Using Data Science for Crisis Management and Response
In this era of Coronavirus and Covid-19, how can we use data science to manage a crisis? In this episode, two experts share their advice for using data science to help with crisis management and develop a crisis response plan.
Anthony Scriffignano has over 35 years’ experience in information technologies, Big-4 management consulting, and international business. Scriffignano leverages deep data expertise and global relationships to position Dun & Bradstreet with strategic customers, partners, and governments. A key thought leader in D&B’s worldwide efforts to discover, curate, and synthesize business information in multiple languages, geographies, and contexts, he has also held leadership positions in D&B’s Technology and Operations organizations. Dr. Scriffignano has extensive background in linguistics and advanced computer algorithms, leveraging that background as primary inventor on multiple patents and patents pending for D&B.
Dr. David A. Bray has served in a variety of leadership roles in turbulent environments, including bioterrorism preparedness and response from 2000-2005, time on the ground in Afghanistan in 2009, serving as the non-partisan Executive Director for a bipartisan National Commission on R&D, and providing leadership as a non-partisan federal agency Senior Executive. He accepted a leadership role in December 2019 to incubate a new global Center with the Atlantic Council.
- Introduction
- What are key Coronavirus and COVID-19 issues we face?
- Government response to the Coronavirus/Covid-19 crisis
- How can we navigate the conflicting data on this crisis?
- Managing data decisions vs. politics
- How can we communicate clearly and accurately with conflicting data?
- Guidance to the public based on data science?
- When will the Covid-19 crisis end?
- Can open data help manage the crisis?
- Advice for business leaders
- Advice for policymakers
This transcript was lightly edited for clarity.
Introduction
Michael Krigsman: Data science can help us navigate our response to crisis situations. Anthony Scriffignano is the chief data scientist at Dun & Bradstreet. Anthony, please give us the tweet length description of Dun & Bradstreet and your role.
Dr. Anthony Scriffignano: I'm the chief data scientist at Dun & Bradstreet. We collect information about businesses all over the world in different languages, in different writing systems. We collect information about how they interact with each other and we try to draw a sense of that for the context of either total risk or total opportunity. Obviously, today's call will be a little bit of both.
Michael Krigsman: Our second esteemed guest is David Bray. He is with the Atlantic Council. David, give us the kind of tweet-length description of the Atlantic Council and your role.
Dr. David A. Bray: We exist to deal with what's called GeoTech, which means the geopolitics of technology. How is technology changing the world? How do our approaches to different world and society issues change how we implement both new technologies and data?
What are key Coronavirus and COVID-19 issues we face?
Michael Krigsman: Let's first take a look at the current Coronavirus/COVID-19 crisis. I know we see lots of economic pain and, obviously, there is health and many other aspects that are very, very difficult, but give us a thumbnail sketch. What are we looking at? Put it in context. Maybe, David, do you want to jump in with that?
Dr. David A. Bray: COVID-19, it's a pandemic, meaning it's a case where there is an infectious—in this case—virus that is on multiple continents around the world, and so you have each nation right now trying to make sure they do infectious control to minimize any additional spread of COVID-19. The challenge is, we're now in the hundreds of thousands in terms of the number of cases in the world. Where in the past, things like trying to actually do better approaches that might be contract tracing and things like that, while we may still try to do those things, we're now really trying to do this idea of social distancing—in some cases, either self-imposed or nation-state imposed isolation—to try and minimize what would be called the peak of the infectious curve to try and at least avoid additional strain on hospital systems, on medical systems. That's going to have the effect of widening the spread as opposed to having a sudden peak. It's going to widen it but, ideally, have a slower infectious rate so that the societies can get the bear with it and address it, address the hospital systems.
Then, as you mention, there are far on effects of it. By doing social distancing, but doing, in some cases, isolation, self-imposed or otherwise, it's impacting economies. It's impacting supply chains. It's impacting geopolitics. It's impacting how we are going to get through this together because I have a hunch that it's probably not going to be over in a matter of weeks. We're talking more about a matter of months, if not longer.
Dr. Anthony Scriffignano: The challenge that we're facing right now is that we understand what's happening to some extent and, to some extent, what's happening is unprecedented. You heard David talk about hundreds of thousands. We don't actually know that because the testing is changing as we're trying to measure the impact.
Unfortunately, this is a nightmare for data science because you don't have any baseline truth. What we call ground truth doesn't exist. Boo-hoo. Get over it. Do something because it's affecting the world.
The other assumption that we have to be very careful about is that this is affecting the whole world at the same time in the same way. Nothing could be further from the truth. As this moves from a business perspective around the world, it will affect different parts of the supply chain. It will affect different parts of the integrated value chain of the world.
As the governments of the world try to look at how they respond to this, they will be either intentionally or unintentionally impacting each other's domain. You get these sort of compound effects of constant disruption.
The world has never seen anything like this and, at the same time, we're trying to use information that's changing to understand what's happening. That's a very, very big challenge.
Government response to the Coronavirus/Covid-19 crisis
Michael Krigsman: Describe the strategies that governments are taking to respond to this multifaceted, multipronged, multilateral set of issues that are happening simultaneously.
Dr. David A. Bray: As Anthony said, this is unprecedented. While there may have been sort of tabletop exercises in the past as to what would be done and things like that, as he mentioned, we think there are at least 100,000 cases but there may even be more. There's also the challenge of a lot of these tests. We have to be concerned about what might be called false-positives as well as false-negatives associated with the tests.
The other challenge with COVID-19 is, when I was involved with response to SARS with the CDC's Bioterrorism Program, SARS it was thought was a period of about three days between a possible infectious exposure and then the presentation of symptoms and the symptoms were very severe. You knew you had it.
Whereas with COVID-19, we're dealing with something that it seems like it looks like it's anywhere between 12 and 16 days between an infectious exposure event and actual presentation of symptoms. Not everyone is going to present severe symptoms. In fact, the data shows that a sizable group of people may actually be fairly asymptomatic. It makes it harder to address this than some of the initial plans that different governments and world leaders may have had.
Things are being adapted on the fly, which is the nature of crisis response. You are always learning. You are always getting better data and information. That helps inform where you should focus.
A lot of them it is. It's about closing down your border for right now. It's about asking people to practice social distancing. Try to minimize their exposure to other people to try and reduce that infectious spread rate.
At the same time, we are spending a lot of effort trying to race ahead. We're trying to better understand the virus and see if we can get a vaccine and also develop better tests. The challenge is, it looks like—and this is, of course, early evidence and I want to caveat not peer-reviewed fully yet, but there is a lot of early indication in terms of articles that are being submitted to journals—there's a case of what's called genetic drift with the virus, which essentially means it's mutating. As it mutates, that means that some tests that are looking for specific fingerprints such as polymerase chain reaction or other tests like that, that are looking for very specific fingerprints, may lose their accuracy if genetic drift or mutations happen because that fingerprint is no longer there, even if it is COVID-19. That's a challenge in terms of getting this out there. The other challenge is if you move to other types of tests like antibody tests, those are only effective when actually someone has produced the antibodies in response to COVID-19.
The world is trying to get a better sense of, as I'm Anthony can relate, what is the baseline, what is the denominator, and then trying to figure out, as we move forward with this, how can we actually also start thinking about how do we rebuild economies? How do we actually work together both locally but, even more importantly, globally? The only way through this crisis is through it together.
Michael Krigsman: Anthony, as David is speaking and as government officials speak, they constantly refer back to, well, we have to see what the data tells us. How do we manage this conundrum?
Dr. Anthony Scriffignano: This is a dilemma that many people face in business every day. You have a point where you have enough information to make a decision. That doesn't mean you have enough information to make a good decision.
The best advice in this type of a situation where we are dealing with unprecedented risk is to constantly revisit what you have to believe in order to make the decision that you're making with the data that you're making that decision upon and, also, to focus on how that affects the other decisions that you're making. For example, if you need to make a decision about the spread of the virus and you know that it will be within a certain range, a certain high and a certain low, and you have no information about whether you're close to the high or close to the low, you assume you're close to the high. You make the decisions accordingly until you have data that proves otherwise because to do otherwise would have a much more deleterious effect on the population that you're trying to protect.
In business, you can't always do that. Everything has a cost. Everything has what we call an opportunity cost. There's a cost to not doing something else.
You have to be able to use methods that are learning in the moment. We don't have historical data to learn from so you have to learn in the moment from the actions that you're taking and the reactions to those actions and have a cyclical kind of an approach.
If you notice at many states, I'm sure probably every state by now, has pretty much a daily briefing where you can listen to what the state government is doing and then we have, of course, federal government responding and telling us what they're doing on a very frequent basis. If you notice, every day there's some reference to what was said yesterday. There's some reference to what we think is true and, therefore, what we're doing. That's good science.
What I'm actually seeing in the public response to this right now is a pretty good, solid, scientific thinking. Not just, "Let's try this. Let's try that." Way too risky to do that in times like this.
Dr. David A. Bray: I just want to amplify, he hit the nail on the head. Science is always learning – the nature of science. There are things that science rediscovers. We used to think in the 1990s that neurons in your brain wouldn't regrow after a certain age. Now we discovered that's actually not true. They do regrow.
We're going to learn more and more about this virus. We're going to learn more and more about its effects. I think good world leaders, whether they're from the private sector or the public sector, will say, "I'm giving you the best that I know today, recognizing I'm going to continue to learn and I'm going to continue to adapt my response to this crisis."
One of the things I also like to tell people is the word "fail" is really an acronym for First Attempt at Iterative Learning. That's really what we're doing here. We are going to always be learning on a regular basis about COVID and this crisis response as we move forward.
Dr. Anthony Scriffignano: I have a simpler way of saying the same thing, David. I said, "Make new mistakes every day."
Dr. David A. Bray: Exactly. Yes. [Laughter]
How can we navigate the conflicting data on this crisis?
Michael Krigsman: We have an interesting question, a very relevant question, from Twitter. Eric Sapp says, "Are there other case studies, examples that can help inform the best response, both in policy and in press coverage, for the lack of reliable data that we're discussing?" He gives examples such as the bad testing, the mixed symptoms, reporting by open and closed nations is very different. What can we do to navigate this conflicting thicket of data flying everywhere?
Dr. Anthony Scriffignano: What the literature will tell you about this, from a leadership perspective, is authenticity will win. The more authentic you can be about what you actually know and what you don't know, the more responsibility you take for the mistakes that you make, the more that you are sharing information, the better off the greater good will experience.
Unfortunately, of course, culturally, not all cultures share with all other cultures and sometimes sharing information can disclose weakness. There are complexities to this that go beyond that. If you look at the best available advice in leadership in crisis, authenticity, communication, collaboration will win every time.
Dr. David A. Bray: Agreed. To build on what Anthony put as an excellent point, from 1960 up until now, open societies that are more open to deliberation and discussion and they separate their private and their public sector have been better than closed societies.
Now, I think the challenge is a lot of that literature has been based on historical events before everybody had a smartphone in their hands. We need to recognize that amid what's happening with COVID-19 in terms of the actual pandemic, we're also experiencing kind of an unprecedented and some have called infodemic. I just call it misinformation or disinformation.
Anthony and I were both involved with the response to 9/11, but 9/11 also when I later did the response with anthrax and even with SARS, there was misinformation. There was misinformation saying that SARS was being used by the U.S. government as some weapon or something like that, or that anthrax was intentional, or that 9/11 was a conspiracy theory. The challenge at that point in time, while there were conspiracy theories and misinformation, that was harder to spread because not everybody had, immediately available to them, a smartphone that allowed them to reshare, retweet, whatever.
I think what's interesting is we've already seen, over the last five years, an increase in misinformation and disinformation in a lot of cases from people that just don't mean to do it. They're cognitive biases or their confirmation biases are kicking in or press outlets have to pursue for-profit strategies and that means sometimes taking an edge that's a little bit more edgy and more selective of what's shared because they're trying to actually play to their viewership.
The challenge you have here is, this is all happening in the background. I do worry for those leaders that are trying to be authentic, will they be overcome by the infodemic that's happening around them?
Dr. Anthony Scriffignano: Let me just add to that. It's not happening in a homogeneous way. We assume that when we look at a newsfeed that we're seeing the news. We're actually seeing news that's been curated for us based on all kinds of algorithms that try to help us. In different parts of the world, different feeds are available, different information is available.
Sometimes it's just the availability of language. Sometimes it's more than that. There are rules about what may and may not be disseminated. Without judging that, we don't all consume the same "truth."
There's a reason why, when you go to court, you swear to tell the truth, the whole truth, and nothing but the truth. Those are three different things. We might have all true information but not all the truth. Not all information that's true stays true over time.
When we read about things, we are constantly using our amazing human brains to try to synthesize and try to make sense out of that. When we get to this infodemic that David is talking about, those processes become overwhelmed.
Now imagine you're a world leader and you're trying to do this. You're surrounded by people who are also trying to do that. They're trying to filter the information to give you what you need to know and they're having the same problem with being overwhelmed with information.
There's very much a constant disruption going on in times like this where the highest level of leadership that you can see from where you are is probably not hearing everything you're saying and you're probably not saying everything you're hearing. You've got to imagine the cumulative effect of that.
Managing data decisions vs. politics
Michael Krigsman: Arsalan Khan raises a really interesting point. He says, "The biggest factor in data is veto, the ultimate authority to accept or reject conclusions. But in times of crisis, we need to be steadfast on what the data is telling us versus what we want the data to tell us." That seems to get right to the heart of many of the political issues that we're facing right now. Who would like to handle that important hot potato?
Dr. Anthony Scriffignano: First of all, that's an example of the confirmation bias that David is talking about. We believe something is true. In any sufficiently large corpus of information, if you go in with a conclusion, you'll find something that supports what you think is true. If you think this is an alien conspiracy and you look long enough, you'll probably find some data that supports that. That doesn't make that true.
The data doesn't necessarily speak. The interpretation of the data speaks. We bring certain bias with us when we interpret the data. Different methods, different techniques have certain preconditions.
For example, a lot of machine learning requires some sort of example, some sort of training. Not all of it, but most of it or a lot of it at least. We can't train right now because we don't have those examples. We have to use the kinds of methods that draw inference from data and then move forward and then make a little bit of a correction and then draw another set of inferences.
Some of these techniques are called Bayesian methods. They're methods that are designed for moving from where you are, forward without necessarily looking too far backward. There are issues with that sort of reasoning and there are reasons why that's not always appropriate.
There's no Swiss Army knife here. The data is important. The preconditions, the critical thinking, the questions we ask, the way we challenge our bias, that's what's going to save us here.
Dr. David A. Bray: I would amplify that one of the thorniest challenges with epidemiology is you never really truly know the denominator. That's in the case like we may have right now countries that are sort of being held up and saying, "Look, they have a low case mortality rate." In that case, they probably have done a large sample, like South Korea. They did a randomized trial both for people that thought they had COVID but also people who didn't, and so they probably got the best towards a denominator as opposed to other countries that are only waiting unless you really are certain that you're presenting symptoms and then testing you. You're going to get an artificially high number only because you're not testing everybody else that may be asymptomatic for COVID 19 but are not presenting and, therefore, not testing. Therefore, you're getting a smaller denominator.
Dr. Anthony Scriffignano: You're testing people's ability to detect that they have the symptoms that they should go and get tested.
Dr. David A. Bray: Exactly.
Dr. Anthony Scriffignano: You're collecting the wrong thing in that case.
Dr. David A. Bray: Right and so, in that case, the data would tell you the wrong thing because you haven't paused and, like you said—
Dr. Anthony Scriffignano: Exactly.
Dr. David A. Bray: –challenge your assumptions, challenge your biases. You just went with the data as opposed to say, "Maybe the data is actually missing something and I need more data."
Dr. Anthony Scriffignano: Very frequently someone will come to me and say, "Can you use this data to prove … ?" I'll usually cut them off and say, "Yeah, but I'm not going to use the data to prove anything. I'll use the data to ask a question. I'll ask a question and try to answer it with that data. But by the way, before I do, I want to make sure this is the right data to address the question that we're asking, so let's get the question right because we might not even be looking at the right data."
To David's point, if we tried to collect the data about positive testing from all over the world and we know that in some parts of the world they're only testing people that are symptomatic and in other parts of the world they're trying to randomize, why would we put that data together and form any kind of conclusion?
How can we communicate clearly and accurately with conflicting data?
Michael Krigsman: We have another question from Twitter, a comment that fits right in. This is from Michelle Batt. Michelle makes the point that the data is changing quickly and people are referencing various versions of the truth, and so how do we clarify and communicate what is the truth? What is the truth anyways when it comes to all this stuff?
Dr. David A. Bray: In any crisis situation, what you believe to be true at that time; I don't know if I would ever say there's ever truth per se. You need to caveat down and say you're always going to be learning; you're always going to be adjusting.
That's why, like I said earlier, where I said this is not peer-reviewed yet. Partly why we go through the process of peer review is, we want to actually have many different eyes look at it as what they believe to be true and challenge those assumptions. It's hard for any one person to do it even though we should.
The challenge is, this is happening so fast. Not only is the data telling us new things that we should reconsider and reevaluate what we believe to be true; the virus itself is changing. Actually, like I said, it has genetic drift. It's mutating. It's going to have different presentations. Who knows? Hopefully, it doesn't happen but we may end up with two different strains. Then it's going to become even a harder problem.
I raise that, though, because I think anyone looking for a definitive truth, I would just put that by the wayside and say, "What do we believe to be true at this time and why?" As Anthony said, and amplifying E.E. Cummings with this too, "Always a more beautiful question asked, the more beautiful answer."
I love right behind you, you're now going down the rabbit hole because we are truly down the rabbit hole, Michael. But really, we should use our questions to guide why we are thinking certain things and that'll help us both to ask what data we need and also what predictions we have. Instead of trying to predict the future, ask the beautiful questions that'll help guide what we need to do to try and have a better sense of the future we're shaping with this crisis response.
Dr. Anthony Scriffignano: There are two different mindsets that we can approach. There are many, but there are two that are important here that we can use to approach a research question.
One is a positivistic mindset where the answer is out there and we're going to go find it. That question had a positivistic tone to it. How do we know which one is true? That assumes that there is a truth out there.
The other mindset is constructivist where we actually form that truth by doing the research and understanding. If we were to study racism today and we studied it 30 years ago, what we would have learned 30 years ago isn't wrong. It's just different in the context of today. That's an example of a constructivist kind of way of thinking.
Right now, we have to be more constructivist. There probably is no ground truth that we're going to ever discover because the thing we're trying to understand is changing while we're trying to understand it. We're all asking different questions and we all have different intent with the research that we're trying to do.
What we have to do is take a more constructivist approach. What are we learning? Exactly what David just said. What are we learning? What is it teaching us? What new questions can we ask from that? Move ourselves forward like that rather than trying to go find the truth. We're not going to find the truth because that truth is changing while we're looking for it.
Guidance to the public based on data science?
Michael Krigsman: Abhinav Aggarwal asked the question, "What guidance would you provide to the public at large from a data science perspective in terms of how we respond and how we manage the current crisis?"
Dr. Anthony Scriffignano: Anything that you think you're learning from a data perspective, first of all, see if you can triangulate. See if you can get that same information from some other non-informed source. You don't want to find two sources that are citing the same study. If you find something that you think is true, try to triangulate on it.
The second thing is, always ask the question, "What do I have to believe in order to believe this number?" I don't want to pick on a country. If the country of Somewheristan publishes data on their infection rates and they say, "This is the infection rate," probably the first thing you have to believe is that the way they're measuring it, measures infection rate, and the way you understand infection rate. The second thing is you have to believe that that information is unperturbed, that they haven't messed with the answer.
The third you have to believe is that it's a reasonably current number, that it's not three weeks old. Are all of those things true? Do you understand the source? There's a reason why when we read literature there's a page that tells you who wrote it, when they wrote it. There are references cited. We actually need to look at some of these things and they're important right now.
Michael Krigsman: David, Anthony just used the phrase, "You have to evaluate and ensure that the data has not been perturbed." David, to what extent do you think that's going on today and can you give examples?
Dr. David A. Bray: There were signs that severe acute respiratory syndrome was happening about five and a half, six months before finally both Vietnam and China said something was going on. That was 2003.
March was when, finally, Vietnam said they asked for international assistance and then China followed shortly thereafter. But as early as late 2002, there were signs because, one, there were people that were trying to do the right thing saying, "Look. You can't attribute this to me, but we're seeing something very odd here. People are getting sick."
At the time, we were calling it an atypical febrile illness. Then later, the other thing is that, at the Centers for Disease Control, we actually had ways of monitoring different things. One was the price of garlic.
The price of garlic was going up more than tenfold, which is seen as a cure-all in those areas. Unfortunately, it doesn't actually work but it's seen as a home remedy cure-all. That's a sign that people were doing a massive demand on it as a certain spike.
Why? Why might that country not have come forward as much as they did? I think, one, it was a sense of, to them, public health is national security. If you look at their history, it's actually somewhat understandable where, in their history, unfortunately, with the opium wars, they were exploited by outside foreign powers that used things like as much as heroin against them that they may not want to reveal that they have an outbreak in their own country out of concerns that other outsiders might take advantage of them.
That's the case where it's not necessarily, as you were mentioning, perturbing. There is sensitivity to sometimes revealing how bad it is because of the geopolitical impacts that it might mean both for your internal domestic stability, but also externally as well. Unfortunately, it is happening and we need to be realistic here that while we're dealing with the pandemic that is COVID-19, there are some countries that see this as a case where the world is distracted, focusing on this, and they may be able to do some strategic power grabs here and there or do things that normally would invoke world response and outrage. Because we're all focused on this, that these other things are happening.
This is the nature of human society. They're messy. They're complex. But as Anthony mentioned, as much as possible, if you're in business, if you're in the private sector, if you're in the public sector, strive to be that one that speaks the truth even if that means you're going to get a lot of arrows and mud thrown your way because, at the end of the day, you've got to be true to yourself and authentic to yourself even if, at the time, people don't acknowledge that or don't want to hear what you're saying.
Dr. Anthony Scriffignano: It's very easy to say the supply chains of the world are disrupted right now. It's very easy to say businesses have closed their doors. But that's the short-term effect that we understand.
What is the longer-term effect of that? Some of the businesses, hopefully most of them, will open their doors. What kinds of clues might we see in the data that suggest that?
When we think about shifting supply chains. Probably respirators are the easy one to think about right now. How many manufacturing operations are trying to shift their focus to something respiratory? What will be the impact of that? This isn't necessarily all negative from a business perspective.
For the world to come out of this and thrive, the economy of the world has to come out of this and thrive. It's not a dirty thing to think about business in the context of something horrible like this because that's the fuel that will help us produce the prosperity that we need to thrive and survive and emerge from this.
It's really important to look beyond what happened yesterday and what's happening immediately tomorrow. Of course, you need to do that. But you need to also have a little part of your focus on presuming we emerge, when we do,. How will we know it's happening? What will we be doing in that? How will we change our posture? How will we shift our focus?
You've got to be doing both if you're a leader. You don't have the luxury of doing one and then the other because the second one has a much longer runway than the first one.
Dr. David A. Bray: The strength of the United States is in times of crisis, whether it's war or warlike events. I would actually say we are at war with COVID-19. Unfortunately, mother nature sometimes is out to get us and so we, as humans, need to actually address what the virus is doing.
Our strength in the United States, but also parts of Europe and North America, is that our private sectors will mobilize when they actually have the opportunity to help with that crisis. It is about the immediate, but it's also about the recovery. I think, in some respects, trends regarding distributed manufacturing, the future of work, network-centric ways of working together, they were already there before COVID-19. What COVID-19 has done is it's actually sort of accelerated them.
If you are a business leader, you should actually say, like, for example, Home Depot now is actually having a service where you can actually order in advance what you want to pick up and it's ready for you to pick up. Then off you go as opposed to trying to spend 30 or 45 minutes gathering it from across the store. I hope that's something they continue post-COVID-19 because that's a value-add for their customers.
When will the Covid-19 crisis end?
Michael Krigsman: Arsalan Khan asks a question. Thank you, Arsalan, for asking this. I also want to know, "When is this going to be over?"
Dr. Anthony Scriffignano: Define "over."
Michael Krigsman: Over means I can go outside without making the assumption that literally every single surface may be coated with or contaminated with an unseen, an invisible poison, that if I touch can make me potentially deathly ill and, when I come inside, I must sanitize everything in order to keep that unseen poison outside. That's over.
Dr. Anthony Scriffignano: For this particular challenge and your definition of over, as a scientist, I would say when there's an effective medical treatment to give you some immunity to that thing that you're afraid of. Unfortunately, that's a very simplified answer because there'll be the next thing and the next thing.
What I would say is that there's an element of system learning here. It's a horrible lesson and it's horrible the way we have to learn it, but the world is learning, hopefully, something right now about how to retreat and protect itself in a way that has never been conceived up before.
I saw a fantastic meme the other day with the satellites around the Earth looking down on the Earth. There was a big, "Sorry, we're closed," sign on the Earth. This hasn't happened at this scale but the fact that it can happen at all, the fact that the Earth can cooperate even partially to this level is pretty amazing to me.
The fact that there is somehow a button that we can press that says, "Everybody, go home. Shut down the supply chains. Turn off the airlines." I didn't know we had that button. It's pretty amazing that we've learned to collaborate this way.
We're having this conversation remotely. We're all in our sheltered in place locations using a platform that was designed for lots of things. It certainly wasn't designed for the majority of the population to try to use it at the same time. Somehow, magically, that's working. Thank you, whoever you are behind the scenes.
I can't believe they're doing the same thing today as they were doing three weeks ago because there's no way they would have anticipated this kind of demand. Yet, here we are.
Dr. David A. Bray: Two things. First, mother nature is always out to get us. Mother nature will always continue to be out to get us. Those who claim, "We could never have foreseen this, we didn't know this was coming," the reality is, there were actually a small group of people, including those of us with the Bioterrorism Preparedness Response Program, that thought it's a low probability, high consequence event at some point in time a natural pandemic will happen.
The challenge is trying to get world leaders to commit funds to have additional capacity, additional laboratory tests, additional ventilators. Until it happens, it looks like it's a waste of government resources. Then it's like, "Well, why haven't you done enough in advance?"
When will this be over? As Anthony alluded to, ideally, it's when we have a vaccine that is readily available to everybody. The challenge is, as you and I all know, influenza is an example of one of those viruses that change every year.
Every year, data scientists working with epidemiologists and public health practitioners guess about five to eight months in advance what are going to be the dominant strains that are going to be present on the planet. They usually mix about three of them together. They do the highest probability ones and sometimes mother nature rolls the dice, she does one of the low probability ones, and that vaccine is no longer effective.
The question with COVID-19 that has not been answered yet is, will a single vaccine take care of it and will we be able to get it out to the entire planet fast enough? We're talking about the entire planet is going to need to get this. The last time we did that was with smallpox and I think that was 1977. We're going to have to figure out how we address that.
Do we discover COVID-19, much like influenza every year, is going to need a new vaccine every year as a way to address it? We don't have enough data yet but, as Anthony mentioned, the world has changed and is going to continue to change. This will probably be marked as both the before COVID-19 era and the post-COVID-19 era.
Michael Krigsman: I didn't hear any dates tossed about on when we're—
Dr. David A. Bray: It's going to be maybe months. It may even be years.
Dr. Anthony Scriffignano: You know, Michael, not to be flippant. When the Mayan calendar was supposedly predicting that the world would end in 2012, I was on a live event that was being broadcast. The commentator at the end sort of flippantly asked me what I thought; was I aware of the supposed Mayan prediction and what I thought of it. I immediately said the Mayans are wrong. She said, "Well, how can you be so sure?" I said because if I'm wrong, nobody is going to know.
Anybody that tells you that they can predict with any degree of certainty when this will be over or in what way this will be over, either they have a really fancy crystal ball, they have a poor understanding of the question, or they're making stuff up. None of those is a good outcome, really, right now.
Can open data help manage the crisis?
Michael Krigsman: A Twitter user makes the comment that open data is essential and can be very valuable in this kind of situation. Can you talk about open data and open data sources?
Dr. Anthony Scriffignano: There's what we want open data to be and there's the reality of open data. Certain types of open data, especially historical data—here are things that happened in the past. Here are demographics, census information, things like that—are great. Fantastic idea.
To say that we're going to take dynamic data, data in motion and somehow make it freely and instantly available to everyone is a very scary idea for a lot of reasons. Very often, data has to be cleaned. Very often, data is wrong. It costs money to produce the data.
Some data open, yes. In healthcare, for example, we have laws about privacy. We have HIPAA. We have all different types of laws around the world.
You can't just suddenly suspend all those. I don't want my medical data necessarily released.
If you anonymize it, there are costs that come with that. There are also risks that come with that.
Yeah, certain types of information, particularly data about things that have happened in the past, what we call data at rest, a good idea. Data in motion, data that are very expensive to create or curate, maybe not so much.
Dr. David A. Bray: Yes, I get the ideal of open data and I do think, like you said, data that are not in motion, that makes sense. Data that's going to be either really expensive to sanitize and clean and then, by the time you clean it, it's out of data anyway because new data has come in or data that does require expertise, as we gave the example earlier, it might that the data is showing something but you don't know the denominator. If you're not an expert in this field, you may reach the wrong conclusions because you don't know to pause and say, "Maybe we're not testing enough," or, "This is mixing something that has randomized tests next to people that are only getting tested if they think they have the virus."
Yes, I get the ideal for open data. All for it. There is also a place for data that's kept private for health reasons or for confidentiality of companies and their IP. I think it's a multisided answer. It's not a single type of data that will solve everything.
Dr. Anthony Scriffignano: A really good analogy is a pharmacy. You should be able to go into the pharmacy and buy certain things without any kind of help. If I want to go in and buy bandages, I want to go in and buy over-the-counter drugs, or I want to go in and buy something like batteries, yeah, you could kill yourself with a battery but you're pretty safe. We pretty much understand what those things are when we consume them and they're reasonably unchanged. They come with directions.
Then there are the other things that are in the pharmacy that are behind the counter that require a prescription that, in some cases, are controlled by regulation. There's a reason for that. You need someone to help you consume those drugs or those products in a way that's not going to be injurious to you.
Data is the same way. Some of it, you could just take it and use it any way you want. Some of it, maybe you need a little bit of explanation. Maybe you need a little bit of help. Maybe it doesn't mean what you think it means and you're likely to draw conclusions from it that are going to be antithetical to what you're trying to do or, in some cases, even really dangerous.
Michael Krigsman: If you cannot tell me how long this is going to last, this situation, however, you define it, then, frankly, what is the value of your millions of dollars' worth of data and data science? Why are we here? I can take guesses also, by the way.
Dr. Anthony Scriffignano: I'm not guessing. I'm making informed decisions based on a known epistemology, based on empirical methods, which means if other researchers did the same thing in the same way with the same data, they'd reach the same conclusions. We call this science for a reason.
We may not know the answer to the question that we want to ask, "When will it be over?" We might be able to ask a question that we can answer, "How will we know when the recovery phase starts?" That so-called second derivative shift, we should be able to see that in certain types of live data.
How will we know when our interventions are having the impact that we intend? By looking at the data. By measuring it. The data science will help us ask really important questions and then using the techniques that are empirical, that are reproducible, answer those questions.
It may not answer every question we want. Getting what you want all the time is for babies.
Advice for business leaders
Michael Krigsman: Anthony, what advice do you have for business leaders for using data to navigate this kind of crisis?
Dr. Anthony Scriffignano: Three things: Number one, communicate. Be authentic. Communicate, communicate, communicate.
Number two: collaborate. Do not try to do this alone. Do not try to do this in a vacuum. There's no way you know everything you need to know and there's no way the next guy knows everything they need to know. We must collaborate.
The third thing is, please think beyond the immediate crisis. Have some portion of your mindshare about what your strategy is going to be as we emerge from this so that we don't fall out of this; we run out of this.
Advice for policymakers
Michael Krigsman: David, what advice do you have for policymakers as we navigate this incredibly complex situation?
Dr. David A. Bray: Be aware there is an absence of trust, whether they're government institutions, whether it's the media. This has been something that was in the making before COVID-19, as mentioned. We're in a period of, the five years, misinformation and disinformation seem to be increasing for multiple reasons.
At the end of the day, we run the risk that we'll discover our own human brains were not ready for what the Internet-enabled us to do.
Now, at the same time, I can't imagine trying to get through COVID-19 without the Internet. Imagine if this was 1960s and we were all relying on HAM radios as a way of trying to communicate and coordinate. This would be very difficult. Recognize communication but recognize the challenges associated with the infodemic and do your part to ideally try to tamp down that fire as opposed to adding to it.
Second on the collaboration, definitely agree the only way through this crisis is through it together. We need to start thinking about not only what we do to empower—at least in the United States—local and state health departments because that's where the action is.
We need to empower front-line first responders. We need to address the fact that there's probably going to start being burnout amongst medical doctors, nurse practitioners, nurses that have to deal with this on a daily basis yet go home to their own families. Whatever we can do to make sure we get ahead of the curve to address the burnout.
Start also thinking collaboratively globally because the world has changed. The world order will not look the same when we emerge out of this.
One of the things we're trying to do with the Atlantic Council and we're working with others is to try to get a sense of where are going to be those disruptions with the recovery involving food, involving other sort of essentials for people, and we're going to find that some countries are going to need assistance. Woe be it unto us if we don't think globally while also addressing our own national needs.
Then lastly, as Anthony talked about, being aware of the fact that we've got to think long-term. When I was with the CDC and I've done this with other efforts as well in Afghanistan, we always had what we called the beta team.
The alpha team is the first responders. They're locked into the immediate. They're meeting the immediate needs.
The beta team, however, is the group that sort of thinks and says, "What else are we missing? What else do we need to do so that the alpha team, when they get to stage 2 or stage 3, has things actually in a better state?" They're able to take the more balcony view of everything that's happening.
That's partly what we're trying to do with the Atlantic Council, the GeoTech Center. I know Anthony is doing it at Dun & Bradstreet as well. We need more people that are coming together as a network beta team that is trying to think about the long-term because, yes, we need to address the long-term, but we've got to get ahead of the long-term as well. Else, we will find that we're going to not have as good of a future as we could moving through this.
Michael Krigsman: All right. A fascinating and important discussion with two brilliant minds. Anthony Scriffignano is the chief data scientist at Dun & Bradstreet. David Bray is the director of the GeoTech Center at the Atlantic Council. David and Anthony, thank you very, very much for taking time and to our audience, and especially to the people who ask questions, thank you so much for joining.
Before you go, please subscribe on YouTube and hit the subscribe button at the top of our website. We'll send you excellent information about CXOTalk. Have a great day, everybody. Watch CXOTalk.com. We have more great shows coming up.
Also, I've been doing LinkedIn live. If you follow me on LinkedIn, you'll see all kinds of other additional, really interesting conversations.
Take care, everybody. Have a good day and stay safe.
