Coronavirus (COVID-19): Press Conference with Michael Mina, 06/12/20

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 11:30 am Eastern Time on Friday, June 12.


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

MICHAEL MINA: Well, there’s a couple of names. I don’t recognize, I think. So I’m an assistant professor of epidemiology and immunology at the Harvard School of Public Health. I’m also a molecular, one of the medical directors at Brigham and Women’s Hospital where I oversee a molecular virology diagnostics. And in general, I study how the immune system interacts with pathogens and how we can utilize information from the immune system to model outbreaks of viruses. And in the context of COVID. I’ve been focusing a lot on testing and all things serology and virology. So happy to take more or less any questions. If I don’t understand or have a background of it, I’ll let you know.

MODERATOR: Thank you, Dr. Mina. Looks like first question.

Q: Thank you, Mike, again for doing this. I think for many of us who hope never to go back into lockdown ever again, one of the really central questions is the relative contribution of the different mitigation measures that have been used to flatten the curve here in Massachusetts, New York, other places that are hard hit. Can you say what do you see?  Do you feel like we have good enough data to hang our hats on on that at all from what we’ve seen so far? And what would you highlight as some of the more important – who’s doing that research?  Where should we be watching for the folks who will be figuring out the relative contributions of, say, social distancing versus masking versus hand-washing versus shut downs, all the other stuff that you can do? Thank you.

MICHAEL MINA: So you’re you’re essentially asking what are the relative contributions of non pharmaceutical interventions to reduce transmission and reduce risk and reduce overall outbreak size? And I think that, so certainly some of us are doing that work. Myself and some folks in my lab and in collaboration with some folks at Princeton, we’ve been modeling a lot of these efforts and other people in the Center for Communicable Disease Dynamics are also doing the same. Hopkins. So there’s a lot of the core sort of epidemiologists in the infectious disease modeling world are certainly, now that we kind of have some grasp of how this virus moves when it’s unabated, there’s a lot of effort going in by all the groups to understand what mitigation strategies are optimal and and how they play together to reduce transmission.

So I think there are sort of two different schools of thought, if you will, about how to stop these outbreaks from occurring. And one is sort of preventing transmission through things like masks and putting plexiglass between people and institutions and things like that. And then the other option is to just really be blasting the community with testing and try to detect any infection once it happens before, you know, before it gets out of hand. One of them is much more expensive than the other. But I think that they were both certainly have a role here. If you ask me my opinion about which is going to be the most effective or which, what’s the simplest, most effective option: mask wearing, I think there’s no doubt about that, that just wearing, if everyone adhered to wearing masks, which we have to be realistic and assume that’s not going to be the case, but if they did, we probably would keep R0 below one and we probably would not have outbreaks run away from us.

What we’re seeing and know our community and our country and our society in the US anyway is not accustomed to wearing masks and the discomfort that comes with it. And so I have a hard time believing that that will be the solution everywhere. But it does lend itself – we’ve seen a lot of data now, if we look at Hong Kong, for example, Hong Kong did have cases. Hong Kong is extraordinarily densely populated, but it really didn’t have any big outbreaks. And that’s probably, I think, more due to mask wearing than exceptional contact tracing. And so in that sense, having mask wearing can be a just, it alone can be a game changer if we as a society collectively decide that we will adhere to to those rules.

I think social distancing and keeping apart that will work. But it comes at an extreme cost. And that cost is both economic. But even if we are able to keep all of our businesses going, I look around at our own at the research institutions. And for those people who are in the labs or doing computational work to research this virus or their normal research, you know, it’s hard to see the long term. It’s hard to see that happening long term in terms of morale. And that’s going to be the same in every institution, every type of business. When you have whole populations sort of staying home, even if they’re working home and keeping the economy running from home, I worry very much that morale will get very low. So I don’t think it’s a sustainable solution. But mask wearing, on the other hand, is quite sustainable.

Q: And when super quick follow up, can you apply, can you give any number at all to that, even a very preliminary one, to the difference that mask wearing makes?

MICHAEL MINA: Probably a lot. If both – I think there’s been some – I haven’t personally included mask wearing into our models that we’ve been running, which have been mostly focused on testing strategies. But if you have two people both wearing masks, I think it will probably cut down on likelihood of transmission by over 90 percent. Even having a pretty, a low quality mask could, I think, really cut down from an infected individual, it might alone cut down on transmission quite a bit just by reducing the droplets that go out to the air. And so I think that it will do a very good job. You know, it it doesn’t take, we don’t have to cut transmission by 90 percent to get R0 below one.

And that’s the important piece. We don’t actually have to prevent every infection. And I’ve been sort of changing as I’ve continued sort of advising and being on all these different reopening groups and advisory boards and things, what I’m trying to push people to start thinking more in line with is looking at this as a marathon, not a sprint, A and B, that we should try to change our approach from this idea that we have to stop every infection to we have to stop every outbreak. And I think when we started thinking about stopping every outbreak, then all of a sudden the tools that we normally utilize for public health monitoring and control become present again and very powerful. We allow a little bit of leeway. We don’t have to have everything perfect, but as long as it’s good enough, we will stop the outbreaks from happening. And I think that’s a pretty important piece here.

Q: Thank you.

MODERATOR: Next question.

Q: Hi. Thanks for doing the call. I want to ask about a batch or grouped testing, and I’m talking about what you take a whole bunch of swabs and run them on one test. I guess the idea is more efficient, saves time and money. Just want to know if you can quickly explain how it works. Some of the benefits and drawbacks and most importantly, whether any businesses, states or countries are doing it already and how it’s going. Thank you.

MICHAEL MINA: Yeah. So we’ve been researching that both theoretically and then backing up with some of our analyses of samples. And it can be extraordinarily efficient. And again, if your goal, there are a few things that that make it difficult but if you can get the logistics in place, you could reduce your costs for the number of tests you have to do tenfold, if not 100 fold.

Theoretically, you could pool one hundred samples into one and and run all of those. And that’s because the viral loads that we’re concerned about stretch across many orders of magnitude. So from 10 viruses per microliter to a trillion viruses per microliter. So if we’re pulling even a hundred samples, if somebody’s viral load is a billion and you pull a hundred samples together, then you’re still going to have a viral load of 10 million in that pooled sample. So it’s still going to be very easy to detect. And if your viral load is so close that diluting out just by 100 fold is going to lose the signal, then you’re not actually running that much. That person probably wasn’t infectious anyway at that point because that would be a very low viral load relative to a trillion. So the point is you can actually pull huge numbers of samples theoretically and still get a very good signal and be able to detect outbreaks.

The difficulty is then if you get a pool of 10 people or 100 people, what do you do with that information the moment you get that result that says positive? Do you then have to call up all 100 people and tell them that they need to go home immediately or put on their mask? And, you know, not not move, you know, whatever it might be. And then you have to kind of go back and retest them all, for example. So it could be a very anxiety provoking effort if you if there’s a long duration of time between telling somebody that the pool was positive to getting them their individual results. And that’s always a problem that we think about when it comes to turnaround time with tests, and especially when you have like this would be called a screening test, which would be sensitive, but not specific to the individual. And then you’d have to go to a very specific to the individual test.

And that window, the shorter you can make it, the better if you’re doing pool testing, because we don’t want to have people living in a state of anxiety for two days while we repeat all the tests. There are institutions, I can tell you from my experience sitting on a lot of committees of quite a number of different universities across the country at this point about reopening, I would say that this this conversation continues to come up in industry and universities and everything else. Pretty much anywhere where you have leadership looking at how to keep their constituents safe, I think that the question of pooling is now arising because people are figuring out how are we going to test this many people and not drive ourselves into the ground financially. Even a place like Harvard with its large endowment has to still be considering this because it can, if you say that you’re going to test 5000 people daily at a test that costs fifty dollars, these are just examples, that ends up becoming extraordinarily expensive.

So I think we’re going to see this continue to become an increasingly utilized tool. I don’t know if anyone that’s using it at this very moment as they’re as their actual policy. I think there’s a lot of experimentation. That said, we have been – my research group and some others at the School of Public Health have been advising different policymakers in sub-Saharan Africa about how to potentially use these pooling strategies for different reasons. I mean, they’re very resource constrained. They only have so many tests that they can run. And so they are very interested in pooling strategies for just to figure out how to get the people tested. And not so much having to do with costs, but just material. And so both of these are happening, and I think we should anticipate seeing pooling become a very important part of the way that we monitor for these viruses in the future.

Q: Thank you.

MODERATOR: Next question.

Q: Hi, Dr. Mina. Thank you for doing this. Just to follow up a little bit on the colleges as they’re trying to figure out their recommendations, including on testing. Is there a certain frequency that would be required to be meaningful? Would you have to test, say, twice a week? Every student, every faculty member?

MICHAEL MINA: So it really depends, again, on how you’re approaching this. If testing if your goal is to prevent every single infection, then yeah, you really want to test every single day. That’s not reasonable. So if your goal is to prevent every outbreak from growing, then there’s a lot of range. I would say testing once a week – again, and I’ll come back to this because I think testing once a week can be a powerful tool – but if you can work with an adaptive strategy, so if you test every one day, every two days, every three days and maybe even every four days, I think that all of those will pretty much stop most spread. But testing everyone, that assumes comprehensive testing of every single person, every one to four days. That is still very, very laborious and difficult to do logistically. Particular places – it’s one thing for, you know, universities and and institutions in Boston to approach it that way because we have a number of testing facilities right here in Boston. But if you’re out somewhere else where testing is hours away in terms of the laboratory, it makes it very difficult.

So what I am beginning to advocate for is to have an adaptive strategy to testing, to control outbreaks where you have some routine surveillance. It’s not comprehensive. It could be done, for various, reasons serological surveillance. Just doing sort of the routine surveillance over time could be with antibodies. It’s cheap. And the important thing with antibodies is that once somebody develops them, they stay. So your sensitivity gets extraordinarily high at the population level to know that an outbreak is starting. The problem with antibodies is that you won’t find those first cases. But if you can switch gears very quickly and say, OK, we’re going to be testing on a routine basis for antibodies every, you know, maybe we test we have a rolling, rolling testing strategy where every person in a community test themself once a month.

And so you get a certain fraction of people on every day. That will give you a lot of power to detect that an outbreak is starting. And then once you detect that, which would probably be one to two generation times of the virus in, then you throw all the PCR and daily testing and comprehensive testing into the mix to find out exactly who is infected. And you can stop it quickly. And that’s where I think we have to start if we want to use approaches like that, which can greatly increase efficiencies and reduce stress of doing all the testing, in particular at times when community prevalence is low. We just have to be willing to tolerate a few infections.

And in many ways, this is the whole idea of flattening the curve anyway. The idea was never to take this acute respiratory infection that has the ability to spread like wildfire and prevent every infection. If this was Ebola, I wouldn’t be saying this. But this is not Ebola. And I think that our goals should be realistic. And if we allow ourselves to say we want to just have a good system that’s good enough to detect outbreaks early and then battle them by throwing in sort of all of the artillery, if you will, then I think we have a lot of opportunity to be flexible and figure out how to do this in a very efficient manner.

Q: Thank you.

MODERATOR: Next question.

Q: Hi. Thanks again for doing these. So picking up on what you just said about the what our goals, you know, policymaker goals should be. I just wanted to get your thoughts on some of the news of the last couple days with Oregon and Utah. And I think Nashville pausing their reopenings in the face of spikes that they’re seeing. What do you make of how policymakers should react to these spikes that we’re seeing in a bunch of different places?

MICHAEL MINA: Well, I think so far a lot of the conversation has kind of been binary, it’s been where we’re opening and we’re just opening. You know, and I know I understand there’s been phases, but it’s sort of been this binary approach. And I think that there is room for adaptation during the process. In terms of them sort of saying, OK, we have to stop all of opening, I think that might be fine if they are having – if they’re seeing that the outbreaks are getting away from them, then  there is room for potentially slowing down or even reversing opening strategies. I don’t think it’s ideal. I think that the economic consequences of sort of fluctuating and closing down society are just extraordinary.

If there is some work arounds to say you mandate very heavily mask wearing, for example, then maybe you can, and maybe people won’t do that for years, but maybe people would do that for three weeks at a time. So if there is a burgeoning outbreak, you can absolutely mandate that people wear masks whenever they’re around other people in the workplace or otherwise, you know, with the exception of being at home. And maybe that alone could really work to reduce some of these outbreaks that do seem to be getting away as reopenings happen. But I think one of the problems is a lot of the reopenings have been happening without the testing and true surveillance networks in place. And so it’s leading everyone.

If I was a policymaker and I started seeing cases going up and I don’t yet have a robust surveillance program in place to know how to stop it, then you have very little choice, either the mask wearing ends up working and that would be great. But in their shoes, the last thing they want is to have a massive outbreak under their watch. And so in that sense, they have to roll things back. All of this is coming from this idea that we spent the last, you know, testing has been front and center of this entire epidemic since it started in China in December and has been extraordinarily front and center in the US. But nevertheless, we still have not gotten the serological surveillance system setup.

We’ve gotten a lot of virological testing set up, and that’s great. And that should be really used when there are outbreaks happening. But we need to setup surveillance systems that will allow a policymaker to know where are the outbreaks coming from and then work with epidemiologists to understand how best to squelch them when they start arising. But right now, most places continue to fly blind and that sort of handcuffs people or ties their hands behind their back and forces them into a position where one of their only solutions is to close down again. And, you know, it’s why some people have been advocating so much that we needed to spend the last few months getting true surveillance systems set up and ready to go but in most places, they just don’t exist.

Q: Thank you.

MODERATOR: Next question.

Q: Hi. Thanks so much. Here in Minnesota, the Mayo Clinic this week said it had developed a test for that they say is will be broadly commercial, broadly commercially available to identify neutralizing antibodies. I wondered if you had heard about that and if you had any thoughts on its significance.

MICHAEL MINA: Yeah. So, it depends on –  I haven’t actually looked at what exactly they announced. A lot of people are saying that they are – so, antibodies bind to different parts of a virus. If you just do a basic ELISA for a coronavirus, you could get antibodies that bind to any different sides of the proteins that are in that test. And why that’s important is a regular binding ELISA will tell you that there’s antibodies that are specific to the protein, but they won’t tell you how they’re affecting that protein or the virus. So as an example for others who might not be familiar with these aspects, if I was a virus and there were antibodies being formed against me and somebody is measuring those antibodies, I could have all the antibodies in the world binding to my shoulders but they won’t really affect my ability to thrive, at least not considerably.

But if there is antibodies that are binding to my face and covering my breathing, then they are going to affect my ability to replicate. And the same thing goes for viruses. An antibody binding to certain parts of a virus may or may not be crucial to preventing that virus from replicating. So neutralizing antibodies do prevent viral replication, usually in this case by binding somewhere that prevents the virus from being able to itself bind to a cell. So if it if your antibody comes in and binds to a virus, then that virus can no longer attach to a cell because that antibody’s blocking it, then that can be a neutralizing antibody. So what a lot of these assays are saying now is that we know so well exactly how the virus is binding to a cell, that we can guess with pretty good confidence that if an antibody, if we take just that piece of that virus that binds to the cell, we put that into a test, and we find that an antibody binds to it, then we can be fairly certain that if that it was a real virus and the antibody was there, that it would be obscuring that virus’s ability to bind. So it’s not exactly a neutralizing assay in the sense that you’re not actually watching the virus not be able to replicate in tissue culture, but it’s just so well correlated with it based on by virtue of having only this very, very specific piece of the protein in the test, that you can make some pretty good guesses that these antibodies will be neutralizing.

Now, whether the Mayo Clinic is doing that, which would still be an ELISA based test, but it’s just extraordinarily specific to the receptor binding domain, whether they’re doing that or they’ve actually come up with some commercially available true viral CEDO type to virus neutralization. I see where you actually have live virus and you’re putting somebody’s serum in there. I’m not sure what exactly what their assays. But either way, these are important technologies to be bringing to market. And I think they’ll be crucial for understanding vaccine responses, for understanding if people are truly protected after they get infected, and for understanding what the duration of those protection levels are after somebody is infected. So these tests are going to be crucial.

Q: Thank you.

MODERATOR: Next question.

Q: Hi. Thanks for doing this. You know, there’s been a lot of concern about that the U.S. is reopening too fast or at least certain states are as we see, you know, states relax stay at home rules, even in places where we’re seeing cases and hospitalizations still rising. I was just curious. Are we seeing any states that arguably are in a decent position to move into the next stages by virtue of having succeeded in building up the necessary testing and tracing capacity and hospital equipment and staffing, as well as seeing falling case loads at this point? Is there a sort of a gold standard or even like a silver standard at this point?

MICHAEL MINA: I think Massachusetts is doing a pretty good job at that. And, you know, there’s a fairly robust contact tracing effort. There’s a number of pretty high throughput laboratories that are able to help facilitate the testing when needed. And people have more or less been pretty compliant with stay at home orders and or with social distancing on their own. And we are seeing people getting back to work. But in general, when you walk around, you see that people are wearing masks, they’re not having big gatherings and huge parties and picnics and things like that. So I think that in general, and as a result, we have continued to see declining cases in Massachusetts. And our goal, I think, would be to get around to two cases per hundred thousand people and it’s continuing to get there, you know, maybe by the end of the month.

So there are examples of where this is possible. And I think that it just requires a little bit of maintaining the status quo in terms of social distancing when you can. Of course, Massachusetts didn’t start opening up a month ago. It just started opening up recently and so we’ll have to see how it goes. We are a pretty densely populated state. And so, you know, there’s a chance that we could see the cases plateau or increase. But my hope is that we won’t see that, that we’ll continue seeing declines. For states that are opening up and taking – there are some states where the culture is just, you know, not to take it seriously, and in some places that’s because the leadership, you know, whether it’s political or otherwise, I think that we’re seeing some some governments of states just not really want to take all this as seriously as maybe they should.

And there I think we’re seeing a trend towards increasing cases and that is worrying. It’s particularly worrying for me at the moment, because if we assume that this virus is following the normal seasonal patterns of coronaviruses, then right now, June, July and August should really be the absolute minimum in terms of when transmission of coronaviruses, seasonal coronaviruses is usually seen. And so it is worrying that even just fight a normal valley that we should be in, if we’re still seeing rising cases in some states or renewed rise in cases in many states, actually, it does make me a little bit nervous that if they if that trajectory remains a) they’ll just keep building up more and more cases, it could overload some of the health care systems. But if that continues into the fall, we might see a massive burst of cases in many of those locations. And, you know, we’re worried about it everywhere. So places that are taking so few safeguards that they’re continuing to have increase in cases even right now, I think that that’s a pretty scary thought.

Q: Can just as a quick follow up, can I ask if there’s any particular states or set of states that you’re particularly worried about for those reasons right now?

MICHAEL MINA: I was looking at the map and some of the charts yesterday. I wouldn’t want to get it wrong. So I’m going to say the individual states and I haven’t been too close to the individual policies that they’re each putting in as much as just kind of monitoring sort of basic trends. But I can, if you want to send me an email, I am happy to.

Q: OK, great. Thanks very much.

MODERATOR: Next question.

Q: Hi. Thank you so much for taking the time to do this. I’m curious what is known about whether these huge protests and demonstrations about race, racial justice have led to new infections or outbreaks in and how and if researchers can figure that out?

MICHAEL MINA: So that’s certainly been a big topic of discussion and of concern. I think that it’s being monitored and what we’re trying to look, for example, for spikes in cases a week or two after the protests, if not more. And so some of these you might not see – a lot of the protesters are younger. And so if cases were happening among those the individuals within the protest, you might not actually recognize that transmission had occurred at the time and even within a generation of the virus. But it might actually take a few additional steps of spread before those cases that might have ever arisen in the protests just sort of spread out into the community and start hitting individuals more likely to end up in the hospital and have or get symptoms that requires them to go get tested.

And so I think we haven’t seen huge spikes as a result. There has been some places with some increases. It’s hard to necessarily disentangle the protests from more or less everywhere has had some simultaneous re-opening efforts. And so we’re trying to be very cautious before, you know, coming out with any claims that it is or is not as a result of the protests. And we might still be in a waiting period before we before we see any of the effects, if any, of transmission within within those settings. But we have seen, for example, areas – we do know that when people congregate, including younger people in college and otherwise congregate, we do see it cases starting to rise. So some of the big universities brought back their sports teams and there were some reports last week or two weeks ago of increased cases among some of these athletic teams. So we do believe there’s a real risk and we’re trying to monitor this to see if anything does show up as a strong signal.

Q: Just as a follow up, will the evidence mostly be timing or will it be based on, you know, detailed contact tracing to try to trace the source back to someone who was at a protest?

MICHAEL MINA: It depends on how sophisticated the analysis is. So you can use both timing and geospatial type of analysis to understand there was a cluster of new cases over here, over here and over here, and, you know, in all three of those locations, there were massive protests in the last three weeks, whereas the odds of similar cases happening in a place where there were no massive protests are much slimmer. So I think we don’t even have to do the very detailed contact tracing to pin it down to types of events, as long as there’s enough of those events that we could start to derive patterns. So if there was only one protest in one city and and at the same time there was a lot of re-opening, then it would be a little bit difficult to discern. But in this case, I think we’ll see enough heterogeneity geographically and temporally between reopening plans and protests and new cases that I think we’ll be able to discern what happened, has happened over the last month or so.

Q: Thank you.

MODERATOR: And while waiting for another question, a pop up, one of the things that I’ve had recently in my inbox has been questions about how we should be interacting with the people that we do know. And we’re going to be seeing family members and that sort of thing. Is there a safe way that we can interact with the other people in our family that we may not be socially isolating with, such as grandparents and grandchildren? Is it okay to hug? Is there a good way of doing that? Should we wash hands beforehand? Do you have any suggestions on how to get together with others in a safe way?

MICHAEL MINA: Well, I think the most important is to be aware of whether the family members you are getting, you know, whether it’s family members or friends, whether those individuals that you’re getting together with say it’s for a dinner party outside or whatever it might be, I think it’s just extraordinarily important to be aware of whether or not those people are then going to go somewhere else and be around vulnerable people. And that that to me, is one of the most important pieces here. If you know that you’re getting together with your siblings and all of them are in low risk categories and none of them are going to go and say hello to very vulnerable people, assuming that your siblings are also in a low vulnerability age group, for example, then I think there’s not a lot of risk to overall if it’s staying as a very small, close community.

And I would still say, though, you know, there’s no need at this point to be shaking hands and hugging, I would say trying to stay away. But if you’re indoors and your family members are coming over regularly, probably if somebody’s sick, it’s going to transmit at some level. I’m guessing that most people aren’t going to be wearing masks, you know, if they’re having dinner with their siblings, for example. So I think it’s just the more important piece is just to be cognizant of who those people are going to go back to and be aware that if they’re going back to care for an elderly individual in their home, then probably if you have any risk at all that them coming over is going to put them at risk of getting transmission, that I would say just don’t do it at this point in time.

But for some people who have – I know that for me, I’ve left my house, you know, less than five times since February probably or March. Most of that’s been because I’ve just been working so much. But I think my risk of being currently infected is so exceptionally low. But if I start having friends come over, then all of a sudden my risk goes up and then my risk of transmitting it to an elderly person that I may be caring for, for example, who’s here would potentially increase. I think just knowing what the potential transmission chains are is probably the most important factor when deciding whether or not to get together.

MODERATOR: Next question.

Q: You know, here in Seattle, we’re the headquarters of IHMe and their modeling has been, you know, very, very high profile, but also somewhat controversial, and the model has shifted dramatically from its original form. I’m just curious what your impression is of that model, particularly in its current iteration.

MICHAEL MINA: Certainly, I would say that it has improved. The first iteration was more of a purely statistical model without what we would call sort of a dynamical component. And I don’t have to go into each of the different types of models. But I think that the current model, which does use some elements which are more traditionally seen in infectious disease dynamic modeling  where you’re actually trying to capture the dynamics of an outbreak and not use a static sort of probabilistic model to know what it might do is a smart decision. I think they’re able to capture a lot more nuance at this point time.

But I think in either case, you know, keeping track of how the models being used, you probably don’t in a situation like this where the prevalence and ability for a new outbreak to occur and sort of spread so quickly is pretty real, then you probably don’t want to place too much weight on sort of what these models are predicting too far into the future. But to understand sort of what the infections have been doing at the population level, I think that at this moment in time, I assume the model has become quite a bit more solid and discernible in terms of how the projections are currently being made and what are the implications for those projections.

Q: Thank you.

MODERATOR: Next question.

Q: Oh, hi, Dr. Mina. I would like to ask the question. Contract tracing in countries like Italy where at least officially there have been no more cases over the last weeks and the situation seems to be stable. So I would like to know if at this stage there is a possibility for the government to use contact tracing to start all over again and then try to identify in a potential new way a contagion ahead of the of the new coronavirus season, which is predicted to be in fall, winter. And you were talking about a predictive modeal in answering the previous question. I would like to know if it would be a kind of predictions concerning potential outbreaks in the fall, which could be even even worse than the one we experienced this spring. Yeah, the two questions.

MICHAEL MINA: So contact tracing, I think now that Italy has really gotten cases down to manageable, very manageable number, hopefully that it’s close to very, very few new cases, that I think contact tracing can be as most efficient at that point. But to make it really reliable, you have to have some way to, or just to use contact tracing, you have to have a good surveillance system set up to capture the cases when they start so then you can go and contact trace and isolate.

And so I think if those two pieces are there, well three, if the cases are very low, surveillance has been set up, whether that’s antibody based surveillance or viral surveillance, and then you have quick, readily deployable contact tracing, I think that’s a very powerful combination of tools to and in situations to be able to greatly reduce potential for outbreaks in the future. In terms of predictive modeling with seasonality, I think the hard part is we still don’t really, so we don’t fully understand the different drivers of seasonality. Is it that people stay at home in the winter? Is it that there is weather changes in terms of the viral behavior, or does the virus die or desiccate in the heat more?

So there’s all these questions that make it very difficult. Without really understanding how to discern them, it does make predictive modeling particularly difficult because all of the data we have so far could be very confounded in terms of the seasonality. It could be confounded not so much by because we don’t know – if we were thinking that it’s weather that’s driving the changes, how do you disentangle weather from people’s behavior of staying indoors more and things along those lines? So it’s going to take some basic biology to really understand this more. Animal experiments in different weather settings, for example and monitoring humans very closely. But I do think that we can start to make some, we can make some educated guesses that can be incorporated into predictive models. But unfortunately, we won’t know until we see it what exactly is going to be the role of seasonality here. But we are concerned that there will be a strong seasonal component.

Q: OK. A follow up question. There is a way to determine whether I mean, which is the contribution of the different factors to the downward number of cases in Italy, other counties that now are stabilizing, whether the contribution of the lockdown measures or whether they find that the weather is changing, how we can actually determine which is the shareholder version of what the two factors were, the one which is controlled by man or the governments or the other measures, and the one which is out of control, which is the weather? 

MICHAEL MINA: So it’s sort of the same question just retrospectively versus prospectively, and I would say that there are some ways to do it where you can look at different programs that were used, different lockdown interventions, when those interventions were put into place, and try to come up with, for example, two different communities, one that put interventions in place here and one that put interventions in place here, but they had the same weather system.

And you can do that across a number of different locations to try to understand what, if any, is the role of weather versus shutdown measures and manmade and sort of non pharmaceutical interventions. And I think that if you can drive those types of, if you can get that kind of data, you can start to better understand what the role of lockdown was versus weather. And I think that there is a lot of papers that are coming out that are doing just that. The lockdown measures tend to be so effective that it could obscure any lesser effect of the weather. And so I think it’s why we won’t necessarily know just what contribution the weather would have had in the counterfactual scenario where lockdown measures did occur. But we can use that heterogeneity to try, if there were some that had a pretty mild lockdown measures versus full lockdown measures, for example, we could maybe try to understand relative contributions. But it’s a very difficult epidemiological task.

Q: OK. Thanks a lot.

MODERATOR: Again, I have another question. There has been some talk in the news lately about the upcoming political conventions and plans for those, whether they should move forward or be canceled or how they should take place. Do you have any thoughts about how large events like that should should take place?

MICHAEL MINA: With care. We need to be able to vote. We need to have, you know, I think that if voting can be done in some way and, you know, a lot of politics should be able to be done from somebody’s home. We have a pretty antiquated system. I don’t think anyone would doubt that about how we’d vote in this country. But given that probably we’re not going to change that between now and November, I think, you know, again, wearing masks to get out  into the political arena. I mean, for voting anyway, I think having large debates, you know, with with big audiences, it just doesn’t need to happen. Having large conventions, I think that those are terrible ideas.

There’s no good reason when people’s lives are at stake, I just don’t think there’s possibly a good reason. You know, if one person – you know, of course, I don’t think it’s a secret that our president is probably going to go and have his large conventions again and lots of people show up and at those places there is there does tend to be a fairly high average age of people participating. And that is very concerning for me, I think. I just don’t support large gatherings and I definitely don’t support large gatherings by the leader of a country where cases still are not under control. It just sends the wrong message. And I would encourage them to happen, for what that’s worth.

MODERATOR: Thank you. Do you have any other final thoughts before we go?

MICHAEL MINA No. I hope everyone has a nice weekend.

This concludes the June 12 press conference.