Like many of you, I’ve been having conversations with friends and colleagues about COVID-19 (aka Coronavirus/SARS-CoV-2) and how–if at all–we should be preparing for it.
My position: you should be prepared to spend at least 2 weeks isolated in your home, and should take other precautionary measures like buying a mask.
Many denounce this position as paranoid and ill-informed, as the disease doesn’t yet show signs of high mortality, particularly for someone like me who is young and in good health.
But not only do I believe this position to be the right one from a risk standpoint, I think I, and others in a similar position, have an obligation to prepare.
Disaster Prepping is a Cost/Benefit Analysis
Before we get to specifics of COVID-19, let’s be clear: all disaster preparation is a cost/benefit analysis.
There is some cost–in time, money, or otherwise–to preparing for a disaster or potential negative event.
We accept many of these costs every day–when we wear seatbelts (effort), when we upgrade to a safer car (cost), when we walk a little slower on icy sidewalks–because we evaluate the cost/benefit ratio to be worthwhile.
For some things, like wearing a seatbelt, this is influenced by both sides of the equation: the data is clear about the increased risk of collisions when we don’t wear a seatbelt, and given that seatbelts are standard on every car, and the effort required to wear one is very low (and there is a penalty if caught without one), the cost/benefit ratio makes sense.
Whenever we are faced with a new potential danger in life, we consider the cost/benefit ratio, and decide whether to take action or not.
We Are Bad at Prediction
Nassim Taleb shook the establishment when he published Fooled by Randomness, The Black Swan, and the following books in his Incerto, as he showed that we consistently fall victim to the narrative fallacy, and are terrible at predicting the future, particularly when it comes to low-probability, high-consequence events.
He illustrated this with the example of the turkey problem: a turkey is fed for 1000 days, living a life of relative ease, each day becoming more confident that his butcher master loves turkeys and that his life has little risk. But of course Thanksgiving comes around on day 1001, and the turkey is no longer safe.
By definition, if we were able to predict future events with certainty, there would be little risk in our world. That this is not the case should be obvious. That we are poor at prediction is not always obvious, but one need merely look at the financial crisis or geopolitics to know this is the case.
In Taleb’s words: “Prediction requires knowing about technologies that will be discovered in the future. But that very knowledge would almost automatically allow us to start developing those technologies right away. Ergo, we do not know what we will know.” - The Black Swan
So, we are bad at prediction...does that matter?
Small Probability Does Not Mean Small Consequences
Indeed, often unlikely events have the most drastic consequences, and humans are particularly bad at estimating probability and severity in the “tails”–the areas of extremely low probability on a distribution.
Much of Taleb’s argument centers around the fact that it is actually these improbable events–the “tail risks”–which have an outsize effect on the outcome.
Examples of potential tail risks include cataclysmic disasters like potent viruses and nuclear war. It’s hard to estimate the probabilities of these events, and imagine the full consequences; there are a lot of unknowns.
Further, is it really possible to calculate the risk when there is a chance of death?
Clearly, as we make that choice every day by stepping outside, and stepping into cars, etc.
But typical methods of probability estimation often break down in cases where death is a significant possibility.
Taleb uses the example of calculating the expected probable outcome of a $1M game of Russian roulette. In this scenario, your expected chance of winning is ⅚, or 83.3%, and so your “expected value” as calculated by an economist is $833,333–clearly you should always play!
But most of us would choose not to, because the lower probability of dying is still too high for us to accept.
As Taleb says, “...if you keep playing Russian roulette, you will end up in the cemetery. Your expected return is...not computable.”
COVID-19 Still Has Many Unknowns
How does this relate to COVID-19?
Well, first of all, death is a legitimate concern. We’ll talk about how likely it is for different populations, but the mortality rate is not negligible.
Second of all, there are still too many unknowns about COVID-19 to do much of an analysis at all.
The fatality rate is rumored to be around 2%, but it varies by population, and many of the cited numbers are questionable due to their source.
Much of the early data fit an almost perfect quadratic curve, which doesn’t align with the estimated R0 of this virus and others that are similar (the rate at which it infects others).
The current estimated fatality rate is higher than seasonal flu, which already kills thousands of people per year (~60K/yr in the US).
The point here, however, is not the specific data that already exists; it is that we are unlikely to have sufficient, reliable data before the point at which we need to be prepared.
The virus is spreading too quickly, and so waiting for more data before making a decision about severity is not possible.
This doesn’t even include other possibilities, like virus mutation, which only add to the unknown risks.
What Do We Know?
The spread of this disease has so far been rapid enough that it is likely to spread across the world.
From The Economist:
“It has become clear in the past week that the new viral disease, covid-19, which struck China at the start of December will spread around the world. Many governments have been signalling that they will stop the disease. Instead, they need to start preparing people for the onslaught”
From The Prepared:
“Our baseline scenario — what we feel is most likely to unfold — is that the virus will spread person-to-person in a sustained way globally, not only in Asia, the Middle East, and North Africa, but also in the US or Europe. We’re expecting disruption to daily life at some point here in the US, and in most places worldwide.”
From Elad Gil:
“Despite the WHO’s assurances that things can still be contained, every epidemiologist I have spoken to thinks the virus has broken out and will spread around the world. Many think >20% or more of humanity will be infected due to a lack of baseline immunity and therefore herd immunity for this disease (as an example, the 2009 H1N1 flu infected 16% of all humans)”
There is reason to believe that this is one of the most severe pandemics we’ve seen so far.
“Global health experts have been saying for years that another pandemic whose speed and severity rivaled those of the 1918 influenza epidemic was a matter not of if but of when.
There are two reasons that Covid-19 is such a threat. First, it can kill healthy adults in addition to elderly people with existing health problems. The data so far suggest that the virus has a case fatality risk around 1%; this rate would make it many times more severe than typical seasonal influenza, putting it somewhere between the 1957 influenza pandemic (0.6%) and the 1918 influenza pandemic (2%).
Second, Covid-19 is transmitted quite efficiently. The average infected person spreads the disease to two or three others — an exponential rate of increase. There is also strong evidence that it can be transmitted by people who are just mildly ill or even presymptomatic. That means Covid-19 will be much harder to contain than the Middle East respiratory syndrome or severe acute respiratory syndrome (SARS), which were spread much less efficiently and only by symptomatic people. In fact, Covid-19 has already caused 10 times as many cases as SARS in a quarter of the time.”
The spread of the disease so far, and the apparent incubation period in which people exhibit few symptoms, yet still spread the disease, lead me to believe that global spread is a legitimate scenario (though I acknowledge we are terrible at prediction).
Further, if state-controlled countries like China have failed to quarantine the virus, I find it unlikely that Western countries will manage either.
This Guardian piece does a good job of laying out many of the common arguments against preparing for COVID-19, and I’ll cite and address some of them here.
Young People Are Safe
“Claim: ‘It only kills the elderly, so younger people can relax’
Most people who are not elderly and do not have underlying health conditions will not become critically ill from Covid-19. But the illness still has a higher chance of leading to serious respiratory symptoms than seasonal flu and there are other at-risk groups – health workers, for instance, are more vulnerable because they are likely to have higher exposure to the virus. The actions that young, healthy people take, including reporting symptoms and following quarantine instructions, will have an important role in protecting the most vulnerable in society and in shaping the overall trajectory of the outbreak.”
This particular argument irks me, because while I fall into the group of people not likely to become critically ill, this argument basically says “yeah, I don’t care about anyone else.”
Do you interact with anyone who is older? Do the people you work with have parents? Young kids? Of course they do. If you get infected and then spread the disease to them, and they in turn spread to their family, do you carry some responsibility? I believe so.
This is the key sentence in the above quote: “The actions that young, healthy people take, including reporting symptoms and following quarantine instructions, will have an important role in protecting the most vulnerable in society and in shaping the overall trajectory of the outbreak.”
You have an obligation to try and avoid being infected due to the consequences for others.
We’ll All Get Infected Anyway
“Claim: ‘If a pandemic is declared, there is nothing more we can do to stop the spread’
A pandemic is defined as worldwide spread of a new disease – but the exact threshold for declaring one is quite vague. In practice, the actions being taken would not change whether or not a pandemic is declared. Containment measures are not simply about eliminating the disease altogether. Delaying the onset of an outbreak or decreasing the peak is crucial in allowing health systems to cope with a sudden influx of patients.”
The key point here is this sentence: “Delaying the onset of an outbreak or decreasing the peak is crucial in allowing health systems to cope with a sudden influx of patients.”
This is related to the counter-argument I laid out above: you have an obligation to others to try and reduce the peak of this disease, so that those who need it can seek treatment.
The Flu Is Just As Dangerous
“Claim: ‘It is no more dangerous than winter flu’
If borne out by further testing, this could mean that current estimates of a roughly 1% fatality rate are accurate. This would make Covid-19 about 10 times more deadly than seasonal flu, which is estimated to kill between 290,000 and 650,000 people a year globally.”
I hope by now I’ve made clear that I don’t think it is a reasonable assumption to think we know this yet, or will know with certainty before we have to make a decision about precautions.
“Most estimates suggest 80% of COVID-19 cases are mild and feel roughly like a flu. Estimates I have seen suggest that roughly 10-15% of cases will be more significant and may necessitate hospital visits (see also) with 1-3% potentially needing an ICU.”
If this indeed were the case, do you think our healthcare system could cope with 10-15% of the population needing to visit the hospital? 1-3% requiring an ICU? I don’t.
If you’re healthy and have the means to do so, I believe you are obligated to try and avoid infection so that others who need it have access to further treatment.
How You Should Prepare for COVID-19
I’ll restate my position from the beginning of this article: you should be prepared to spend at least 2 weeks isolated in your home.
This is in anticipation of the virus reaching where you live, and having to participate in either a voluntary or imposed isolation.
The Prepared does a good job of laying out what’s required for various scenarios, and here’s their basic list:
- A source of potable water (storage is preferable, but if space is tight then get a filter)
- Extras of any essential medications you’re taking
- Toilet paper, feminine hygiene products, soap, detergent, and everything else you need to keep your body and your home clean
- Nitrile gloves and an N95 mask
- Food for ~2 weeks
The part to emphasize here is that almost all of these things are items you already have or will use.
There is very little required to spend ~2 weeks at home if you think ahead (assuming the taps continue running).
I dislike accumulating or owning anything that isn’t regularly used, and I try to adhere to a zero-waste lifestyle, so some of these items are tricky. That said, most of the required items for 2 weeks at home are things you would use anyway.
- Stock up on various basic medications (I don’t take any prescriptions, but would stock up if so)
- Toilet paper, detergent, cleaning products–get products in larger quantities than usual (I tend to buy in bulk anyway)
- Nitrile gloves, Tyvek suit, N95 respirator–most of these items I had anyway, from painting that I had done previously
- 2 weeks worth of food–stocked up on canned goods, pasta, rice, sardines, frozen cooked food–things I normally eat anyway, I just stock larger quantities
- Water filter–I have a Lifestraw I use hiking and camping, but otherwise am assuming that water service will remain–I also have various water bottles, CamelBaks, etc.
And that’s it! When it comes down to it, the cost of preparing is low–another reason I’ve been advocating preparing in this particular case.
How Startups Should Prepare for COVID-19
Elad Gil’s post has more details about how you can prepare personally and how startups should be preparing.
Here are some of the items for startups:
- Encourage hand-washing.
- Wipe down work areas regularly.
- Encourage flu vaccination.
- Zero tolerance sick policy (Anyone who is sick, or starting to feel sick, should take a sick day or work from home).
- Curtail travel and conferences and move to video calls.
- Curtail visitors from other countries.
- Plan for the remote work contingency.
A Final Word
In the end, preparing for disasters comes down to the cost/benefit analysis.
In my view, the costs here are few (some effort, some up-front cost, but mostly things you would use anyway) for the potential large downside risk.
As Tim Ferriss recently outlined in a blog post:
“I am not panicking this time, either. That said, I am curtailing unnecessary travel and group interactions for the next 2–3 weeks to see how things shake out, particularly given the asymptomatic “incubation period” of up to 14 days.
Might that be an overreaction? Might I be misinformed? Totally. But then again, how many head-on car accidents have I had? Zero. I nonetheless put on my seatbelt every time that I drive, and we have great data on traffic fatalities. Do you have a fire extinguisher in your kitchen? Would you accept $100 to get rid of it? $1,000? I wouldn’t. As unlikely as a kitchen fire may be, the extreme known consequences of an out-of-control fire easily justify a fire extinguisher, even if it gathers dust forever. It’s cheap disaster insurance, just like having emergency stores of water in the garage.”