I am not an expert on the coronavirus at all. But like many of you the topic has been distracting me lately. In this post I give my personal perspective and propose that we could use web apps, mail-ordered tests, open data and concert bands to aid our response.
The experts know what they are doing.
I’ve noticed online that many journalists and bloggers seem to think it is their duty to emphasize horror scenarios and demand stronger action. I think that’s counterproductive and cooler heads should prevail. I find it reassuring that experts have been preparing for decades for another zoonotic outbreak.
Zoonotic means that a virus jumps from animals to humans (often when we eat their meat) and that’s problematic because we have no pre-existing immunity to such an animal virus. This kills 2.7 million people per year on average. This might surprise you, but let me give some famous examples:
- The 2002-2004 SARS outbreak probably came from people butchering civet cats sold on animal markets in Guangdong (China). SARS is so closely related to what we now informally call “the coronavirus” that the official name of the current coronavirus is SARS-CoV-2.
- HIV jumped from chimpanzees to humans in the 1930s.
- The 2009 swine flu killed 400,000 people and came from pigs. (An earlier outbreak of the same virus killed 50 million people in 1918.)
- Ebola most likely comes from people eating bushmeat in Africa.
- Creutzfeldt-Jacob probably comes from eating meat from cows infected with BSE (bovine spongiform encephalopathy aka mad cow disease).
The 2012 MERS coronavirus outbreak came from contact with camels (e.g. eating their meat or drinking their milk or urine).
Experts are not unprepared
So although this outbreak is bad, the experts are not unprepared and I think we should put some trust in them. But of course, it’s always good to learn and if you want to understand more, I can offer some pointers that I found helpful to me as a novice. I think the report from the Imperial College COVID-19 Response Team arguing for a robust package including intermittent social distancing and intermittent school closures until the vaccine is found, looks impressive.
However, the Science Advisory Group for Emergencies (aptly abbreviated SAGE) who are arguing that we need to target herd immunity is not exactly a group of novices either. And here is a direct exchange on the subject between two other experts. At the end of the day, this Twitter thread by an expert resonated most with me.
My personal perspective
In 2018, almost 10,000 people died of the flu in the Netherlands (I guess 150 per day during the peak) and the numbers where similar across several other years. And I didn’t even notice. I thought that was simply part of life. Of course, I’m not saying that COVID-19 is like any ordinary flu because as we all know, we don’t have any immunity against it yet and it will probably spread to 60% to 80% of the population before it dies out naturally. And it would be awfully nice to keep casualties to a minimum until a vaccine arrives. But I do wonder how far we should go when it comes to surrendering our life to this virus.
That is not because I don’t care about the elderly. On the contrary. As I write this, I’m thinking about my mother who I loved dearly. During the last years of her life, my mother often wondered whether it was really a blessing that the medical profession could keep her alive despite her increasing deafness and all her other debilitating and painful failings. Sometimes she said when I visited her: “This was such a wonderful day. Wouldn’t it be great if I didn’t wake up tomorrow morning?” And all I could do was hug her. Thinking back, I think she might have chosen physical contact over safety.
That said, everybody agrees we need more testing and we need to flatten the curve. Here are some underutilized ideas that might contribute something.
Testing: There is an app for that
Companies make you fill out web forms because it saves them time. It’s an extremely efficient way for them to gather information. Where is the app that gathers information from all of us about anything that concerns the coronavirus?
I could imagine a survey that helps people self-diagnose if they’ve got COVID-19. It would be unreliable compared to a real test. Nevertheless, if you aggregated that information and applied machine learning it could tell you a lot about the spread and impact of the virus. And in these dire times the number of people who would cooperate is bound to be huge. It might even make people feel good because it gives them something useful to do.
You could also use it as an intake form to make it easier to treat people remotely (combined with the test described below) or simply to minimize the time spent at a hospital if you choose to test them there.
Who is tasked with building that app? (Shall I help? I used to manage ICT projects and would be willing to lead the project for free.)
Distribute mail order ‘pregnancy tests’
Medical doctors are learned individuals whose decisions can determine life and death. We revere them and they try to be careful and not make mistakes. And that’s good. But it can also lead to erring on the side of caution and reticence in asking for help. And as far as testing for coronavirus is concerned, we don’t need caution. We need speed!
Why are these tests still done at a hospital when that takes up too much valuable time of the nurses and helps spread the disease? Many experts claim that we could soon have tests that are simple to do and can be done in 20 minutes. Another claimed it could work more like a pregnancy test, in the sense that you can mail order them and do the test yourself. Says another: “it’s really just logistics, which can be solved.” This is not only a lot cheaper and easier to scale up, but also safer since someone with symptoms won’t end up spreading the disease to the hospital wards.
What I would like to see is a big X prize like a bonus for the first company who comes up with that test, a logistics initiative that is started in parallel with this, and a law that would make it possible to implement the test right away. I understand that’s not how we usually do this, but speed seems to be of the essence here.
Use open data and data science
I was glad that some scientists at my Eindhoven University of Technology are trying to put their data science skills to good use in order to predict the number of COVID-19 cases in various countries a few days in advance. But that’s scratching the surface. There should already be an international data repository that anybody can use. Where all (anonymized) data is pooled, so scientists worldwide can try to add their brains to this challenge.
There are many machine learning competitions and given the right dataset, some young guys and gals could come up with amazing results. Machine learning is very good at learning from large piles of poorly understood data. I’m sure it could tell you a lot about what to plan for and who to test. For example: during what period are people with symptoms most infective? Who do they infect? At what kind of places and through what forms of interactions? What kind of weather? What age and demographic gets infected but doesn’t know it? I’m not saying we can make the disease go away or develop tests or vaccines using data and machine learning. But we can learn an awful lot about who gets it (or will get it) and when. We can even see which policies work.
Right now, the available data is abysmal. The CDC in the US removed the daily number of people tested from it’s website ‘to avoid a mismatch with state numbers’. The Dutch RIVM (National Institute for Public Health and the Environment) isn’t tell us anything about the test numbers at all as far as I know. China is manipulating it’s performance like crazy. (You only have to look at that gigantic spike of COVID-19 cases on February 12 to know that.) So let’s stop the red tape and add open data and machine learning to the arsenal we use in our battle against the coronavirus.
Give people who are immune a concert to go to and put them to work
It’s really important to also use tests that find out who has recovered from the disease. This can be done with new tests that look for antibodies. Of course, that in itself is important as a source of information. The Netherlands is planning to do a weekly test on the blood of 10,000 regular blood donors, just to see how many people get infected and how immunity evolves over time.
But you could also use this test to find people that are immune and put them to good use. Doctors and nurses who have recovered from it should be the ones with the most client contacts. Young people who have recovered could bring groceries to the elderly. What else could we do with an ever-increasing group of immune hosts in a time that requires social distancing from everyone else? And wouldn’t you like to know that you are safe now for other people and won’t need to avoid parties anymore?
Make it a public Moon Shots Program
The Apollo program was so successful because it was chock-full of bright young people sharing the same inspiring goal. I think the fight against the coronavirus is ideally suited for a program like that. And in these modern times, we could harness the internet as a force for good. One that can efficiently utilize and coordinate the goodwill in society so as to make this program more successful. From EV companies quickly producing medical equipment, to Italian engineers 3D printing respirator parts for free.
And that’s my 2 cents on the coronavirus. Tomorrow, it’s back to electric vehicles and renewable energy.