Coronavirus (COVID-19): Press Conference with Thomas Tsai, 07/24/20

You’re listening to a press conference from the Harvard T.H. Chan School of Public Health with Thomas Tsai, assistant professor in the Department of Health Policy and Management and an affiliated faculty of the Harvard Global Health Institute. This call was recorded at 11:30 a.m. Eastern Time on Friday, July 24th. 


THOMAS TSAI: Good morning, everyone. I was just smiling because I was telling Nicole that I’ll try to be a little bit more upbeat than I was on my last one of these where I felt like I needed to give myself a hug after reviewing the sobering information about COVID-19. I just want to start with saying a few things. One is that the pandemic still continues to rise based off of the Harvard Global Health Institute Risk level dashboard. Now, 11 states are in the tipping point with a daily growth rate of over 25 new cases per hundred thousand. What’s concerning is that it’s not just the same states, but it seems to be expanding to other counties in other states. So I think the need for vigilance remains paramount. 

The second point is that the Harvard Global Health Institute, a large collaborative effort with partners from other research teams at HSPH to the Harvard Safra Center, and a broader coalition in Massachusetts around tests trace and supported isolation released a similar risk level that corresponds the incidence number of new cases to potential recommendations for school reopening, what we’ve called pandemic resilience for schools. So, happy to talk a little bit about the implications of that. I think it’s important to understand that we need to have nuance around schools and it’s one not one size fits all. And based on the emerging data on transmission risk for different age schoolchildren in elementary school versus middle school versus high school age children, we may have to develop a more nuanced way of looking at individual risk for those categories, but overlaid upon the over line community transmission risk in the green, yellow, orange or red zones. Suffice to say that the key to reopening schools is really getting control of transmission in the community. 

The third piece is that when it comes to testing, perfect may be the enemy of good. Michael Mina from the Harvard School Public Health has done really great work and has been a real advocate over the last several weeks, and really educating all of us, me included, about the way to think about testing is that perfect is the enemy of good. Sometimes higher frequency testing with tests that are less sensitive, may actually have a better role in controlling the pandemic than relying on less frequent testing with PCR based tests, that may be a more highly sensitive based off what we are learning more and more about the transmissibility and the timing of transmission for COVID-19. And the last bit of optimism is that there is encouraging information now coming from the negotiations between the Senate Republicans and the White House around funding for testing, and we’ve been providing some technical assistance to that work and those considerations. So I think there’s now broader agreement that we need to continue federal funding for this, and the negotiations next week will likely focus on how to operationalize that funding to make sure the dollars are actually getting to the states in a way that’s actually supporting increased capacity. So I’ll stop there. Hopefully, that was a little bit more upbeat than last time and happy to take any questions. 

MODERATOR: Thank you, Dr. Tsai. All right. Looks like our first question. 

Q: Hi. Thank you for doing this. My question is about universities returning in the fall and what role testing should play in that. I know some universities are planning on testing all of their students once they get to campus. Given that we’re already seeing some pretty long turnaround times, do you think this is a wise use of resources? Are there other approaches that would work without having to test a bunch of people who aren’t showing any symptoms? 

THOMAS TSAI: That’s an incredibly important question and on the minds of many parents and and faculty around the country. Testing absolutely plays an incredible role in a surveillance strategy for universities reopening. I have a two year old daughter, you know, the risk of COVID-19 transmission in daycare, and she just returned to daycare this week in Boston, is very different than the risk of transmission in a teen age or in a college age students. So I think a surveillance strategy is important for epidemic and infection control. But I think also in creating the confidence among the students, among the parents and among the faculty in order to return to, you know, in onsite education. 

I think we can all agree that getting education and onsite education to all our schoolchildren and college students is incredibly important, especially as younger children for their educational growth as well as their social development. So we need to, you know, recognize that this is an essential goal of policymaking. And if we can come to that realization that we can really bring around the policies that we’ll need to reinforce that. And this is where we need to think beyond testing in very discrete categories. In July, the testing resources that we have available is very different than what we had available in March. Even though we’re having the testing delays and some of our commercial labs, such as Quest and LabCorp, there’s a couple ways forward. One is we need to think beyond just the the traditional capacity for providing tests. And there’s capacity available beyond the commercial laboratories. There’s health care systems. There’s the academic research groups, the Broad Institute at Boston and Cambridge between Harvard and MIT is a very great example of how an academic research center has really been able to step in and fill a void to produce tests to support the surveillance and diagnosis of COVID in the greater Boston community. There’s also opportunities to increase testing from the biotech and repurposing some of the supply chain and manufacturing lines for biotech to support testing. 

The efforts are underway now across the country in Massachusetts. You know, one example has been Ginkgo Bioworks. The last one which has been happening in other countries, in Europe especially, there’s less of it in the United States, is also repurposing some capacity in veterinary labs. So there’s over six categories of testing capacity that we can be focused on. Where we really need to move is on the policy side, a dashboard of the load balancing, the capacity utilization, across those categories to think about how can we use all that capacity to help generate surveillance testings? So move beyond just the testing delays in Quest and LabCorp, and really think about the capacity across the entire spectrum. And the second answer to that is also we need more rapid point of care testing. And again, that was more of a wishful thinking back in March. But now we actually have commercially available point of care tests. One example of that is the Centers for Medicare and Medicaid is now deploying point of care antigen testing to nursing homes as part of the broader HHS initiative around testing. So I think that’s an example of providing cheaper point of care testing and doing more routine surveillance is important. And one incredible place to use that is at our universities. 

MODERATOR: Did you have a follow up? 

Q: Yeah. Just really quickly, I am wondering, how would a strategy differ between trying to do surveillance and then trying to do more universal testing of the student population? Would surveillance be more just like doing a random percentage of students? 

THOMAS TSAI: Yeah. It’s not just about the testing itself, it also depends on what the schools are doing to mitigate the risk of transmission. So it’s that, plus are the schools also enforcing masking and social distancing? And if they’re not, they may want to do more frequent surveillance testing. I think everybody agrees that it’s important to test before students come onto campus. And if the capacity is available and the hope is to really change the conversation, to stop letting capacity constraints dictate our guidelines and really just the guidelines drive the supply of tests, is to do that on a more routine basis for the students. So it’s both a combination of a more broader population testing as students come onto campus, but also doing more focused surveillance once they’re on campus. 

The other part is that this doesn’t happen in isolation. This also happens in the context of contact tracing. So if there are cases that occur on campus, you still fall under the rubric of a contact tracing framework where all the contacts of those individuals would need to be traced and then also tested. So it’s the combination of all those different testing strategies. But I think that’s the nuance we can be talking about now of not just should we be testing students, but really, the schools and the states really need to develop very robust guidelines about what their strategy is going forward and the contingency plans for testing, should a case occur while somebody is on campuses? 

Q: Thank you. 

MODERATOR: Next question. 

Q: Hi, yes, thanks for taking my question. What kind of struck me a little bit about the testing dynamic is certain states, it seems most that are located in the Northeast, including Maine, have passed the mitigation threshold from Harvard Global Health Institute, and at least in Maine’s case, is well on its way to the suppression. So I guess I have two questions. From a testing strategy standpoint, when a state moves from mitigation to suppression, how does that change what strategies you use for testing and what types of populations that you might test? And then my other question is, with the explosion in cases in other parts of the country, with this testing and tracing strategy by state, can you still have a pretty good shot of keeping the virus suppressed or contained? Or is that not very likely? Thanks. 

THOMAS TSAI: Yeah. So two questions there. I’ll tackle the first one first, which is the question around as states hit their suppression target, does their strategy change? The capacity for testing definitely does change. And in some ways, the incident cases is really driving the strategy, so unfortunately it is the opposite way around. And what we’re seeing in the red states like Texas and Florida, for example, when the testing times are getting longer, especially in Texas and some counties, there’s a return to only testing symptomatic individuals, and a lot of anecdotal stories of people being turned away. In fact, last week, CVS sites around the country in some of these harder hit areas because of the turnaround time to get a result back, have asked individuals to not come in for testing anymore at those sites. So that’s deeply concerning. But for the states are doing well. I think it’s still important to maintain the testing capacity because as we’ve seen, the pandemic can turn on a dime, and these states still need to maintain the capacity for tests, but doesn’t mean they necessarily need to be doing the same density of testing as in the states that have a wider pandemic. 

In the states, Maine, for example, or some of the states in the Northeast, they’re now sort of in the situation that South Korea has been in the last several weeks and months. Once you get the pandemic under control, so it’s no longer just mitigating, you can actually suppress the pandemic. So that means relying on contact tracing to identify, in a very more focused and surgical way, the number of tests that need to be done, the contacts of those individuals. In the ideal situation, we would have 100 percent test positive rate, meaning that we would know exactly who has COVID, and be able to test all those individuals and only those individuals. That’s not realistic. No real world. So you need some basic level of surveillance, which is almost like an early warning system. But when those number of cases do pop up, then you have enough capacity to test broadly across their entire critical context all their contacts, all their working environments, for example, to make sure that you can suppress these individual cases before they become an entire cluster. That allows you to have a more effective contact tracing strategy, which is challenging in states where there’s a high number of cases, delays in testing, and long turnaround time in testing. Do you mind just repeating what the second question is? 

Q: OK. The second part of the question is with the raging number of cases in other parts of the country, can you still, as a state, have a shot of suppressing the virus if you have enough testing? Or is that a very big risk?

THOMAS TSAI: Sure. So I think states can’t go at it alone. We’ve seen the challenges with that approach when it came to PPE or just earlier in the pandemic, we’re seeing the challenge of that now. Individual states, especially the states that may be in the best position to purchase tests, are also states where they have the lowest burden of tests right now. Each individual state may not have the resources to be the market maker, to be able to purchase enough tests and procure enough tests at bulk to be able to drive the prices of testing down and also signal to the market that there is this ongoing demand for testing. So if you are about a company that you need to retool to produce tests. That’s why there needs to be a regional collaboratives. And there are proposals out there, including around the ideas of interstate compacts, which are legal frameworks to coordinate in a way that also allows them to receive congressional and federal dollars to support those efforts. There’s also more informal collaborations and consortia that can occur between states working together. And I think one really powerful idea to incentivize that is as this next round of stimulus package for COVID-19 before Congress breaks, is to really tie the financing for testing to states in an incentivized way so that states work together. Let’s say three or more states work together to purchase tests in bulk, that opens up additional resources from the federal government to help offset the cost of that. We know a lot of state budgets are already overtaxed from increased public health spending and decreased revenues because of the economic recession and unemployment resulting from the pandemic. So that’s where I think the federal support to really incentivize and enable states to work together is incredibly important. 

Especially as individuals are now traveling between states, there needs to be thoughtful planning around testing, as well as mitigation strategies across states. So in the dream world, there would be a pandemic testing board that could help coordinate this, that can happen nationally, that needs to happen regionally and it’s already happened regionally. There are coalitions of governors in the Northeast and the Midwest and on the West Coast working together to try to coordinate some of these efforts in an informal way. But if there’s a way to support that work and invigorate that work with federal dollars, because as we know, this one regional epidemic to quickly turn into a national pandemic. 

Q: I’m sorry. Just one more little thing. Is there a way you can just give some examples of suppression level testing versus mitigation level? Like maybe testing teachers or other groups? If you could just give a few examples of how states do that. That would be much appreciated. 

THOMAS TSAI: Yeah. The best example is in the health care setting. You know, I’m a surgeon at Brigham and Women’s Hospital, thinking of critical context, and I think the health care setting is one such trivial context. You can learn about how effective the testing strategy is in hospitals as an example, and you can apply that to to other industries. And the other example is, you know, the White House has a staff every single day. You know, the NBA, Major League Baseball, NFL are all testing the players on a very frequent basis. So we know that high frequency testing from a surveillance standpoint can help mitigate the risk of community transmission. The goal is to make that the same level of care that the politicians have, that the celebrities and sports and athletes have, make that available for all individuals. And I think it’s just as important to have that surveillance testing available for teachers, for students to not just control a pandemic, but also to create the confidence to be able to return. So I think that’s an example of context, where we know that surveillance testing can be effective and we can apply that a little bit more broadly. 

The fundamental issue is getting control of the pandemic in the community, because if the risk levels go from red twenty five hundred thousand, to orange, which is ten to twenty four cases per hundred thousand, to yellow, which is between two to nine cases, to green, which is less than one case, per hundred thousand, then the corresponding level of surveillance also decreases. So that’s the idea of moving from where you need to go from mitigation, because your hands are forced, you just can’t test widely enough when the pandemic is out of control, versus suppression, where you can test broadly because, you know, you’ve already suppressed the community level transmission and you’re test positive rates end up being very low, which is a good thing because most of the individuals are no longer actively transmitting COVID. 

Q: Thanks a lot. 

MODERATOR: Next question. 

Q: Hi, thanks very much for taking my question. I have two questions. One is not testing related, and the other one is. I want to go to the non testing related one first. The governor of Massachusetts, just a few minutes ago, announced travel restrictions that’s going to require anyone coming from all but a handful of states with low transmission to quarantine for 14 days. They have to fill out a form. They have to be in touch with the state and they have to basically isolate themselves for 14 days or face a fine. And I’m wondering, other states have tried similar things. What is known about the effectiveness of this type of measure? 

THOMAS TSAI: That’s a really important question and saw as a side comment, I like these calls because I feel like I learn more, as much as you guys are, probably. I’m learning breaking information in real time, which is always good for me to be on top of it. I think it’s absolutely the right approach. And it’s not just about the quarantining for the 14 days and testing the symptoms as part of the governors efforts. It’s also if you’re able to produce a negative test result within 72 hours, I believe, of entering the state. So that sort of speaks to traveling between states and the need for testing. And one of these critical contexts is not just an occupational category or an age group, but if you’re traveling from a high risk area with high levels of community transmission, going to a low risk area, I think that’s also sort of a geographic critical context where you want to also be doing surveillance as well. So I think that’s the right approach. Other states have adopted that. I don’t have any firsthand knowledge of how effective that that approach is. But it makes epidemiological sense. It makes a lot of intuitive sense that this is something they want to do. It really focuses the efforts on not just testing as an isolating approach, but sort of testing in relation to all the other interventions like support, isolation or quarantining and masking. So I think this is an example of how you have to tie together all these different strategies in order to control the underlying community transmission, which is not a static concept. Community transmission can be very dynamic, going county to county, city to city, state to state. 

Q: OK, so no one’s actually studied whether this works, but you’re saying it just kind of makes sense. 

THOMAS TSAI: People may have studied this. I don’t have any information or firsthand knowledge about this. And I think this is where the contact tracing data will be helpful, because this will help provide information to the contact tracers. And I suspect some of the states will probably have a bit more detailed information if any contact tracing has tracked people that have come from other states. 

Q: So is it really possible, though, in our society to kind of put up a gate like that? I mean, Massachusetts has a lot of borders. It’s a small state. I’m wondering, is this a pipe dream that you can really prevent the illness from getting back into the state? 

THOMAS TSAI: Yeah. I don’t think it’s a gate. I think of it more as a turnstile. You know, and obviously, in New England, people live in New Hampshire and work in Boston, and they may go hiking in Vermont and to the beach in Maine. So this is why states can’t go at this alone. You can have all the states enacting restrictions on all their neighbors. And this is why it really needs to be a coordinated regional collaboration, because, especially as people go back to school, go back to work, the travel is going to be more interstate and more dynamic. This coordination really needs to be on a regional level. You can’t have the states working at opposing ends, you know, this is about decreasing community level transmission because, you know, if Maine has low transmission rates, people coming from Maine represent a lower risk than people coming in from South Carolina, for example. So, again, this is where region collaboration is incredibly important. And I think these travel advisories would be even more effective if they were announced and coordinated on a regional level, whether formally through the interstate compacts or just, you know, more informal regional collaboration. 

Q: OK, great. Thanks. So my testing question has to do with an incident a week ago when 113 Rhode Islanders, and I don’t know how many people from other states were victims of a laboratory error in New York. They were told they had positive COVID tests and it turned out to be a false positive. And I’m just wondering, actually one of my relatives was one of those people, so it kind of ruined our vacation. But even worse, there was a group of people who work in the nursing home who got these false positives and they missed a week of work and had to spend the time fearing that they had exposed their patients. And I mean, it’s better than having COVID, but not a good thing. I’m wondering how common are false positives? This was a PCR test, which is supposed to be pretty accurate. Was this a very freaky, unusual situation or is it something we see from time to time? 

THOMAS TSAI: Yeah, I think that’s a great question and the false positive rates are going to be more common when the community prevalence is low because we think about this in terms of what we call the pretests probability. If 100 percent of the community has COVID, there’s zero false positive because, you know, everybody has COVID. That false positive rate, or what we call it in epidemiology, the positive predictive value or negative predictive value test, that’s very much driven not just by the characteristics of the test, you know. You know, the sensitivity, specificity, which we often hear about those in some ways intrinsic characteristics of the test itself, but the positive predictive value and negative predictive value also factored in what the underlying prevalence of the test is. And, ideally, you want those to be very high. But the fact of the matter is, if the baseline prevalence is very low in the community, small differences in the sensitivity and specificity that tests can actually result in a very, very high false positive rate. So that’s nuanced, but I think that also means that we have to communicate effectively on the public health side, medical side, you know, the the risks of that. 

We need to start moving from just thinking about testing as these one off or monolithic events, and thinking about how do we later on tests with different characteristics in terms of false positive and false negative rates so that we can screen with tests that may be less sensitive. They may not pick up the extreme ends of the cases, but then use that to screen in individuals for live the more sensitive tests. Right now, we’re using the PCR tests, which have very high specificity and sensitivity. But that also depends, that’s true in the laboratory setting, but that depends on where an individual is in their course of their disease. If the viral copies are very low, they may not be very transmissible. And on the other extreme, if they’re, coughing and have chest pain and a fever and there’s actively replicating virus, then, even a poor performing test can pick that up. So that’s a very long way of saying it’s not uncommon. We also have to communicate with the consequences of this are. As a surgeon in the hospital, false positive results in my world, I mean, there’s a whole slew of medical interventions that come along with that which may result in harm. You don’t want people getting unnecessary surgery. 

In this context, right now the harm can be emotional, psychological, social. At the same time, the false positive results, for the most part, haven’t led to a sort of additional treatment harm. So that’s a very small silver lining. It doesn’t add any measure of comfort for the people who had false positive results. But I think we need to move beyond. We’re talking about tests in a very blunt way and start communicating that, especially in areas where transmission may be low, there is a risk of false positives. 

Q: So are you saying that we should stop looking at tests as kind of an ultimate answer, but just like a piece of information that might be right and maybe needs to be confirmed every time or, how should people regard the tests, given that they’re imperfect? How do you react to it?

THOMAS TSAI: Yeah. I think that the level of nuance and understanding of the viral transmission has changed since the early half of the year. So I don’t want to downplay the importance of testing. That’s all I’ve been talking about for the last three or four months. Tests are really important. But what we’re also learning is that the implications of tests can be different. And now it’s also confusing because you have lots of different tests out there. Early on there’s a couple of different PCR tests. Now you have a plethora of PCR tests. And before, in March, people were being tested when they’re mostly symptomatic. So the false positive rates were likely lower because you’re only testing people with symptoms. So now, we are testing asymptomatic individuals. That’s incredibly important to casting a wide net because it goes to catch those presymptomatic individuals early, and before they’re turning into clusters and responsible for a super spreading event. But that also means that you may be truly testing people that don’t have not just symptoms, but don’t have the actual virus, and then there may be some false positive there. Again, this gets more complicated and such as we move from a medical diagnostic framework towards a surveillance and public health framework. So, yes, the information is still incredibly important,. 

Q: But it’s important on a population level more than on an individual level. Is that a way to put it? 

THOMAS TSAI:  Yeah. Talk to your doctor. 

Q: OK. All right. Thanks very much. 


MODERATOR: Next question. 

Q: Thank you very much. Just to give a bit of context to my question, I’m looking at testing more at the global level, including the states, of course, but more internationally. From my discussions with a few epidemiologists, it seems like there are starting to be two camps that are emerging. On one side, some epidemiologists are pushing for really wide scale testing to try and test as many people as you can. And in the U.S., we’ve seen that that’s led to testing delays. And you were mentioning earlier some ways that we could leverage existing capacities to mitigate some of those testing delays. But then there also seems to be this other camp that is saying, let’s scrap the idea of as many tests as possible or more testing, and focus more on more targeted testing and trying to identify potential pockets of infection and going after those specifically as opposed to more widespread community based testing. And I was wondering what your thoughts were on those two sort of strategies. 

THOMAS TSAI: Yeah, I think the issue is you can’t do the latter without doing the former. You know, in the ideal situation, we would do very focused testing and if we had a crystal ball, test those individuals that we would know that would be associated with a super spreading event, if we knew that ahead of time. The problem is we don’t know that ahead of time. That’s why you still need community level testing to provide the information to be able to do that more focused testing. So I think that’s where I’m in the middle of that is, you know, there’s what’s ideal and what’s sort of practical. And I think that’s a challenge. I think the two camps would probably agree on the goal. The problem is, how much you’re factoring in the tradeoffs between the constraints that you’re faced. Does that answer your question? 

Q: So I guess in your mind, the ideal situation, if we had, say, unlimited resources and contact tracers and people who could do testing, the idea would be to test as many people in the community as possible?

THOMAS TSAI: Yes. Sort of in this phase of the pandemic. And again, it’s not just about sort of indiscriminate testing, that’s why tying it to the risk levels is incredibly important in terms of the, you know, what we’ve been working on and trying to crystallize the risk into the green, yellow, orange and red risk levels. So, you know, the green levels where there’s low risk of testing your testing threshold can be lower because you’re not going to pick up a lot of positive cases. You have to do enough and to be able to know that there are positive cases out there and making sure that you’re getting accurate information. But if you’re testing a lot of individuals in green counties and green states with low incident number of cases that results in the situation mentioned where you can get false positives. 

If you’re only testing symptomatic individuals, for example, like you are doing in March and April in states with a very high burden of cases, then you’re not really in control of the pandemic either. You’re just really diagnosing and confirming individuals with COVID. You’re not testing, in broadening circles to be able to contain and eventually suppress the pandemic. In part, the testing capacity numbers that we’ve been putting out in collaboration with the Harvard Safra Center and others, it’s a capacity that allows you to have a strategy, and I think the strategy is important. So if you haven’t test, you don’t have a strategy. You’re just reacting to people showing up at the hospital, needing ventilators and just confirming that they have COVID. So the capacity lets you have enough tests to do contact tracing, because if you don’t have enough tests to test those contacts, that doesn’t fulfill the entire strategy. And that’s why it’s now TTSI, not just T, it’s trace, test, supported isolation. It’s not just trace and stop there. 

Q: OK, great. Thank you. 

MODERATOR: Next question. 

Q: I’m interested in what a good statewide testing strategy should look like, especially for a state like Georgia, you know, where there is still widespread community transmissions and we are experiencing record numbers of new cases. And what key benchmark should be included in a testing plan. How do you determine or assess success or effectiveness and what other sort of key elements should be involved in a testing plan for a state like Georgia? 

THOMAS TSAI: What’s really needed is the key performance indicators of testing. I think we’ve been talking a lot about the number of tests, but what we’re seeing again is that there is a lag in getting the results back to individuals. So it’s incredibly important to measure the key performance indicators in a good testing process. So one of these would be how long it takes to get the results back to individuals. I think Oklahoma is the only state dashboard that I’ve seen that’s actually reporting their turnaround time for testing. And just this week, Dr. Tom Frieden’s group, Resolve to Save Lives, had put out some guidelines on what states should be reporting as part of not just COVID cases, but also their testing strategy. And that’s very much aligned with the work that we’ve been doing at the Global Health Institute to kind of converge around a set of key performance indicators. 

Another one is the proportion of test positives that are coming from contact tracing or surveillance as opposed symptomatic individuals, that ties back into some of my comments from the earlier questions, that you want to be making sure that if even if you have a high test positive rate, that it’s high because you have a strategy and you’re identifying the likely individuals to have COVID. So you want a high proportion of greater than 80 percent, ideally over 90 percent of the test positives testing positive because there’s an adequate surveillance and contact tracing, not just from symptomatic individuals, because if we’re just testing symptomatic individuals, from a public health perspective, you’ve sort of fallen behind. So these are all really the key performance indicators. 

The other one that I’d really love to see, and this goes back to my opening comments, is that we have all these different categories of testing capacity and we’ve been really focusing on commercial labs in a lot of the conversation in terms of the delays in testing results back. In order to do load balancing, we really need a dashboard that looks at capacity utilization as part of the Ma TTSI, the Massachusetts TTSI collaborative. So we’ve mocked up a dashboard of what the testing capacity would look like. We need to look at the delays and test results across these categories of tests so we can get a better sense of what the capacity for tests is across these different areas. Speaking to folks in the business world, this is a pretty common concept in manufacturing. But I was just on a call this morning where we were thinking that, you know, in some ways, our public health effort has really become a manufacturing effort. And that’s why on the federal response side, it’s maybe not enough for HHS or CDC to be solely driving some of the funding around testing, we may need to involve Department of Commerce or, you know, individuals from the business sector to help sort out some of these supply chain issues, which are not public health issues in and of themselves, but have important public health consequences and are manufacturing issues. So I think that’s where we need to to connect the dots. 

Q: That actually interests me, because that sounds like something that would have to be handled on the national level. I mean, can that even be incorporated into a state wide plan, especially for a state like Georgia, which you can’t rely very heavily on commercial testing? 

THOMAS TSAI: Right. Because supply chains can be national or regional and not individual states. And, you know, even though a factory may be located in a state’s distribution, maybe a much more broad. And that’s why you really need that national level and if not national and regional level coordination. And part of that coordination is also procuring reagents and tests and supplies and swabs in bulk, which is hard to do if you’re a county health board or even on a state level, because there’s this disconnect and mixed messages for so many manufacturers about what they’re what the demand for testing is going to look like. I mean, it’s hard to retool your supply chain if there are shifting targets about what the demand for the tests are going to be. So, this becomes really, a very nuanced, business and supply chain and logistics question. Like I said, there’s huge public health consequences, but they don’t need a public health solution. They need business solutions around that. 

Q: Thank you. 

MODERATOR: All right. Next question. 

Q: Hi Dr. Tsai, thank you so much for doing this and thank you for taking my question. I just was curious, generally looking at the benchmarks on the map, I was wondering how when coming up with this, the benchmark for suggesting a return to stay at home orders was twenty five new cases per 100,000 people. And why new cases were used as the metric as opposed to some other factors like infection rate or positive test rate. 

THOMAS TSAI: So that’s a great question, because the cases are hard, because they are very much influenced or potentially bias by how aggressive you are testing. But that’s why we have the separate dashboard’s for what the test positive rate should be. So we will have a separate testing dashboard that looks at the testing processes that test positive rate and then the risk level dashboard. The White House in their governors report, their dashboard, which was leaked, I guess a week ago, which probably should be a public thing, but I’ll save that comment for something else. They look at the combination of both the number of new cases as well as the test positive rate within 10 percent. I believe they’re cut off for a red category or a critical category is over 100 new cases per hundred thousand per week and a 10 percent test positive rate. We decide to split that out because we wanted to be able to look at that in separation because, again, it really depends on what the testing strategy of that state is. Are they just trying to catch up with a growing number of cases in a community versus they have a really robust surveillance network that’s doing that? And do we consider looking at mortality, which is obviously the most objective measure of COVID-19? 

The issue with mortality, as we’ve seen over the last few weeks to months, is that it is a very much a lagging indicator. And there’s actually even delays in how mortality is reported. So the mortality rate does influence some of our testing targets in terms of trying to understand what the mortality rate means, in terms of what the underlying community infections are, and differentiate between infections in the community versus cases which are confirmed by positive tests. So the long story short is as part of a dashboard was a collaborative effort across different research teams around the country to agree on a common vocabulary. And the new cases was in some ways the most intuitive for the public to understand. It’s the one that’s reported the most number of new cases per day. So we wanted to, again, not let are perfectly be the enemy of good. And that was one of the main reasons that that drove sort of the categories of response. 

MODERATOR: Do you have a follow up? 

Q: No, that’s all. Thank you much. 

MODERATOR: Next question. 

Q: I was interested to hear you speak a little bit about what steps can be taken to incentivize new production of some of the underlying supplies needed for testing. I think you touched on this a little bit, but what type of specific investments could the federal government or state governments do to help existing companies here in the U.S. or trying to procure from abroad those supplies that are in short supply?

THOMAS TSAI: Thanks. And I think it’s helpful to go back to where we were in March, where the federal government used the authorization from the Defense Production Act, DPA, to increase the purchasing of ventilator’s, as an example, and then help to coordinate a lot of the supply chain shortfalls around PPE, the personal protective equipment, masks and respirators. What been sort of disappointing is that there hasn’t been that same focus on using some of the existing authorization on the federal side in terms of the DPA around testing to help ensure both the supplies and reagents needed for testing. But even more important, is to just signal to the market and demand for testing by pre purchasing the tests directly, either through funds transfers or a voucher program or grants to the states. So I think that’s incredibly important. In order to change the capacity or increase the capacity from just the silo or bucket of commercial labs and expand it into the academic centers, for example, or the biotech world, then you need to incentivize that through, essentially, bulk purchasing and bulk procurement and even prepaying in the tests. 

I think that’s incredibly important. State budgets are very strapped right now. I think that’s an incredible way the federal government could do that. We’ve spent trillions of dollars in economic support, but the best way to support the economy is to beat the virus. It’s in some ways it’s penny wise, pound foolish. We’re arguing about the merits of spending a few billion dollars here and there on testing and contact tracing and public health efforts. Yet we’re constantly trying to plug in the holes economically where we know what the issue is, we know what the problem to solve is, it just requires collective will to do that. 

Q: Thank you so much. 

MODERATOR: OK, great. Do you have any final thoughts before we end the call? 

THOMAS TSAI: No, thank you, Nicole. I think a lot of the conversation today has been around the need to increase our capacity for testing. So really, three themes. As one, we all agree we need to increase testing, but I think the strategy moving forward is thinking about the types and the strategy of tests, not just the PCR tests, but antigen based testing at the point of care. Two is to really think about creating new capacities. So not just augmenting existing capacity, but really trying to expand the capacity through noncommercial testing sites. And the next point that’s really emerged, I think, is the importance of regional collaboration. And that’s true for testing, that’s true for coordinating travel, masking and sort of distancing policies. I think those are sort of the important points to the public health side and policies side really need to focus on the next couple of weeks. 

This concludes the July 24th press conference. 

View more press conferences