Latest / Public Health Review Morning Edition / 1119: How PopHIVE Is Rewiring Public Health Intelligence
Transcript
- 0:00This is Public Health Review Morning Edition for Tuesday, May
- 0:035th, 2026. I'm John Sheehan with news from
- 0:06the Association of State and. Territorial health officials.
- 0:09Much of the data public health leaders need already exists, but
- 0:13it just isn't as accessible as it could be.
- 0:16Today we'll hear about a new platform aiming to unlock the
- 0:19full potential of population health data.
- 0:22Doctor Anne Zink, a senior fellow at the Yale School of
- 0:25Public Health and former ASTO president, tells us about Pop
- 0:28Hive, or the Population Health Information Visualization
- 0:32Exchange. Born out of frontline
- 0:34frustration during COVID-19, Pop Hive brings together de
- 0:38identified data from across healthcare, public health, and
- 0:41even non traditional sources like Google search trends and
- 0:44home healthcare monitoring devices into one open
- 0:47interactive tool. Later, Doctor Jen Layden, Senior
- 0:51Vice President of Population Health and Innovation at ASTO,
- 0:54will talk about other data sharing, public private
- 0:56partnerships and tools like Pop Hive that are improving early
- 1:00detection of threats and empowering public health
- 1:03decision makers before the next crisis begins.
- 1:07Pop Hive stands for Population Health Information Visualization
- 1:11Exchange. But the simple way to think
- 1:14about it is it's just de identified data that you all as
- 1:17state health officials, as state leaders can use today for free.
- 1:21I like to think about it as the gift to my former self.
- 1:23I was incredibly frustrated by not being able to see
- 1:26particularly de identified data, spending a lot of time having
- 1:30really meaningful but challenging conversations with
- 1:33Ehrs and TEFKA and Hies and all sorts of other things.
- 1:37And those all play a really important role in this ecosystem
- 1:41of data and data exchange that we need.
- 1:43But the number of times, like particularly like during COVID,
- 1:46that I went to our world and data or I went to the New York
- 1:49Times or I went to the Johns Hopkins website to just have a
- 1:52better sense of what was happening around me and to see
- 1:55how we compared to other states, how we were doing.
- 1:59It felt like there needed to be better ways to have this
- 2:01information. This was really accelerated then
- 2:03when I had a grant to decrease diabetes and my team could only
- 2:07see the Burfers data. And that data is great.
- 2:09But in Alaska, that is 527 people that they interviewed on
- 2:13the phone and that's limited. And as a clinician, I knew that
- 2:16there was data sets such as hemoglobin A1C that are required
- 2:21for me to be able to test, to be able to monitor someone's
- 2:23diabetes. That gives me a lot of detail to
- 2:26the data outside of just has someone been diagnosed with
- 2:29diabetes. It tells me how how sick they
- 2:31are with diabetes, how poorly controlled their diabetes is.
- 2:34That's a really important metric if I'm going to use millions of
- 2:37dollars to address chronic diseases or other challenges.
- 2:40So when I stepped down from the state, I was trying to, you
- 2:44know, think about what I wanted to do to try to make population
- 2:47health better. And part of that was thinking
- 2:49about how to make state health officials, epidemiologist world
- 2:52a little bit better and was lucky enough to spend some time
- 2:56at Yale School of Public Health. And together we built this
- 2:59project called Pop Hive. And we've been able to sit on
- 3:02top of all sorts of really interesting and great data sets
- 3:05and sources, making them publicly available.
- 3:07The biggest kind of change and game changer for us was being
- 3:11able to sit on top of the Epic Cosmos data, which really has
- 3:14not been available in the past, publicly available to be able to
- 3:18see that sort of data. But then we can also sit on
- 3:21everything else from home thermometer testing data to
- 3:24crisis line AI generated synthesis of what people are
- 3:28stressed about and what they're texting about.
- 3:30So instead of worrying about specific data standards, which
- 3:33are really important if I'm looking at line level data and
- 3:35I'm really trying to understand the TB case, it's able to really
- 3:39ingest data from from different ways, different avenues and be
- 3:43able to understand context of population health in a different
- 3:46way. Wow.
- 3:47And tell us a little bit more about how Yale fits into the
- 3:51picture and why academia is so important to to building these
- 3:56kinds of tools. Yeah, I mean, I live in Palmer,
- 3:59AK. Like, I don't live in New Haven,
- 4:01CT, I've never, I mean, I went to Med school, I went to
- 4:04residency, but that was kind of my last experience honestly,
- 4:07with academia. But what I've really found and
- 4:10loved working in this partnership between the private
- 4:13sector, the public sector and academia is being able to align
- 4:17different skills and resources to be the best version of
- 4:19ourselves. So sometimes I feel like Yale is
- 4:21like a Mary Poppins bag. I'm like, I need an
- 4:24epidemiologist to help to understand this or who is a
- 4:26modeller or who's an AI expert to help with this.
- 4:29And suddenly, like, I don't know where they come from, but Dan
- 4:32Fine, some great person who's there and they're all connected
- 4:35to this other entire network of academics who are then
- 4:39incredibly helpful. So for example, Burfers data,
- 4:42the CDC has done a great job of trying to model that data to
- 4:45understand well, how representative is this or not?
- 4:48But that question really hasn't been asked in the same sort of
- 4:50way of electronic health records data.
- 4:53How much, how much representation do I need?
- 4:55What is good representation? How much can I use this for a
- 4:58community? Can I make real decisions based
- 5:00on real dollars, based on this data?
- 5:02But there's people at NYU doing that work, there's people in
- 5:05Philadelphia doing that work, there's people at Yale doing
- 5:07that work, there's people at Emory doing that work, and they
- 5:09all know each other. And so then one of the really
- 5:11fun things that's happening with our partnership with ASTO is we
- 5:14can come back and we can ask them how representative is this
- 5:18electronic health record data? How much can I trust it?
- 5:21And then you get these modellers who really understand the
- 5:24representation looking at different data sources and can
- 5:26put that to action. Yeah.
- 5:29And you, you described this a little bit already, but like,
- 5:32tell us some more about the kinds of sources that are going
- 5:35into this tool and, and what kinds of gaps.
- 5:39It's, it's illuminating. Yeah, I, I mean, it's one of my
- 5:42favorite parts of this whole project, to tell you the truth.
- 5:44Sometimes it feels like a little Easter egg hunt.
- 5:46I just feel like there's always some other data set or source
- 5:48out there and we're very nimble in how we built it in.
- 5:53So some things that we did initially was to take big public
- 5:56data sets that were already available.
- 5:57So for example, CDC has a lot of data on wastewater, on a
- 6:01syndromic surveillance, on, you know, different rates that are
- 6:05happening from from infectious diseases to to chronic diseases,
- 6:08et cetera. So ingesting that there is more
- 6:10and more data available from Medicare that you can be able to
- 6:13ingest looking at Medicaid data. So looking at claims and billing
- 6:16data, we've been able to sit on top of previous work.
- 6:20So the Carnegie Mellon group, the Delphi group out of Carnegie
- 6:23Mellon had previously looked at and been working with CDC in
- 6:26many states to look at near real times claims data.
- 6:30So you know, when I'm in the emergency department working,
- 6:32I'm charting, but that is actually on the back end using
- 6:35technology to claim to bill and claim in real time.
- 6:39And that is another signal that can be used on top of things
- 6:41like ADT feeds to understand very quickly what's happening
- 6:45in, in a community or in every, anything from infectious
- 6:48disease. So like what I'm putting in my
- 6:49chart. So they use claims data.
- 6:51So we sit on top of their claims data.
- 6:53We've been having conversations about looking at larger claims
- 6:55data to understand the cost and the implications as a, as
- 6:59another way to help build the healthcare, public health joint
- 7:03infrastructure so that we're using our dollars as wisely as
- 7:06possible to improve public health.
- 7:08But then the other really funny thing about it, you know,
- 7:10there's always this joke that like Google knows if you're
- 7:12pregnant before your family does because of what you're putting
- 7:14in there. There's a lot that we're
- 7:15constantly putting into social media and to asking questions.
- 7:18So we have a partnership with Google.
- 7:20We can look at Google Trends data we are working with, as I
- 7:24mentioned, home thermometer and blood pressure cuff monitoring.
- 7:26I think particularly as as we get more of this individualized
- 7:29personalized data, you know, patients bring into the ER to me
- 7:33all the time on a piece of paper what their blood pressure looks
- 7:35like all written out. But what if we had that as their
- 7:39home blood pressure monitoring and they're pregnant and we can
- 7:42see signals in that much earlier to see when is this a lack of
- 7:45access to care? What are the challenges in
- 7:47getting someones blood pressure well controlled at different
- 7:50stages of their life? So I think all of that
- 7:52information and then honestly with AI, our ability to sit on
- 7:55top of conversation, there's data, but then there's
- 7:59information. And I think AI was really
- 8:01allowing us to sit on top of information in a different sort
- 8:04of way. So for example, if people are
- 8:06using a text chat line to discuss their frustration or
- 8:10even what they're searching in, in different searches, AI can
- 8:14kind of pull out of that trends and themes.
- 8:17Pop hive is only de identified population level data.
- 8:21And so we only sit on it's all there's no way for me to even,
- 8:24you know, someone could hack all of our data tomorrow and there's
- 8:26still no way to cross it because everything is de identified
- 8:29before it even comes in. And we're not trying to pat
- 8:31patient match it. But what we're trying to do is
- 8:34to help to show where data comes together and where it's
- 8:36different. So for example, someone says,
- 8:38I'm not sure I trust CDCS data right now.
- 8:40Well, when you look at syndromic surveillance data from CDC, it's
- 8:44lining up really closely and very well with everything from
- 8:48the Delphi claims data to other data sets that we have from Epic
- 8:50Cosmos. So you're like, OK, well, that
- 8:52looks like it's pretty consistent.
- 8:54And our goal is to be totally transparent about the data.
- 8:57Data sets. You can go play with it.
- 8:58You can click on and off. You can compare them.
- 9:01We're working on an AI model so you can ask it lots of questions
- 9:04because we really believe that community, local health
- 9:07officials, state health officials have the best context
- 9:10to put this context into use and to augment it with what they
- 9:14understand about community, augment it with their own data
- 9:17to be able to make data for action.
- 9:18Wow, you just went through a lot of different aspects to this
- 9:22this program. Partly how people can can trust
- 9:28the information going into it by cross referencing it between
- 9:31various sources to how how you can use AI to sort of add new
- 9:39levels of complexity and context to these like disparate datasets
- 9:44that you have. That's.
- 9:46That's a lot of it's a lot of power in one place.
- 9:50It's been fun. I mean, I think that there has
- 9:52been this model for good reason where we had an individual
- 9:55patient that got sent up to the healthcare, the local health
- 9:58authority then set up to the state, then got upset to the
- 10:00federal government. They understood it or analyzed
- 10:03it and set information down. But maybe I'm biased because of
- 10:06my experience in a state like Alaska, you know, when you're
- 10:08close to the ground, you have a different understanding of
- 10:10what's really happening. And you know, UDVI advoc Alaska
- 10:13is really different than Atlanta, GA.
- 10:15But in the meantime, there's been a lot of consolidation of
- 10:18what's been happening within the healthcare and within tech.
- 10:21And so I think if we can be able to take that consolidation, for
- 10:24example, like Epic data, Google Trends data, and we can push it
- 10:28down back to state, local, community, individual level,
- 10:32then people can put that into context in a different way to be
- 10:35able to put action. This is never going to replace
- 10:38your TB case control. I could be able to look at this
- 10:40is never going to be the end all be all for all data that's
- 10:44needed in public health and healthcare.
- 10:46But I think it's an an additional super useful tool in
- 10:49understanding population level data that can compare states,
- 10:53that can compare data sets and sources, and can be both
- 10:56hypothesis generating as well as just ease of use so that you
- 11:01don't have to. I'll never forget when a very
- 11:04large hospital system was begging CMS to regulate them
- 11:08better because it was so hard to work with all of our different
- 11:10jurisdictions. And that was an eye opening
- 11:13moment to me about how hard it can be to work with all the
- 11:15different states. And so if we can take some of
- 11:18that burden off by at least for some of the data, trying to get
- 11:21that back out to States and local communities as easily as
- 11:24possible, rather than having to have 50 different agreements,
- 11:27then we can just accelerate this data for transformation.
- 11:31And now I'd like to bring in Doctor Jen Laden, Asto, Senior
- 11:35Vice President of Population Health and Innovation.
- 11:39Dr. Laden, I think that was a pretty compelling argument for
- 11:44public health and, and private partnerships.
- 11:48What an incredible tool that sounds like.
- 11:51And, you know, the mind reels it.
- 11:53The the opportunities for clinicians for for public health
- 11:57officials to sort of like use these tools to inform their
- 12:00decisions. Yeah, no, absolutely.
- 12:02And it's hard to follow and so, but happy to be here and really
- 12:06excited to have ASTO be partnering with Yale and the Pop
- 12:09Hive effort. You know, I think broadly, ASTO
- 12:12has, as part of their strategic plan highlighted, emphasized the
- 12:16importance of public private partnerships, not just to
- 12:19support the work that ASTO does, but really importantly to
- 12:21support the work of our members. So our state and territorial
- 12:24health agencies. And I think POP Hive is 1 great
- 12:28example of how we're doing that, how we're making data available,
- 12:32pulling all the data together in one place so that not just
- 12:37health agencies, but others that rely on the data can access it
- 12:40as well. I think there's some other great
- 12:42examples. Doctor Zink mentioned the, the,
- 12:45the US data tracker that John Hopkins working with, we're
- 12:48working with that effort. And I think it just it shows the
- 12:51way in which public health can be working with academia, it can
- 12:55be working with the private sector to tackle and and solve
- 13:00some really hard challenges that better serve our communities.
- 13:04And it can almost be overwhelming sort of the amount
- 13:07of data that sounds like our our at members fingertips.
- 13:12How do how can add a stone members sort of dig in to some
- 13:14of these, some of these tools, some of these partnerships?
- 13:18Yeah, and there's a couple ways. One, it's just to familiarize
- 13:21yourself with what's there. I think the tools become better
- 13:25the more folks use it, the more people provide guidance and
- 13:31thoughts of what would make it better.
- 13:33So I think one of the things that we've done recently is
- 13:36launch a group of, of of sessions to engage members
- 13:40across the state and territorial health agencies wanting their
- 13:42input to help shape and inform continuing evolution of what the
- 13:49data on the website looks like and what it serves.
- 13:52And Drive Zink touched on this a little bit, but can you talk
- 13:55about how you know, expanding or increasing use of these tools
- 14:00can actually reduce that administrative burden?
- 14:03Yeah. So, you know, in the data
- 14:05technology space, there's, you know, it's a constantly and
- 14:08rapidly evolving field. And in public health, you know,
- 14:12we often don't have the resources that the person power
- 14:15of financial resources, the technical, you know, backbones
- 14:19to do what's needed with the data, right.
- 14:23There's a, as you mentioned, an influx of, of data and it's so
- 14:27critical to the work of a public health, whether it's identifying
- 14:31an outbreak, a novel infection, you know, understanding the, the
- 14:36impact of, of climate, for example, on, on health, disease
- 14:39and health outcomes. But it sometimes takes time if
- 14:42we're just relying on ourselves to, to make that data
- 14:47meaningful, understandable, and then to disseminate it out.
- 14:51So it's such a tool like this, it really puts it at the
- 14:53fingertips. So it's it's bringing in
- 14:56different areas of expertise, bringing in additional person
- 15:00power to help us make data more understandable.
- 15:03So to me, I see this as an mentioned a tool, a tool that
- 15:07can be used not just by state health agencies, state health
- 15:10commissioners, but by the public as well to understand what's
- 15:13happening in their communities and surrounding communities.
- 15:18And Doctor Zink mentioned in the case of Yale, being able to have
- 15:22access to these. Different experts in different.
- 15:24Fields and and how this is the sort of spidery network of
- 15:28experts that know everything. I'd like to talk a little bit
- 15:30more about sort of the role of institutions, especially
- 15:33academic institutions, in developing these tools.
- 15:36Yeah, absolutely. Having been in the data space
- 15:38for for several years now, not just at here at ASTO, but at the
- 15:42federal level, at the state level, city level, I think one
- 15:45of the the the biggest takeaways I have is that public health
- 15:49can't do it alone in the data tech space.
- 15:51We won't have enough financial resources, but it's we don't
- 15:55have the technical capacity expertise.
- 15:58It's hard to hire, hard to retain the right expertise.
- 16:02And yet we are surrounded by folks that have expertise that
- 16:06can be so valuable, the private industry and healthcare and the
- 16:10vendor industry in academia, people that want to help solve
- 16:14really challenging problems. And that can be part of this
- 16:17solution. And so as I mentioned this
- 16:19strategic priority, a strategic plan of ASTO that that concept
- 16:23of public private partners ships is something that's really
- 16:26important to us in a priority. And certainly the benefit to
- 16:30health agencies seems pretty obvious to me.
- 16:34What are what are the companies getting out of this
- 16:37relationship? Yeah, it's a great question.
- 16:40I think that depends on what the public private partnership is.
- 16:43I can't speak for the Popeye per SE, but I'm sure that there's
- 16:48huge value in what they see as participation in this.
- 16:52From my experience working with private industry and others, you
- 16:57know, they one see the value and importance of public health.
- 17:01Many of them were helping with different solutions during other
- 17:05start of COVID and recognized the challenges we have in the
- 17:10data space and the data ecosystem.
- 17:12And they don't want us as a nation, as our communities to
- 17:16continue to struggle. So I think there is some genuine
- 17:20desire to help solve these problems and in their
- 17:22recognition that they can. I think at the same time as, as
- 17:26we're working on, for example, the Public Health Data
- 17:28Consortium, our, our model there is that we need to have a, a
- 17:32common united mission that we're all aiming for.
- 17:35And that's a improving data access and quality.
- 17:38And, and that's something that is of benefit to them because
- 17:41they want access to data for the types of customers that they
- 17:45serve. For example, working with Epic,
- 17:47you know, they serve hospitals and they need better types of
- 17:50data to understand the patients on the populations that they
- 17:54serve. So they see value.
- 17:56And I think that's a key with public, private academic
- 17:59partnerships is, is having a shared value, shared mission
- 18:04that you can unite, unite on. My perhaps how you approach it
- 18:07might be different, but I think that really ties groups together
- 18:11on these complex issues. And I'd like to close with sort
- 18:16of the thought that it seems, it seems to me like these, some of
- 18:20these tools and these partnerships were developed at
- 18:24least in the aftermath of COVID and certainly with lessons
- 18:27learned from the pandemic. And I'm wondering now that these
- 18:32tools sort of exist in the world, is the potential to at
- 18:35least, you know, mitigate the next emergency?
- 18:39Yeah, it's hard to, you know, mitigate some of these public
- 18:42health threats that pop up. But I think what we can do is,
- 18:46is help us be prepared to identify threats faster, to get
- 18:50information out there faster. I think one of the challenges we
- 18:53saw during COVID was in continue to see is diminished trust in
- 18:57public health. And part of that is getting data
- 18:59information in a way that people see the value of it.
- 19:02So I think every time there's a lesson learned and in how we
- 19:07respond in public health, similar to really any profession
- 19:11or a thing that we do, there's that responsibility to say, how
- 19:14can we do it different, How can we do it better?
- 19:17And some of these challenges that are really complex, how do
- 19:20we bring in different areas of expertise that can help us with
- 19:24these? Efforts.
- 19:27Doctor Zinc, any final thoughts? Yeah, I guess I wanted to add
- 19:30one thing about what is the value to others.
- 19:33I think as a state health official, I thought a lot about
- 19:35this being a gift to my former self.
- 19:37What has surprised me is the gift that this has been to so
- 19:40many others. And as Jen had mentioned, how
- 19:42many people are motivated and excited to be able to solve
- 19:46health challenges in partnership with us.
- 19:48So for example, the testing data that we have the NIH is using to
- 19:53better model what's happening with RSV, what's happening where
- 19:57cases are are happening across the country is helping, you
- 20:00know, clinical trials better understand what new therapeutics
- 20:03are needed in the future. I also think getting this
- 20:06information back to clinicians, you know, I remember taping up
- 20:09in our bathroom wall health advisories from our state and
- 20:13well, it's one way to get information out.
- 20:15But it's much easier if RSV's on the rise or we're seeing
- 20:18something new. IT comes in open evidence or up
- 20:21to date or some tool that I'm using at the bedside or Epic or
- 20:25Cerner. Where I'm actually documenting
- 20:27on my patient chart so that I can understand as a clinician.
- 20:30And to your question earlier, is this going to help reduce the
- 20:32next public health emergency? Every emergency starts with one
- 20:36case or one place. And the more that we're able to
- 20:39address issues and challenges early, the earlier we are from
- 20:43keeping this to be a large problem.
- 20:45So I, as an individual frontline clinician, when I have tools in
- 20:48front of me that allow me to detect measles quicker, to
- 20:51educate patients faster, to be able to report it to the state,
- 20:54quicker to know what's happening happening faster.
- 20:56I can help be a partner to my public health officials to
- 20:59keeping measles under control or being able to prevent the next
- 21:03pandemic. So I think we need to really be
- 21:05thinking about not just how do we respond, but how do we every
- 21:09single day address health challenges that are arising.
- 21:12And these data and tool can make a big difference for that.
- 21:16Doctor Jen Leyden and Doctor Anne Zank thanks both.
- 21:18Of you so much. Thank you.
- 21:20Good to. See you.
- 21:21Awesome. Thank you so much.
- 21:22Doctor Anne Zink, a senior fellow at the Yale Law School of
- 21:25Public Health and former ASTO president and chief medical
- 21:28officer for the state of Alaska. Doctor Jen Layden is ASTO senior
- 21:32vice president of population health and innovation.
- 21:36State and territorial health agency leaders are invited to
- 21:39join ASTO and the Pop Hive team for virtual engagement sessions
- 21:43to explore the Pop Hive platform, share feedback on the
- 21:46utility and applicability of the data, and help shape how it can
- 21:49generate value for public health programs and decision makers.
- 21:53Learn more and register through the link in the show notes.
- 21:58The flux within HHS has ongoing impact on the process for
- 22:01issuing public health guidance, approving and altering vaccine
- 22:05recommendations, and more. Congress remains interested in
- 22:08understanding the impact of reforms occurring within HHS,
- 22:12including in hearings centred around agencies and nominees.
- 22:15Find Asto's federal health policy update on the changes
- 22:19within HHS, as well as responses from legislators at the link in
- 22:23the show notes. Join ASTO for the webinar
- 22:26Thriving under pressure. Building resilient dialysis
- 22:29systems and teams. Understanding the structures,
- 22:32processes and practices that influence patient safety during
- 22:35times of stress is essential to informing priorities and
- 22:38targeting interventions that strengthen resilience.
- 22:41This one hour session convenes frontline clinicians, workforce
- 22:45leaders and public health partners to examine practical
- 22:47evidence informed strategies that enhance both system and
- 22:51worker resilience in the outpatient dialysis setting.
- 22:54The link is in the show notes. This has been public health
- 22:57review morning edition. I'm John Sheehan for the
- 23:00Association of State and Territorial Health Officials.