Making Outbreak Models

Listen to this episode

S1: Following recording may or may not include instances of words being said that the FCC would find me for their long arm could ever reach.

S2: Slate Plus members, I’m here to remind you to take the sleep survey. It’ll be open through April 1st. This is your chance to tell us what you think about sleep plus and sleep. It’ll only take a few minutes and you can find it at slash survey.

S3: It’s Monday, March 23rd from Slate’s The Gist. I’m Mike PESCA. State governments are stretched thin with the governors of California, Washington, Illinois and New York begging the federal government for help. What can be done? The answer, when you may have heard, is the Defense Production Act, as in war.

S4: The government can compel businesses to produce materials needed for a fight. It’s so analogous to the current situation. It’s almost not an analogy, but an equation. Three days ago, Donald Trump made a big show of triggering the Defense Production Act, at least saying that he was invoking the act.

S5: I invoke the Defense Production Act. And last night we put it into gear. We moved the National Response Coordination Center to the highest level of activity. I mean, if you if you take a look at what we did, the level of activation has been increased to a grade 1 level, which is the highest level.

S4: A Grade 1 level of activation, by the way, is a reference to a FEMA agency. So what may have sounded like he was utilizing the Defense Production Act was actually him saying he was ready to use it. That announcement was prompted by repeated questions from the media the day prior. Why haven’t you actually use the Defense Production Act? What is your reticence to invoke the Defense Production Act that you have done?

S6: Yeah, if we find that we need something that we will do that. And you don’t know what we’ve done. You don’t know whether or not we’ve ordered. You don’t know if we’ve invoked it. You don’t know what’s been ordered, what’s not to. I know.

S4: Well, now we do know nothing was produced due to the Defense Production Act actually being utilized, as Illinois Governor Jay Pritzker made clear this afternoon in a press conference.

S7: I have called for him to invoke the National Defense Production Act and actually utilize it, because I know he’s invoked it, but not utilized it in any industry.

S8: The president doesn’t want to use it. He thinks our needs are being met with the mere threat of use. Yesterday in a press conference, Peter Navarro, the extreme Sino antagonist who authored the administration’s tariff strategy, told us all why the Defense Production Act was not being utilized.

S9: What we’re seeing on a purely voluntary basis based on the leadership of this administration. We’re seeing the greatest mobilization in the industrial base since World War 2.

S8: Navarro is articulating a philosophy, a philosophy of laissez faire capitalism tied up with an extreme fawning over the president’s wisdom. Right as the president was watching him from less than six feet away. By the way, not good for distancing. Navarro is saying that the mere threat of the act is enough to meet the needs of the moment. Well, actually, to be clear, the threat of using the act plus the imminent strength of the man at the top.

S9: Now, what I can tell you so far is that the Defense Production Act here has given me quiet leverage. When you have a strong leader, you can take a light hand initially quiet leverage.

S8: At another point, he said, quote, We are getting what we want without the heavy hand of government. As per Jay Pritzker, the situation with the invisible hand of capitalism and not the so-called heavy hand of government is not going well.

S7: In one case, they told me I was competing with FEMA to acquire ventilators. So I’m competing against the federal government to get ventilators for the state of Illinois. And the federal government is not distributing ventilators to the state of Illinois. I’m literally working against, you know, a competitor. I just want to give you the second example, if you don’t mind. I called another manufacturer of ventilators and he pointed out to me that while I would be competing with countries other than the United States, as I put an order in, and that I better put in as big an order as possible in order to put myself higher on the list of priority to get ventilators from that manufacturer. Now here I’m competing with my own federal government. I’m competing with countries outside the United States to get things that we need to to keep our people safe and healthy.

S8: Pritzker did note that he talked to the president today and the president seemed to want to help and said, let me see if I could get that for you. Well, okay. One model of actually addressing a problem is an ideology driven free market purist, making sure that the government stays its hand lets it be seen as potentially heavy. The other is actually using the power of law to ramp up manufacturing rather than seeing if the needs of 50 states can be met by the president personally following through. As far as requests are phrased nicely and he doesn’t get distracted, I think I know which one I prefer. And which one we need? Because the opposite of the heavy hand isn’t necessarily an effectively light touch. It may just be losing grip or dropping the ball. On the show today, I give the spiel over to this extended interview with a researcher who is modeling possible courses that the outbreak can take. His work was featured. In fact, it was pretty much the basis of a front page article in The New York Times this weekend titled Corona Virus Could Overwhelm US Without Urgent Action. Estimates say Jeffrey Shaiman is an epidemiologist at Columbia University who plotted out the possible course of infection.

S10: We talk about his methods, his takeaways and his recommendations for what could mitigate what’s already an emergency, but could soon be a full blown crisis.

S1: Some of the most interesting and frightening studies have been conducted by Jeffrey Shaiman, who is at the Columbia University Mailman School of Health. He studies infectious diseases. And perhaps you’ve seen the big New York Times study with an interactive graphic that shows how and where the virus can spread if left unattended or even if attended with some, but not severe mitigation. Mefford efforts. We’re going to talk about all of this with him. Hello. Thanks for joining me.

S11: Thanks for having me onboard.

S1: So I think I’ve seen a couple of your studies and I’m not sure if there was one big study or if you’ve come out with a couple. But a couple of days ago, they were citing work you’ve done that essentially showed how the disease spread through people who weren’t showing symptoms. And then I saw a couple days ago a big study that shows how the disease, how covered, can spread throughout the society if we don’t do anything. Were these two separate studies that just landed one, two, or were people picking up different aspects of the same study or the related studies?

S11: But there are two different studies. And the first, which came out on Monday, it’s in the journal Science. We looked at China specifically and we examined what was going on in China in particular before they implemented strong control measures on January 23rd. So on January 23rd, the Chinese government imposed travel restrictions and control measures and adopted some very aggressive contact tracing that they started in Wu, which is the epicenter of the outbreak. And they expanded outward to much of the country, so much so that at its peak, I think about three quarters of a billion people were essentially on lockdown. But we focus on the period right before they implemented that, because that shows the time when the virus is circulating somewhat naturally as it would in a society that is not reacting to it. In other words, how the how the virus is going to transmit and move around if left to its devices. We were particularly alarmed when we saw the data there by how rapidly it spread geographically and internationally. You know, it moved very quickly from the epicenter in one hand to other places in China and then started hopping to other countries. And that’s very concerning, because that means it’s very mobile, it’s very capable of getting around. It’s an emergent pathogen to which the whole world is essentially susceptible to it because we’ve never experienced it before. It has sustained human to human transmission properties and sufficiently high force of transmission. You may hear the term the basic reproductive number is sufficiently high that it’s going to grow exponentially. But there have to be other factors in place that allow it to move very rapidly from place to place. And what we suspected was that even though they were reporting cases, there were a lot of undocumented cases, people who were still going around and doing their normal routines, but were shedding the virus and bringing it to new areas in their communities and even getting on planes and trains and going to new parts of China or even internationally.

S12: Right. And this was taking place during the Spring Festival. So was that or the spring period where there was a lot of travel, there is a lot of travel built into the Chinese culture at that point.

S1: Is there reason to believe that there was any more or less traveled than would normally be happening?

S11: We use travel data to inform the model that we use to make these kinds of inferences and estimate what fraction of people were undocumented infections. What fraction of the total infections were undocumented. And we informed it of actually an analogous period from twenty eighteen rather than the twenty twenty data which weren’t available at the time we did the study. However, we did have another proxy that shows movement in China that was available in real time. It wasn’t as comprehensive. So couldn’t give us the absolute numbers. But what it showed was that travel and movement between cities and around in China up until January 23rd proceeded as it normally would. And then it just falls off a cliff right at that point.

S12: Right. Is there reason to think or not think that travel, Sara, to be so American centric? But that’s what we’re concerned about. Travel in America, spread in America would essentially follow the Chinese patterns.

S11: Well, you know, it’s not so much the fact of the amount of travel. And I would have to say that the amount of travel in China around the lunar new year is the largest migration in the world. So we’re not talking small numbers. We’re talking enormous numbers on a scale that dwarfs what happens in the United States. That time of year sees an insane amount of travel, the likes of which are not recorded anywhere else in the world. So it is very large there. But what really matters is the level of connectivity which is going to allow the virus to be brought to new places. So we actually took the model that we had applied to China where we estimated that prior to. Travel restrictions, only one in seven infections were being confirmed, and after the travel strip restrictions, we estimated about two thirds were confirmed because they were doing very aggressive testing and contact tracing for infections. But we transferred this model to the United States. We adjusted it. We use the travel patterns that are available between counties that we have, as well as population data in the case count datas. And we repeated the estimation process for the United States, much larger model now because instead of looking at three hundred seventy five cities, we look at over 3000 counties and their rates of movement and commuting and random travel between them. And we use that to inform this. This is the study that’s in the second publication, which it really came out in The New York Times. It’s a little unconventional for publishing scientific findings initially, but it was their data set that allowed us to do this because they had been compiling confirmed case data on a county by county basis and on a daily basis for the entire United States.

S12: Right. And I noted that since there was no time to peer review it in that article, they quoted other scientists who basically said, yeah, this comports generally with what we’ve been finding. So I guess that’s kind of a quick and dirty peer review.

S11: That’s a very quick and dirty peer review. And, you know, the reality is, the specific findings, because they’ve been out in The New York Times, may not be eligible for peer review. We’ll see. We’re going to post it on to med archive. And I may approach a couple of journals to see if they will peer review. What we’ve done, though, some of them may be hesitant because they’re going to say essentially and legitimately, you know, you already have it as out in The New York Times. But what you want. And honestly, peer review at this time in a time of crisis is not what, early after we’re not after actually trying to get, you know, publications. We want to get information out there. And obviously, The New York Times is a very effective way to do it.

S13: Now, what our findings were for that study is that it’s one in eleven instead of one in seven going up to March 13th in this country. And that is that’s even less. And it’s actually not surprising given the limited amount of testing that’s been performed in this country, that the model corroborates that and says one in 11 people who have it are actually being publicizes, have it in it or noted as having it right there, being confirmed there, being laboratory confirmed that they actually have this R.S. Ovie to virus.

S12: Right. So multiply by 10 or 11. The numbers we’re getting now. I want to just go back because in the Chinese data study, you use 10-cent location based services. And that seems very efficient, especially if as a researcher you use, you know, anonymized data, which you would expect the government give to you. Is it harder to get the data from the United States is the equivalent services like Facebook? Have they been able or willing to give it?

S11: Well, so it is harder in the model that we use for the times. What we did is we we really were interested in travel between counties. And so what we did is we used commuting data that was actually documented by the census. It’s very comprehensive in that way, saying who lives in one county and works in another. And then we have to scale to that also random traveling, which we have to infer in the process in a somewhat crude overview way for the whole country. The thing that’s going to be really critical about getting things like location based service data, mobile phone records or the like, is that particularly going forward now, given that there are different responses going on in different components of the country and they’re going to be different levels of compliance to them, it would be really important that we’re able to understand how much person to person contact within communities and between communities has diminished and how it varies across space and time in the United States.

S13: And to get it that we do need something like cell phone records on location with g._p._s location. Facebook has this thing about people coming into contact with one another. But the reality is I’ve approached them. They’re very friendly and helpful, but they require a data licensing agreement that I’m not allowed to sign at Columbia University, even in this context, unfortunately, on the Times and you enabled this to happen there.

S12: Interactive’s and then people could look at it and say, OK, how far and how wide will the virus spread if nothing is done? And thankfully, things are being done. And the answer there is game over everywhere all the time. Right. Then there is a middle toggle where you can say, OK, what if some social distancing is employed? And the answer for that is it’s still over 75 percent of the United States gets it by August 1st. So what does I guess it all rests on what the sum means, kitten. Is there a lot of some and a little sum that could change those figures?

S11: Well, you know, you have to ratchet it up even further. The third scenario, obviously, has it, which doubles that initial amount of intervention. And in that, you see you do see a great reduction. We’re going to be going at this in some sense in a trial and error approach.

S13: We have to come at it hard. We have to shut down businesses. Unfortunately, if we want to control this. And that has enormous. Economic costs to it. It’s psychologically and emotionally stressful for everyone. The government has to come in with a plan, probably something along the lines of what Britain proposed, where they’re going to pay 80 percent of everybody’s salary up to some limits so that people can pay the rent and that they can buy food and they can get basic necessities and pay their medical insurance. Of course, these are things that are really vital that we maintain more so than bailing out an airline company. You know, want to give them. Personally, I don’t think you should be giving the money to an airline company so that they can pay their employees, give it to the people directly. The airlines can house and park their planes. There may be some financial need that we have to impose on them later so that they can restart and reboot things. But that’s not the main concern right now. But we need to actually go at this aggressively. But we shouldn’t be doing it in a bubble just to say, all right, we’ve got to do everything. We have to look at other countries around us and see what has worked and what has not worked. And they’re already telling examples of that. In particular, China and South Korea have been able to completely bend the curves around in the face of exponential growth, which is what we’re facing right now. And the measures they employed and in China, they were particularly draconian, but in South Korea, they were very invasive as well. They were tracking people’s credit cards who were infected to find out where they shopped so that they could quarantine the people who had been in those stores as well, that we need to do that along with aggressive testing. If that’s the level we need to get to for this, that’s something we have to consider, too. As a society, even though it infringes on our civil liberties.

S12: So I’ve read in your study you talk about death rates when you calculate the death rates. Is that taking into account where the ventilator in ICU capacity is now or is that projecting out? What if that capacity is impinged?

S14: We don’t actually factor in any sort of deterioration of of of health care delivery. We just put in a simple multiplier about this is what the case fatality rate will be amongst confirmed cases. So right now, it’s very simplistic. We are looking at trying to actually weave in some of those elements that you’re talking about right now, looking at ICU bed capacity and seeing if we can actually project what’s going to go on and how hospital systems will need to adapt and bring more beds and ICU units on board.

S12: So my inference from that answer is that if anything, the case fatality rate in your studies will be low because we might not have the capacity we have now.

S11: That is true. That is true. But remember, we have to assume what the case fatality rate is. And there also is the possibility that somebody is going to come out with an effective point of care. Therapeutic. I wouldn’t I wouldn’t hold my breath for it happening really quickly. You know, you hear a lot of noise and chatter about some of these other drugs that were actually designed for other diseases being tested. There’s very preliminary evidence that maybe they have some help, but we’ll see that that might actually lowered a bit. But we can hope. But we have to work towards it.

S12: Face masks in public, not the N95 that that they’re hospital workers need, but the kind of lower kind that you buy off the shelf that may provide some perfect protection that may help you not spread it if you are a carrier who doesn’t know it. Do you recommend them?

S11: Well, you know, there is sort of a cultural practice of doing this, particularly in East-Asian societies.

S14: It’s sort of an indication of solidarity and recognition. And it’s a form of politeness, particularly if you yourself may have the sniffles to wear a mask. It may if somebody is sick, prevent them from spewing the disease out into the public in places where people can use them. But the reality is people tug at their masks all the time. They take them on and off. And in doing so, they get the goop that might be deposited on the mask onto their hands, and then they can rub their eyes and they can transfer it to other materials and then people can pick it up and they’re not airtight. So if the virus is transmitted through an airborne means, it’s not going to prevent it. I also want to just point out that there’s a lot of dogma being traded around, that this is droplet transmitted or this is only on surfaces. We don’t know. That is all I really say bunk. And people are overstating something that we don’t know. Nobody has done the experiments on this virus to really determine what its mode of transmission is. And the reality is for respiratory viruses in general, we don’t know how they’re transmitted. We know how they can be, but we don’t know what the dominant pathway is. Is it indirectly because people deposit droplets on surfaces and keyboards and doorknobs that other people transfer through their hands and then they eat a cheeseburger or pick their nose or rub their eyes and transfer it to themselves and in fact, themselves? Is it airborne, tiny droplets that come out that remain aerosolized and the turbulence in the air that people then inhale and they can go deep into their lungs and set up an infection? There is a droplet because somebody coughs or sneezes or we all spray when we’re speaking and it hits people in the face and they get the infection that way. You’ll people you’ll hear people talking about it often very much dismissing the airborne route. And really. Sizing droplet, but the reality is we don’t know. There’s been some experiments finally published this week. Not finally. I mean, we’re very early in this by a group from the NIH, Princeton and UCLA that looked at the survival of this virus in laboratory settings on various surfaces and aerosolized airborne. There’s a special type of apparatus for doing it. It’s a bit of a pain. And they only extended the test out three hours. But the virus was viable three hours out in the air.

S15: And they also found it was quite viable, out multiple days, up to 80 hours on surfaces. It decayed at different rates, on different surfaces. They just didn’t test the airborne survival out that far because it’s more difficult.

S12: Anthony Foushee has said that he has acknowledged that we don’t have the testing where we want it to be. But he has said we actually don’t need it right now, because if we follow these advisories to essentially shelter in place and not spread it, it doesn’t really matter if you test or not. You could kill the virus. And he’s right. And maybe if I were in his position, given that there is no testing, I would say that. But it’s more than just the best case scenario that we’d like testing. It seems that without testing, it’s really unreasonable to think that these shelter in place dictums or advisories are going to hold. I mean, the places like I read about that town in Italy voe it really seems to be. But I want you to talk about the combination between testing where, you know, if you have it and really strict quarantine measures and those pretty much need to act in unison. Don’t. Thank you.

S13: I agree with your assessment there. They need they’re needed in unison. They have to be used to reinforce one another. And particularly the testing has to reinforce the isolation and quarantine measures. You really have to ratchet down where you see pockets of it emerging. And remember, there are areas of the country that have yet to have this actually come up. There are counties in this country that are not yet experiencing this virus. And those in particular could be very well-protected if the testing can be done and it can be identified early so that we don’t have this explosive growth happen and it gets out of hand there. I don’t agree with Anthony Foul’s she’s saying statement. I heard it before and I don’t agree with it. I kind of understand why he might be saying it as a figurehead and somebody who’s really caught in a political context given the realities of the situation. But I would hope that they are working to ramp up our testing capacity so that we can use testing effectively, because particularly in China and South Korea, that’s what they use to help enforce the isolation and quarantine and directed really well.

S12: Communities that aren’t taking this as seriously as others. If you go through in the Times article mentioned, which Taib been looking into, many different states have either mandates or recommendations. And if you look at Texas, for instance, Euston in San Antonio are taking it very seriously. But if you go to Frisco, Texas, which is a large city of 200000 people, I’m not saying that they’re just having parties left and right, but the restaurants are maybe spacing customers from each other. My question is, Will, is it possible that the communities that don’t take it seriously because they haven’t had covered 19 yet are going to be untouched and they will say, oh, see, we were right.

S14: I think it’s unlikely. I think what you might see is it’s it’s you know, it’s it’s possible, I would say low probability that some individual communities may avoid this unscathed. But you have plenty of examples of communities that didn’t do anything for which the virus emerged in their community and it ran like wildfire and it got out of hand.

S15: And they wish they had taken more proactive steps to protect the population up front. We have to be precautionary about this. We have to pop practice public health prevention here. And it is painful. And I get that. I get that. We don’t want to shutter stores and businesses. And we are facing something that is going to have an enormous economic cost. And I think there are justifiable debates to be had that about whether or not is it worth it. I think you should have that conversation. And that’s what they’re basically saying if they’re not coming out. That, you know, is are the economic ramifications of 20, 25 percent unemployment having to do this for a prolonged period of time while you hopefully wait for a vaccine that’s effective to be developed and deployed so that you don’t have a health care system collapse in the health care services can be delivered. Can we ask people to live in their homes for a very long period of time and restrict their movement and at the same time risk economic ruination of people, people not being able to feed themselves and a prolonged depression coming out of it because it’s difficult to restart an economy. I think that’s a perfectly legitimate conversation that should be had. I think some of the measures the government is now already considering. With these stimulus packages and their ability to print money are needed, but they may have a lot of long term consequences. That is very difficult to actually estimate. And I you know, I’m not an economist, so I can’t speak to that directly. I don’t know if anybody can speak to it directly because we really don’t have an analog situation for making a reliable prediction.

S1: Jeffrey Shaiman is a professor of environmental health science at Columbia University’s Mailman School of Public Health. He was the lead researcher on the study that The New York Times wrote up as Corona virus could overwhelm us without urgent action. Thanks so much. Thank you for having me.

S8: And now, not the schpiel, but a clip of Governor Andrew Cuomo exhibiting empathy and connection in a time that requires literal disconnection. You have more time with family.

S16: And yes, I get tramaine cramped quarters can be difficult, but it’s also the most precious commodity for myself. This young lady, Kyra, is with me. She would never be here otherwise. You know, I’m I’m dad. All right.

S17: The last thing you want to be when you’re in Kyra’s position is hang out with the old man and hang out with dad and hear bad dad.

S18: Joe Chenault, come to the holidays. Come when I give them heavy guilt.

S16: But I’m not going to be with Kyra literally for a few months. What a beautiful gift that is. Right. I would have never had that chance. And that is precious. And then after this is over, she’s gone. She’s flown the nest. You know, she’s going to go do her thing. But this crazy situation, as crazy as it is, gave me this beautiful gift.

S8: It’s a good it’s a good sentiment. When I personally consider before Cuomo mentioned it. It’s one that our president seems utterly incapable. Love me. Did you know? Fun fact. The president actually has a young child at home. So maybe as a father, Donald Trump sits on the edge of the boy’s bed at night and explains that things are going to work out and it will be tough, but we’ll get through this together. That interaction does seem utterly impossible, doesn’t it? But what I liked about what Cuomo was saying was the words themselves, but also that it was a political leader saying it and doing yet not by stepping aside from his role of governor, but actually embodying it. It was in his capacity of leadership. It didn’t show weakness. It created an empathetic bond, which was a comfort in itself. OK. Let us all shelter or at least beautiful gift in place for the next few days. Let us rename are lockdown order operation beautiful gift. Operation Beautiful Gift shall continue indefinitely. Take care tonight.

S19: And that’s it for today’s show, Priscilla Lobby is the just associate producer, or rather she has had occasion to enable her associate producer status, but she never has had to avail herself of its full powers and responsibilities. Daniel Schrader is the justice producer. He invokes the Defense Production Act whenever each one of those off brand yellow sunshine cookies that aren’t Oreos. And he says, damn it, if the government were in charge, we’d have some creamy wayfor consistency. The jest as I pass out tonight for my fourth shot of grain alcohol, please remember to lower me onto the couch using quiet leverage.

S20: Quiet beverage room per adepero du Peru. And thanks for listening.