Pandemic Statistics
Yesterday I made a facebook post that sparked a good discussion:
A family friend wrote an “open up America!” post. Here was my response:
If we open up completely right now, figure 1/4 of Americans will get COVID-19. It’s about 6% fatal. Here’s the math: 300,000,000 * .25 * 0.06 = 4.5 million dead. That seems a high price to pay for freedom. Thoughts?That said, I -hate- staying home and hiding from this thing and I know the loss of productivity is staggering. I’m watching my daughter’s development change for the worse and I hate it.
My friend Tim posted an interesting article and I riffed on it here:
And then, here’s some studies that says my estimates could be way off. I’m happy to be wrong. See Reason article 1 and Reason article 2
“A hundred deaths out of 48,000-81,000 infections corresponds to an infection fatality rate of 0.12-0.2%,” they report.* That’s about the same infection fatality rate the Centers for Disease Control and Prevention (CDC) estimates for seasonal influenza.”
So maybe if we open up completely, the math looks like this: 300,000,000 * 0.25 * 0.002 = 150,000 dead total.
I wish we knew which way we were headed.
And here is today’s post in the form of a (very long form) question:
So, how dangerous is COVID-19? Should we get back to our lives?
Looking to Sweden’s death toll may be useful to see what may happen in the US. Sweden didn’t shelter-in-place.This page (https://coronavirus.jhu.edu/data/mortality) shows Sweden having 32 deaths per 100,000 people, (that is a mortality rate of 0.032%) which isn’t the highest mortality rate; why don’t they have a higher mortality rate with their open policies??
If 0.032% mortality is what we can expect in total, and people actually recover, then there is a strong argument for us to just go back to work, yes?What am I missing? With a 0.032% mortality rate, the US would be expected to have a total of 102,400 deaths (calculated by 320 million people * 0.00032). That’s… honestly… not bad. That’s in the realm of a “really bad year with influenza”. There are lots of reports out there that say comparing COVID-19 to the flu isn’t reasonable. Why not? In all seriousness, why not?
Let me try to enumerate the important factors of the pandemic. If we can figure out all the important factors and then address them, maybe we’ll get somewhere:
– Concern about a very high mortality rate among individuals that are older and have co-morbidities
– Concern that COVID-19 leaves people permanently injured
– Concern that people don’t build immunity… and subsequently, that the mortality rate will climb insidiously over time.
– Fear. Fear that there is a new disease that we can readily identify but cannot treat.
– what else? What, specifically, am I forgetting in this list?
Well, based on the fact that we’ve passed 80k deaths in the U.S., and that’s before widespread reopening of the country, I’d say the low estimate is a pipe dream.
We’re still in the realm of not enough data. It’s still too hard to get tested, and without universal testing, all of our numbers are, at best, educated guesses. Our best numbers come from other countries that are doing better than we are at comprehensive testing like South Korea. They’re looking at a mortality rate of 2.4%, and they have the testing to say that’s a pretty hard number. Even then, they are seeing additional outbreak spikes after starting to reopen the country.
Our way of life has been utterly disrupted, and will be until there’s a working vaccine. But, if we open up now, life won’t go back to normal and we’ll have lots and lots more people dead. You can guarantee that if we open up now our healthcare system will collapse rapidly. At that point the mortality rate is going to start climbing aggressively.
It’s a really stressful time right now. Everyone is going a little crazy as the things we took for granted stop working or are inaccessible. On an individual level, there doesn’t feel like there’s a lot we can do but hang on. And that sucks.
It’s not at all like a bad flu season. According to Scientific American (dang, I wish I could link here!), reported numbers for flu are greatly exaggerated, while those for COVID-19 are underreported. To quote that article:
Moreover, COVID-19 has caused a spike in deaths from other causes, as people are afraid to go to hospitals and hospitals are overwhelmed even if they get there. According to the New York Times, the number of excess deaths this year over other years has been over 100,000 already–and that’s even after the lowered mortality from less driving. Also, you’re comparing a whole year of flu to about two months of COVID-19–and COVID-19 hadn’t had a chance to spread much in the early part of those two months.
And all of this occurred in spite of lockdown orders. The idea that lockdown orders didn’t cut down on the number of deaths seems far-fetched.
Different countries are going to have different mortality rates, regardless of social policies. Italy, for example, has substantial trade with China, so its death rates were much higher than those of other European countries. Greater mobility within the country also contributes to the spread. So we can’t assume that Sweden is worst case. That’s why we compare its rates to those of other Scandinavian countries, not to the world at large. And its rates have been much higher than other Scandinavian countries.
Plus, even Sweden did not just stay completely open. While there were no official lockdown orders, officials recommended social distancing and extra hand-washing–and Sweden has a history of paying a lot more attention to such recommendations than we do in the US. Plus, Sweden has universal health care and paid sick leave, so people were less likely to go to work sick or to die of COVID-19 if they got it than they would be here.
So no, we can’t assume that opening up the country would lead to only a “not bad” number of deaths. And that’s even leaving aside the number–including children–who end up with permanent disabilities from it.
Thanks to both of you for your thoughtful comments! This is all so weird, it’s good to talk about it!
Free, you can leave a link by just not including the “http://” part of the link. I’ll fill it in for you. (that simple spamproofing stops 80 spams/day!)
I’ll write a thought-out reply to your comments in a few days. Thanks again!