Thread by @jasoncrawford

Why are many people worried about the new #coronavirus, when the regular flu kills tens of thousands a year?

The concern is that it could be something like the 1918 influenza pandemic (aka the “Spanish Flu”).

It's the huge spike on this chart of US mortality rates:

The 1918 flu is estimated to have ultimately infected over 500 million people—almost a third of the world's population at the time—and killed an estimated 50 million. ncbi.nlm.nih.gov/pmc/articles/P…

Of course, #COVID19 is nowhere near that bad yet. The concern is that (a) it's spreading rapidly, and that (b) it may have a high case-fatality rate.

I am *not an expert* on these matters, and our knowledge on this is evolving rapidly, but here's what I understand so far:

Re spreading, the contagiousness of a disease is measured by something called R0 (pronounced “R-zero” or “R-naught”). This is simple: it's the average number of new (unvaccinated, not previously infected) people a patient will infect.

Unfortunately we don't really know the #COVID2019 R0 yet. Estimates range from 1.4 to 4.0 (see below). That may sound like a small range, but after 10 rounds of spreading, that's the difference between under 100 people infected, and over 1 million.

worldometers.info/coronavirus

Note, R0 isn't an inherent property of a disease—it depends on the environment. And “basic R0”, the maximum potential R0, is different from “effective R0”, which is partially in our control. Good hygiene, social isolation, etc. will improve it. sph.umich.edu/pursuit/2020po…

Regardless of the R0 estimates, it's clear that the disease has spread quickly in certain populations. Outbreaks have happened around the world, including Italy, Iran, Japan and South Korea. And total cases have far surpassed SARS: nytimes.com/interactive/20…

And remember, these are *confirmed* cases—and you can only confirm a case by *testing*.

Unfortunately, thanks to mistakes at the CDC and a tangle of FDA regulations, the US isn't testing many people at all right now.

This has to get fixed—and fast.

https://twitter.com/davidalim/status/1232822521344483328

Re fatality: Like R0, we don't really know the case-fatality rate. The current estimate from the WHO is around 2%. But we can't be sure because we don't know how many people have gotten the disease.

worldometers.info/coronavirus/co…

In general, early estimates of fatality rates are often overestimates, because of sampling bias—you only know about the most severe cases, because those are the ones that show up in the hospital. The mild cases stay home, or don't even realize they were infected.

So, one big hope for #SARSCoV2 is that it's actually way less fatal than we think—closer to the regular flu, which is closer to 0.1%. But again we just don't know yet.

Putting these two factors together, here's how #COVID19 compares to other diseases and pandemics, from this very good @nytimes explainer: nytimes.com/interactive/20…

Note that #COVID19 is not nearly as contagious as measles, and not nearly as deadly as MERS, but the combination of contagiousness and deadliness could turn it into a global health disaster.

Also, we have a vaccine for measles, but not for #COVID19. And yes, researchers are working hard on one, but it's not going to be available this season. (And it's not clear how long it would remain effective—the seasonal flu mutates so fast that a new vaccine is needed each year)

Coming back to the “Spanish Flu”: The 100+ years since that pandemic have seen enormous leaps in our ability to deal with disease. So there are good reasons to hope that—with intelligent decisions and swift execution—the #COVID19 epidemic won't be nearly as bad.

First, and most importantly, we have antibiotics now! A huge factor in influenza mortality is opportunistic secondary infections—especially pneumonia. But bacterial pneumonia can now be treated with antibiotics, saving many lives.

Second, we have a much better understanding of how diseases spread and how to slow them down.

Back in 1918, we barely knew what viruses *were*. We didn't understand RNA/DNA, and we didn't have any microscopes powerful enough to see them.

And think of all other the inventions since then that will help:

• Genetic sequencing to understand the virus
• Statistical methods and software packages to analyze the data
• Global communications to coordinate response
• Mass manufacturing of sanitizers, etc.

Even something as simple as basic hygienic practices, such as covering sneezes and washing hands, weren't nearly as widespread in 1918, as I mentioned in this post: rootsofprogress.org/draining-the-s…

So what should an individual do?

1. Don't panic
2. Wash your hands; don't touch your face
3. Postpone meetings, or video chat
4. If you can't do (3), at least no handshakes
5. If you can't do (4), do (2) right after
6. Buy a few weeks of grocery/drugstore items now to allow (3)

I don't think you need to wear weird masks, buy a bunch of food you're probably not going to eat, or go into full lockdown. But if you want a fairly comprehensive prepper checklist, I found this helpful: theprepared.com/wuhan-coronavi…

Anyway, stay informed. I made this list of helpful folks to follow. Thanks especially to

@ScottGottliebMD for authoritative, informed statements
@balajis for being early and for lots of intelligent discussion
@chr1sa for daily updates on statistics

twitter.com/i/lists/123343…

*Hopefully* this all blows over and isn't as bad as anyone fears—or swift action contains or at least minimizes the disease. But it's good to be aware and prepared.

Stay safe out there, everyone! And if you see me—wave or fist bump, OK? 👊

PS: I think one thing that makes this hard for people to talk about is that most people want binaries and want certainty. Either it's definitely going to be a disaster, or it's definitely going to be fine.

We don't have either right now.

There's a big range of possibilities for the outcome, from “very little real impact” to “worst pandemic in modern history”, with the extremes both being *very* unlikely, and most of the probability in the middle.

But that's hard to talk about.

Even the previous tweet will probably get pushback from people saying “how dare you suggest it's nothing!” and “how dare you suggest it's a disaster!”, ignoring the fact that I said “very unlikely” on both extremes.

Probability distributions are hard to grapple with.