Coronavirus (COVID-19): Press Conference with Michael Mina, 06/08/20

You’re listening to a press conference from the Harvard T.H. Chan School of Public Health with Michael Mina, assistant professor of epidemiology and a faculty member in the Center for Communicable Disease Dynamics. This call was recorded at 11:00 a.m. Eastern Time on Monday, June 8.

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


MODERATOR: Dr. Mina, do you have any opening remarks?

MICHAEL MINA: No. I’ll take questions.

MODERATOR: All right. Great. Looks like our first question.

Q: Hi, Dr. Mina. My question is in respect to the governor’s announcement several weeks ago that he had set goals for the number of daily testing that would be happening, and that would be viral testing. And he said that he wouldn’t – the administration’s goal is to test about 45,000 people a day by the end of July and 75,000 a day by the end of December. The question is, do you think that’s realistic, particularly in light of the fact that the number of daily tests being done since he made that announcement has declined from sort of the low teens, 13,000, 12,000 twelve down to 9,000, 7,000, 6,000. So, yeah. Do you think that’s a realistic goal at this point? And do you expect him to reach it?

MICHAEL MINA: I think that the goal to have the capacity and ability to do it should be there. Whether or not it happens currently will depend on demand. And as we see that the incidence of the virus is decreasing in many ways, as anticipated, it hasn’t been known how much it would necessarily decrease, but I think that, you know, potentially this summer and then the normal seasonality of this virus is taking some effect, we might anticipate seeing continued reductions in demand if incidence continues to reduce as a result of both weather and social distancing.

But I think that having the capacity to do that type of testing is absolutely essential. My anticipation is that in the fall, if things continue to be controlled throughout the summer, my anticipation is that in the fall we will probably see a resurgence of this virus that matches sort of what we normally see with coronaviruses, which is usually that the virus gains an incredible amount of momentum throughout September and October, peaking usually sometime in October, November, December. So at that point, if the virus returns, I think we want to make sure that all of that testing is in place and we can monitor outbreaks as they as they begin again. Certainly, we still have plenty of transmission ongoing.

So, we want to ensure that we have testing that remains available. And I think one of the things we can continue focusing on now in this state is to keep increasing access to testing for for anyone who feels that they need it. I think that’s still a difficult test to receive if you don’t have the right sort of Know-How of where exactly to go to get tested. So using some of that capacity currently to ensure that everyone who feels they need a test should be able to get it, I think would be a wise choice right now. So I guess it’s not quite the answer you’re looking for, I suppose. But I do think that building capacity is absolutely essential right now.

Q: Thank you.

MODERATOR: Next question.

Q: Yeah. Thank you so much for taking the question. I want to ask you a bit about like immunity, which you’ve talking about a lot and where this relates to vaccines. So say everything goes well. We get a large study that proves that it’s effective at preventing people from being infected, at least for that duration when we do the study. How long do we expect that protection to last? And when will we know the answer to that question? And then the third thing there, does the answer for that change depending on what sort of platform or technologies are being used?

MICHAEL MINA: So we hope that the protection will last a very long time. We expect, I think, that the protection probably will not be extraordinarily durable and very well might only last six months or a year, for example, before a booster vaccine needs to be given. We know that this virus occurs seasonally and we know that it’s likely that people get this same virus more than once. Hopefully not with disease the second time, but that repeat exposures of the seasonal coronavirus can serve, for example, as booster infections, if you will.

So I don’t think I’m not personally holding out too much hope that the vaccine will lead to really two to very durable immunity. But I do hope that it leads to some immunity sufficient so that if somebody becomes exposed, they will not have disease. And any exposure that does happen can actually serve to just boost the immunological memory response to make it more durable. The type of studies that will have to be undertaken to understand this are going to be long term efficacy studies where we continue to follow both people’s risk of of infection as well as their immunological responses and antibody responses over time. And essentially that will be – to understand what the durability of the immune response elicited by a vaccine is, we will have to monitor closely what happens, not just in the short term after somebody gets a vaccine.

We can assume that in the months after a vaccine is provided that people will have a spike in their antibodies, in their serological response. And we’ll then assume that some of that will wane fairly readily. The question is, will it wane to levels that are below the threshold of protection or will it wane and then kind of plateau off to a to a steady state that is protective? And that’s a difficult question to answer without just waiting and seeing what the population level dynamics look like. So we’ll have to monitor lots of people of different ages, different co-morbidities, and follow them in time to understand which, if any, lasts longer than others, for example.

So I don’t think we’ll have a good answer for just how long the immunity sustains itself, or probably months or years as we continue to monitor people after in the same way that we have to do this for any vaccine that comes about.

Q: Great. Thank you so much.

MODERATOR: Next question.

Q: Hi. Thanks. I wanted to ask about how many actual coronavirus cases we think there are compared to the confirmed or reported number. I have seen some estimate that the actual number in the US is like five times or even 10 times as high. So that, you know, if right now we have 20,000 thousand new confirmed cases a day, you know, do we have some good idea or estimate of how many actual new cases a day? I mean, is it really 100,000 new cases a day or even higher than that?

MICHAEL MINA: So at the peak growth rate of this virus, when it was growing quickly, I think we can expect that the underreporting was quite high, at least tenfold higher during the early stages of the epidemic. During the very early stages, it was much higher than tenfold because we had no testing. And so we had no confirmed cases, and yet now we know that there were probably many, many thousands that had spread before we really got any testing available. So over time that number has come down. I would say we’ve continued to increase testing and the epidemic has continued to slow its course in terms of growth rates. It’s, of course, plateaued to sort of a steady state, unfortunately at somewhat of a high number.

But I think that now we know that somewhere around eight or nine out of 10 infections lead to asymptomatic infection. So we hope that contact tracing of symptomatics is at least helping to identify some fraction of those individuals. My guess is that testing has become sufficiently robust, despite it still being, I think, somewhat inadequate on a national level in terms of access. I think it’s become sufficiently robust in most places to enable detection of symptomatic people. And so I don’t think that it’s currently 10 times higher. But it is very likely to be at least a couple of fold higher, so two or three and maybe five times. It’s difficult to say the actual number without serological data.

We do know, of course, we look at New York City. We know New York may be 20 percent or more  serologically positive, which suggests, of course, that the underreporting from confirmed cases from viral tests was very high, that the underreporting was very high. So it’s difficult until we continue to get more and more serological surveillance data to really understand that number. But in general, we’re finding that the population prevalence has been somewhere between around two to five percent in most places in the country. And so we can do the math on that. But I don’t anticipate, for example, that it’s tenfold off at the moment.

Q: OK, thanks. And do you think I mean, are there policy implications from that, like if if the number really is a lot higher? I mean, does it get to a point where there are just so many cases out there that we can’t really do contact tracing and the contact tracing is not a, you know, good thing we should be trying for? Or should we be really trying to do, like, diagnostic testing of these cases, no matter how much there are, and then trace them?

MICHAEL MINA: I think that contact tracing is the most powerful when cases are more or less controlled or at least in lower numbers and you can actually get to all of the individuals who need to be traced. Once pathogen replication and the population is increasing exponentially, then it’s very difficult. But once you’re in an equilibrium, which is essentially where we are in the States right now, you can then at least allocate sufficient resources to make sure that you have enough contact tracers if it’s doable.

But certainly the most effective contact tracing happens when case counts overall are low. And then you can actually try to make programs and policies that are driven by contact tracing to detect and find groups of people who are infected and and isolate them, quarantine them versus when the outbreak is sort of out of control and spreading unabated, then you have to take the drastic types of measures that we took over the last few months in order to control it and then also have some element of contact tracing during that. So it’s in general more crucial when you have smaller numbers and actually capturing most outbreaks is feasible.

MODERATOR: Next question.

Q: Hi, how are you? I was wondering if you could comment on today’s study published in the journal Nature that says about 4.8 million confirmed cases or about 60 million US coronavirus infections were prevented as a result of regional and national lockdown measures.

MICHAEL MINA: Well, I have I did notice the paper, but I have not read it yet. So I don’t think I can really comment too much on it, although I would agree that without having done the math or seeing the paper yet. I think that probably a lot of measures have served an important role in our society to prevent and mitigate unabated spread of the virus, probably saving many, many millions of lives in doing so, at least from the infection. Of course, all of that needs to be balanced against economic consequences as we look to the future and what policies lay ahead. But I can’t speak too much more to the specific paper.

MODERATOR: Any other follow up questions?

Q: Yeah. I mean. OK. Well, looking at that, since I can give you the number that you know, that they predicted, about 4.8 million confirmed cases were prevented. Do you think that’s what experts believe the fall will look like as the US reopens? Are we looking at millions of more infections?

MICHAEL MINA: I think that absent of – there is a risk, yes, that that we will have millions of additional infections, and I think that should be expected that we will have that. I think that mask wearing and having systems set up to identify when local transmission is occurring in somebody’s community and then sort of flipping switches on and off to say, OK, for the next two weeks, everyone has to be an incredibly diligent about wearing masks and trying to social distance as much as possible. And then in the interim, when there’s not transmission locally, for example, people can sort of go about their daily lives while remaining cognizant of the fact that there’s a virus circulating, I think we’ll have to sort of balance that.

But I do expect that things like mask wearing and being more aware will overall help to diminish cases. But when we look at the seasonality of this virus, the entirety of the epidemic on U.S. soil has occurred during a period of time that the virus is usually waning. If we consider that the seasonality of this novel coronavirus will match what we know from other seasonal coronaviruses, we usually know that by the time February and March come around, the virus is sort of going back down in numbers usually without any without any sort of other community involvement, just as a natural course of seasonal seasonal decline. And then it usually comes roaring back in the fall. So we actually have yet to experience this particular novel coronavirus during a season when it is usually most transmissible.

And that’s a scary thought, I think. But given that knowledge, I think that we should probably plan for and expect that we could potentially have millions of additional cases in the fall. And we should just plan accordingly to to ensure that we are preparing for it and hopefully warding off those cases by putting in place preventative measures like mask wearing that still allow people to function and society to function without causing collapse of the economy, but while also mitigating spread.

Q: OK. Thank you.

MODERATOR: Next question.

Q: Hi. Thank you so much for hosting us and for taking my question. Just on that subject, I wanted to ask you a little bit of, you know, when states were reopening and it was viewed as somewhat preemptive, there was a lot of talk of a sort of second wave of the virus happening as a result of that. I’m hoping you can address whether the US has seen that now. It seems like from what you’ve been saying, that you’re expecting more of that kind of activity in the fall. And I’m hoping you can also address whether there are any states that you think I’ve seen as sort of second wave or where cases are spiking in a way that’s concerning.

MICHAEL MINA: So there are states where cases are continuing to increase. And, you know, whether those are entirely linked to reopening plans or other events is difficult to say at the moment. I think certainly we will learn a lot, for better or worse, from the protests that are happening, the protests are, without making any sort of comment one way or the other on the protests, I think that they are serving in part as a natural experiment that we can at least look to to understand from infectious disease epidemiology lens to look for hints of whether or not these types of social interactions at this moment in time and during the course of the year is going to necessarily result in increased cases.

So if we if we know that many tens of thousands or hundreds of thousands of people across the country are gathering and we don’t end up seeing spikes of cases, then we will, at the very least that will give us some inkling that perhaps the season and the weather is actually on our side here. And, of course, people do seem to be wearing masks in a lot of these protests and so disentangling that is a little bit difficult. But so far, we haven’t seen massive shifts in increases. That said, there have been a number of sports teams, for example, I know on college campuses and university campuses where the reintroduction of students to campuses is now resulting in the spread of the infection among players. And I think there’s been some news reports about that recently as well. So that gives us some idea that that it’s not foolproof, that the weather is not going to stop everything. Masks won’t stop everything.

We are seeing increases both on campuses where people are returning. We’ll probably see some areas where increases happen in businesses and then in states that open up very quickly. The question is, how much will that, how big will those potential increases be? And do we need to sort of change course as a result? Or can we expect and almost actually hope to see some small amount of transmission? And that sort of gets back to the very beginning of this epidemic and pandemic when when all of the discussion was surrounding flattening the curve. That still exists. And that’s an idea that we still want low levels of transmission in order to slowly be building in a controlled way through herd immunity.

And so I would say that just having transmission alone isn’t the end of the world. But what we don’t want is for that transmission to get out of control. And we just have to continue monitoring the situation in states that are reopening, and places where there are protests, and businesses that are reopening, and treat all of these through a very scientific lens. I think that they’re all needed opportunities to be able to watch, as long as they’re happening anyway, then we should be using them to learn as much as we can about the virus so that we can ward off major outbreaks later on.

MODERATOR: Do you have a follow up question?

Q: No. That’s great. Thank you so much.

MODERATOR: Next question.

Q: Oh, hi. Thanks so much. I actually wanted to follow up, if I could. We are seeing states right now with both increasing numbers of cases, increasing percent positive rates and increasing hospitalization rates. Do you think it’s possible, actually, that whatever we’re calling a second wave or not, that we may already be in the midst of it and that further that the way this happened in March and April, where the entire country went through a spike at the same time, is not necessarily what a second wave will look like that we may, in fact, see Utah spike, ahead of, whatever, New Mexico and things like that? So those are my two questions, sort of are we in the middle of something right now that maybe we don’t recognize yet? And then second, is that a pattern that we should anticipate?

MICHAEL MINA: Yeah, I think that we should not anticipate, at least during the next few months, that we will see these in an extremely punctuated outbreaks. So the waves might be more rolling waves rather than spiking at waves, if you will. And that makes them a little bit more difficult to understand or at least difficult to detect. We can expect that some states might have different, that those rolling waves will come across different states at different times, and that’s pretty normal. We’re a very large country and we know that some of these waves get started by super spreading events that are localized, for example. And so I think that we should definitely anticipate that some states as we open up, I do think will see increases.

But I also see that there is going to be this potential countering force, which might be the weather. June, July, August is usually the the valley in terms of seasonality for coronaviruses. So we might be getting lucky if, you know, in terms of as states are opening that the increases in cases would be a lot higher if it were currently September, October. And so what that means is, I think in individual states, you know, a lot of these new cases might be driven not so much by a homogeneous sort of increase of a wave across the state that’s sort of not going across the border, but rather it’s probably being driven – at the state level, it looks very rolling, but it might be more spiky, smaller outbreaks within that state owing to individual events that lead to communities spread.

And so I think that, you know, we might not see a massive outbreak across the whole country all at once. And even during this this last phase of it, we saw essentially just that. A number of states did have pretty large outbreaks, but even among those that had the largest outbreaks, they happened all sort of in somewhat of a staggered fashion. And it wasn’t until quite a bit later, for example, that we started really seeing upticks in cases in the middle of the United States and more rural areas. And that’s something we can – we see that with other infectious diseases as well. We see sort of cases roll across the United States, not necessarily all exactly simultaneously. So it’s not an unprecedented thing to see that kind of heterogeneity across states.

Q: So if I can just ask them, what’s your definition of a second wave? What would it take for you to call something a second wave?

MICHAEL MINA: Probably something that approaches what we saw already. I think that having small rumblings of new cases, even when they’re increasing the overall number, I don’t know that I’d call those necessarily a second wave as much as sort of the first wave bouncing back and just kind of hitting the equilibrium in some of these states. We want to, of course, see numbers go down. But if numbers continue to be sustained and increasing for a matter of weeks, I think that I would start calling that potential evidence of a second wave. It might not get huge if the states are now – I think that there is a lot of machinery or mechanisms in place within states to squelch the epidemic if it starts to occur.

And that might be everything from closing down schools for a couple of weeks to messaging across the population to wear masks and to sort of closing down the whole state again if needed. So we’ll probably start to see reactions to any small ripples and increases over time. They might not be the drastic type of responses that had to happen in the last couple of months. But there will be reactions. And if we remember back to February and March and even into April, there was almost no sense of wearing masks.

In fact, we were getting the opposite message. The CDC was putting out the opposite message saying masks aren’t useful if you’re not infected and things along those lines. So we really were not working as a community, I would say, to mitigate spread. And so now I think any sense of a second wave might actually be squashed pretty quickly through something as simple as everyone putting on masks. If we though, I would say that the definition for me of a second wave would be to start seeing exponential increases in cases for a sustained period of time, maybe a few weeks. And then I think I would get very worried that we’re witnessing second waves in certain communities.

Q: Thank you.

MODERATOR: Next question.

Q: Thank you very much for taking my questions. I have one and then a follow up. My first question is, are there any states in particular that are doing a pretty good job of testing, tracing and isolating? And if so, are any of them approaching, at least approaching the effectiveness of the South Korean model?

MICHAEL MINA: To be honest, I haven’t followed the individual testing, tracing policies of other states well enough to really intelligibly answer that question. South Korea and some other place have done really extraordinary jobs. I think that it would be hard to approach that given how many cases continue to occur across the whole of the United States. I would say that there have been a few places where testing and and contact tracing have been rolled out. Massachusetts has pushed for it with the help of Partners in Health pretty widely. And even here, it’s been hard to keep up. So, I am not quite sure just how just how well it’s working across the United States. And I don’t know of places that it’s necessarily rivaling the successes that we’ve seen in South Korea.

Q: OK. Thank you much. My follow up question is, you know, if if a state has the testing capacity and tracing capacity to routinely test congregate care facilities like assisted living homes and nursing homes, should they do surveillance tests without there being even a positive case, would that be helpful or is that not necessarily the right thing to do?

MICHAEL MINA: I think there is efficient ways to test. I do think senior senior living centers and nursing homes remain the areas at greatest risk for heavy burdens of mortality in their populous and high morbidity if these viruses get in. And we know there are some that have done very good jobs at just completely preventing any infections on their premises, and that’s been through extraordinary measures to not allow any visitors essentially, nor to allow people to leave. In other cases when the virus does get in, it often unfortunately burns through the facility and can cause massive destruction and death.

So I think that continuing to do surveillance testing through a number of different avenues can be extraordinarily important in these centers. And whether that’s through pooled viral testing to increase efficiencies or through serological testing through sort of longitudinal serological testing, where you’re testing everyone at baseline and then following them up and looking for any new antibody positive individuals can also be a very efficient way to continue testing people throughout the marathon of this pandemic. And so I think that each nursing home in each state will come up with their own policies surrounding it.

But I do think that continued surveillance of these settings that may be reliant in terms of the frequency of what that surveillance looks like, should probably be informed by the community prevalence overall. So if you have a very, very low community prevalence, you may not have to do it quite as extensive surveillance within the nursing homes or the senior living centers because the risk of entry of the virus from that community remains low. Where if you have a high amount of population prevalence out in the wider community in which that center sits, then I would suggest very frequent sampling to ensure that any potential outbreaks are being squelched before they can get out of control.

Q: Can I just – to make sure I understand you. So you’re saying that if there’s a high prevalence in the wider community, then at these assisted living centers, even if there’s not a positive case yet, you would recommend comprehensive surveillance testing of the residents and staff.

MICHAEL MINA: Yes, I think that’s accurate to what I was saying.

MODERATOR: Next question.

Q: Hey, Dr. Mina. Thanks for taking questions. Dr. Woloshin at Dartmouth had an interesting piece last week – I don’t know if you saw it in the New England Journal – about the accuracy of PCR COVID-19 tests and sort of unknowns about clinical sensitivity, false negatives. I know we’ve touched on this before. I’m just wondering – it’d be great to get your perspective on what do we know right now about the accuracy of the molecular tests and specifically how many false positives, you know, how many cases we might be missing? And is there something that FDA should be doing differently or that, you know, other other parts of the health care system should be doing to to get more information about that accuracy?

MICHAEL MINA: I’m less concerned about false positives, because acting, at least in in the current setting, acting on a false positive probably isn’t a bad thing. It’s probably not going to harm anyone. And you can always do repeat testing to see if if somebody was actually positive or not. If they get positive on one day and then they’re negative for the next two days, then there can be an expectation that it was a false positive. Now, that said, if the policy is just the moment you have a positive that you get sent home or you’re not able to work for 14 days, that is a false positive then for that individual that can be quite debilitating in a number of ways. So I think it’s important to understand those.

We do know that false positives happen, particularly on different, on certain instruments, you can have contamination. There’s a lot of ways that false positives can occur. And it is important to continuously do quality controls on your methods to ensure that you’re not getting contaminants that would be causing false positives. False negatives are a more difficult thing to tease out because, at least when used for screening, with false negatives, you probably would not go back and test that person again unless they become ill. And with nine out of 10 people not necessarily showing much symptoms, you probably wouldn’t go back and test those people again. So it’s very difficult to to know exactly what the number is. But in general, the assays themselves are very, very accurate. If the virus makes its way in any sufficiently high concentration to get picked up by the PCR, it will turn positive. But the swab itself can oftentimes be misleading because the swab just might not be picking up the virus.

And so that’s always a tradeoff. And I liken the tradeoff very much to the difference between an MRI and an x-ray. We all know that x-rays aren’t perfect, but they can often be good enough if you’re pretest probability, meaning the probability or likelihood that somebody actually does have lung cancer, for example, if that’s a very low probability that maybe an x-ray is sufficient to screen that person and decide that they can come back in another year. Versus some people, if you have high suspicion tthat they have some some problem, then maybe even if the x-ray is negative, you send them to get an MRI just to be sure. So those similar types of tradeoffs exist with infectious disease testing. No test is 100 percent accurate. We know that the swab in the way that we look for this virus is not 100 percent accurate at all. The swab often can be negative despite a positive individual.

That said, most people who are transmitting the virus will be positive on the swab. Most of the false negatives that we get are occurring in people who are likely at the tail end of their infection, maybe are still RNA positive bed, no longer even have infectious variants inside of them. And so whether or not it’s actually important to capture those people from a clinical perspective, it might not be clinically that relevant anymore. So I think along with this question of are there a lot of false positives, false negatives, I think with the false negatives in particular, we really have to have a little bit more nuanced dialog surrounding what does a false negative mean. If somebody is three weeks post-symptom onset and they have a bunch of residual RNA that is lingering potentially for six weeks, four of those six weeks might not be transmitting virus. It’s just RNA that’s sitting around in the nasopharynx. So if you were to miss those because they’re very low viral counts or very low RNA counts, is that a true false negative? It’s a false negative, technically speaking. But is it truly a clinical false negative or is it actually an inaccurate result?

And so a little bit of a philosophical question about laboratory testing, but I think there is a lot of confusion today about what defines something as a false negative or a false positive. Well, false positive is not too unclear. But what defines something as a false negative, I think we need to be a little bit careful. There’s a lot of people who might have single molecules of viral RNA that stick around for much longer than six weeks, but of course, those might not be detected if it’s just a single molecule that’s around. So we have to sort of define all of these. And that’s why on tests, we usually define a positive or a negative based on some limit of detection. And instead of saying, we usually don’t say this person is free of virus, we say this test has failed to recover any positive virus or if there is virus, it’s below the limit of detection of the test. And then it becomes the clinician’s duty to understand how to interpret that.

Q: Thank you, that’s really helpful one. One more test question, if I may. We here have started to hear from kind of a new group of test makers talking about developing antibody tests for what they call neutralizing antibodies, suggesting this is going to better get us at that question of immunity of who’s really able to neutralize this virus. I just open it up. It’s hard to evaluate, you know, how much credence to give these tests. I haven’t read a lot about it in the research. Is that going to be a really useful tool beyond the many antibody tests that are already out there?

MICHAEL MINA: Neutralizing antibody titers are what we look out to understand if somebody is actually protected. And it’s not always 100 percent. But if you can say that somebody – if you say that a drop of somebody’s serum can neutralize virus on a petri dish, that’s important to know. If you can dilute that drop of serum 10000 times and it still neutralizes that virus, then that’s even more important to know. That means that person has a lot of neutralizing antibody, for example. So we look at these dilutions series and try to understand, okay, somebody who does have antibodies, how many times can you dilute out their serum before that diluted serum no longer serves to neutralize the antibodies. And that gives you some idea of the strength of somebody’s antibody mediated immune response against the virus.

There’s a lot of other things that interact with the immune system that could also serve to neutralize the virus. T cells being one of them, other antibody producing cells that might only get activated when somebody gets reinfected. So even if somebody doesn’t have a high neutralizing antibody titer at rest, if they were to get re-exposed, they might have a burst of new B cells develop from their immune memory pool and that then sort of create a whole lot of shorter lived a neutralizing antibody. So it’s not the end all be all metric but what we do with it, what the most important and crucial piece to use from neutralizing titers, because it is a laborious assay to perform. You actually have to have some sort of live virus that you neutralize on a petri dish. Doesn’t necessarily have to be the true coronavirus, you can use pseudotyped viruses to make it safer. But essentially, it’s not a quick and rapid serological test. So what we do with it instead is we try to correlate antibody titers from regular serology, from things like ELISA-based tests.

We try to figure out, is there a correlation when somebody is above a certain value on their ELISA test, does that correlate with a protective level of neutralizing antibody titers? And so we actually look to make those correlates. That’s what we call correlates of protection are oftentimes driven by those correlations between the two. And so it can be very important as a tool to gather enough data and make enough correlations to say, OK, we know that anyone is above an antibody level of random units of 400 should usually, 95 percent of the time, has sufficient antibody to neutralize this virus at this level. So we develop those correlates of protection and then we can map those on to ELISAs later on and sort of use that as our road map to understand what an ELISA means. So it is very important, but it’s not something that we’ll be introducing and running on millions and millions of people. We will probably use it to inform our ability to better interpret ELISA-based assays.

MODERATOR: Next question.

Q: Thank you very much, Dr. Mina. I have a question about all these asymptomatic cases, because many places, including nursing homes and assisted livings, are screening staff every day, sometimes twice a day, by basically screening them for symptoms. When we’re opening up, people are coming back to work, many of them are being screened for symptoms. How useful is this?

MICHAEL MINA: Well, it’s not bulletproof. It’s more useful than nothing for sure, as long as – well, there is a chance that it can backfire. As long as people are understanding that not having symptoms does not equal not having this virus, it’s better than nothing. There is, of course, the opportunity if the correct interpretation is not put in place for people to act inappropriately, just because they don’t have symptoms. So if somebody doesn’t have symptoms and they don’t recognize that that doesn’t necessarily mean that they’re not exposed and infecting others with this virus, then they could end up acting in a way that could cause them to spread the virus to large numbers of people. So I think all of this needs to be considered as one small piece of a larger program to prevent outbreaks, especially in nursing homes and facilities like it.

It’s useful in that it should serve to prevent some people who are infected from coming to work. And so it doesn’t mean that it will prevent everyone who’s infected from coming to work. But we hope, for example, that it could be that people who transmit the most might also be the most symptomatic. So if we can prevent those people from coming to work, then that’s a great use of these sort of syndromic tools for surveillance and monitoring. But I would say that we don’t have – but it’s certainly far from 100 percent safe as a way to prevent infections. And it should be considered as one piece of a larger puzzle of how to keep these communities safe.

Q: A follow up question – is the rest, is the other piece actually repeat testing?

MICHAEL MINA: I think that’s one of them. I think that, you know, I personally have a hard time considering this idea that we will test every single person every single day for the indefinite future. I think that there have to be better, more measured approaches to this. It could be, though, that we do have pretty significant efforts to pool testing, to have everyone essentially put a Q tip in their nose every day, drop it in a bucket, and all of those samples get run together as a single pool. And then, you know, that would increase efficiency of testing and I think could be a manageable program for facilities and for institutions.

I also think population surveillance, if prevalence is low or during times when population prevalence is very low, I think that less frequent testing can be can be deployed. It could be antibody based testing. It could be viral testing. In some ways, antibody based testing when prevalence is very low can be more powerful because you can oftentimes miss viral a person’s viral positivity. If you are testing every two weeks or every three weeks in a fraction of the population, you might entirely miss an asymptomatic person during their positivity stage. But you probably won’t miss them after about 10 days post infection when their antibodies come up, because then their antibodies just stay up for good, so you don’t risk that transient nature of the virus. But both of these tests have their positives and negatives. And I think that testing will be a very big part of of dealing with this coronavirus, particularly in nursing homes and the like and in the future.

Q: Thank you.

MODERATOR: Next question.

Q: Thank you, Dr. Mina. Thanks a lot for taking the time. I was wondering if you could speak directly to the protests and the effect that that might have on spread, particularly, you know, how worried you are and if there are any steps that we should be taking or if there are steps that you’ve noticed that people are taking in response to what could be obviously a super spreader event.

MICHAEL MINA: Well, I think that from an infectious disease epidemiology lens, I agree that these run the risk of becoming super spreader events. I think, as other people have noted over the last week or so, that these are happening outdoors is up is one good step to to sort of reduce transmission. It does seem like being outdoors in the way that airflow is distributed, of course, helps prevent prevent viruses from lingering in concentrated patches, for example. Certainly wearing masks as much as possible will help.

But, you know, I said it last week and I’ll say it again, I think that there are some things that that society at different times decides are more important than other other seemingly important things. And controlling this virus has been extraordinarily important over the last few months and probably I think the way that we controlled the virus can’t be completely untangled and unlinked from the way that these protests have developed. In general, that’s a different question. But I think that society is kind of choosing that it’s an important time to seize this moment and try to effect real change that has other public health consequences, hopefully beneficial in the future, in particular, for young black American men.

And so it’s it’s difficult to say if this is right or wrong from an infectious disease perspective or from an overall public health perspective because I think that there are hopefully going to be positive consequences that come from these protests in the long run. But I think that to best prevent these protests from becoming super spreading events, then people just have to be as cautious as possible, recognizing that none of it’s going to be perfect. But if people feel like they are sick, probably I hope that those people will sit it out and not go and protest if they think that they’re sick or at least go get tested first. And I know that friends of mine and neighbors where I live who have gone to the protests, are now asking me how they can go get tested so they can be sure that they weren’t exposed during the protests. And I think it’s a good idea. If you’ve been around thousands of other people for a few days, I would say go get tested, try to figure out what your status is at that moment in time so that you know, you know, whether or not you run the risk of spreading into your family and things like that.

Otherwise, it’s a difficult thing, I think. Again, the low prevalence that we’ve gotten to, at least a lower prevalence in some areas than we were at. So that’s good. And I think that will help prevent massive spread at this moment in time and again, I think and hope that the weather and being outside will help. But I don’t think – I do expect that we will see some outbreaks occurring as a result.

Along those lines, I would say that for people who have gone to the protest, just be very careful not to, especially if you are younger and robust and your odds are that you won’t get severely ill if you do get infected, but I would highly encourage anyone who is going to the protests to try to otherwise self isolate and quarantine for  a week or two, two weeks in particular, before going and seeing any family that might be older.

Q: I’ve got a one related follow up. I did see on Friday – I came in a little late to this call, so it may have been brought up – but on Friday, there was a spike in cases that went up to, I think, 28,000. And I’m aware that that could reflect an increase in testing capacity. But my question is about the geographical distribution of that testing and whether you believe we’re testing the right people. I think a national number, you know, increasing capacity plus the increase in positive test could obscure where the virus is growing. And if you’ve noticed that or if that number affects how we should read national testing percentages and increases in cases.

MICHAEL MINA: Yeah, I’m a big proponent of breaking down this virus to as geographically isolated locations as possible and evaluating it sort of from the – either from the top down or from the ground up doesn’t matter, but just taking one number and using that for the country can very well be misleading. In Massachusetts, we have a large number of tests available per populace, in other states, not so much. And so I think that it can obscure.

Where the testing is happening is certainly not homogeneous across the country. We hope and it’s likely that more testing is available where the virus spreads more readily, which is usually metropolitan areas, which also happen to usually have more testing available than more rural areas. But that doesn’t mean at all that the distribution of tests is equitable across the country. And it also suggests that the distribution of tests can certainly, or the heterogeneous distribution of tests, can certainly obscure actual spread and results there. And so I think that we need to be very cognizant when we’re reading reports to understand, you know, are we missing whole groups of people as a result of lack of testing in certain areas. And surely across the country we are missing entire communities, I’m sure. And so all of our decisions have to recognize the limitations of the tests that are available anywhere, but in particular the geographic spread of these as we take a broader picture of the United States as a whole.

Q: Thanks a lot, Dr. Mina. I appreciate that.

This concludes the June 8 press conference.

Michael Mina, assistant professor of epidemiology (June 5, 2020)

Barry Bloom, professor of immunology and infectious diseases and former dean of the school, and William Hanage, associate professor of epidemiology and faculty member in the Center for Communicable Disease Dynamics (June 3, 2020)

Natalia Linos, executive director of the Harvard FXB Center for Health and Human Rights (June 2, 2020)