., basically do everything possible to prevent a further spread of the virus). The earlier and the more we do that, the less likely are new infections, and the faster we reach that magic number of less than 1 new infection for each existing infection.   

    

What makes the whole situation slightly worse is that the numbers we hear are the number of identified cases and for this there are two things to consider: the quite long incubation period (there is a delay of about 5 but up to 14 days from getting the disease to showing first symptoms) and the fact that even after the incubation period 80% of cases are quite mild. Both of this implies that a lot of sick people are not getting tested and thus do not show up in the statistics.

So at each point in time there are many more unidentified than identified cases in the population. For the current situation this means that unfortunately there are already a large number of infected people walking around Tuscany and inadvertently spreading the virus. This again should emphasise the urgency of the situation. So please start to act now and do not wait until we reach the numbers of Lombardy and other similarly affected regions. The only difference between them and us is that the virus got there earlier than here.

As mentioned above, the often repeated argument that for about 80% of cases the Coronavirus is basically like a regular flu is true. But there are two replies to this: First, for the other 20% it is not like the common flu at all and unfortunately it can get really bad. And, second and more importantly: if you're young and healthy and do not have a lot to fear, good for you! But please have the basic human decency to consider also the fate of those not as fortunate as you, in particular people with pre-existing conditions (and this can include children) and the sick and the elderly who are most at risk. For them the extra years that they might or might not enjoy make all the difference in the world. So maybe think of your parents and grandparents first.

  

Even if the epidemic can not be stopped anymore (and indeed for Italy, and probably the planet as a whole, this point seems to have passed) slowing it down is absolutely essential because it means more time for the health services to react and thus better chances of appropriate treatment for everyone that will be affected. So again, the earlier people react the better.

In this worst case of a largely undiminished spreading what should happen eventually is that more and more people will get the virus at some point or another and either recover (the vast majority of us) or pass away (about 0.5-2% *).

In the positive case that recovery would provide immunity (which unfortunately for Corona can not yet be confirmed) this would actually be very similar to the vaccination scenario described above. In fact, vaccinations are nothing else than deliberate inductions of very mild cases of the disease. But the difference to the vaccination scenario would of course be that in this worst case many people would have died, so we should really try to do everything in our power to not let it come to that.


A lot of us are in some way negatively affected by this epidemics. The sudden drop in people visiting Tuscany has devastating consequences since many jobs and incomes depend on tourism. This is obviously very unfortunate and we should do everything possible to alleviate the situation.

But as I try to explain in this post, the solution should not be just to ignore what is happening around us with the virus and to insist on going on as before. For sure we are already at the point where any unnecessary travel should be discouraged. It just contributes to a faster spreading of the virus.

So instead, better try to find other ways to help each other: Support businesses in Florence and Tuscany in general (in particular buy locally, don’t use Amazon), look out for people in need. If you want to go even further, please spend a week in a nice spa hotel in Florence or visit an agriturismo in the beautiful countryside nearby, hire a local guide etc., in short try to replace the missing tourists from far away.

If everyone everywhere would do that, we would have more winners or at least not that many losers from this situation and without any of the grave consequences of excessive travel. If the worst happens, seek help, self-isolate or help people that self-isolate (someone healthy needs to do their shopping etc.).


Currently Tuscany is about two weeks behind Lombardy on the curve. This statement actually means that inviting people to come to Tuscany in two weeks is the same as sending them to Lombardy now. Would one really want to do that?

Or think about China. What is happening there right now could also happen here in the future (hopefully not but how can we be sure?). In fact, the relative numbers of cases are already higher in Italy (Italy 6.4 cases, China 5.6 cases per 100.000 people **).

But since now we’re talking about the future, it’s still in our hands. The one thing to prevent us from reaching the high numbers of Lombardy pretty soon (or of China in the more distant future) is to start acting now so that the epidemic curve in Tuscany has a milder (less steep) course than those ones. We actually have the advantage that we have been warned, so we should not ignore these warnings.


Finally, I would like to stress that there's no need to panic (and in particular don’t take this text as an appeal to buy masks, deplete supermarkets or act in a xenophobic manner towards anyone). Life goes on and hopefully people still can enjoy many of its brighter sides (https://www.youtube.com/watch?v=jHPOzQzk9Qo).

However, it's completely reasonable to worry and take precautions. Extraordinary situations require extraordinary measures, at least temporarily. The point about closing the schools (which seems to anger many of my friends and colleagues) is that the risk of children bringing the virus home from school is right now just getting so big that we shouldn't take that risk anymore.

The children would almost certainly be fine but they would endanger their older relatives. That would be especially true here in Italy (particularly the south) where many grandparents live in the same household as their grandchildren.

In epidemics like this one, instead of doing nothing it is certainly better to prepare for the worst and be pleasantly surprised if things turn out better. And there certainly is still hope for some more optimistic outcomes. But it really depends on how each of us deals with the situation. The earlier and the better we react now, the sooner we can all get back to normal.

   

* According to current estimates the mortality rate is about 2% for all but around 15% for people 80 and older (both numbers should decrease considerably when also more mild cases are taken into account, current estimates for the final number are close to 0.5%).


** Italy: 6.4 cases per 100.000 people (as of March 5:    3858 / 60.4 million)

China: 5.6 cases per 100.000 people  (as of March 5:  80430 / 1.435 billion)

   

    

Some references for further information:

This is where I tracked the number of cases in Italy, where I got the “80% of all cases are mild” number, the estimate for the incubation period and also the mortality estimate of 2% (for comparison, for the common flu this value is 1.3). The estimate for the expected final value of 0.5% is from a calculation of Prof. Craig Feyed, a well-known epidemiologist who used the “Diamond Princess” cruise ship in Japan as starting point.


This is where I found the number 2.2 for the estimated number of new infections per infected person.


This is an understandable description of a basic epidemiological model. The green curve in the graph shows a typical behaviour for the number of infections in an epidemic in which catching the disease and recovering provides immunity. As mentioned above it is not yet clear if this is the case for Corona (also for the flu immunity only lasts a certain time and then you can catch it again). So no one can know exactly how the Corona curve will develop in the future but it is probably safe to say that currently we are still rather at the beginning of the rising part.

     

Three disclaimers:

1. I am not a medical doctor so I cannot give any medical advice, please look elsewhere for this, for example here:

In case of (suspected) infection, of course better speak to a doctor as soon as possible (preferably first by telephone, do not just go to the hospital).

2. The whole text is written after quite extensive research and to the best of my (current) knowledge. But please refer to official sites for more in-depth knowledge and the latest updates.

3. I’m aware that this is a controversial debate and many people speak about hysterics, fear-mongering and spreading panic but this should not prevent us from a reasonable discussion based on facts and careful extrapolations. If someone has evidence-based counter-arguments to some of my thoughts I am very open to listen and would potentially acknowledge that you have convinced me of a different point of view. Then this would be my new recommendation based on my then best knowledge. This is how science works, recommendations get updated based on new data and information coming in.