Contagious Conversations / Episode 25: You Can't Fix What You Can't Count
Transcript
Claire Stinson: Hello, and welcome to Contagious Conversations. I'm your host, Claire Stinson. Every episode we'll hear from inspiring leaders and innovators who make the world healthier and safer for us all. Contagious Conversations is brought to you by the CDC Foundation, an independent nonprofit that builds partnerships to help the Centers for Disease Control and Prevention save and improve more lives.
Joining me today is Dr. Daniel Jernigan, deputy director for public health science and surveillance at the Centers for Disease Control and Prevention. Dr. Jernigan provides leadership for CDC's Data Modernization Initiative and serves as an advisor to the CDC Director. He joined the CDC's Epidemic Intelligence Service in 1994, and that was the start of a career at CDC for more than 25 years. Dr. Jernigan has led the agency's responses to dozens of contagious disease outbreaks and flu pandemics in the United States and around the world.
In this episode, we talk about the importance of data modernization and its impact on public health. Dr. Jernigan also discusses the Lights, Camera, Action National Summit Series convened by the CDC Foundation and other partners to advance recommendations for a modernized U.S. public health system. Welcome, Dr. Jernigan.
Dr. Daniel Jernigan: Thanks. Great to be here.
Claire Stinson: We're really pleased you're a part of the conversation today. So, let's start today by talking about your background. You joined the CDC's Epidemic Intelligence Service program, also known as EIS, in 1994. Why did you want to be a disease detective?
Dr. Daniel Jernigan: Well, back then I was in medical school, and like other people that have come to CDC, I learned about CDC and the Epidemic Intelligence Service from somebody called Steve Thacker, who came to visit our medical school. And from him, understood the incredible kinds of work that you could do. I was very interested in internal medicine and in infectious diseases, but the interesting part of medicine was the diagnosis.
The clinical management part of it was not as interesting for me. And so, when I learned about the program and I heard from a few other doctors that had done the program at Baylor College of Medicine, where I was, in Houston, they always had a gleam in their eye about the program, talking about it as the best thing they ever did. And so, for that reason I came and did a one month student elective, which is available to professional students to come and work for just a short period of time at CDC. But through that program, I got to go on an outbreak to New York City with some strange kinds of bugs that were causing disease up there. And then after that, joined in the respiratory diseases group and was able to do a lot of respiratory disease outbreaks from that time.
Claire Stinson: So, while influenza has been your focus throughout your career, it is only one part of your expertise. Tell us about your current role at CDC.
Dr. Daniel Jernigan: In my current role, I took on the new role of Deputy Director for Public Health Science and Surveillance. And so, after serving as the Director of the Influenza Division for several years, I moved into this space to address a number of the big concerns and issues that the agency is trying to fix. As the Deputy Director, I'm over the National Center for Health Statistics, over the Office of Science, over the Office of Laboratory Science and Safety and the Center for Surveillance, Epidemiology and Laboratory Services.
The big activity right now is stewarding the Data Modernization Initiative, which is a multi-billion dollar activity to try and address a number of long-standing problems in surveillance and data capture that need to be addressed and need to be fixed in order to have public health to a place where it's really modernized and able to do what it needs to do for major emergencies.
Claire Stinson: Right. And would you say that your work at influenza really prepared you for your current role and also to provide learnings to the COVID-19 pandemic?
Dr. Daniel Jernigan: Yes, absolutely. A number of us were involved in a pandemic in 2009 due to the influenza H1N1, that was an influenza virus that jumped from transmitting in pigs to transmitting in people. And when it did that, there were a fair amount of the population that did not have any immunity to it. Older people actually did have immunity, and so were relatively spared from the significant problems like we've seen with COVID. But during that response, we were able to put together task forces and groups that were able to address things like communications and community mitigation, laboratory, vaccine development, things like that. So, all of that in response to that 2009 emergency really helped with us to be prepared for subsequent problems that we've also addressed like Zika and Ebola.
Claire Stinson: Fascinating, thank you for sharing that. So let's turn to COVID-19. What have we learned from the COVID-19 pandemic that is helping us improve the quality sharing and use of data for preventing chronic and infectious diseases?
Dr. Daniel Jernigan: COVID was not the first, but certainly the largest to try and collect information on every case that occurred. We recognize that we don't get information on every case, but the attempt during COVID has really been to get as much information as possible. And so, that requires that anyone that's tested, that result can be captured; not only a positive test, but a negative test as well, that means that there's lots of data that's being collected.
And for that data and then all of the contacts that had been with those cases that might turn out to have illness, especially early on in the response, we needed to know who all those contacts were and talk with them and give them guidance on how to prevent further transmission. That's a lot of data. And we learned that we don't have the people to do the work that's needed, we didn't have the electronic data systems that could handle that volume of information, and the data that we were collecting and the interpretation and information that was being generated was not fast enough.
So, COVID appearing at the time it did, people are very used to having immediate data about everything on their cell phone, and so, we need systems or needed systems that really could collect that data from multiple different sources, get it in as quickly as possible, be able to handle the volume of it, be able to analyze it and do things like forecasting and prediction with that data, and get that information out as quickly as possible and visualize it for the country so that they could take their own actions to prevent further transmission. So, those needs, we have on a much lower basis regularly, but during a time of crisis, the overall system needs to scale up quickly and be able to handle that. There are a few challenges that COVID had to address that are really a part of long-standing problems that we've had in the United States. The first is really that the U.S. has a federated public health structure and diffuse authority. And what do I mean by that? The U.S. Constitution actually doesn't mention public health in it. And so, public health is the responsibility of states. And so, the overall U.S. national capabilities and authorities to do public health are relatively minimal.
We work, at CDC, generally through our collaboration with our state health departments and local health departments in order to do work. During a crisis when things need to happen quickly, or new kinds of data need to be collected, there were not a lot of ways to collect that information through authorities. And so, we worked with the Secretary of Health, worked with the Centers for Medicare & Medicaid Services, and we were able to identify ways to have nationally collected information about various different parts of COVID, but that kind of thing only could happen during an emergency. And so, some changes in that space are needed.
In the U.S. we don't have a single payer for insurance, we have multiple different insurers and different kinds of companies that provide care. Those companies don't necessarily speak easily with each other, with data sharing, and also they cross state lines where different authorities occur. Equally, public health is distributed too. In Massachusetts alone, you have a health department for almost every city, which just means that it's extremely distributed and difficult to have a coordinated common approach when you have so many different authorities with different leadership, mayors, governors, et cetera, all having their own different approaches to how to address these major concerns.
Like I mentioned, a second thing is that we do have a decentralized healthcare system. The United Kingdom, other countries, have single systems, single capability to collect data, and the ability to really understand things because they're already in place collecting that information. So, in the U.S. there is not that kind of approach. We'd love to get there, some things are happening to make that change, but it certainly was not there at the beginning of the COVID response.
A third thing is the underinvestment in public health data capabilities. And so, that just means that there has not been funding toward infrastructure for surveillance, data collection, et cetera. So, there just has not been that capability to keep the systems running, let alone modernize them. The U.S. government actually supports most state health departments with federal dollars to keep them going.
A fourth thing is that the... and this is a problem that the federal government has largely generated, and that is, money comes from Congress for specific programs, let's say for influenza. And that money turns into siloed data systems and siloed funding and siloed staff at the state health departments that are doing that work. And those silos are able to collect information for that particular pathogen or program, but they're not able to talk with one another or be able to scale up during a problem. And so, those siloed systems at the beginning of this pandemic definitely led to less impactful timeliness, completeness and quality.
The last thing, just there is a balance too within public health where some data you can collect very quickly and get that information out, but that information may not have the full breadth of findings that you need to change policy. So, I might be able to tell you that there's 2,500 cases. I'm not going to be able to tell you what the racial and underlying conditions and how many were fully vaccinated, that kind of information takes collection from multiple different places to come together. And so, we have to understand that we can have accurate and very complete information that's useful for policy changes, but that may not be possible in the time that you want it to be there for counting cases for situational awareness. So, those two things are in balance all the time, getting things fast enough to know what's going on as trends, but also needing to wait a little time in order to get that more rich information that helps us change our policies. So, those are five things that we know were problems in place prior to the pandemic that were amplified and clearly made it more difficult for us to respond once the pandemic started.
Claire Stinson: So many challenges and so many lessons learned. So, let's take a step back. Dr. Jernigan, why is data so important in public health?
Dr. Daniel Jernigan: Data is the currency of public health. You can't fix what you can't count. In order for you to know what the problem is, you need to enumerate it and quantify it. And once you have that information, then you know whether things are getting worse or getting better. With that information, we're able to say what trends are happening, who's not getting the care that they need, how to improve the process of getting more information so that we can then take that. But that data, for instance, comes into a health department, just essentially a raw material for the work of public health. Once the data is there, it gets clarified, it gets fixed, and it gets additional data added to it, and it turns into information that is a determination of what the problem is, how bad it is, who's being infected, et cetera.
And so, with that, state health departments then can use that information to help change the activities. So, if it's a behavioral problem, like hypertension, having that information, knowing which groups are most affected, allows you to focus your preventive services to those that need it the most. With that information, you can then use it to change big policies, change funding, other things that are needed in order to better address those public health problems. So, data is that raw material that comes in, turns into information, and then eventually turns into actions that we can do public health work with.
Claire Stinson: That's a great way of explaining that, that data is the currency of public health. So, with that in mind, as part of the Lights, Camera, Action national summit series, the CDC Foundation and its co-hosts are exploring ways to build a stronger, more equitable and more resilient public health system for the future. And in the most recent summit, the discussion centered on creating an interoperable and modern data and technology infrastructure. Why is this important?
Dr. Daniel Jernigan: Well, it's important so that we can get the public health ecosystem up to a place where it should have been and begin using new technologies to bring it to a place where it needs to be. So, if you are a usual citizen, you have a cell phone, not everyone has a cell phone, I recognize that, but for those that do, these smartphones are really changing the way your life unfolds. So, just in healthcare alone, we see that medicine is moving to this decentralized, on-demand, at-home kind of approach. So, whereas before you would make an appointment, you'd wait two weeks, and then you'd go in and then you'd wait another couple weeks to get a laboratory test and... it would take time.
Now, you can talk with your doctor through text or through patient portals, order a rapid test for things, some of them require you to use your phone for, you can have a telemedicine interview with your doctor, you can now download your entire electronic health record onto your phone. Your phone can tell somebody that you have been vaccinated, and eventually your phone will be able to connect in with other parts of the healthcare ecosystem so that you can get results from your lab tests, or you can set up appointments.
And also, the healthcare system can send you information about how to treat your hypertension better, or your diabetes, or you can monitor it with things that you wear. So, that's where medicine is going, and public health isn't even on the periphery of that. It was not a part of a lot of the changes that have been happening to make medicine different.
Claire Stinson: Wow. So it sounds like public health really does need to be brought in to that modern data ecosystem, a really important topic right now. We'll be right back with Dr. Daniel Jernigan.
The CDC Foundation is convening a national summit series on the future of public health in collaboration with the Association of State and Territorial Health Officials, the National Association of County and City Health Officials, Big Cities Health Coalition and other public health partners to advance recommendations for a modernized U.S. public health system. The Lights, Camera, Action summit series includes four virtual convenings, leveraging recommendations across a variety of research. Learn more and register at futureofpublichealth.org.
And now back to our conversation with Dr. Jernigan. So, for quite a bit of the public health data, there is a lag time from when data is collected, analyzed and reported. How is CDC working to address this challenge?
Dr. Daniel Jernigan: Yes. So, some of those lags are kind of biologic that the viruses or the bacteria don't let you go any faster or some of the technologies don't. For instance, right now you hear about all the variants that are emerging and are causing disease, Delta, Omicron. The way that we know a COVID virus is an Omicron variant is by doing next-generation sequencing, which is just a way of looking at the inside of that virus and determining the sequence of the different proteins that are in there.
So, with that, it helps us to know what's circulating with variants. It takes about at least a week to a week and a half, or longer, from the time you put a swab in somebody's nose to the point where you can say it's an Omicron. So, that's not a data lag because of slow systems, it is part of the biology of it.
And so, there are some others like that, but deaths. We know that when cases start increasing rapidly and people want to know how fatal is this, you won't know for a while, because it does take some time before the infections continue to worsen. People get hospitalized and then subsequently die. And even once they have died, there are systems where we collect that death data that generally can take up to 10 days before they make it all the way through.
We are now working with our National Center for Health Statistics through funding that we're providing to state health departments, so $200 million, to get the death reporting systems to go faster. Won't make the actual time from when somebody is sick to when they die, but from the time that they have died, we are working to get the reporting much faster. So, right now our goal is to get 80 percent of all those reports to be within 10 days. And so far, even since the beginning of the pandemic, we have now up to around 67 percent of those. So, we're making progress at getting that data faster so that we can follow what's happening with deaths more quickly.
There are other issues, too. For instance, we have implemented electronic laboratory reporting of all COVID negatives and positives in the United States. And that translates to around 1.5 million lab results per day being sent to state health departments. So, those results can come quickly. They can determine that somebody had a test, but verification that that test is in fact a new case takes a little more time. And so, we're working to try and connect different sources of data to help add to that lab data so that states can automatically say whether it's a case or not. Without that additional data, you don't know if that lab result is just the third or fourth result that that same patient has had versus it being the representation of a new case.
So, technologies can help with that time reporting to CDC about the case, but even if we get them fast lab results, the state health departments still are...it takes time for them to put that information together to get to CDC. So, there are different parts of the life cycle of data as it gets collected, as it gets stored, as it gets interpreted and analyzed, as it gets reported and visualized. Each of those steps, we are looking at to make sure for all of these different data sources that we're addressing everything we can to try and make the whole process faster. We're doing that for COVID, but we want these gains to also be used for all the other reportable diseases and other diseases that the public health establishment is following.
Claire Stinson: This is such an important topic and conversation right now. I'm so glad to hear that CDC is so focused on addressing these challenges. I know that there has been frustration with lag time, with test results, with COVID, and things like that, but there is such a need for accuracy here. Let's switch over to speak about the CDC Data Modernization Initiative. You mentioned it earlier, it is a multi-year billion-plus dollar effort to modernize core data and surveillance infrastructure across the federal and state public health landscape. The initiative offers unprecedented opportunities to generate real time, complete and actionable data to prevent disease, promote wellness and assure prosperity. What concrete actions would you prioritize during the next three years to help assure those investments translate into long-lasting improvements to the health data system?
Dr. Daniel Jernigan: There's a lot to do, but we have recently published on the website and presented at the Lights, Camera, Action CDC Foundation meeting that we had this week. There's five areas where we are working and prioritizing. The first of those is what we call building the right foundation. By that, we mean getting the data from those electronic data sources coming automatically to the states and to the CDC so that you can get the information quickly.
A second component of that foundation is that once you get that data, it needs to be in a place where you can use it and share it easily. So, what that means is, building the right foundation is, using electronic laboratory reporting, electronic immunization reporting, electronic case reporting, and electronic death reporting, and electronic messages from 70 percent of the emergency departments in the United States. So, getting that data in and then getting it to a place where it can be used in these modern ways, ways that you're used to doing all the time, if you have iCloud or OneDrive, or whatever, you can get to your cloud information from whatever your PC or phone is, we need to be able to have the capability that public health data can get into that kind of cloud infrastructure as well. So, building the right foundation is assuring we are getting rapid electronic automated information from the data sources like the electronic health record, and getting it into the cloud's platform. At CDC, as well as at state health departments, that's a big lift, that's a real infrastructure-y kind of thing that has to happen while we're working on all the other parts.
A second part of the priorities is taking the data and accelerating it into action. And so for that, we want to not just have the data sit there in the cloud, but have it be able to predict what's going to happen and forecast what's going to happen, use important analytic tools to understand trends, be able to detect health equity concerns, or be able to understand how climate from one data set is able to affect chronic diseases in another data set, so that acceleration of data into action, especially so that it could be in a response-ready system, one that can scale-up quickly to a national response or be in the same kind of approach for day-to-day business. So that second component, that second priority is accelerating data into action, especially through the development of new standards that allow us to do that connection with the healthcare ecosystem.
A third priority of the five is developing a state of the art workforce, essentially making sure that we can bring on smart people to help us, but also to make our people and myself smarter so that we're able to use data science in ways that we weren't before.
A fourth of those five is to support and extend partnerships in order to have policy changes. Technology is great, we have a lot of it, it can change the way we do things, but it won't get implemented if we don't have new policies, new data use agreements and new regulatory components to help us with that.
The fifth of those priorities is managing change and governance. What we're talking about here of using enterprise-wide services or of working with a number of partners in different ways requires that we think about things differently. And so, managing that change and having governance that allows for incentives and for teeth in order to make sure that these things happen. That's the fifth of those priorities that we're working on now.
Claire Stinson: Fascinating, and all really important actions that you would prioritize during the next three years. So, Dr. Jernigan, there's a divide of sorts in public health: those with deep understanding of the standards, technology, and implementation for data modernization; and those that are considerably less familiar with those things. How does that impact cohesion and progress in public health?
Dr. Daniel Jernigan: It's that way historically. Our public health partners at state, territory, local and tribes have variations in the funding and resources that they have, and in the capabilities that they have. Some are very small and just don't have the bandwidth or don't have the staff to do some of these things, and others are very large state centers that do have that capability. And so, through a number of new grant programs that we have, for instance, the new workforce grant is $3 billion over three years to bring us to a state of the art workforce. That's going to be distributed among 108 jurisdictions. And so, that is a way to try and level the playing field to get some help to those states.
Through some of these approaches, we also have direct assistance, that is, that they can request staff that the CDC can make available to help them with the implementation of a lot of these things. There's also the need to work closely with the public sector, with industry, with vendors and with service providers. There are public private partnerships that can be transformative in this space. And so, we want to be sure that we are reaching out to them, connecting them to those that have the greatest need, and transferring that technology and that knowledge, but also developing innovations that can then be applied across the public health ecosystem through the use of open standards and the use of solutions that can be used in many different places.
Claire Stinson: Thank you for sharing that. A recent national commission report from the Robert Wood Johnson Foundation on equity-centered data noted we have an opportunity now to create a data infrastructure that is centered on equity, and that creates fair and just opportunities for everyone. What do you consider the single biggest challenge to achieving an equitable data infrastructure?
Dr. Daniel Jernigan: There are a number of challenges there. I think one that is an important one to start with is the representativeness of our own workforce. And so, as a part of the grants that will be coming out, there's a clear need for us to try and identify how we have a diverse and equitable workforce that can help the public see that our workforce represents the general public and is able to then think of things differently and come up with solutions that are going to be more broadly applicable to the whole society.
There are a few places where health equity in the collection of better information are focuses for us. We do want to have more complete and higher quality public health data. The more you can collect and the more you can develop the systems to capture race, ethnicity and social determinants of health information as a part of that primary data collection, the better off you'll be in order to have that data to apply.
There are also ways that we need to do communications to the general public. Some people when they're asked to fill out the information at their healthcare encounter, don't understand some of the categories for race, ethnicity or other questions. They want to know, why do I need to know what neighborhood I live in? Or why is it that you want to know if I use public transportation? There is a need for us to tell the general public why that information is important to collect, and also what it means. So, that's one thing. So starting at the very front before someone has a healthcare encounter. And then once that data is collected, to make sure that we're collecting all of it and are able to put it into a format that can be used by different parties to see the race, ethnicity and social determinants of health distributions.
You know, we want more representative data. Of course, we want more open and accessible data as well. So, there's lots of data that CDC collects and even data that's outside of the public health system that can be quite useful like Google social mobility data, or other social vulnerability index data, things like that, that people can use to understand better where these disparities are occurring. So, getting that data available is good. But one interesting area that DMI will be working on is artificial intelligence, machine learning. Those capabilities use algorithms. And those algorithms are just simply math that's being used to try and tease through a bunch of data and come up with an interpretation. Often, those algorithms themselves can introduce bias. And so, we have some efforts through our new Center for Forecasting and Outbreak Analytics at CDC, where we're working with academics to see how are we introducing bias through the analysis that we have? And so, that's an area as well where we want to make sure that the way we look at the data isn't introducing bias there as well.
Claire Stinson: So it sounds like lots of challenges, but lots of opportunity here. And it's great to hear that CDC is working to achieve an equitable data infrastructure. So, Dr. Jernigan, we're focused on data here in this conversation. You said that data is really the currency of public health. Looking beyond COVID-19, how do we show value to the public when it comes to the importance of data?
Dr. Daniel Jernigan: Well, one way is by providing them the valuable part of that data. So the interpretation of it. And so, as we collect information, we want to be sure of a couple of things. One is that the data is available to other researchers or available to the public, available to the media so that they can use that information themselves and turn it into useful media reports or in documents, et cetera, to take that data and also put it into actionable bits so that a person can see how they can do things differently themselves in order to have an overall better public health outcome, which makes their lives better, but also makes others better.
So, the data itself is not useful if it sits in a tower somewhere, it needs to be turned into information, it needs to be made visual, and it needs to be able to be ingested by multiple different parties so that as that data comes through, it can lead to a whole number of activities that will help public health to do its job and help individuals understand where they fit in the overall public health mission.
Claire Stinson: Thank you so much for being a part of Contagious Conversations. This has been a fascinating, an important conversation, on the importance of data modernization.
Dr. Daniel Jernigan: Great. Thanks a lot.
Claire Stinson: Thanks for listening to Contagious Conversations, produced by the CDC Foundation, and available wherever you get your podcasts. Be sure to visit cdcfoundation.org/conversations for show notes. And if you like what you just heard, please pass it along to your colleagues and friends, rate the show, leave a review, and tell others. It helps us get the word out. Thanks again for tuning in, and join us next time for another episode of Contagious Conversations.
###