Epidemiologist forecast about Coronavirus’ impact on the U.S.
[Author’s note: the following notes were taken 19 March 2020 during a briefing with Dr. Bar-Yam. These are my rough notes, not a complete transcript. Where noted, I have provided updated information.]
Dr. Bar-Yam has been working on pandemics research since 2006, using a mathematical approach that enables us to understand how contagion networks work, and how they can be stopped. He is currently studying the COVID-19 virus spread. (Update: Read the last four paragraphs on page 4 of his working paper on Coronavirus.)
Coronavirus cases start out mild, then they suddenly become severe. People don’t anticipate this, so when the suddenness comes on, it “comes out of nowhere.”
(Update: the map below (March 24, 2020) compares today’s cases with yesterday’s cases. The darker the red color, the higher the spike in day-over-day increases. Source: https://www.endcoronavirus.org/map/english)
In Italy, Coronavirus seemed to come out of nowhere:
- Twenty percent (20%) of all new cases in Italy are severe.
- They’re making military triage among patients and many people are dying.
- The reported death rate is 8%, but it may be higher than that.
China locked down when there were 800 cases and they ended up with 80,000 cases. China sent 42,000 health care workers and built 14 extra hospitals and they still couldn’t keep up.
The United States waited until we had thousands of cases, so we know we’re going to be overwhelmed. If we assume a 10X spread every week (1.3X per day), we can expect 90,000 cases in the next week and 900,000 in the next two weeks. Of those, 20% or about 200,000 will require hospitalization.
We assume at least five times more people have the disease (median 3–7X) than we have known cases.
(Update March 24: New York’s confirmed cases are 28%, many times higher than Italy’s original confirmed cases.)
If we step back from this for a moment, we can see that there’s no way to stop this. We have to take extreme and radical action. Waiting one more day makes everything a lot worse, and the more we wait, the worse it gets. Act now. This is a global and national emergency. Everyone is in trouble.
What works is a complete lockdown.
If we do a complete lockdown and everyone isolates at home with necessities, then the outbreak will be over in 3–4 weeks.
China and South Korea proved this to be true. The reason is straightforward. Two weeks is the max incubation period, so within 2 weeks, we can know everyone who’s sick. After that, only those who were together with the sick would be sick. So after 4 weeks, things decay dramatically and after 5 weeks it’s gone. (Update March 24: Science and health reporter Donald G. McNeil reports that if we did a complete lockdown, this whole mess could be over within weeks. Here’s a step by step guide to locking down the Coronavirus epidemic.)
After it’s dramatically suppressed, there’s an exponential decline. At that point, we can do standard contact tracing of those who are sick and in isolation. Anyone who’s telling you it will take 12 months to figure this out is wrong.
There will be a lot of tough times especially in the near term emergency phase. We must have very good execution during non-normal conditions.
[Author’s note: During this briefing, the following questions were asked of Dr. Bar-Yam and his responses are noted. Again, this is not a formal transcription, but my attempt at catching the most salient points.]
Q: Is there a new strain?
A: As far as we know, there hasn’t been enough of a differentiated genetic strain for there to be a pronounced, additional strain of the virus. If there is a slightly different strain, it doesn’t change our approach because the lockdown should be the same.
“If everyone in the country decided to self-isolate, no further action would be required by (national or state) leadership.” One of the things that’s really important that I didn’t mention earlier: the key thing (about national coordination) is to separate the country locally. There are some parts of the country that have less of the disease v. more of the disease. In the near term, the disease is everywhere and we need to have a national lockdown.
Q: What’s the expected rate of adherence to the lockdown?
A: In order to stop the outbreak, we have to impact the “spreading rate” (it’s currently assume that each carrier spreads the virus to 3–4 others, but the actual number may be larger or smaller). We need to have 1 divided by that number. If we get 90% compliance, for example, we may be successful. But the goal would be for the decay rate (post-peak) to be sharper. In China and South Korea, there will always be lack of compliance and some enforcement. The best thing is do this now as a society and take the actions needed to ensure that people who aren’t understanding instructions or choosing not to follow them are addressed.
Q: What about behavioral economics (BE)?
A. We have very strong reasons to believe that BE has a limited approach because people are now making decisions under very different circumstances than those used to during BE research.
Q: Is fear as a clear motivator? “Don’t kill your grandmother” is a simple, clear message.
A: The people who really know how to influence behavior is advertisers; talk with advertisers and marketers. They know how to sell potato chips, they can help motivate people.
Q: Contact tracing — how are we going to do this?
A. China had a massive contract tracing effort. Singapore did the same thing: find people who have the disease; ask them whom they’ve been in contact with; find those people, and trace them, too. China approached people walking into stores and had them scan QR codes and then followed them throughout the store. But all of that comes later. Lock down is what solves the near-term problem. Public health authorities don’t know how to do lock down; they know how to do contract tracing. The fact that we don’t have enough testing is a huge problem.
Q: What is Plan B?
A: There is no plan B. My concern is that we won’t go on lockdown until our hospitals are overwhelmed.
Q: How can we do testing?
A: Doctors participating in EndCoronavirus.org are reporting that only serious cases are being tested. People who have minor symptoms are being sent home. So ramping up 5 or 10X in testing is not enough. There are companies that are trying to change from DNA test to other tests but the real question is how many tests we can do.
Some people believe we should test people who come into E.R. That might not be the right approach. If we can catch people at the beginning of the onset of symptoms (like we did with Ebola, and then isolated them), that’s how you stop the spread. Respiratory exercises are very important, to keep breathing right. 80% of cases in China were within family units, so we have to keep family units isolated together. Could we use the hotels with kitchenettes and restrooms, to spread people out who are infected? There are steps we can take to really cut down the number of people who get infected.
Anesthesia masks used with filter can be used. Military can help with separate facilities required by different levels of care. The key part is that these facilities have to be identified and built out. Universities that are now empty could be used as backup hospitals. And that requires amazing scale and complexity. A lot of “small scale” things and careful decisions need to be made at the local level. Large scale things are done well by the military; but health care people don’t know how to do large scale things like setting up and maintaining at stadium-level. Dr. Bar-Yam has a slack channel to see what’s happening at large scale.
Q: What’s the relationship between fixed facilities and tents (to help quarantine those who have mild to moderate symptoms)?
A: We have to create facilities for low and medium cases that aren’t being dealt with at all by the current hospital system. People with low or medium cases are currently being sent home. So then people go from moderate symptoms to urgent symptoms very suddenly; can they get into hospitals then? We need facilities to care for those who have low to moderate symptoms, so we don’t send them home to infect others. No one is doing this yet. We have to do this today and have stuff set up by the weekend.
The overall behavior of this disease is very clear: People get it by touching each other or breathing each other’s air. If you’re not six feet apart or you share surfaces with others, you’ll get infected soon. Don’t touch surfaces in public spaces (door knobs, elevator buttons, handles). Talk to everyone about self-isolating.