Lauren Ancel Meyers' lab at UT Austin analyzes data and models scenarios for how things could play out with COVID-19. So what can the data tell us about where the pandemic is headed next?
Meyers, a mathematical epidemiologist, joined KUT's Nadia Hamdan to talk about what goes into forecasting what's next with COVID-19 and how that could change.
This is the second episode of Now What? — a weekly livestreamed series in partnership with UT and the Dell Medical School, focusing on what the coronavirus pandemic means for our lives now and in the future.
The video will be available on Facebook and YouTube, via podcast and at KUT.org.
This transcript has been edited lightly for clarity.
KUT: What do the most recent models say about the spread of COVID-19 in Central Texas? Have we peaked? And if not, when might the peak come?
Meyers: We are looking at on a daily basis data for the entire five-county, Austin-Round Rock MSA. And what we see is that when the pandemic started spreading in Austin, cases started rising, hospitalizations started rising, we started seeing a few deaths, and they sort of got to this high point and they've stayed there. We haven't seen hospitalizations or cases or deaths start to subside. And in fact, with the relaxation of social distancing measures, we wouldn't expect that to happen at this point.
What we're hoping is that people are taking precautions when they're going out in public; they're wearing face masks, they’re keeping their distance so that what we don't see in the coming weeks and months is a large second wave or a surge in cases and hospitalizations.
KUT: And can you talk to me a little bit more about the second wave? Because I know some people are a little interested in understanding, you know, when was the first wave? And how do we define the first wave versus the second wave?
Meyers: So, in Austin, if you go on to Austin or Travis County Public Health website or city website, you can look at the data yourself. And what we track is the hospitalizations in Austin. It gives us the most reliable indication of how quickly a disease is spreading. And in Austin, we didn't really see what you would consider a wave. We didn't see cases come up and then go away. What we saw was this thing started spreading.
We slowed the spread by going into the “stay home-work safe” order. But it never really started to go away. And then we open things up. And so, in Austin, you don't see a wave that comes and then goes. You see a wave that sort of comes and now we have persistent community transmission of the virus. So the idea of a second wave is a bit of a misnomer.
I know I just said a second wave, but what we're concerned might happen and what we're worried about and planning for is that … we're going to see a second, more alarming increase in COVID transmission, in hospitalizations and in deaths, that lead to just a much more critical situation in our health care system.
KUT: And have we peaked? And if not, when might the peak come?
Meyers: We haven't really peaked, I would say. People are always talking about flattening curves. You know what? You've got to decide what you're tracking. Are you looking at cases? Are you looking at hospitalizations? Are you look at deaths? But these things go hand in hand, and as I just explained, in Austin, things started climbing, but they never started receding. We didn't wait to the point until we got to the point where there was no more disease to open things up. We waited until things had sort of slowed. The cases were no longer skyrocketing, but there was still disease in the community. And so, when you talk about something peaking, usually you're thinking about something that rises and then falls. But things never fell in Austin. And so, we have not peaked in the sense that it never went away. And what might happen in the coming months is we may see more of a rise in cases and hospitalizations and deaths before we see things going away.
KUT: And you kind of just touched on this, but a good question would be, what does peak actually mean? I mean, is it just the standard definition of what we all understand as a peak? Or is it kind of different when we're talking about these kind of cases?
Meyers: The idea of a peak is I think what a lot of people think intuitively. It's like a mountain or a hill. Something goes up and then it comes down. And in some places around the world, in some cities where there was very fast transmission and then the cities intervened with social distancing measures and other measures. And they held onto those measures for a long time, you know, places like New York and Wuhan. We really did see very, very large increases in pandemic transmission, much, much more rapid to much more alarming extent than we ever saw in Austin. And then those measures really did lead the virus to start to slow and cases to decline. And so in those cases, in those cities, if you look at their data, you really do see kind of a peak that starts to subside.
In other places like Austin and other cities where measures were taken really very early on and then relaxed, what we saw was that the virus slowed but never fully went away. So we don't see this sort of true peak structure that you see in other places where the pandemic kind of grew more explosively before strict measures were put in place. And really what we call epidemic curves, can take any shape you can imagine. You can imagine you slow spread. Then you relax things and then get a second summit, a second peak. You hold things in check. You could have a plateau.
Almost any geological structure can appear depending on the measures that are taken, how long we hold onto those measures, and how much people adhere to those measures. In general, the more we do to slow transmission, the more quickly the virus will recede in our communities. The more we allow the disease to spread, the faster the rise will be of the epidemics in our communities.
KUT: Whenever the peak does hit, what are the risks that still exist post-peak?
Meyers: I think there's a lot of different ways that an outbreak or an epidemic can peak. An epidemic can peak if you don't do anything. If we just had let COVID spread, first of all, it would've been catastrophic. We would have had many more people hospitalized than our hospitals could have cared for. We would have seen lots of people dying. But if you really allow an epidemic to just run its course without intervening, then epidemics on their own will eventually peak for viruses that immunize people — where after you're infected, you're no longer susceptible. So if you let a virus just run without any mitigation, then eventually, once enough people become immunized because they've been infected, then the virus will slow down and will eventually stop spreading. This is the concept known as herd immunity.
Once a sufficient number of people have been immunized who've been infected, there's no longer enough susceptible people around for the virus to keep spreading. In that situation, if you really allow a virus to just run its course, then a peak will happen. After that peak, it will not spread as much. The risk will be lower in that community that you'll become exposed because there just won’t be as many people infected.
However, epidemics and outbreaks can also peak because of intervention measures. So in a in a city like New York, which really had a lot of virus, a lot of disease and then a very strict stay home measure, a very strict sheltering in place order, it peaked. And then it started subsiding.
Well, the population hasn't yet reached the point where there's a lot of immunity built up in the community. So, if measures are relaxed or if they're relaxed here in Austin, then we still sort of have the same risk on the ground. The virus is still spreading. Most people are completely susceptible, not immunized because they haven't been infected. So, in a sense, nothing has changed. And so we have to continue to figure out ways to prevent transmission and to protect ourselves other than going back into a stay-home order.
How do we do that? We can do things like religiously wearing face coverings when we go out in public. There's a growing amount of evidence that by covering your mouth and your nose, you can protect yourself from being infected. But really importantly, if you happen to be infected but don’t know you're infected, you can prevent yourself from spreading disease to others. Also, keeping your distance when you're out in public. If you feel a little bit sick — you wake up with a mild fever, you have a cough — err on the side of caution. Stay home and stay away from other members of your household.
The risk is really still the same as it was back in March, because this virus still spreads just as fast as it did when it was first introduced into Austin and into communities around the United States. It can still spread silently. This is something that we really understand well now that people who are infected are contagious before they feel symptoms in many cases. And so you could feel perfectly fine, go out and about your usual daily activities and actually be infectious and potentially spreading disease to others.
Or you can bump into someone who looks perfectly fine and not realize that they are actually infected. Because it spreads quickly, because it spreads silently you may be past a small peak because you've intervened, but that doesn't mean you're past the risk.
KUT: You mentioned this at the beginning this talk of flattening the curve. Remind us what that means. Have we done it? It sounds to me like we haven't. Can you give us a little more understanding of flattening the curve?
Meyers: I just mentioned a minute ago what happens if you just let an epidemic run its course. If you just don't do anything. If we had allowed COVID-19 to spread unmitigated in Austin, we would have seen a very large wave of transmission that eventually sort of peaked. Along the way, a lot of people would have gotten very sick and died. That is what the curve looks like unflattened. You may have seen images of unflattened curves.
Flattening the curve is the idea that if we take measures that slow transmission, whether they are really strict stay-home measures or us taking precautions to prevent transmission when we're out in public, and staying home if we're not well, then the virus will spread more slowly than if we didn't take those measures.
What happens to an epidemic is it rises more slowly and it peaks later. At the peak, the number of infections, hospitalizations, and deaths is just not as many. So you go from what would have been a curve that shot up very steeply and peaked very highly to a curve that rose more slowly, didn’t reach its peak until later and that peak is lower. You may end up with the same number of people infected or close to the same number of people infected or hospitalized or dying in the end — hopefully not quite as many — but there is a benefit to buying yourself time.
The more time we have before we hit that peak, the more time we have to take measures that might even slow the spread more, to ensure we're protecting our most vulnerable populations, to ramp up testing and tracing and isolation that help us to contain clusters when they occur. The other benefit of flattening the peak is that we may be able to flatten it enough that we never have too many cases requiring hospital care. We never exceed the capacity of our health care system. So flattening the peak buys us time. It buys us capacity and it may ultimately save lives.
KUT: Your group has new research out about the timing and impact of implementing social distancing. What does this new research show?
Meyers: This is a paper that's just coming out in Emerging Infectious Diseases. In this study we collected data from 58 different cities in China that experienced small outbreaks, small epidemics in January, February, just as the pandemic was emerging out of Wuhan to cities around the world. What we looked at in each of these 58 cities was how many days was there between the first detected case in the city and the first social distancing measure. Many of these cities very rapidly implemented strict, strict measures. They closed restaurants. They closed schools. They told people to stay home. So how many days between the first case in the city and the enactment of those kinds of measures?
The second thing we looked at was how many days between those first measures and the city successfully bringing the epidemic under control. The way we measured that is how many days until the reproduction number of the virus is below one. Once you get to that point, it means that the virus peak will decline. We will see the peaks go away. What we asked is, was there a relationship between how long it took them to intervene and how long they had to battle the virus until they finally got to the point where the epidemic peaked. We found that for every day that a city delayed enacting their first intervention, they had to battle the virus for two and a half more days. Translating that to perhaps something in the United States, if we are able to accelerate interventions by a week, that might mean that we can we can shorten the time that we're in a strict measure by two and a half weeks.
So, it really just shows that the timing of when we take action can really make a difference in terms of how long we have to enact costly measures — things that are economically or socially negative. It also might just reduce the number of people that end up getting infected and dying from the virus.
KUT: So it sounds like you're giving a specific recommendation for what cities and other entities should take from this research. “Start earlier,” I'm guessing would be a recommendation you give.
Meyers: Yeah, that is a recommendation. But keep in mind that some of these measures are costly. We're not saying that we should just lock down today and err on the side of really early recommendations. But we're saying if the situation is changing and you see increasing risk, things have to change. Either individuals have to adhere better to the measures or there have to be stricter measures. You need to do that sooner rather than later in order to mitigate that increasing risk.
That's one study looking at data from China. But what we're working on on a daily basis with the City of Austin and Travis County hospitals is really thinking about when should we implement stricter measures or relax measures so that we can ensure that we're managing the risk in our community without going overboard, without spending unnecessary time under really restrictive measures.
KUT: I wanted to ask a question now from Facebook, because you had spoken about this earlier. It says, “Has any new data been introduced that gives us a definite answer as to whether or not durable immunity that would allow for herd immunity to actually develop truly exists? There have been concerns of reinfection still happening and worsening.”
Meyers: That's an excellent question. It is something that we do not know definitively. However, I would say that given what we understand about coronaviruses, and other viruses out there, there's sort of a tentative consensus that there would probably be at least short-term partial immunity following infection that would to some extent mitigate transmission once a substantial number of people were infected. So the answer is we don't know definitively. Most experts agree that there will probably be some degree of herd immunity as the number of cases builds up in the city.
KUT: Has there been a difference in cases or hospitalizations since Texas started reopening some businesses and loosening stay-at-home rules? Are you seeing any big changes in that?
Meyers: We're looking in Austin most closely. So I can speak to some of the data from Austin. The data are very, what we call, noisy. If you look from day to day cases jump, cases fall, cases, jump, cases fall. Same thing with hospitalizations. What we're not seeing is a decline. We're definitely not seeing a decline in cases and a decline in hospitalizations. We may be seeing a slow upward trend. Suppose on May 1, when orders were relaxed, suppose that on that day people changed their behavior in a way that increased the rate of transmission in our community by a little bit or maybe by a lot.
It will take a little while for us to really see that strongly in the data. If transmission starts to increase over a few weeks, the people that got infected over those two weeks, a few of them may end up in the hospital or dying, but they won't end up in the hospital for another 10 days. And if they die, they won't die for another three or four weeks. So we don't see alarming upward trends in the data yet, but that doesn't mean that it's not spreading more quickly. We may not see that in data for a couple more weeks. We know the risk is not subsiding. How much the risk is increasing, how quickly it's spreading — the jury's still out. We will be able to know more within a few weeks.
KUT: Dr. Mark Escott, the health authority here in Austin has recently said that there has been an upward trend that he's starting to see in terms of cases and hospitalizations. Earlier this week, we had our largest day over day change with the number of cases. I'm wondering, have you been monitoring that alongside the city and kind of seeing an upward trend yourself?
Meyers: Yes, I am looking at the same data that the city is providing. The city is doing an amazing job of making the data available not just to researchers, but to the public. They now have two dashboards that provide a lot of the key indicators that we're all looking at: the cases, the hospitalizations, possible admissions, deaths. We are looking at all of those alongside the city. Yes, everything Dr. Escott has reported, you can see for yourself on the dashboard.
The reason we want those estimates is because we want to be able to provide really data-driven guidance as to: What are the risks? Are they increasing? When should we change measures? The data that we track most closely has been made available on this second dashboard the city has just put online. It's called “Key Indicators for Staging.” The data are the daily number of hospital admissions for COVID-19 across all hospital systems in the Austin-Round Rock area. We don't look specifically at how many are today. We actually look at a seven-day rolling average. So, what is the average number of new daily hospital admissions over the last seven days? That helps us, what we call “smooth,” so we don't see these ups and downs and jump to conclusions. I would say two days ago, we saw the highest number of daily admissions ever in Austin, for COVID-19, but the day before, it was a very low number.
Back to the question, are we seeing an increasing trend? The answer is maybe slightly. … Data will take a week or two for the increase in cases, to end up translating into an increase in hospitalizations. …The case data is great, it really tells us what's happening on the ground; however, because we're changing the number of people that are tested, we're changing which populations we're offering testing in — and this is true all over the country — sometimes if you see a jump in cases, it could be due to an increase in testing effort or a decrease could be because you're testing in a community that just hasn't seen much of a virus yet.
So we find that the hospitalization data give us the most reliable indication how quickly the virus is spreading in the community. However it's lagged by a couple weeks from when that transmission events actually happen.
KUT: Given the current path of COVID-19 cases and the spread, where do you see Central Texas headed? Obviously, you talked about that second wave and now you're talking about hospitalizations. So, where do you see us headed right now?
Meyers: I think it's a big unknown. We understand a lot about how this virus spreads, but what we don't have a crystal ball for is the choices people are going to make, what their behavior is going to look like in the coming weeks and policies that might come down the pike either from the city, state or the federal government.
The way this virus is going to spread, how quickly it's going to spread and who is going to be at risk is absolutely going to depend on the decisions we make on a daily basis. As things have opened up, if we continue to be cautious when we go out in public and do the things we talked about earlier — cover our faces, keep our distance, and stay home when sick — the extent to which we as individuals do those things, we can actually prevent the virus, the pandemic from spreading quickly.
If people get more relaxed, if people are taking more risks, then the virus will spread more quickly. In one scenario we could see a very large wave occur over the summer, maybe peak in the late summer or even in the fall. In another scenario, where people really are being cautious, we might not see a tremendous wave, a second pandemic that threatens our health care system. So, it really depends on our behavior in the coming weeks and months.
KUT: With a lot of loosening of restrictions at the state level I think a lot of people have been going out into the community in droves. We're seeing larger groups of people. Mayor Adler had some videos from bars where people were really crowded that first Memorial Day weekend where everyone went out once the restrictions on bars were loosened. I'm wondering, what would you tell people who are feeling a little more loose and a little more relaxed when they go out?
Meyers: I understand this is very difficult because it's sort of a silent threat. It didn't hit us hard in Austin. So many of us don’t know people who ended up in the hospital or died from this virus. So first, the thing to emphasize is the reason that didn't happen is because we went into this unprecedented, strict stay-home measure that stopped this virus from spreading. But as we look at the data from around the world in places that were slower to take action, we see that this is not a garden variety virus. This is not your typical flu. This is a really deadly virus. It's not just people who are old or have health conditions who end up dying and end up in the hospital for this virus.
The thing to emphasize is that even though we may not have seen it personally at home, there's a reason that the world has taken such drastic measures. It’s because it really is a deadly virus. And unfortunately, it spreads so quickly and it spreads silently. You might be next to someone in a bar who looks perfectly healthy, but they could actually be infected and they could spread disease to you. And even if you're not someone at high risk, even if you may just have a very mild case, you might end up spreading it to someone that you care a lot about, who is high risk, who ends up in the hospital.
So unfortunately, it's a really severe virus. It spreads really easily, sometimes silently. And it really is kind of upon us to make decisions to protect not only our own health, but our family's health, our neighbor’s health, our community health. Yes, things have loosened up and you can go out, you can see your friends, you can have fun, but we really all should be taking steps to limit transmission when we do go out. Keep your distance, wear a face covering and please don't go out if you or someone in your family has even mild symptoms that we know are associated with this virus.
KUT: How do you and the UT COVID-19 Modeling Consortium create these models?
Meyers: The models we create and analyze are really based on, at least in my own team, on about 20 years of being experts in this field. I've been building models of epidemics, of pandemics since 1999 when I was a postdoc and I was working with the CDC. I got to UT in 2003. Since I've been here at UT, I have worked with public health agencies around the world to help respond to the SARS epidemics, to the 2009 H1N1 influenza pandemic, to the West African Ebola epidemics and the emergence of Zika in the Western Hemisphere a couple of years back.
These models come out of my own work – a couple of decades of just understanding, of learning, of researching how virus is spread and how human behavior impacts the spread of viruses. And my work rests on centuries of research and epidemiology and infectious disease dynamics and the actual tools — the actual mathematics and computer science, the engineering that goes into them.
In the UT COVID-19 modeling consortium, every day we are meeting on Zoom, working with dozens of researchers not only from around the world, but from many different disciplines that are bringing different perspectives, different tools, different kinds of expertise that have actually allowed us not just to build the kinds of models we've been building for last 20 years, but really to very rapidly innovate so that we can we can build models that actually look at daily cell phone geo location traces from millions of people and use that to inform our estimates for how fast this disease is spreading. We’re using science and information and knowledge that rest on centuries of research and we're also using very innovative tools that are coming from collaborators across many different disciplines.
KUT: Well, I could ask you a dozen more questions about the modeling research, but I also want to make sure we also ask one important question, which I think is something that is on a lot of people's minds. Many people are curious about whether or not school could or should happen in-person this fall. Does any of the consortium's modeling work speak to forecast what might happen if schools open and how, if possible, to open safely?
Meyers: We have certainly done research not just during the COVID-19 pandemic, but for other threats looking at how contacts that happen in school, how do they impact the transmission of a virus not only in schools but into the community. And what we know for many different respiratory viruses is that children can easily spread viruses in schools and then they can bring them home and the kind of sources that that propagate virus spread in communities. That's what we know for viruses in general.
We have certainly done research not just during the COVID-19 pandemic, but for other threats. Looking at contacts that happen in school, how do they impact the transmission of a virus, not only in schools but into the community. And what we know for many different respiratory viruses is that children can easily spread viruses in schools and then they can bring them home and the kind of sources that that propagate virus spread in communities.
And so, you know, that's what we know sort of for viruses in general for COVID, there have been some questions raised regarding how frequently children actually become infected, how infectious children are once they do become infected. And there's still a lot we don't understand. And that's partly because in many countries around the world, the COVID-19 pandemic has happened at times where schools are either already closed or they were closed very quickly to respond to the pandemic. So we don't have much direct data on how quickly the virus will spread in schools once they're opened. Regardless, what our models tell us is that if schools are open and there are not steps taken to reduce the risk of transmission, then the opening of schools can increase the rate of spread in the community at large and can lead to future waves of transmission.
But what they also tell us is — and this is in any context, in any industry or school or health care setting — that if measures are taken to limit the number of contacts that people have per day, so if kids just aren't having as many contacts with other kids, and to limit the risk of transmission when people do come in contact with each other by putting desks further apart, coming up with some sort of strategies so that kids aren't going to the washroom at the same time or walking into the school or leaving the school at the same time. Steps can be taken just to simply minimize the number of close encounters between people, and if steps are taken to reduce risks when those close encounters happen, like wear face coverings.
And also if we think through: how do we ensure the health of our vulnerable populations? You know, children who have underlying conditions, teachers and staff who have underlying conditions or maybe live in households with people with underlying conditions. Anything we can do to especially ensure their health and safety will be critical to opening schools in a safe and effective way.
KUT: I'm wondering, how do you personally process the data points, especially when you're charting hospitalizations and deaths, knowing that those are real people who had friends and families? We talk a lot about data. But at the same time, we're talking about people and we're talking about lives. So can you tell me a little bit about what that's like for you?
Meyers: It's very stressful. I have to say, to be honest. Even though I just said earlier I try not to, as a scientist, look at the day-to-day ups and downs because I know that that's not a reliable signal. Every day. I kind of wait apprehensively for the daily update. My stomach sinks when the numbers go up. And, you know, I'm thinking about people, and I have a little bit of relief when I see the numbers go down. And so I think it's sort of always there in the background. This is actually what motivated me to go into this line of work and motivates a lot of the work we do today. And I think a lot of people in the consortium feel the same way. A lot of people who have kind of signed up to help and to lead projects, these are people who really didn't know much about epidemiology.
They're experts in different fields who brought different kinds of tools to the table, and so we really like to help. And I think that's what's motivating a lot of people. They want to do whatever they can to bring science and to bring data to the table to help decision makers make whatever decisions they can make to ensure that we're saving as many lives as possible while also being cognizant of other economic and social costs. There's different kinds of interventions, and I think that's why we're all trying to work around the clock to do what we can do.
KUT: I do also want to ask, especially given this moment, are you concerned at all about the effect the spread of the virus may have due to all of these protests happening within the city? Obviously, you know, it's not just here. It's across Texas, across the country. And we're just seeing a lot of people packed together. Yes, many of them are wearing masks. But still, in terms of social distancing, that's not really happening. So I'm wondering, you know, are you at all concerned about that?
Meyers: Yeah, I'm very concerned. Absolutely. I mean, the more I see people wearing masks, I feel a little more relieved. This virus spreads when people come in close contact with each other. There's respiratory particles and spitters, whatever, coming out of their mouths. And so the more people are coming in contact with each other and the less they're actually protecting themselves, the more opportunities there are for transmission. And when it's large numbers of people, you run the risk of what we call super spreading events. If you have one infected person who is not wearing a facemask and is shouting loudly, and there's a lot of people around them in close physical proximity, there's a lot of potential risk of exposure. And so I'm really hoping that these don't amplify transmission. But it certainly is increasing the risk in our cities.
KUT: Is there anything else before we conclude the interview that you'd like to say right now?
Meyers: I guess I would just say that I feel really fortunate to live in Austin and to be a scientist in Austin. I have colleagues around the world and around the country who do what I do. And I don't think there is any one of them that lives in a city where the policymakers, the city leaders, all of the leaders of the major health systems and the experts, the scientists, the community leaders are all really working together on almost a daily basis to try to understand the threat, to try to keep tabs on it and to make policies for the months ahead that are really kind of geared towards protecting lives, protecting safety but also thinking about people's social and economic well-being.
If you found the reporting above valuable, please consider making a donation to support it. Your gift pays for everything you find on KUT.org. Thanks for donating today.