originally published April 8th 2020
Fear of the novel corona virus certainly has ramped up quickly. During a normal flu season people look at me like I’m a lunatic for asking if we can please not shake hands. Now when I’m out walking on a windy day, people in masks give me a wide berth and a disapproving look for not having my own mask.
Many knowledgable and rational medical professionals and scientists give me their dire view of the situation and far from considering the destruction of the world economy a mild over reaction, they believe we aren’t doing enough and the fools who are not properly socially distancing on hiking trails will surely be the doom of us all. This disease is 10X as deadly as flu!!! Well maybe 5-10X, we shall see, but it seems that 5-10X deadlier is the difference between “Erik is a no hand shaking loon” and “We must create the greatest of depressions to save ourselves!”
And hey, I’m certainly not arguing for doing nothing. Remember I’m the weirdo who won’t shake hands during flu season. Welcome everyone to my wonderful world of giving a crap about infectious disease control! I just marvel that people have two speeds for this issue- 0 or 100. They simply cannot grok that maybe this situation calls for a 25 or a 40.
Allow me to offer a less scary way of looking at this pandemic. When we use data and analytics to inform and justify our emotions, perspective is key. Some ways of viewing the data will rationalize your terror (10x deadly as flu! 80% of the population will get it!). Some other ways of looking at the same numbers might make you feel a little more relaxed. Maybe even give you some space for concern about the less well off members of society whose lives might be ruined. Be conscious of the fact that shelter in place and unemployment also have terrible effects on large numbers of people. Those effects just don’t have their own convenient statistic…yet.
Here are two links to a paper published on The Lancet about expected fatality of this new virus in different age groups. The first is the paper and the second is a WebMD article which calls out some key numbers. I hope one or both of them remain live.
This Lancet paper gives estimates of the fatality of the virus if you get infected with it (i.e. not necessarily showing symptoms- everyone). See the scrollable table last column “infection fatality ratio”.
They quote 4.28% for age 70-79 and 7.8% for 80 years and older.
Compare to the following
This is a mortality table for the US from the social security administration. It shows the proportion of people at each age who are expected to survive to see the next year.
I am going to be generous and make 4.8% the rate for 70 years old even though if that is for 70 and up it is likely a little high for an actual barely 70 year old. The number is likely more correct for a 73 or 74 year old. I’m going to look at the male columns of the SSA table because men are more likely to die from COVID-19 and I don’t feel like doing extra math.
The SSA table shows that out of a population of 100,000 male humans born 70 years ago, we expect 72,843 to be alive today.
My question then is – without COVID-19 when would we expect to lose that 4.8% of 70 year olds anyway? 72,843 minus 4.8% is 69,346.
According to the SSA table that is the number of of those 70 year olds we expect to make it to 72.
If we do the same calculation with 74 year olds (which makes a little more sense if 4.8% is the average for people in their 70s) you start with 65,419 74 year olds and 4.8% less is 62,279, which is a little less than the number of them we expect to make it to 75 (63,274) but far more than we expect to make it to 76 (60,998).
My point being that when reacting to the likelihood that COVID could kill a certain number of people in that age group, the average layman has no idea how dangerous it is just being in that age group. Be honest- before you read the above you didn’t know within a factor of ten how likely a 74 year old was to die during the next year.
The 80s look even less unusually deadly: 7.8% of those infected die if they are over 80. Let’s assume that >80 skews close to 80. The expected mortality shown in the SSA table at age 83 is 7.8%.
The wiseguys among you might claim we should add the two mortalities together. That is, if a regular 75 year old has a 3.6% of not becoming a 76 year old normally, now he has the added risk of 4.8% if he gets the ‘rona and that’s 8.4%! Of course we know there will be overlap. The intersection of those two sets won’t be the empty set. It won’t be the same set. Will there be 50% intersection? I say at least 50%, and given that COVID-19 is mostly killing people with underlying conditions and kills more the worse those are, I think it is fair to assert the the intersection is closer to 100% than 50%
I admit any minimizing of any death based on cold calculations is emotionally distressing. Right now many of us are focused on the potential for people, especially our beloved grandparents, to die in large numbers. The average non medical person, in a normal non-COVID-19 year, doesn’t have a conscious sense of how many elderly people die in an average year so of course these numbers seem horrible and unprecedented. The emotional response is that we must take any action, no matter how painful, to stop this.
But we need to overcome emotion and cognitive bias and make the best policy decisions to balance the needs of all of society. We cannot ignore the side effects of the lockdowns to people’s mental health and economic well being. We need to balance effects and so have a balanced policy. All medicines have side effects. The more dire the condition, the worse the side effects we are willing to risk. The acceptable side effect portfolio of a pancreatic cancer drug is much worse than for a nasal decongestant.