INTRO

PIPER

Hello and welcome to another episode of Inform Me, Informatics! This podcast has been on a bit of a hiatus, so if you still have our feed on your pod-catcher, or you found this podcast some other way—thank you for still listening!

In our last episode (think way, way back to when that came out), I spoke with our director Vivian Singletary on how the pandemic had made certain public health gaps a priority overnight—in that case, the need for speedy and accurate contact tracing. But it’s not just contact tracing that rose to the forefront over the last three years. The COVID-19 pandemic has brought into stark relief how many shortcomings still exist in public health infrastructure. 

One way that the CDC is addressing these shortcomings is through the Data Modernization Initiative, or DMI. DMI is a multi-year, billion-plus dollar strategy that funds core public health functions, from strengthening early warning systems on emerging health threats to boosting data reporting capabilities, building national data standards and equipping a robust public health workforce. This hugely ambitious initiative channels federal funding to individual public health agencies around the country to make progress toward core data modernization goals. Every agency is tackling the challenge with a collaborative mindset to limit the need to reinvent the wheel while also recognizing that there is no one-size-fits-all solution —which makes sense! What works for the Texas Department of State Health Services likely won’t translate directly to the Massachusetts Department of Health, or to the Puerto Rico Department of Health.

The Public Health Informatics Institute has been involved in a support role in DMI for the last couple of years, and last year, we launched a collaboration with CDC and the Council for State and Territorial Epidemiologists (CSTE) on a new initiative called DMI Stories from the Field, a project centered around exploring these diverse approaches to modernization and asking: what’s working? What isn’t? What approaches may be usable to other agencies, and which successful processes may be unique to the health department implementing them?

As part of this work, we had the opportunity to sit down with two outstanding modernization experts from the Ohio Department of Health: Chief Data Officer Jonathon George and State Epidemiologist Kristen Dickerson. Jonathon and Kristen shared with us how beginning a maternal health project all the way back in 2017 paved the way for them to get a running start on critical data modernization work and eventually weather the deluge of pandemic data.

Jonathan got us started by explaining a little bit about how Ohio has handled data in the past.

         JONATHON

So prior to the InnovateOhio Platform, data was typically limited to use within a specific program area. Most of the analytics being done and shared were static reports that focused on really, you know, specific descriptive statistics. So while there were instances of data sharing, it certainly wasn’t the norm and typically required individuals to manually update shared documents and spreadsheets. So I’d say definitely a lot of spreadsheet sharing and really manual processes going on.

Much of the day-to-day analytics consisted of hand reports that were created from transactional data systems. And when a new report was desired, typically, programs had to rely on IT or they would hire vendors to do special custom projects for them. Yeah, this is often time-consuming and can obviously become really expensive. So the launch of IOP really created an opportunity to have all of our data on a single platform and in a similar format. So data pipelines from our source data systems are automated to update on a set interval, most of them daily. So programs don’t need to rely on this manual reporting, you know, or updating spreadsheets anymore.

Because these data feeds are automated, we can create data products that are current and do not require somebody to manually download a report out of a system, clean the data, connect to the clean data, and update the reporter dashboard. Rather, we can automate all of that on the InnovateOhio Platform, which, you know, certainly, we’ve got to go through the first step of actually doing it the first time. But at that point, then, we’ve got that data feed automated, and it’s updating on a regular basis, which is nice because you’re creating things once and then kind of letting things run from there. You know, this helps to keep our stakeholders and staff informed, and it opens opportunities to analyze data in a lot of new and exciting ways.

PIPER

The Innovate Ohio Platform ended up officially launching in 2019, and as you can imagine, having this system in place a year before the COVID-19 pandemic ended up being very fortunate timing.

         Jonathon

In the early days of the COVID-19 pandemic, you know, we, understandably so, were faced with a barrage of questions and the need to get information out to leaders and to the public as quickly and accurately as possible. Well, it was difficult at the time, especially early on, and the problem was exacerbated by the need to first figure out where to get the data once a question was asked. So one way to figure out who the subject matter expert was to send an email or call them, you know, whenever a question comes up with their area of expertise. Obviously, this is inefficient, it’s prone to error, requires a great deal of person power, you know, to keep those lines of communication open. What’s more is that, often, the subject matter experts for a particular area were busy with multiple pandemic response activities.

One solution was to give individuals working on our COVID-19 data response team access to the system databases. But doing so still required multiple logins, manual data quality processes to get the data in the format that was needed. And the list goes on. In addition, running analytics on transactional data systems impacts that transactional system. It can cause low performance on both the data collection and the analysis side. So we realized that one of the first things that we needed to do is get all the relevant data into one place. This provided the COVID data team with a single place to access all the relevant data and develop automated processes. So for example, data transformation, data cleaning, data formatting, all of that could be automated. And then we can create the necessary dashboards and reports, and now those reports and dashboards are now hooked up and being fed by those automated processes so they’re updating on a regular cadence.

You know, once we had an approved data product, it was already set up to refresh at regular intervals. So it frees up the subject matter expert resources to answer additional questions and analyze data. The result was a much more efficient process and a large number of data resources for COVID response in the public. This is a real-world example of how much we could accomplish with IOP.

So I have to say, the biggest impact that IOP has had so far has been our ability to get information out to the public for COVID-19. I think that was certainly our biggest accomplishment with it, and it’s shown the most impact there. Looking back on how things were before IOP, I can’t imagine our being able to really automate and publish the amount of information we did without requiring a massive increase in resources. You know, the time it has saved with being able to automate processes, rather, it’s allowed us to shift our focus to really getting the information out and analyzing the data in ways that we simply didn’t have time for before the existence of IOP.

So that’s not to say that we didn’t have an amazing team of people doing a lot of great work during the pandemic. Quite the contrary. But we are able to complete a task and then shift our focus to new challenges and focus less on maintaining and updating dashboards and reports.

PIPER

Kristen also shared the epidemiologist perspective on how IOP empowered the state agency to not just respond to COVID, but to carry over this work into use cases beyond.

KRISTEN

So our IOP actually has allowed us to focus more on the big picture instead of focusing on the minute details and treating every data project or every response differently. So we’ve been able to shift our focus to building out frameworks rather than looking at each analysis as a unique challenge.

When mpox first arrived, we were able to use some of the COVID reports and adapt them to being able to report out mpox. From there, we actually developed a basic framework for reporting on all infectious disease outbreaks using our data that we’re collecting from our disease surveillance system. We were able to extend this to the recent measles outbreak that we just closed in Ohio as well. So rather than starting from scratch, we’ve developed queries and processes that just require very small adjustments to report out on entirely different infectious diseases. This not only streamlines the creation process but the communications and approval processes on our end as well. It leads to faster dissemination of the information to interested stakeholders and actually allows us to get more resources for response involved faster. The faster that we’re able to share our data, the faster we are at responding to public health issues, which actually does relate to better health outcomes in the long run.

One of the biggest impacts of IOP, it actually has given our leadership at both the state and the local level the ability to make faster decisions, because we’re getting the information and the data to them faster. Because of that, they’re able to implement really sound public health intervention sooner, which, again, improves those health outcomes, which is what our end goal is.

PIPER

And it wasn’t just public health leadership and the boots-on-the-ground public health workforce benefiting from the availability of all this data! Jonathon also shared with us how IOP enabled Ohio to make public dashboards available to the community.

JONATHON

One of the goals was to be transparent. I mean, here’s this new disease that’s coming out, and everybody wants to know what’s going on, what direction is this going, what are the dangers. So, you know, getting this information out was critical. You know, we have a ton of data, and we want to keep the public informed on all things public health in Ohio, not just COVID-19. But COVID was definitely probably the first big example of being able to put that out. The way we really did it, we had a team of folks that really worked kinda day in and day out, going through this data, working with the program who has that expertise, and then creating these datasets, creating these data flows to create dashboards off of, and then getting that information posted up on our coronavirus website to make sure that the public’s informed.

 We want to be as transparent as we can, you know, with our data. So IOP and the DataOhio Portal has really made that possible.

             KRISTEN

When you start to give people the data, then they start to become more engaged, because it’s not just a bunch of numbers on a page anymore. It’s in an interactive dashboard where they can go to their county and they can compare their county to another county. And then it becomes their data, not our data, and it’s, like, giving them ownership of their data. And so they become more engaged, but in doing so, they also become more aware of the diseases that are around them and, like, what’s happening.

So, then, it’s not just allowing leadership to make better decisions on public health intervention. It’s actually giving community members down to, like, the individual level the ability to make informed decisions themselves and also ask questions to try to understand more about the disease process or how it’s spreading or what puts their community more at risk than another community. By sharing this data and the information, it does engage the population and the community, but it does come with more questions. But I think that’s a good thing, because it shows that people are showing an interest, and they want to know. And they want to make decisions, and they want to base it on information that they’re seeing and the science that’s behind it.

JONATHON

So the ODH, in partnership with our IOP team, we developed the social determinants of health dashboard from really publicly available data sources to provide a geographic look at over 100 different metrics that impact health. So by itself, it’s informative with respect to economic vitality, neighborhood and physical environment, healthcare access and quality, education access and quality, and social and community environment at both the census tract or county level. Some are only available at the county level, but several of the measures are available at the census tract as well. However,by itself, it does not show the relationship between these factors in public health issues.

So one of the great things that IOP has made possible is that we can more easily combine data, specifically, in this case, social determinants of health data, with data from other public health programs. So for example, you know, perhaps we wanna see which census tracts in Ohio have the lowest immunization rates for child immunizations. Right now, we could create a dashboard and see rates by geographic location. But we don’t necessarily know why the rates are low. We can just see that, like, this one, you know, this rate is lower than another census tract that’s maybe nearby.

Using social determinants of health data, we can overlay or combine that social determinants of health data with our immunization rate data and gain more insight into some of the potential barriers these communities face. As the example with immunization, so perhaps we learned that transportation and access to healthcare providers is a barrier. From there, we can do a more focused assessment of the factors that are influencing those low immunization rates and hopefully develop impactful interventions based on what is learned. It allows us to really take a more community-focused approach to public health and should lead to new ideas in a healthier Ohio.

KRISTEN

from a local level, a lot of our local epidemiologists and the health commissioners are engaging with this dashboard, and I know that we have used the dashboard to inform decisions about vaccine distribution and also reaching out to health departments where there may be a lower immunization rate to say, “Hey, it looks like there’s an opportunity here. Can we support you in any way? Would you like staffing?” Because we have some of that ability to do that through the COVID funding that came through vaccine just to support more.

I would say, though, when you’re talking about the social determinants of health and health equity, it’s not just public health’s job. It actually is the job of all the agencies. Like, Education plays a role, Job and Family Services play a role, and Medicaid and Medicare play a role. So there’s so many different factors that factor into health equity and social determinants of health. That’s why the move that Ohio is making on this IOP platform where all state agencies are going to be expected to put their datasets, it will allow us to do some major projects that could actually impact social determinants of health upstream. Instead of just seeing a low immunization rate and reaching out, maybe we can start to partner with other agencies and work across statewide enterprises to really hit these health problems head-on together.

So I think that’s the integration. It’s not just with our social determinants of health dashboard. It’s the ability to look across agencies’ different datasets because, before, it was rather siloed. But putting it all on one platform allows us to look at things in new ways.

PIPER

The value that IOP has brought to Ohio’s public health efforts has been immense. But, of course, launching and using a platform of this scale isn’t always smooth sailing. Jonathon and Kristen shared some of their challenges and lessons learned with us. Kristen got us started by talking about the critical importance of change management on a project this ambitious.

KRISTEN

From my perspective, the biggest challenge that we had for data modernization was taking those first initial steps and kind of being the guinea pig, because most of the data related to COVID-19 was coming out of our infectious disease unit, which is where our immunization program is housed. So data modernization does mean change. It means it’s changing the process, the way we do things, the way we report things. And to anybody, change can be very scary. So there was a little bit of hesitation from the team in wanting to move forward, and that’s why it’s really important to communicate and engage people from the beginning, because then you bring them along with you, and they’re working with you, and it’s not just a directive that’s coming down. So it’s really important to start that communication early and bring them along with you and realize that modernization is change, and it can be scary, and just kinda keeping that in the back of your mind any time that you go to communicate about data modernization to teams.

There was a little bit of hesitation from my team just adopting because it’s new, and putting the information in a dashboard was new to them, and sharing at that level was new to them, because, in infectious diseases, you’re used to reporting numbers annually and having time to clean it. So it did take a little bit of a push. And then once they saw how they could automate the numbers and it reduced the strain on the team, and they were able to free themselves up to actually do epidemiology instead of just, like, some of the manual data processes that they were doing and the cleaning, it made them buy into it. And then, all of a sudden, IOP becomes this great tool, and now that we have this, we can go do X, Y, and Z that we never had time to do before.

So I think that COVID was a challenge, and it was stressful. But there’s a lot of good things that have come out of it, and we’ve been able to develop some processes and tools that can be applied not just to infectious diseases but to other things. What we’ve done can be shown as a success, and this is where we started, and this is what we were now able to do. And because of IOP and having those tools, we were freed up to answer the questions, to do the epidemiology, to start to investigate clusters, to try to prevent outbreaks from happening.

So I am very thankful that we have IOP and the tools that we have because, at first, it was stressful, trying to put everything onto it and getting teams to kind of buy into it, but once that was finished, it has really alleviated stress in mpox and in measles and the other infectious diseases as we start to stand up dashboards and public-facing or even internal dashboards just for our own use. It really makes it easier because, then, we’re not reporting through email, like a sitrep, every day. It’s just there, and people can look at it. And it frees stuff up to do more important work.

JONATHON

So I think there’s always a bit of uncertainty when it comes to change, especially when it falls in tech and tech-adjacent fields like data. While the IOP employs a change management team to help communicate and facilitate change at a high level, it’s important that ODH does so on really more granular levels. To do so, we want to work directly with programs to help them transition their data and analytics work to IOP. So rather than just providing them with resources to learn the tools and skills necessary to navigate that platform, we want to really get in the trenches with them to help determine how we can best help them. I think it’s incredibly important to incorporate their real-world day-to-day activities and training. This way, they’re not only learning new skills, but they’re accomplishing something that’s relevant to their daily activities. I think this is gonna really help kinda show real-world examples and show them the value that IOP can provide.

So really breaking down that fear of change and of that tech stuff by just getting in there and kinda doing some of the work with them and showing them, like, “Hey, look, this may look scary or may seem difficult,” but I think, as we go along, I think people are gonna start to realize the benefits of that, and I think it’s gonna pay big dividends for us in the long run. It already has.

PIPER

After reflecting on the successes of IOP so far, we asked Jonathon to look ahead and share with us a glimpse of what’s next for IOP and Ohio.

JONATHON

So we are in the early stages of the data catalog process. That’s one of the big things for our DMI implementation. And we’re really kind of set to really start rolling that out on a larger scale in the next couple of months. So at this stage, we’re really piloting kind of the metadata collection process and have been meeting with the IOP team to really give feedback on the plan and get a basic understanding of what the software can do. That said, people are probably sick of me saying this, but I think that data catalog is probably the single most important foundational element that we’re doing in the first phases of our DMI implementation.

The analogy I like to use is to imagine that you’re going grocery shopping and there’s no signs and no organization to the grocery store. There’s just food on shelves everywhere, and you have to wander around, you know, trying to find things that you might need. But you really aren’t sure what’s available, you don’t know what all food is there, and there’s no easy way to figure it out. Similarly, you know, while some organizations, or some organization, I should say, exists on IOP, the data catalog allows you to not only define what data elements you have but to track your data lineage, you can link policy and relevant documents, you can develop common queries, analyze the quality of your data, and much more. The data catalog really serves as your inventory and allows you to put all the information about your data in a single place. And I really truly believe that this is going to lead to more innovation, enhance data quality, and really greater efficiency both now and in the future.

PIPER

We wanted to know what advice Jonathon and Kristen would have for other agencies looking to follow a similar path to modernization. Jonathon was happy to share some key words of wisdom.

JONATHON

These are actually the things that I need to hear as well. So this is not just what I’m telling people. It’s kind of follow-my-own-advice type of stuff here. You know, first off, everyone is in the same boat. I think it’s really helpful to talk to others that are going through the same process. I can’t tell you how many times meetings with leaders from other states has really helped me to understand that what we are doing is not easy. It requires a great deal of time and effort. It really does help though to know that you’ve got peers in the same boat, and it also helps to listen to see what your other peers are doing. It can give you new ideas and help kind of break down if you’ve got kind of a roadblock or something. It really helps to get that perspective.

Second, I would say, do not let perfection get in the way of progress. This is a process, and you will not get it perfect no matter how hard you try. So instead, try new things. Talk to your peers. Talk to your customers. Work to make things better than they were yesterday. That is really the goal. Can we improve on something to make things better than they were?

Some other advice I would have, you know, ideally, you should include members of the executive leadership on your steering committee. So, as you’re moving through the process, whether it’s developing a plan or actually implementing initiatives, I think it’s really important to get their feedback, you know, to make sure that we’re aligning with agency perspective and to know when to pivot when something is not working out and how to go about it or how to change when you need to.

Third, communication is key, working with leadership to ensure that everything aligns with the agency goals, making sure that you’re listening to the leadership to say, “What are the needs of the agency regarding data and analytics?” And then try to come up with some solutions that can make that possible.

Make sure that you get your ideas out there early and often. From there, find out which programs are ready to move forward and work with them to gain some momentum and actually show some results from your modernization initiatives. And then, from there, utilize those successes to get additional buy-in from other agency programs.

Be open to feedback even if it’s not what you want to hear. The goal is to learn and to make things better for everyone, your employees as well as the public. I mean, you know, we’re all here to drive things forward, and everybody has a voice. So it’s important to listen and communicate.

PIPER

And finally—we normally ask all of our guests, “how do you define informatics?” but we mixed it up a little bit this time by asking Kristen and Jonathon how they define data modernization. Kristen took her stab at the question first.

             KRISTEN

To me, data modernization means that we get to take time to look at our current processes in our systems that we have now but then also know where we want to go and then try to work on the path to get there. The ultimate goal for all public health departments, I think, would be to improve health outcomes. So we all know kinda our end goal.

Data modernization also means that we need to start looking across the dataset continuum and look at the whole person. So maybe it isn’t just data that comes into public health. Maybe it’s data that Medicaid has. Maybe it’s data that our prison system has. And how can we look across datasets to do better for the public and to do better for people?

JONATHON

What does data modernization mean to me? So, to me, data modernization is always changing. It really means building out a foundation by developing an agency strategy and organizing all the pieces of the puzzle. The COVID-19 pandemic drove many of our data modernization efforts forward out of necessity, but it was often done in really the most efficient or the fastest way possible. And in many cases, for the first time in ODH, so this means that there really wasn’t a lot of documentation or clear procedures for how to replicate some of the processes. Data modernization starts with getting our strategy put together and communicated throughout the department. From there, we need to update and create policies and processes to ensure that we can maintain momentum and clearly outline expectations for our future direction.

Ideally, it’s making all things data and analytics more approachable, efficient, and transparent for both our internal staff and our external stakeholders. 

PIPER

Many thanks to both Jonathon George and Kristen Dickerson for taking the time to speak with us on Ohio’s data modernization journey. If you want to explore this and other stories on data modernization from PHII—including written stories, videos and data modernization tools—check out phii.org/dmistories. Want to learn more about the Innovate Ohio Platform? Visit innovateohio.gov/platform. Both these links are also in the show notes.

This podcast is a project of the Public Health Informatics Institute, which is a program of the Task Force for Global Health. Visit phii.org to learn more about all of our informatics work! You can also find us on Facebook and LinkedIn.

As always, the music used throughout our show was composed by Kevin MacLeod. Thank you to Jessica Cook, PHII alumna and rockstar communications consultant, who conducted the interview you just heard. Thanks also go to informatics and DMI subject matter expert Sarah Shaw, fellow communicator Shawn Eastridge, and PHII project managers Santita Hooper and Mondie Tharpe. This story could not have been told without you! We’re also grateful for the partnership of CDC, the funder for this work.

I’m Piper Hale, and you’ve been informed!

[BUTTON]

JESSICA

Oh, well.

JONATHON

Oh, no, we lost her.

JONATHON

You know what, hold on one second, Jessica. I think Kristen just walked in the room. So one second.

JESSICA

OK!