Coronavirus (COVID-19): Press Conference with Yonatan Grad, 04/8/20

You’re listening to a press conference from the Harvard T.H. Chan School of Public Health with Yonatan Grad, the Melvin J. And Geraldine L. Glimcher Assistant Professor of Immunology and Infectious Diseases. This call was recorded at 11:30 AM Eastern Time on Wednesday, April 8.

Previous press conferences are linked at the bottom of this transcript.


YONATAN GRAD: I’m happy to field questions. We’ve been getting some questions on our preprint on anticipating the trajectory of the epidemic under a variety of different social distancing durations and levels of success. I’m happy to talk about that.

Another question that we’ve been thinking about is how to do serological surveys and then how to interpret those results, given that we anticipate they’ll probably be done with convenience samples at first. So, we’re happy to talk about some of the thinking around serological testing and surveillance as well. But I’ll leave it open to questions.

MODERATOR: OK, great. Thank you, Dr. Grad. Our first question –

Q: You can always count on me to ask the Massachusetts question. The one for today is, are we in the surge? We’ve heard politicians talking about the surge. And how does an epidemiologist define when a surge has begun? Thank you.

YONATAN GRAD: You know, I am not sure what they mean by the surge. So, I mean, certainly, we expect, for epidemics, there to be exponential growth at the start of an outbreak. And that is consistent with what we’ve been seeing.

I imagine – although, I don’t, of course, want to put words in anyone’s mouth– that what they’re worried about with the surge is the number of cases that are being hospitalized relative to – or the pace with which, the rate at which we’re seeing hospitalized cases relative to the number of hospital and critical care bed availability or relative to the number of beds available.

So, I imagine that’s what they mean by the surge with the anticipated crest of the wave being the peak of the epidemic, but I’m not sure that that’s what they mean. So, it’s hard for me to address exactly what that is.

But, from my perspective, thinking about the epidemic, what we expect to see is an exponential–  a continuing exponential increase in the number of cases, consistent with those diagrams that I think everyone has seen where we see this kind of bell-shaped curve for the epidemic.

Q: OK, thank you.

MODERATOR: Next question.

Q: Hi, I’m with The Hill newspaper down in DC. I wanted to ask about the West Coast. California and Washington state seem to be doing better or not as bad as some people feared.

I wondered if you think that’s accurate. Do those states seem to be doing better? And what might explain that? Have they just been very good at their mitigation? Or what’s going on over there?

YONATAN GRAD: The data that I have seen is consistent with what you’re describing. I think that seen the same things that other people have seen that suggest successful, at least in some areas, successful flattening of the curve. What could explain that? Yes, I do think successful mitigation efforts, which refer both to the timing, so being able to do it early after first cases were observed, and the extent.

So you may have seen, in The New York Times, there was – looking at cell phone mobility data, there was a figure that indicated that the average distance traveled in the Seattle area dropped over a month from somewhere – I think it was 3.8 miles down to 61 feet. So, in terms of physical distancing or social distancing, depending on which term you prefer, that seems to have been rather successful.

So I think linking the success of the mitigation efforts together and the social distancing and other interventions together with the timing is particularly – could be particularly instructive. This underscores the importance of a lot of work that’s currently going on where people are trying to really formally investigate the impact of these different intervention strategies, their timing relative to the presumed or the known rise in cases or initiation of cases in a particular area.

How well do you do these interventions work? The variety of different interventions from school closures to stay-at-home recommendations, what have they – how well have they done in actually reducing mobility? And then how well has that reduction in mobility translated into reduction in cases and reduction in deaths?

So, being able to look across the US where we’ve seen very fragmented responses, but also very different timing of introduction of epidemics, being able to interrogate all of those data together will enable us to really get a sense of which mitigation strategies were successful. But, going back to your original question, it really does seem like this may be due to early response, early mitigation efforts, and ones that were successful in reducing individuals’ mobility and kind of promoting this notion of social distancing.

Q: Thanks.

MODERATOR: Next question.

Q: Yes, thank you for taking my questions. I actually have several questions, but I’ll stick to two that are related and then see if any other hands show up and come back to the others, perhaps. A lot of talk about the curve and, in fact, it’s, of course, several curves happening in rural areas and counties and small towns and big cities and so forth sort of in – waves is not the right word, but at separate times.

So, I wonder if you could talk a little bit about or a lot about how these curves are likely to play out in terms of both timing and severity. So, for example, will a relatively sparsely populated county see a similar curve as New York or another hot spot, even though the overall numbers would be a lot lower? And then, related to that, could you also talk about when we might see the overall peak in the United States, given all these different curves happening across the country?

YONATAN GRAD: So, you raise an important point alluded to in my last comments and in that question as well. There is no one single curve being experienced by the country or by the world. What we’re seeing is local epidemics, which have staggered start times, so to your point.

The shape of those curves, in the absence of any intervention, is influenced by the contact patterns within those locations. So, in places where there is a lot of contact among people, we might expect to see a steeper upward slope and a steeper downward slope and higher numbers, something that reflects a higher R0 or basic reproductive number.

In places where, just by virtue of how they’re structured, we see more social distancing, that’s the equivalent, basically, of seeing a flatter curve. And that may just be, as you were again alluding to, in rural counties, there are different patterns of interaction. So, the shapes of these curves may be different.

The shapes of the curves will also be different because of the different timing and type of mitigation efforts and the differences in the success of those mitigation efforts, so recommendations – not just the recommendations, but how well they’re actually followed by populations. So that will also very much influence the shape of each of these individual curves.

So, it’s hard to say exactly what a curve will look like with mitigation efforts. That depends on, again, how well they’re actually adhered to and what they are. And, to the same point, trying to predict for the country as a whole what the cumulative shape of this curve will be, as we try to add up all the different waves in all the different communities, very hard to do in the absence of much better understanding of the timing of the initiation of the epidemic in these different places, the nature of interactions, and the impact of whatever mitigation efforts are being put in place.

So, I’d say everyone wishes that they could have an answer to that, but I think it’s a moving target, particularly as interventions are changing and being introduced, and people are following them to different extents.

MODERATOR: OK, it looks like we’ll go to the next question. Please, go ahead.

Q: My question is, today, we are seeing the lockdown lifted in Wuhan and, immediately, high amount of travel in and out of that city. I’ve heard some people say they felt this was not a good idea because it could possibly lead to a second wave, given there is still no vaccine, and, therefore, more shutdown orders, waves of them. I wanted to ask you your thoughts on that and then also ask you, for us, how likely do you think it is that we may see multiple waves of lockdown orders, not just the one we are in?

YONATAN GRAD: One of the concerns that we have about this notion of whether we’ll see one wave or multiple waves centers on the fraction of the population that remains susceptible because of mitigation and containment efforts. The more success we see in containing the epidemic, the more susceptibles remain in a population. And those susceptibles are then a population in which we could see a second wave emerge.

So, one way to view this is what would happen if we did nothing. If we did nothing, we would see the epidemic go through a population, and we would see a large fraction of the population infected. And, if infection confers protective immunity, then we would expect that there would just – the number of susceptibles would be depleted, and the epidemic would come to a conclusion because there was sufficient protection in the population. This is the herd immunity concept.

As we introduce mitigation and containment efforts, the extent to which they’re successful will prevent the epidemic from going through the population to that extent, meaning that there will be some individuals in the population who remain susceptible. Those then – the size of that population would then – or that population would then be at risk for subsequent waves. So that is just, basically, the underlying concepts here.

In Wuhan, it’s not clear to me what fraction of the population remains susceptible. Unless there was a lot of unreported infection or very high rates of asymptomatic infection that we’re not aware of right now, it would suggest that still a substantial portion of that population remains susceptible.

So, it suggests that to prevent another wave or outbreaks will take a lot of vigilance for surveillance and then case identification, contact tracing, quarantine and isolation, consistent with the types of efforts that China put in place initially. I think they’ll have to maintain them, given the extent of the population that’s susceptible.

This is similar to what has been observed in Singapore. So, in Singapore, they did a very good job of containing the outbreak, but then they’ve had to deal with additional outbreaks, as people have been coming in and introducing – we think introducing COVID-19 again. So, they’ve started to see additional outbreaks that have required recrudescent mitigation efforts.

Could we see the same thing here? It’s quite possible. I think that, formally, the same processes are– this is just how epidemics work. So, it could be that, if we are very successful, then, again, those places that are most successful and that have susceptible populations still will be at risk for additional waves.

How we manage that I think will be very dependent on what tools we have at our disposal. So, if we have at that point a vaccine, we can get to the point – we can get to herd immunity. We can actually prevent additional waves by vaccinating a lot of the population.

If we have therapeutics, then that is another possibility because we could help reduce – that is another thing that we could have in our arsenal that could shape how we respond because maybe that will help reduce the stress on hospitals and critical care beds if the therapeutics either prevent progression to severe disease or are able to help people recover rapidly.

Otherwise, with non-pharmaceutical interventions, so the social distancing and stay-at-home orders and so on, masks, those would then be what we’d have to turn to. So, for places like Seattle or Washington where it seems like their containment efforts have been successful, as we talked about earlier, I would worry that the extent of the susceptible population means that they will be at risk for another wave.

Q: Yeah, I was just going to say that the numbers didn’t make sense to me because, with 83,000 cases, but 11 million people in that city – I have a lot of friends in Singapore. And it’s much easier to control there. They take people directly from the airport to hotels. It’s also an island. They see very easily how people are coming in and out. There are not as many egresses as Wuhan.

One more question, will, one day, you believe that, when a vaccine is developed – because, obviously, there’s a lot of effort for that– this will join sort of the arsenal of mumps, measles, rubella that this will just be something standard that we all get or young kids get?

YONATAN GRAD: I think that is certainly one plausible scenario. And that is – I think it’s hopeful to imagine that we will come to a vaccine. It is not a certainty. And I think that is just a caution that we need to keep in mind.

There are a lot of very smart people working very hard on trying to come up with a vaccine and many different approaches being taken, but it’s not a given that we will get one. So, again, I very much hope we will, and I know that people are putting maximal effort towards that end, including developing the infrastructure such that, if we do come up with an effective vaccine, it can be mass produced and distributed quickly.

And I think those are all the right things that we need to be doing, but it’s also important to recognize it may not – we don’t know whether we will be successful yet.

Q: OK, I thought it was one of the things that, if enough resources and time, it would likely happen, but I realize now that that may not be the case. Thank you so much.

YONATAN GRAD: I think – I think the operative word in your statement is likely.

Q: Yeah.

YONATAN GRAD: We still don’t have a vaccine for HIV.

Q: Yeah, right.

YONATAN GRAD: So worth keeping in mind that, even something where decades of effort had been poured in, we have not gotten a vaccine. Now a coronavirus is a very different kind of situation from HIV for a variety of reasons. So, I’m not saying that that should be our guide.

And there are many viruses for which we have come up with effective vaccines. You mentioned measles, mumps, and rubella. Those are three. So, it certainly can be done, but, again, I am just being cautious by saying it’s not a given.

Q: Sure, thank you so much.

YONATAN GRAD: You’re welcome.

MODERATOR: All right, and just want to check in with you if you have another question or if not.

Q: I do. I do, thank you.


Q: Yeah, I messed up my mute button before there, but that was a fascinating discussion. So, actually, it was among my questions, and I’m going to do a follow-on on that immunity discussion. Will, at some point, we be able to figure out how many people have actually been infected? So, say, in three months or six months or a year from now, can this be accurately determined by random testing of people who were never diagnosed or something like that so that we have a better answer as to the questions that you just spoke to?

YONATAN GRAD: Yes, so people are working very hard on this, so trying to come up with a way to do population serological surveys. So serological tests look for antibodies to SARS-CoV-2. There are serological tests that exist already. They were one of the first things that people were working on because of their importance in identifying people who have been infected.

And so, they may not have as good utility as a diagnostic, like a PCR-based test, but, because the immune system develops antibodies to infectious agents, they could be very useful for asking whether people have been exposed to SARS-CoV-2. And those types of tests can be used in doing population surveys to estimate the fraction of the population that has been infected.

And, of course, as you scale that up, the more people you test, the better view you’ll have of the population. You can imagine testing everybody to see, for each individual, whether they’d been exposed or not. But, of course, we have various methods for trying to estimate the fraction, given a sampling of a subpopulation.

Now that depends on a few things, first, that you develop antibodies. So, we assume, because this is what we see with many other pathogens, that, if you’re infected, you do develop antibodies. But for how long do they stick around?

And how much does this depend on the extent of symptoms? So, could it be that people who have very mild infections or, potentially, who are asymptomatic, do they develop as robust an immune response, in terms of both the amount of antibodies that they generate and then the duration with which those antibodies stick around?

Assessing the presence of antibodies is not necessarily the same thing as assessing immunity so, actually, protection from being infected again. So, it could be that, if you have enough antibodies generated after an exposure and infection with SARS-CoV-2, it could be that, above some amount of antibodies, you are protected against getting re-infected. We do not know, at this point, if that’s the case or what that number might be. But that is another really important issue to address.

And then, relatedly, we expect to see waning of immunity because that is what we see, again, from our knowledge of other infections that protection generated by infection or by vaccination can wane over time. So, you may be most strongly protected immediately, but then it may go down with time. This, for example, is why, for some vaccines, you need to get re-vaccinated regularly.

So, there are I think still some really important questions around serological surveys and our interpretation of them that will need to be done. But, getting back to your original question again, yes, we think serological surveys will be a critical tool for understanding the extent of spread within a population. And we really hope that those will be able to be rolled out in the near term.

In fact, I’ve seen some initial small studies from various locations where they have already started doing this. So, hopefully, even in the next few weeks, we’ll see serological surveys that can expand our understanding of the extent of spread and the fraction of the population that may have been asymptomatically infected.

I’d also add onto that we need these serological surveys now to address these questions in part because of the challenges we initially had in doing diagnostic testing. If we had large-scale testing from the start, we would probably have a much better sense of what the answers are to those questions, even without serological testing. But now we really need these serum surveys or serological surveys to answer these questions.

Q: OK, quick follow on that, so what are the odds, if this is even possible to say, that this virus mutates enough that it changes all of that? And how would it change that picture?

YONATAN GRAD: I can’t speak to odds about that. But your question gets at one of two reasons why we could see, at a population level, the equivalent of waning protection. It could be that it’s just the waning of the immune system to a given strain. And it could be that what is circulating changes, like what we see with influenza where, from season to season, roughly, we see change in the parts of the virus that we develop immune responses to that help protect us.

So, is that going to be the case here? I don’t know. I think that is going to be one of the things that we’re going to track very closely.

Q: But let me press you just a little bit on that. So, we had MERS, and we had SARS 2003. Historically speaking, do we have any sense at all that coronaviruses tend to mutate at higher or lower rates or possibilities than influenza?

YONATAN GRAD: So, there’s mutation rate, and then there’s change in the antigenic parts of the virus, right? So, mutation rate is different from antigenic change where the antigenic change is the change in the parts of the virus that we develop immune responses to.

The mutation rate is about what we see for RNA viruses generally. That does not mean that the antigenic –  the rate of change of the antigenic sites is going to be the same as flu. For SARS-CoV-1 and MERS, I think that the question is it’s a little bit tough to answer for those because they haven’t spread so widely in the population that there might be pressure from host immunity that selects for novel antigenic strains. For example, I think that, overall, the numbers for both of those has been relatively small.

For the other coronaviruses, I think the numbers are higher, but I haven’t seen the genetic or, relatedly, antigenic data to suggest whether there’s antigenic change. However, there was a study using one of the circulating coronaviruses that causes cold-like symptoms that – I believe this was a study in military recruits where they challenged them with the virus and then challenged them again with the same strain a year later.

And they were able to see that, a year later, people were still able to be, after an initial challenge and infection, were able to be infected again. So, it suggests that at least the immunity to the circulating coronaviruses may wane fairly quickly.

And then, for SARS, people looked at titers, sort of the extent of antibody response to SARS-CoV-1– or the duration, excuse me, the duration of that response and saw that the antibody levels went down I think it was after around three years.

So, we don’t know whether that decline in antibodies also means a decline in the extent of protection because no one was going to challenge people again with SARS-CoV-1 to see, OK, the antibody levels went down, but are they still protected. That kind of study won’t get done, but at least there is evidence out there that immunity may, for other coronaviruses, wane over time.

Q: Thank you so much for all that. I appreciate it.


MODERATOR: OK, next question. If you would like me to ask, what is your sense of when we’ll be able to return to somewhat normal life? And how do you see that playing out? Do you see it being gradual, a gradual process? And what activities do you think we will able to resume first?

YONATAN GRAD: These are questions on everyone’s mind. And I think no one has an answer, either from the perspective of political will or from the perspective of public health as yet. So, a couple of colleagues, Michael Barnett, Caroline Buckee, and I, had an op-ed in The Washington Post last week where we underscored the importance of using this time to try to understand which mitigation efforts were successful in slowing the spread of SARS-CoV-2 and then trying to use that information to help shape the decision-making around how to safely emerge.

But I think that, right now, it’s very hard to make guesses about when we’ll be – what will be our threshold for re-emergence and what the staging for emerging from all our different social distancing will be. I think, again, it’s going to depend on location, given how fragmented the experience of this epidemic is by location and then by different types of interventions. So, I think it will really depend on a variety of factors and may not be uniform across locations.

This concludes the April 8 press conference.

Marc Lipsitch, professor of epidemiology and director of the Center for Communicable Disease Dynamics (April 7, 2020)

Benjamin Sommers, professor of health policy and economics (April 6, 2020)

Michael Mina, assistant professor of epidemiology (April 3, 2020)